Climatic

tic

Ch
M.

American Geophysical
Washington,

Union

ng
I. Budyko

Published under the aegis of the AGU

Translation

Board:

Helmut

E.

Landsberg, Chairman; Ven Te Chow,

Andrei Konradi, William H.K. Lee,
John Lyman, Frank T. Manheim,
David H. Miller, and Ned A. Ostenso.

Translation 1977 by the American GeophysicalUnion

Russian Edition 1974by Gidrometeoizdat,Leningrad
ISBN

0-87590-206-5

Library of CongressCatalog Card'Number: 77-81868

Printed by
Waverly Press, Inc.
Baltimore,

Foreword

At the beginning of the 1960sthe opinion was voiced that noticeable, world-wide changescouldtake place in the climate in the very near
future, due to increasedindustrialization and energy production. Since
such changes would have a significant bearing on all fields of human
endeavor, the task of utilizing this forecast in economicplanning became urgent. Shortly afterwards, research on anthropogenic climatic
variations commencedin several countries and the results were regularly discussedat national and international scientific meetings. These
studies confirmed the pressing need for developing methods to predict
future climatic variations causedby man's economicactivity.
It has been recently establishedthat it is theoretically possibleto
influence global climatic conditionsand to prevent undesirable changes
in climate by modern technical means.
A foreknowledge of changes in global climate resulting from the
influence describedabove has becomea practical necessity.
Thus it is essential to study the mechanism of climatic changes
taking place at the present time, as well as those that occurredin the
distant past.
Our previous book Climate and Life" was in part devoted to this
subject. Since its publication, however, much new research has been
conductedemphasizingthe urgency of this problem. This made it necessary to publish an additional monographon climatic changes.
The author wishes to expresshis sincere gratitude to all who have
read and commentedon the manuscript of this work.
Leningrad, 1974.

Some small changesand additions elucidating the works carried

out in 1974-76 are introduced

into the American

edition

of the book.

Leningrad,

Figuring the nature of the times deceas'd;

The which observ'd, a man may prophesy,
With a near aim, of the main chance of things
As yet not come to life; which in their seeds,
And weak beginnings, lie intreasured.
Shakespeare,Henry

Contents

Foreword

Chapter 1. Introduction 1
1.1. The Purpose of This Book 1
1.2. Sources of Information

on the Climate

1.3. Climates of Different Epochs 8

1.4. Factors of Climatic Changes 17
Chapter 2. The Genesis of Climate 31
2.1. The Energy Balance of the Earth 31
2.2. Semiempirical Theory of the Climate

44

Chapter 3. Contemporary Climatic Changes 72

3.1. The Warming During the First Half of the
Twentieth Century 72
3.2. Causes of Climate Warming 80
Chapter 4. Climates of the Past 98
4.1. The Quaternary Period 98
4.2. Pre-Quaternary Variations

111

Chapter 5. Climate and the Evolution of Living Organisms

5.1. Ecological Systems 132
5.2. Climatic
5.3. Climatic

Factors in Evolution
143
Factors in Human Evolution

161

Chapter 6. Man's Influence on Climate 171

6.1. Changes in Local Climate 171
6.2. Changes in Global Climate 184
Chapter 7. The Climate of the Future 197
7.1. The Outlook for Climatic Changes

7.2. The PossibleEcologicalCrisis 219

7.3. Climate

Conclusion

Modification

245

Summary 246
Bibliography 248

236

197

1 Introduction

1.1 The Purpose of This Book

The subjectof climatic changeshas attracted the interest of many

scientists.Their works have been chiefly devotedto the compilationand
study of empirical data on climatic conditionsduring variousepochs.
These studiespresent extensivematerial dealing with the climates of
the past.
Fewer results have been obtained, however, in studying the causes
of climaticchanges,althoughscientistsworking in this field have long
focusedtheir attention on these elements. Due to the lack of precision of
climatic theories and the inadequacyof specialobservationsnecessary
for this purpose,many obstacleshave arisen that have not yet been
overcome,in studying the causesof climatic variations.
Hypothesesexpressedin researcheson the reasonsfor climatic
changeshave very often remainedunconfirmedby numerical calculations or observationaldata, other contentionsevoking all sortsof objections. That is why there is no commonlyacceptedopiniontoday on the
causesof climatic changesand fluctuations, for the current epochas
well as for the geologicalpast.
Nonetheless, the mechanism of climatic changes is currently of
much greater importancethan ever before.Several studieshave found
that man's economicactivity has begun to affect climatic conditionsona
global scaleand that this influenceis increasingvery rapidly.
This makes it highly essentialto find ways and meansof predicting
climatic variations that would so deteriorate
as to make them hazardous for mankind.

environmental

conditions

It is obviousthat such predictionscannot be based exclusively on

empirical data about past climatic changes,though suchdata couldbe
useful for extrapolating currently observedclimatic variations for a
projectionof future conditions.This methodof forecasting,however, is
suitable only for very limited time intervals, sincethe numerousfactors
affecting the climate are highly variable.
In as much as the human economy is increasingly influencing
atmosphericprocesses,a physical theory of climatic changesmust be

CLIMATIC

CHANGES

used in order to developa reliable method of predicting the climate of

the future.

The achievementsof modernclimatologyresearchare great, but it

must be bornein mind that the quantitative modelsof the meterological
regime are highly approximate, and the principlesthey are basedon
involve various important limitations.
Obviouslyall available empirical data on this subjectare of particu.
lar importance in developingand verifying approximate theories of
climatic

variations.

This likewise applies to the study of probableconsequences

from
changesin the global climate that may be expectedin the foreseeable

future.

Prospectsfor sheddinglight on the mechanismof climaticchanges

are enhancedby the rapid developmentin recentyears of a new branch
of sciencecalled physicalclimatology,which applies the methodsof
theoreticaland experimentalmeteorologyfor studyingclimaticconditions. Many researchreportsin this field explain variousregularitiesin
climate genesis, i.e. the conditionsof climate formation in various
regionsof the globe.
It may be assumedthat the use of modern methods in physical
climatologywill facilitate the study of the causesof climatic variations
on a wider basis than the majority of researchwork carried out in this
field to date.

This bookis devotedto theseproblemsand is basedmainly on data

acquiredby the author and his colleaguesat the A. I. Voeikov Main
GeophysicalObservatory.
Chapter i givesa brief descriptionof climatic conditionsin the past
and a survey of various opinionson the causesof climatic changes.
Chapter 2 presentsa descriptionof a semiempirical theory of climate,
which is applied in Chapters 3 and 4 to clarify the causesof natural
climatic changesof the past and present. In Chapter 5 we discussthe
influenceof climatic changeson the evolutionof living organisms.
The final chapters of this book are devotedto the effectsof man's
economicactivity on the climate at the present time and in the future.
Thesechaptersdeal with the possibilityof an ecologicalcrisisarisingas
a result of the anthropogeneous
variations of climatic conditionsas well
as with ways and means of preventing climatic variations that could
prove hazardousto man's economy.
1.2 Sources of Information
Climate

on the

We glean the greatestamount of knowledgeabout climatic

I INTRODUCTION

tions on our planet from what has been called the era of instrumental
observations,during which systematicmeasurementsof a number of
meteorologicalvariables were made by means of standard instruments
at a large number of stations.It is difficult to name the exact beginning
of this era becauseit took a long time to set up and develop the vast
network of meteorologicalstations which now exist throughout the

world.

Although the prototypesof most meteorologicalinstruments were

developedduring the seventeenthand eighteenthcenturies,the number
of meteorologicalstationsconductingcontinuousobservationsremained
very small up to the middle of the nineteenth century, especially in
tropical countriesand sparselypopulatedregionsat other latitudes. By
the latter half of the nineteenth century, a more or less numerous
network of such stations was already functioning on all continents
exceptthe Antarctic. Now there are over 10,000suchstationsand their
programs include measurementsof temperature, humidity, precipitation, pressure, wind velocity, and several other meteorologicalelements. There also exists a systemof actinometricobservationswhich is
of vast importance for studying the climate and its variations.
Today the actinometricnetwork comprisesabout 1,000stationsthat
conduct observations of incident
at the earth's surface.

solar radiation

and its transformations

In recent years, meteorologicalsatellites have facilitated the study

of the climate by supplying very important material, including data on
cloudsand the radiational regime at the outer boundary of the atmosphere.
Mention must be made of special research being carried out by a
limited number of stations, such as observationsof the compositionof
the atmosphereand of the solid and liquid particles it contains, atmospheric aerosol. These observationscommencedonly a few decadesago
and their importancefor the study of climatic changescannotbe overestimated.

Yet it must be noted that the existing system of meteorological

observationsthroughout the world is, in many respects,inadequatefor
studyingclimate and its variations. For instance,there is comparatively
little information about atmosphericconditionsover the vast part of the
earth's surface coveredby oceans.There are at present relatively few
weather ships (and those situated chiefly in the North Atlantic) that
conductobservationsof all important meteorologicalelements in certain
regions. Data on the meteorologicalregime of other regions of the
world'soceansare suppliedmainly by passingships. This information is
inferior to that obtained by meteorologicalstations on the continents,
both in quantity and

CLIMATIC

CHANGES

Many meteorologicalparameters vital for climatic studies, such as

turbulent heat flow in the atmosphere, evaporation, heat consumption
during evaporation, etc., are not recordedat the majority of meteorological stations, although methods for doing so have been developedand
are being applied by a small number of research stations.
No direct methods for observing horizontal heat advection, heat
release during condensationof water vapor, and other climatic elements
have as yet been developed.
Despite its limitations, the world's system of meteorologicalobservations has yielded a great deal of material in the past century enabling
scientists to study the variations of the atmosphere's thermal regime
near the earth's surface,precipitation, pressure,winds, and many other
meteorological elements.
Earlier data for the meteorologicalstudy of certain regionsare also
available, but the territory covered by these observations, carried out
during the first half of the nineteenth and particularly during the
eighteenth century, comprisesonly a small part of the earth's surface.
A certain amount of information on climatic conditionsprevailing
in the period of two to three thousand years before the era of instrumental observationsmay be found in the writings of contemporaries.These
findings vary widely in character; some were obtained in the courseof
several decadesand involve a detailed descriptionof day-to-dayweather
changes, while some present brief notes of great catastrophes that
attracted universal attention, such as droughts, hurricanes, torrential
rains, etc.
Since this sort of information refers only to a very limited number of
regions, it may be applied mainly to the study of climatic changes in
those particular locations.
Considerably more important for the study of the climates of various epochsare data on the natural conditions of the past. Since such
processesas the formation of sedimentary deposits,the weathering of
rocks, the developmentof reservoirs and the existence of living organisms dependedon atmosphericfactors, available data on these processesallow us to estimate climatic conditionsduring given periods.
It is very difficult indeedto interpret data on the natural conditions
of remote epochsto clarify the climatic regime, though someof these
data are of fundamental importance. One such difficulty arises due to
the need for applying the principle of uniformitarianism to these studies. This means acceptingthe assumptionthat in the past the relations
betweenthe climate and other natural phenomenawere similar to those
existing today. While this approachis by no means beyondcontention,
the great variety of natural phenomenadependingon the climate allows
us to independentlytest this hypothesis,by reconstructingthe

1 INTRODUCTION

conditionsof the past in accordancewith divers paleogeographicindications. There is no doubt that the general laws and patterns of climatic
conditionsduring the geologicalpast, as establishedin paleogeographic
research,hold true, although certain results of these investigationsare
in dispute and require further confirmation.
An essential complementto the paleogeographicdata in studying
climatic conditionsof the past are materials on the paleotemperatures
obtained as a result of analyzing the isotopic compositionof organic
remains. The task of estimating the precision of these data and interpreting them correctly is fraught with great difficulties, but these are
gradually being overcomewith the developmentof methodsfor paleotemperature research.
The study of past climates involves the application of information
on the structure of sedimentaryrocks, and on geomorphologicalproperties, as well as data on fossil flora and fauna.

The former approach is based on the currently well-established

relations between the lithogenetic and the climatic factors. For example, under the conditions of a hot and humid climate, an intensive
chemical weathering of rocks takes place leading to the decay of chemically nonresistant minerals. Chemical weathering is less intensive in a
hot and dry climate but the destructionof rocks by winds and temperature fluctuations begins to play a seriouspart. In a cold climate, chemical weathering is even less marked and physical weathering predominates, so that the chemically nonresistant minerals are preserved.
Humidity has a telling influence on the amount and structure of
sedimentary deposits. In dry regions the amount of such deposits is
scarce,while in humid regionsit is much greater, and includesalluvial
depositsto a great extent.
The processof coal formation is closely connectedwith climatic
conditions, which means that data on fossil coal can be used for reconstructing past climates. The interpretation of these data, however,
presents certain difficulties becausethe dependenceof coal accumulations on climate during different periods of time varied greatly in
accordancewith the changing character of vegetation from which it was
formed. For instance, while many of the coal beds of the Devonian
period were formed under dry climatic conditions, coal accumulation
during the Permian and Carboniferous times was associatedwith comparatively humid climatic conditions.
When studying the climates of the past we also have at our disposal
data on sediments of limestones, dolomites and salt-containing deposits
which are valuable aids in reconstructingclimatic conditionsof ancient
reservoirs.

In somecases,data on the structure of sedimentaryrocksis used

CLIMATIC

CHANGES

estimatingseasonalclimaticconditions.
The bestknownexampleof this
is the studyof claysformedin regionsnear the peripheryof continental
glaciations.
Duringthesummermeltingofglaciers,thestreamswashed
fromtheir surfacesa largeamountof coarse,fragmentedmaterial,and
duringthe coldseason,a smalleramountof fine-grained,argillaceous
matter wasdeposited
in theseregions.The stratifiedstructureof varved
claysenablesus to estimatethe durationof their formationperiod.
In orderto studythe regime of atmosphericprecipitationand the
formation of the ice cover, we have wide recourseto geomorphological

data,includinginformationonthe positionofthe oceancoastline.These

allow us to estimate the water loss and gain by the oceandue to the
formation and melting of continental galciation.

A major sourceof informationon the developmentof glaciersare

data on changesin the surfacerelief underthe influenceof glaciation.
Anotherimportantaid in determiningpast climatesare data on the
altitude of the snow-linein the mountains,as its positiondependson the
temperature and precipitation.
Pertinent informationon humidity conditionsmay be gleanedfrom
data on ancient lakes and river valleys. For instance,tracesof numerouslakes and rivers on the territory of present-daydesertsattestto the
drastichumidity changesthat occurredafter the epochin question.
Data on the level oscillationsof land-locked reservoirs, the Caspian

Sea, for example,allow us to estimatethe water inflow into this reservoir and, consequently,
the amountof precipitationin the basinof its
feedingrivers during variousperiodsof time.
For the studyof humidity and thermal regimesof the past,information on the nature of fossilsoilsmay be used.Tracesof permafrostin the
soilis of importancefor reconstructing
zoneswith a coldclimate.
Likewiseimportantfor paleoclimaticinvestigationsare data on the
nature of the erosionprocess,which also dependsgreatly on climatic
conditions,and particularly on humidity.
Data on the geographicaldistribution of living organisms,and
especiallyof vegetation,which is highly dependenton climaticconditions, are alsoa majorcontributingfactorin the studyof pastclimates.
The applicationof this methodyieldsmore reliable resultsfor the
not-too-distantpast whenthe vegetationdifferedonly slightlyfrom its
modern forms, and climatic conditionsaffectedits distribution in the
same way as today.

For reconstructing
the climatesofmoreremoteepochs,
it is difficult
to applythe relationbetweenthe vital activityofpresent-day
vegetation
and meteorological
factors,and this limits the use of evidenceon the
distribution of plants in paleoclimaticstudies.

Onegreat achievement
of the last decades
wasthe useof a

1 INTRODUCTION

for analyzing fossilpollen and spores,from which we get a pretty good

idea of the vegetation coverof a given region. This methodis alsoeasier
to applyto epochsin which the vegetationwas similar to that of today.
Annual growth rings on trees tell us a great deal about the climatic
history of the past. These rings reveal short-term climatic fluctuations
while their structure is an indicator of general climatic conditions.For
example, weakly developedgrowth rings on trees in carbonaceous
swampsindicate the absencein this caseof distinct seasonalvariations.
It is more

difficult

to utilize

material

on fossil fauna

since the

dependenceof the geographicaldistribution of animals on the climate is,

as a rule, not as pronouncedas that of vegetation. Nevertheless, these
data, and particularly data on poikilothermal (or cold-blooded)
animals,
are a valuable complementto other methodsfor the study of climatic
variations. The interpretation of data on the distribution of animals as
well as of vegetation for the purposeof estimating climatic conditionsis
basedon the assumptionthat the influenceof climate uponthe life of the
past organisms was similar to the influence upon the related forms
existing today.
In paleoclimatology
it is of specialimportanceto applymethodsfor
determining paleotemperaturesfrom the content of the oxygen isotope
Osin the fossilremains of aquatic animals. It has been establishedthat
the ratio of O s to O in shells of mollusks

and other remnants

of sea

organismsdependson the temperature under which these organisms

lived.

The determination of paleotemperaturesfrom the isotopiccontent

of specimensunder study made it necessaryto develophighly sensitive
mass spectrometersand to solve a number of other technical problems.
Beginning with the middle of our century, these methods of direct
measurements of paleotemperature gained wide recognition in the
study of recent climatic conditionsand thosethat existedover hundreds
of millions of years ago.
When evaluating the resultsof studyingclimatic changesby means
of all the above describedmethods, it shouldbe mentioned that, with the
exceptionof a relatively short recent period, nearly all the data on past
climatesrefers to the air temperature at the earth's surface,to that of
land and bodies of water, and also to the moisture conditions on the
continents.

Several other climatic elements may also be gleaned from paleographicdata: in somecaseswe candeterminethe directionof prevailing
winds from the forms of fossil dunes and barkhans; but this information
is comparativelynegligible.
Data on the air temperature near the earth's surface is of paramount importance in studying climatic conditionsalso in the era

CLIMATIC

CHANGES

instrumental observationsas well, for two reasons:firstly, becauseof

the closerelation between the air temperature and the componentsof
the heat budget which greatly affect the climate, and secondly,because
nearly all other elementsare registeredat meteorologicalstationswith
a relatively lower precisionthan the temperature, thus limiting the use
of these data in the empirical study of climatic variations.
1.3 Climates of Various Epochs

Modern Climate. For comparisonwith climatic conditionsof the

past, let us briefly describemodern climatic conditions,with the emphasis on two meteorological elements-air temperature at the earth's
surface and the amount of precipitation.
Table I represents data on mean latitudinal temperatures at sea
level for January and July and on atmosphericprecipitation over the
entire year. We seefrom this table that the mean temperature varies by
approximately 70Cfrom the maximum value at the equator to the
minimum value at the south pole. The reason for the temperature
variations is discussedin detail in the next chapter; here we only
mention that due to the sphericalform of the earth, the solar radiation
incidentuponthe outer boundaryof the atmospherevariesgreatly with
latitude.

At high latitudes where the air temperature is below the freezing

point for water during the whole or greater part of the year, there is a
permanentice cover.In the Arctic the main part of this coverconsistsof
Table 1

Air Temperature and Precipitation

Temperature

Latitude
(N)

Temperature

Pre-

(oc)

cipi-

(C) PrecipiLatitude
tation
JanuJuly (cm/yr)
ary

90-80

-31

80-70

-25

-1

(os)

January

ta-

tion

July

(cm/

yr)

19

0-10

27

25

147

26

10-20

26

24

129
85

70-60

-22

12

97

20-30

25

18

60-50

-10

14

72

30-40

2O

14

92

50-40

-1

20

78

40-50

12

102

97

40-30

11

26

77

50-60

30-20

19

28

70

60-70

20-10

25

28

117

70-80

-8

10-0

27

27

192

80-90

-13

-12

67

- 30

25

-42

1 INTRODUCTION

comparatively thin sea ice, while in the Antarctic it consistsof a thick

continental glaciation covering almost the entire surface area of the
continent.

Along with the strong meridional variations, the mean temperature at the earth's surface in the majority of latitude zones varies
significantly with longitude. These changesare causedmainly by the
distribution

of continents

and oceans.

In middle and high latitudes the air temperature in the summer is

considerablylower over the oceanthan over the continents,while in the
winter it is higher. This is due to the high thermal capacity and
conductivityof water, causingthe oceansto absorb a great amount of
solar energy in the summer and to discharge it during the winter
season.A telling effect on the thermal regime is producedby ocean
currents, particularly by warm surfacecurrents that transfer a large
amount of heat during the cold seasonfrom low and middle to high
latitudes, where it is released and contributes to the maintenance of

comparatively high air temperatures.

Oceans have a strong climatic influence on continents in middle
and high latitudes. Here the so-calledmaritime climate prevails, characterized by a small temperature range throughout the year. In the
nontropicalparts of continents,where the influence of the oceanis less
noticeable, seasonal temperature variations increase sharply, correspondingto the conditionsof a continental climate.
As to the distribution of the zonal mean precipitation, we find a
primary maximum in the equatorial region, a decreasein subtropical
latitudes, secondary maximums in middle latitudes, and a decrease
with increasing latitude to the polar region.
The variations of the zonal mean precipitation are governedby the
distribution of the average air temperature and the peculiarities in the
atmosphericcirculation.
Other conditions(includingrelative air humidity being equal, the
total precipitation increases with temperature since the amount of
water vapor consumedin the condensationprocessalso increases.It is
obviousthat if no other factorshad any seriouseffect on precipitation,
its latitudinal distribution would have a single maximum in the lower
latitudes.

However, for falling precipitation,the field of vertical air motionsis

of paramount importance becauseit affects the water vapor transfer
acrossthe condensationlevel, leading to the formation of clouds and
precipitation.
General atmosphericcirculationis closelyrelated to the geographical distribution of semi-permanentpressuresystems,among which the
most important are the low pressure band near the equator,

10

CLIMATIC

CHANGES

pressureregionsin the tropicallatitudesoutsideof the equatorialbands

andin subtropical
latitudes,andregionswith a high cyclonic
activityin
middle latitudes.Sincedescending
air currentsprevail in the high
pressurezones,the amount or precipitationin this region decreases,
producingtwo minimumson the curveof the latitudinal distributionof
precipitation.

The considerable
intensityof ascending
air currentsin the equatorial andpartsof middlelatitudesis responsible
for increased
precipitation in theseregions.
The earth'slargestdesertslie in the subtropical
highpressurezone,
wherethe amountof precipitationis very low. The precipitationdeepin
the continents in middle latitudes also decreases as the wind currents

carryvery little watervaporfromthe oceanthat far, thusleadingto a

decrease
in the relative air humidityanda weakeningin the condensa-

tion processes
of water vapor.

Hence, the zones of humid climate on the continents are situated

mainly in the equatoriallatitudesand in regionsof oceanicclimatein

the middleand high latitudes.Conditions
of insufficienthumidityprevail in hightropicalandsubtropical
latitudesandin regionsofcontinental climate.

Time Scale. In discussingavailable empirical data on climatic

fluctuations,we shall deal with four time intervalswhosespansincreaseas we go downinto the ages.The first of theseintervals refersto

the periodof instrumentalobservations

a little morethan onecentury
old.The secondinterval, the Holocene,coversan intervalfollowingthe
end of the last glaciation,that is, a little morethan 10,000years.The
third interval, the Pleistocene,lasted about 2 million years. When
studyingthe Pre-Quaternary,or fourth interval, we shall concentrate
mainly on the last 200million years, for which data on climaticcondi-

tions are more reliable.

Note that the duration of eachinterval is 100-200times greater

than the previousone. This divisioninto intervalsis dictatedby a
decreasein the amountand accuracyof data on climaticconditionsfor

more remote epochsin comparisonwith lessremote.

When discussingpast climatic conditions,we should take into
accountinformationon the sequenceand durationof variousstagesin
the geologicalhistoryof the earth. Somedata on thesestagesare given
in Table 2.

The Tertiary periodis usuallydividedinto twoparts:the Paleogene

(whichincludesthe Paleocene,
EoceneandOligocene),
andthe Neogene
(which includesthe Mioceneand Pliocene).Table 2 doesnot give the

namesof the epochsin the Pre-Cenozoictime, and theseare not referred

to in the

1 INTRODUCTION

11

The Paleozoicera was precededby the Proterozoic,which lastedat

least two billion years. The Post-Proterozoic
time is often called the
Thanerozoic.

The geochronological
scalepresentedin Table 2 is basedon successive changesin lithogenesis,flora and fauna establishedduring the
studyof the earth'scrust.The divisionof the earth'sgeologicalhistory
into periodswhich takes these changesinto accountis inevitably
fraughtwith a marginof uncertainty;therefore,discrepancies
asto the
mostreasonableconstructionof the geochronological
scalecanhardly be
avoided. For instance, a consideration of the Holocene as a separate

epochof the Quaternaryperiod,is oftensubjectto dispute,becauseits

duration comesto less than 1% of the total span of the Quaternary.

Without pausingto discussthese discrepancies,

which are of no
importance
in the studyof pastclimates,we wouldlike to pointoutthe
limited precisionof the data givenin Table2 on the durationof various
Table

Geochronological Scale
Duration

Period

Quaternary

period (million years)

Epoch

Cenozoic

of

Absolute age
of beginning
of period
(million
years)

era

Holocene
Pleistocene
Pliocene
Miocene

Tertiary

Oligocene

64

66

132

Eocene

Paleocene
Mesozoic

era

Cretaceous

66

Jurassic

53

185

Triassic

50

235

Paleozoic

era

Permian

45

28O

Carboniferous

65

345

Devonian

55

4OO

Silurian

35

435

Ordovician

55

49O

Cambrian

8O

CLIMATIC

12

CHANGES

geologicalperiods and, consequently,on their absolute age. The probable error of this dating, in any event, is not lessthan several per cent of
the values given here.
Contemporary Climate Changes. For a brief description of climatic conditions during various epochswe shall resort to material
contained in a series of monographspublished in recent years (Bowen,
1966; Flint, 1957; Inadvertent Climate Modification, 1971; Lamb, 1973;
Mazkov, 1960; Rubinshtein and Polozova, 1966; Schwarzbach, 1961;
Sinitsyn, 1967; Velichko, 1973).
The most significant climatic change in the period of instrumental
observationscommencedat the end of the nineteenth century. It was
marked by a gradual increase in the air temperature over all latitudes
in the northern hemisphere, during all seasonsof the year, while the
greatest warming occurredin high latitudes during the cold half of the
year.

The warming was continuousthroughout the 1910sand 1920sand

reached its maximum in the 1930swhen the mean air temperature in
the northern hemisphere registered in increase of about 0.6 as compared to the end of the nineteenth century. In the 1940sthis warming
gave way to a coolingprocesswhich has continued to recent time and
resulted in lowering the mean air temperature by 0.3-0.4.
Although data on modern climatic changesin the southern hemisphereare not as clear as those for the northern hemisphere,there are
groundsto believe that, there too, a warming occurredin the first half of
the twentieth century.
In the northern hemisphere the rising air temperature was accompanied by a decrease in the surface area of the polar ice cover, a
recessionof the permafrostboundary into higher latitudes, a northward
expansionof the forestsand tundra boundaries,and by other changesin
natural

conditions.

Of great importance was the marked change in the regime of

atmospheric precipitation during the warming. The amount of precipitation in several regions of insufficient humidity dropped even lower
during the warming, particularly in the cold period of the year. This
resulted in a decreaseof river runoff and a drop in the water level of
several

land-locked

reservoirs.

In the 1930sthe drop of the water level of the Caspian Sea, caused
mainly by the reduced discharge of the Volga River, was particularly
noticeable. At the same time the warming brought more frequent
droughts in vast territories of intracontinental regions at middle latitudes in Europe, Asia and North America. This climatic changeaffected
the economiesof many

1 INTRODUCTION

13

The Holocene. The Post-Quaternary climatic fluctuations comprise several large climatic fluctuations.
The W(irm glaciation reached its maximum development about
20,000years before our time. In several thousandyears the area of this
glaciationwas considerablyreduced.The succeedingepochwas marked
by a comparatively cold and humid climate in the middle and high
latitudes of the northern hemisphere. About 12,000 years before our
time, there was a significant warming (the AllerSd epoch), which,
however, soon changed to cooling. During these climatic fluctuations,
the summer air temperature in Europe underwent a change of several
degrees.
Later, the warming resumed and the last big glaciations in Europe
and North America disappeared 5,000-7,000 years before the present.
During this period the post-glacial warming reached its peak. It is
assumedthat about 5,000-6,000years ago the air temperature in middle
latitudes in the northern hemisphere was about 1-3 higher than it is
today.
Changesin the atmosphericcirculation apparently occurredsimultaneously with the northward shift of the polar-ice boundary. The
subtropicalhigh pressurezonealsomovedto higher latitudes, resulting
in the expansionof drought areas in Europe, Asia and North America.
At the same time precipitation increased in low latitudes, over the
region which are desert today. During this epoch the climate of the
Sahara was comparatively humid and it possessed
rich flora and fauna.
Later there was a tendency toward cooling which was particularly
noticeable in the early half of the first millenium of our epoch. Parallel
with the change in the thermal regime, the precipitation regime also
varied, gradually reaching its present condition.
A noticeable warming took place at the end of the first and the
beginning of the secondmillenium A.D. During this time the polar ice
retreated into high latitudes, making it possible for the Vikings to
colonizeGreenland and leading to their discoveryof the continent of
North America. A later cooling led to a fresh encroachmentof the ice
cover, and the Greenland colony, no longer able to communicate with
Europe, perished.
The cooling which commenced in the thirteenth century and
reachedits culmination at the beginning of the seventeenthcentury was
accompaniedby the growth of mountain glaciers, wherefore it is sometimes referred to as the Little Ice Age. Later another warming and a
recessionof glaciers took place.
In the eighteenth and nineteenth centuriesclimatic conditionswere
more or less the same as they are

14

CLIMATIC

CHANGES

The Pleistocene. During the Pleistocene the climate differed

greatly from the previous conditionsof the Mesozoicera and the Tertiary period, when the thermal zoneswere barely distinguishable. In
the course of the Pleistocene the cooling in middle and high latitudes
intensified, bringing about the formation of large continental glaciations. We know very little about the number of these glaciationsor their
ages; it is generally assumedthat the first of them originated about
700,000 years ago.
Researchwork carried out in the Alps in the last century revealed
evidence of four major European glaciations: Guntz, Mindel, Riss and
Wiirm. It becameknown later that each of these glaciations had several
stageswith intervals of recession.The Alpine glaciations coincidedwith
the development of the ice cover on the plains of Eurasia and North
America, stretching over vast areas in high and middle latitudes.
It would seem that the periods of glacier advance and recession
comprisedthe lesser parts of the total duration of the Pleistocene,while
the comparatively warm interglacial epochswere of longer duration. In
the courseof the warm epochsthe continental ice coverremained only in
mountain regions and high latitudes.
It has been establishedthat the advanceand recessionof glaciersin
Europe, Asia and North America were more or less synchronous,and
this also applies to the ice ages in the northern and southern hemispheres.
The continental ice cover expanded during the ice ages mostly in
the regions with a humid oceanicclimate, while in the comparatively
dry climate of northern Asia it occupieda very small area. During the
most severe glaciations the continental ice cover in the northern hemisphere reached an average latitude of 57N, while in certain places it
extended as far as 40N. Its thickness over a considerablepart of the
area came to several hundred meters, even reaching a kilometer or more
in some places.
There is no dcbt that during the expansionof continental glaciation the boundary of the polar sea ice also migrated to lower latitudes.
This considerablyincreasedthe total area of the permanent ice coveron
the planet.
During each advancement of glaciers the snow-line in the mountains which were beyond the glaciated areas droppedseveral hundred
meters, and sometimes a kilometer or more.

In addition, the area of permafrost expandedsignificantly during

the ice ages. The permafrost boundary shifted several thousand kilometers into the lower latitudes.

A characteristicfactor of the ice ageswas the lowering of world

1 INTRODUCTION

15

level by 100-150 meters as comparedto its present one; this was the
result of enormousquantities of water being "lockedup" in the thick
continental glaciations. In the warm interglacial epochssea level rose
severaldozenmetersas comparedwith the presentlevel. The fluctuations of sea level changedthe surfacearea of the oceansby severalper
cent.

The climate in the ice ageswas marked by a noticeabledropin air

temperature over the entire globe. This temperature drop averaged
several degrees and increasedwith latitude. During the interglacial
periodsthe temperaturewas higher than today, though this warming
was not as marked as the cooling.
The picture of precipitationchangesassociatedwith glaciationsis
not soclear. Available data indicatethat glaciationsexerciseda varying
influence on humidity in different parts of the globe. This testifies to
changesin the atmosphericcirculation systemdue to glaciation and is
related to fluctuationsin the temperature gradientsbetween the poles
and the equator.
At the height of the glacial developmentthere was a decreasein
precipitation over many continental regions, apparently due to the
reducedevaporationfrom the oceansurface,an extensivepart of which
was coveredby ice.
It should be noted that the first continental glaciations at high
latitudespresumablytookplaceas far back as the Mioceneepoch.Their
centerswere situated in mountainousregions.
The development of large continental glaciations not connected
with mountain systemsis a characteristic feature of the Pleistocene
period.
Pre-Quaternary Time. Our knowledgeof the Paleozoicclimatic
conditionsis extremelyvague.Apparently, during the major part of this
era (which existed 570-235million years before present), the climate
throughout the globe was very warm and humidity conditionson the
continents extremely diverse. By the end of the Paleozoic era and
aroundthe Carboniferousand Permian periods,a glaciationdeveloped
that spreadover a vast area, the major part of which is not tropical.It is
difficult to estimate the geographicalpositionof this glaciationat that
time becausethe continents and the earth's poles have since shifted
greatly. Climatic conditionsin other regionsof the globewere typically
quite warm during this Permo-Carboniferousglaciation. The thermal
zonesin the Permian periodbecamepronouncedand the regionsof dry
continental climate vastly expanded.
The Mesozoicclimate (235-66 million years before present) was
rather uniform. Most of the earth resembledthe tropicstoday;while

CLIMATIC

16

CHANGES

high latitudes the climate was more moderate,although very humid,

and fluctuated only slightly with the seasons.The humidity conditions
on the continentsduring the Mesozoicera were apparently more uniform as comparedto recent times, though there were zonesof insufficient and excess moisture.

At the end of the Cretaceousperiodthe hot zonesdiminishedwhile

the dry regionsexpanded.During the transition to the Cenozoicperiod
there were no marked climatic changes.The secondhalf of the Tertiary
period(from middle of Oligocene)was characterizedby the beginningof
a temperaturedropthat was more pronounced
in the middle and especially in high latitudes.From that time on a new climaticzoneemerged
in the higher latitudes and gradually expanded, its meteorological
regimeresemblingthe climaticconditionsin the middlelatitudestoday.
In this zonethe air temperaturedroppedbelow0Cin winter, resulting
in the formation of a seasonal snow cover. At the same time, the

continental climate became more extreme in the central parts of the

continents.

The coolingdid not proceedat an evenrate; at certain periodsthere

were warm spellswhich, however,did not affect the generaltendency
toward the formation of thermal zonescausedby the temperature drop
in high latitudes. This processacceleratedduring the Pliocenewhen the
first large continentalglaciation, still existing today, emergedin Antarctica.

The climate of the Pliocene was warmer

than ours but came closer

to it than that of the Mesozoicand the early half of the Tertiary period.
A studyof successive
climatic variationsover the periodsfor which
more or less reliable data are available, leads us to the conclusionthat
regimes with strongly pronouncedthermal zones are not typical of
climatic conditions on our planet.
The great temperature gap between the poles and the equator,
which has existed since the end of the Tertiary period and increased
particularly during the ice ages,is typical for only the small fraction of
time that has elapsedsince the onset of the Paleozoicera. During the
last 600 million years there was only one large Permian-Carboniferous
glaciation (besidesthe Quaternary glaciations), and its duration was
also small as comparedto the time that elapsedsincethe beginningof
the Paleozoic

era.

At the same time we cannot be certain that the big difference in

temperature betweenthe polesand the equator existed when this glaciationdeveloped.(This questionis dealt with in more detail in Chapter
4.)

Data on climatic conditions in the Pre-Cambrian time are not very

reliable. For instance, there is no consensusof opinion on the

1 INTRODUCTION

17

tion of the till-like rock which some scientistsconsiderto be depositsof

large g]aciations. Though we may assume that g]aciations occurred
during the Proterozoic era, their position relative to the earth's pole at
that time is totally unclear.
Thus, the climatic conditions of the last comparatively brief interval in the earth's geologicalhistory with distinct thermal zonesprobably
have no close analogy with any past climates. They differ drastically
from the conditions that prevailed in the courseof the greater part of the
last 600 million years, when hot or warm climate prevailed at a]!
latitudes, when there were small temperature gradients between the
polesand the equator, and a small variation of the annual temperature
beyond the tropical zone.
1.4 Factors

of Climatic

Changes

The reasons for climatic changes may be divided into two groups.
The first group coversexternal factors that undoubtedly influence the
climate, although opinions differ as to the scale and character of this
influence. The secondgroup covers the factors whose influence on the
climate is not beyond dispute; some authors acknowledge and others
deny the role of these factors in climatic changes.
The factors of the first group include changesin atmosphericcomposition, in the form of the earth's surface, and in astonomical elements
defining the position of the earth's surface relative to the sun.
Changes in Atmospheric Composition. Contemporary investigators devote special attention to the climatic implications of two variable
componentsof the atmosphere, i.e. the concentration of carbon dioxide,
and the amount of aerosolejectedinto the atmosphereduring volcanic
eruptions.
Over a century ago Tyndall (1861) pointed out that since carbon
dioxideand water vapor absorblong-waveradiation in the atmosphere,
the changesin CO2 concentrationmay lead to climatic fluctuations.
Later this problem was studied by Arrhenius (1896, 1903) and
Chamberlain (1897, 1898, 1899), who assumed that the change in the
amount of atmospheric carbon dioxide could have caused the Quaternary g]aciations.
Arrhenius studied the absorption of atmospheric radiation and
suggesteda quantitative model for determining the temperature at the
earth's surfaceas a function of atmosphericproperties.By applying this
model, Arrhenius deduced that a 2.5 to 3-fold increase of carbon dioxide
raises the air temperature by 8-9C, while a decreaseof 38-45% lowers
the temperature by

18

CLIMATIC

CHANGES

On the basis of geologicaldata Arrhenius found that the carbon

dioxide left in the atmosphere today is only a small fraction of the
amountconsumedby carbonaceous
sedimentsduring their formationin
the past. Furthermore, Arrhenius came to the conclusionthat the
concentrationof carbondioxidein the atmospherecouldhavechangedto
a large extent and that these changesgreatly affectedthe air temperature thus facilitating the emergenceand disappearanceof glaciations.
Chamberlain'sworkswere mainly devotedto the geologicalaspects
of this problem. He noted from a study of the balance of carbondioxide
in the atmospherethat the inflow of carbondioxidefrom the lithosphere
changedsignificantly with the level of volcanicactivity and other factors.

The consumptionof carbondioxideduring geologicalprocesses

was
also modifiedto a large extent, depending,in particular, on the surface
area of rocks subject to atmosphericerosion. With the increase of the
surfacearea the consumptionof carbondioxidecausedby the weathering processincreased.Chamberlainassumedthat the glaciationsoriginated as a result of the intensive processof mountain formation and the
rising of continents which led to an increase of erosion and of the
weathered rock surface,and, consequently,to a decreasein the concentration of carbondioxidein the atmosphere.To supporthis hypothesis,
Chamberlain carried out calculations;these, however, cannot be considered a completequantitative model of the given process.
Later the carbon dioxide hypothesisof the origin of Quaternary
glaciationsbecamedoubtful becauseit was discoveredthat in the 13-17
/m band there is a considerableabsorptionof radiation by water vapor
and not only by carbon dioxide, thus reducing the influence of carbon
dioxide amount on the thermal regime.
By taking this effect into consideration,Callender (1938) obtained
smaller values of temperature variations at the earth's surface with
fluctuations of the carbon dioxide concentration, than did Arrhenius.
Accordingto Callender, a doubling of the carbon dioxide increasesthe
air temperature by 2C,and this dependenceweakens with the growth
of carbon dioxide concentration.

Callender suggestedthat the warming of the climate in the first

half of the twentieth century was connectedwith an increase in the
carbondixoide concentrationin the atmospherecausedby man's indus-

trial activities.

Callender provedthe importanceof consideringthe absorptionof a

long-waveradiation by water vapor, for correctlyevaluating the effectof
carbon dioxide concentrations on the temperature. However, Plass
(1956)and Kaplan (1960)did not take this into accountwhen carrying
out similar

i INTRODUCTION

19

Plass has obtained comparatively large temperature variations at

the earth's surface:the temperature increasesby 3.6 when the carbon
dioxide concentration doubles, and decreasesby 3.8 when the concentration drops to half of its present amount. This has enabled him to
revert to the carbon dioxide hypothesis on the origin of Quaternary
glaciations.Plass suggestedthat the changesin the carbon dioxide
concentrationsin the atmosphereand oceanwere of an auto-oscillatory
nature.

Plass claimed that when there is a small initial

decrease in atmos-

pheric carbon dioxide, the temperature at the earth's surface drops,

resulting in the growth of the polar ice mass.This causesa reductionin
the volume of ocean water, which leads to an increase in the atmos-

pheric carbon dioxide, and consequently,to a new warming. This, in

turn, causes the ice to melt, the volume of water in the ocean to
increase, and the amount of carbon dioxide in the atmosphereto drop,
whereupon a new coolingcommences.
Kaplan showedthat the cloudcoverconsiderablyweakenstemperature variations at the earth's surface causedby carbon dioxide fluctuations, as comparedto the estimatesgiven by Plass.
Kondrat'ev and Niilisk (1960), as well as MSller (1963) stressedthe
necessityof taking into accountthe absorptionof long-waveradiation by
water vapor in calculatingthe atmosphericthermal regime. The values
of temperature variations obtained by Kondrat'ev and Niilisk were
much smaller than those obtained by Plass and Kaplan.
MSller pointed out that the variations in air temperature are generally accompaniedby a changein absolutehumidity, while the relative
humidity remains more or lessconstant.The increaseof the absoluteair
humidity with temperature enhancesthe absorptionof long-waveradiation in the atmosphere, and this leads to further temperature rises.
MSller has established that the temperature increase at the earth's
surfacecomprises1.5when the concentrationof carbondioxidedoubles
while the absolute air humidity remains constant, whereas with constant relative humidity this increase is several times greater. At the
same time

MSller

noted that

the influence

of carbon dioxide

on the

thermal regime can be compensatedby comparativelysmall changesin

the absolute air humidity or cloudiness.
A detailed investigation of the effect of carbon dixoide on the air
temperature was carried out by Manabe and Wetheraid (1967). They
pointed out the inaccuracyof the calculationsmade by MSller, who
estimated the thermal regime variations in the atmosphereonly on the
basis of the thermal balance of the earth's surface, without taking into
accountthe thermal balance of the atmosphereas a whole.
Manabe and Wetheraid calculated atmospheric temperature

20

CLIMATIC

CHANGES

files, taking into considerationthe absorptionof long-waveradiation by

water vapor, carbondioxide and ozone.In their work they relied on the
profilesof relative air humidity basedon empirical data. It was assumed
that the temperature distribution was determined by conditionsof local
radiation balance, if the vertical temperature gradient did not exceed
6.5C/km. The given value of the temperature gradient was considered
to be the maximal possibleone due to the limiting effect of convective
processes
on the growth of the vertical gradient.
Manabe and Wetheraid found that for average cloudinessand constant relative air humidity, the doubling of carbondioxide concentration
increasesthe temperature at the earth's surfaceby 2.4, while one-half
of the present concentrationdecreasesthe air temperature by 2.3.
In subsequentworks of Manabe and Wetheraid (1975) the effect of
variations of CO concentration on temperature at different latitudes
was calculated. Using three-dimensionalmodel of the atmosphericgeneral circulation they foundthat by doublingCO concentrationthe mean
air temperaturerisesapproximatelyby 3,this value beingconsiderably
increased

with

increase in latitude.

Air temperature fluctuations at various latitudes for a large range.

of carbondioxide concentrationswere calculated by Rakipova and Vishnyakova (1973). They used a two-dimensionalmodel of the atmospheric
thermal regime which took into accountthe radiation heat flows, the
release and consumptionof heat during condensationand evaporation,
the turbulent heat exchange in the atmosphereand the heat exchange
between the ocean surface and its lower layers. In calculating the
radiative flux, the influenceof water vapor and ozonewas considered.
Rakipova and Vishnyakova calculated the air temperature variations at different

latitudes

and altitudes

as a function

of the carbon

dioxide concentration, which varied from zero to five times its present
level.

These calculations were carried out for the cold and warm parts of
the year, under conditionsof clear skies and average cloudiness.They
found that cloudiness, like the warm and cold periods of the year,
exerciseda relatively small effect on the relation between air temperatures and carbon dioxide concentrations.

It should be noted that, contrary to Manabe and Wetheraid, Rakipovaand Vishnyakova assumedthat the absoluteair humidity remains
constant, which means they neglectedthe effect of evaporation changes
with surfacetemperature. This, accordingto Manabe and Wetheraid,
leads to a certain underestimation of temperature fluctuations stemming from carbon dioxide concentrations.
All the foregoinginvestigations,of course,contain many simplifying approximations.In mostcases,this problemis lookeduponas a

i INTRODUCTION

21

one and doesnot take into accountthe possiblechangesin the general

air circulation resulting from air temperature variations.
Of interest

in this connection

are the latest works of Manabe

and

Wetheraid, Rakipova and Vishnyakova, which imply that average temperature variations calculatedfor various latitudes do not differ greatly
from temperaturescalculateddirectly on the basisof data averagedover
the planet as a whole.
It is important to note that in most of the foregoing works, the
feedbackbetween the atmosphericthermal regime and the surfacearea
of the polar ice cover was not taken into account.This effect was dealt
with in part in recent work of Manabe and Wetheraid, in which the
changein the area of snowand ice cover is consideredas a function of
the atmospheric thermal regime.
As a result there was obtained a greater variation of air temperature, especially for higher latitudes, with the fluctuation of the carbon
dioxide concentration in the atmosphere.
The influence of the changing polar-ice surfacearea on the thermal
regime of the atmosphere during fluctuations of carbon dioxide was
analyzedby Budyko (1972a,1973);theseresults are describedin Chapter
4.

This effect, it appears, is of great significance in clarifying the

regularity of climatic changesduring the last part of the Tertiary period
and Quaternary times.
A number of works have been devoted to the study of the possible
effects of atmospheric aerosol on climatic conditions. One of the first
hypotheseson the relation between the atmosphere'stransparencyand
the climate was put forward by Franklin in the eighteenth century.
Later, P. Sarasin and F. Sarasin (1901) presented the hypothesis
that the Quaternary glaciationswere causedby diminished atmospheric
transparency during periodsof increasedvolcanicactivity. This premise
was later supported by others (Fuchs, Patterson, 1947;et al.).
Humphreys dealt with this conceptin greater detail in the first half
of the twentieth century, claiming that after volcanic explosive eruptions, cloudsof small particles form in the stratosphere,which, he said,
consistedof volcanic dust. He estimated the effect of these particles on
the radiation regime and concludedthat they can significantly screen
the radiation incident on the earth's surface, without, however, causing
any appreciable effect on the long-wave radiation emitted into outer
space.

Accordingto Humphreys, the amount of particles ejectedinto the

atmosphereover the northern hemisphereby the eruption of the Katmai
volcanoin 1912was sufficient (if these particles remained in the atmosphere long enough) to bring about a drop of several degrees in

22

CLIMATIC

CHANGES

temperatureat the surface.Humphreysfoundfrom observations

that

after intensive volcaniceruptions, temperature dropsof several tenths

of a degreeactually took placeover vast areas. He claimedthat this

relatively small temperaturedrop was occasioned
by the fact that the
particlesof volcanicdusthad remainedin the atmospherefor a rather
short time.

Later Humphreys'viewswereconfirmedand elaboratedby Wexler

(1953,1956),who foundthat the finest dust particlescan remain in the
stratospherefor severalyears after volcaniceruptions.
Accordingto Wexler, the slowclearingfrom the atmosphere
of the
volcanicdust resulting from the major eruptionsthat took placeat the
beginningof the twentiethcenturycausedthe warmingthat characterized the 1920s-1930s.

An important role was attributed to volcanicactivity by Brooks

(1950).He compiled,on the basisof volcanicrock thicknesses,a summary of volcanicactivitybeginningwith the Paleozoicera andcontinuing to our times.

Brooksadvancedthe opinionthat the increaseof volcanicactivity

servedas one of the factorscontributingto the developmentof continental glaciation.
Among the most recent studiesof the effect on the climate of

volcanicactivity, mentionmustbe madeof the workpublishedby Lamb

(1970).It containsa list of explosivevolcaniceruptionssince1500which
considerablyreducedatmospherictransparencyfor short-waveradiation. Like Humphreysand Wexler, Lamb expressedthe belief that
volcaniceruptionsconstitutea majorfactorof changesin the climate.
This problemis studiedin worksby Budyko(1967,1969a,1972b),
and the results are set forth in later chapters.

Changesin the Form of the Earth's Surface. The problemof the

effectsof thesechangeson climaticconditionswasinitially taken up by
Lyell (1830-1833),
whocompared
climaticconditions
in twohypothetical
cases for the structure of the earth's surface. These were' 1) that the

polarzonesare occupied
by continents;and 2) that they are coveredby
oceans.In the first case,accordingto Lyell, the climate in high latitudes
must be significantlywarmer than in the secondcase.Lyell therefore
came to the conclusion that the evolution of the form of the earth's

surfacewas accompaniedby strongclimatic variations.

Voeikov (1902),Ramsay(1910),and Lukashevich(1915)attributed

great importanceof the changingtopography

on the climate.Voeikov
thoughtthat this contributionto climaticchangeswasgreaterthan that
of astronomicalfactors.Ramsayand Lukashevichsharedthe opinion
that the coolingwhich causedglaciationsoccurredas a result of the
rising of continentsand the curtailmentof the oceansurface

i INTRODUCTION

23

absorbsmore solar energy than the terrain. They both reached the
conclusionthat, during the developmentof glaciations,the air temperature droppedmainly in high latitudes, with an insignificant decreasein

low latitudes.

The effect on climatic conditions produced by changes in the

earth's surface was studied in greater detail by Brooks (1950). He
showedthat the coolclimate at high latitudes is causedprimarily by the
presenceof polar ice, sinceit doesnot absorbmuch solar radiation due to
its high albedo. Brooks also emphasized the role of ocean currents,
which manifestly reducethe differencein temperature between low and
high altitudes.
According to Brooks, a climate typical of most of the earth's

history,with warm or hot conditionsat all latitudes,prevailedwhenthe

continentswere low-lying and widely separatedby oceansand seas.The
elevation of the continents, especially in polar regions, led to their
coolingoff and the developmentof glaciationsthat later expandedinto

the middle latitudes.

Brooks also believed that the Permo-Carbonifer-

ous glaciation originated in the tropics, due mainly to the high level of
continentsand the changein the oceancurrents in these regions.
The fact that Brooksattributed great importanceto oceancurrents
is borne out by his contentionthat a slight intensification of the current
in the Bering Strait couldlead to the melting of muchof the pack ice in
the Arctic Ocean.

Albrecht (1947)was of the opinionthat the elevationof meridional

mountain chains, particularly the Rocky Mountains and the Andes,
contributedgreatly to the coolingof the climate during the Quaternary

period.Accordingto Albrecht,these mountainsblockedatmospheric

circulation to a great extent, decreasedthe amount of winter precipitation in the continentalregionsin the .middleand high latitudes and, as a
result, lowered the inflow of condensation heat.
A numberof researchershave discussed
the influenceof changesin
the relief of the oceanfloor on climate (Ewing and Dunn, 1956, 1958;
Rukhin, 1958).Theseauthorsexpressedthe opinionthat the emergence
of underwater mountain ridges in the northern part of the Atlantic,
stretching from Scotland to Iceland and Greenland, resulted in the

developmentof the Quaternary glaciations.This bottom relief sup. pressedthe Gulf Stream'sheat transfer and led to the coolingof the
polar latitudes, where cold and dry air massesbegan to form. The
interaction

of these masses with warm

and humid sea air resulted in

abundantsolidprecipitation, which led to the developmentof continental glaciers.

Another problem discussedin several works was the influence on
climaticchangesof continentaldrift and the motionof the earth's

24

CLIMATIC

CHANGES

The hypothesis of continental drift attracted great attention in the

1920safter the publication of Wegener's noted work. On the basis of this
hypothesis KSppen and Wegener (1924) reconstructedthe climates of the
past. Later this assumption evoked strong objections, but today it is
often referred to, in a somewhat revised form, in geological studies.
Polar motion has been the subject of much discussionin connection
with paleomagnetic data (which testified to the position of magnetic
polesin the past), and also in connectionwith ordinary paleogeographic
data.

It is very difficult to clarify this problem becauseof the complexity

of comparing paleomagnetic findings on different continents with each
other and with paleogeographicdata which are often contradictory.
It has been proposed that the location of the poles during the
Cenozoic and the greater part of the Mesozoic eras was close to their
present one, but a substantial change in this location occurredin earlier
epochs. This assumption was applied by Strakhov (1960) in his reconstruction of past climates. Other authors also accept the possibility of
marked polar motion in a later epoch as well.
Changes in Astronomical Elements. The French mathematician Adhmar (1842) was the first to point out that changes in the
elements of the earth's orbit could have been the cause of glaciations.
Later this concept was developed by Karol' and other authors, but
Milankovich (1920, 1930, 1941) subjectedit to a most detailed analysis.
In order to explain climatic changes during the Quaternary period,
Milankovich applied the data on secular changesof three astronomical
elements: the eccentricity of the earth's orbit, the obliquity of the earth's
axis to the ecliptic, and the precession of the equinoxes or, in other

words,the changein the date of the equinoxesin relation to the date of

the closestapproachof the earth to the sun. All these elements change
with time due to disturbances caused by the moon and other celestial
bodieson the earth's motion, the periodsof these changesbeing: 92,000
years for eccentricity, 40,000 years for obliquity, and 21,000 years
[shouldread 26,000years, Translator] for the precessionof the equinoxes.

Although the fluctuations of these elements do not affect the

amount of solar radiation incident on the upper boundary of the atmosphere, they do influence the amount of total radiation received by
separate latitudinal zones in various seasons.

Milankovich has developedan approximatetheory of the radiational and thermal regimes of the atmosphere, from which it followed
that the air temperature at the earth's surfaceduring the warm time of
the year, in middle and high latitudes, can vary by several degrees,
depending on the ratio of these astronomical

I INTRODUCTION

25

As a rule, this change coincidedmore or lesswith the temperature

changeduring the coldpart of the year (with a reverse sign). It is known
from empirical studies, however, that the glaciation regime depends
mainly on the temperature of the warm period (when the glaciers melt).
The thermal

conditions

of the cold season affect the ice cover to a smaller

degree.
By applying the said astronomical elements to studies of the
Quaternary period, Milankovich found that the intervals of tempera.
ture drop during the warm seasons,in the zone where the glaciations
mainly developed (60-70 latitude), correspondedto the ice ages, while
those of temperature increase correspondedto the interglacial warmings.
The first comparison of Milankovich's results with Quaternary
climatic conditions was made by KSppen, who asserted that these
results agreed well with the history of Quaternary glaciations (KSppen
and Wegener, 1924). Later, similar comparisonswere made by many
other scientists, whose conclusions,however, were contradictory. This
evokeda lively discussionas to the influence of astronomicalfactors on
the climate, and these arguments are still going on.
In the course of this discussion, several attempts were made to
compile quantitative modelsof glaciation development that would allow
us to estimate the changes in the size of polar ice covers as a function of
the earth's position relative to the sun.
In several cases these attempts led to the conclusion that this
influence is insignificant. But this apparently was due to the fact that
these models did not take into consideration

the feedback between the

ice cover and the atmospheric thermal regime. When this was done, a
definite correlation could be seen between these phenomena (Budyko
and Vasischeva, 1971; Budyko, 1972b;Berger, 1973).
It should be noted that Berger and other authors (Sharaf and
Budnikova, 1969; Vernekar, 1972) carried out new computations of the
secular variations of the astronomical elements that influence the distribution of solar radiation in various latitudinal
zones.

The probableeffectson climate of fluctuations in the solar constant

have been the subject of discussionin several papers. Unlike the meteorologicalsolar constant (i.e. the flow of short-waveradiation at the outer
boundary of the troposphere), which undoubtedly varies with he changing transparency of the stratosphere, the long-term variations of the
astronomical solar constant (i.e. the energy emitted by the sun), cannot
be considered established.

Many attempts have been made to determine short-periodvariations of the solar constant

on the basis of astronomical

observations.

Several authors have come to the conclusion that these exist; the

CLIMATIC

26

CHANGES

constant increases somewhat during periods of average solar activity

and decreases during periods of low and high activity (i.e. during
periods of very low or very high sun spot numbers).

Other authors(Allan, 1958;AngstrSm,1969)believethat if there

are any changes in the solar constant, they are within the margin of
error of present-day measurements.
It is very difficult to prove any changes in the solar constant
correspondingto the various climatic conditions in the past. Nevertheless, this hypothesis has been repeatedly referred to in several investigations, the best known of which are those by Simpson (1927, 1938)and

pik (1953,1958).

Simpson contendedthat the solar constant changed recurrently

during the Quaternary period and that the amount of precipitation, air
temperature and evaporation correlated with its behavior. Simpsonwas
of the opinion that the warm and humid interglacial periodstook place
during the maximum radiation, and that the cold and dry periods
occurred during its minimum.
Simpson assumedthat continental glaciations originated during
the periods of increasing solar radiation, when the temperature gradient between the pole and the equator increased, as did the atmospheric circulation, thereby enhancing evaporation, cloudinessand precipitation. Increased cloudinesskept the temperature from rising and
glaciers began to develop in high latitudes due to the increasedamount
of precipitation.
When the radiation increased to its maximum, the temperature
rose and, as a result, the glaciers melted. The subsequentattenuation of
radiation led to a repetition of the glaciation cyclewhich stoppedwith a
considerabledecreasein radiation when there was not enoughprecipitation to maintain the glaciation processas a result of insufficient evaporation.

(pik contendedthat the variations in the solar constantwere

caused by irregularities of solar radiation and exceeded20%, the radiation maximum being in the Tertiary period while the minima were
during the Quaternary ice ages. On the basis of his quantitative model

of the thermal regime, Opik determinedthe temperaturevariations

corresponding to these fluctuations of the solar constant, and decided
that this made it possibleto explain the main regularities in climatic
change.
Many astronomers do not share the oft-repeated opinion that the
sun is a variable star (Mitchell, 1965). The question of the changing
solar constant has been discussedrecently at a special conference(The
Solar Constant and the Earth's Atmosphere, 1975) at which there

1 INTRODUCTION

27

noted the absence of sufficient data to answer the question about these
changes.
Several articles have discussedthe probability of changes in solar
radiation reaching the earth, caused by the solar system passing
through a nebula, whose matter either screens the solar radiation
reaching the earth, or falls on the solar surface under gravity force thus
increasing the solar radiation. This hypothesishas not won wide recognition and is rarely referred to in climatic studies.
According to a more popular hypothesis, the solar radius, in the
course of the sun's evolution as a star, gradually diminished and its
luminosity increased.
For example, Schwarzschild (1958) established that during the

earth's history, solar luminosity could have increased60%. This correspondsto a 1% gain in the solar constant over 80 million years. Such a
changein the solar constantcouldnoticeably have affectedthe climate
of remote ages.
Another aspect of the influence of solar radiation on the climate is
related to changesin solar activity. Many works apply available data on
solar activity to the study of various atmospheric processes.It is assumed in these works that the flow of solar radiation

in the ultraviolet

and several other spectral bands can vary substantially, while the solar
constant remains stable. The effect of these fluctuations on physical
processesin the upper atmosphere is well known. However, its influence
on processesin the troposphere that determine weather variations is not
so clear. The assumption has been voiced that the change in solar
activity, characterized by the number of sunspots,has exerciseda great
influence on the climate and causedthe Quaternary glaciations (Huntington and Visher, 1922). It is difficult to discuss these hypotheses
becausethere is no commonly acceptedopinion on the relations between
solar activity and tropospheric processes,and also because of the absenceof reliable data on solar activity in the past, before solar observations began.
Internal Factors. Let us describe more precisely what we mean
by external and internal climatic factors before we proceed with our
study of climatic changes under constant external conditions.
We have already mentioned the external factors determining the
climate, including solar radiation, the structure of the earth's surface,
and the atmospheric composition,the latter depending on the interaction of the atmosphere with the lithosphere and with living organisms.
Sometimes parameters of oceanic conditions, sea and continental
ice, the properties of atmosphericcirculation, and even a meteorological
element such as cloudiness are considered as external

climatic

28

CLIMATIC

CHANGES

We have repeatedly mentioned that the atmosphere'scirculation, which

is characterized by fields of pressure and winds as well as cloudiness,are
as much internal climatic factors as the field of air temperature and
precipitation. All these elements are interconnected and none of them
can be consideredindependent of the other factors determining climatic
conditions.

We must also mention among internal climatic factors the parameters

of oceanic

conditions

and

of the

ice cover

over

land

and water

reservoirs.

It is obvious that the genesis of the climate, in a broad sense,

embraces not only atmospheric but also physical phenomena in the
hydrosphere and the entire complex of hydrometeorological events over
the land surface, including the development of glaciations, if any.
Thus, the relief structure which determines the surface area and

the depth of seas and oceans,as well as the surface area and altitudes of
continents, should be considered a climatic factor, and is called the

underlying surface. Changes in the underlying surface, for instance,

glaciations, must be consideredamong the climatic elements which are
determined by the same external factors (incident radiation, relief
structure and atmospheric composition), although in particular cases
they can be looked upon as factors of climatic genesisat a given period.
When considering the atmosphere-ocean-ice cover system, we
must ask whether climatic conditions in this system can change when
external

climatic

factors are constant.

Lorenz (1968) studied this problem by analyzing equations of the

theory of climate. He showedthat two solutionsof these equations were
possible. The first solution gives one type of stable climate which is
obtained as a result of averaging instantaneous values of fields of
meteorologicalelements over a long-range period. This type of climate,
called "transitive" by Lorenz, is single-valued under the given external
conditions.

The secondsolution presents several types of a stable climate for

the same external conditions ("intransitive"

climate). In this case the

climate can change significantly without any visible external causes.

This subject is further elaborated in the next chapter. Lorenz has
expressed the interesting idea that the climate in our times can be
quasi-intransitive. This means that a climate which is determined over
a long but finite time interval can depend on the initial conditions and
therefore be different for various time periods, under constant external
factors. Whether the present-day climate is quasi-intransitive or not is

still opento ques.tion.

The possibility of auto-oscillatory processesdeveloping in the atmosphere-ocean-icecover system is also of vital importance in

I INTRODUCTION

29

standing climatic changes. The simplest example of such an autooscillatory system is presentedby the level of oscillationsin a drainless
reservoir. As shown by the authors, and Yidin (1960), the level of such a
reservoir cannot be stable even in the absenceof climatic changes, if its
shores are vertical, i.e. if its surface area does not depend on the level.
The reason for this instability is that the sum of random annual oscillations of the componentsof the reservoir's water balance increases indefinitely with time.
Since the shoresof natural reservoirs are sloped,the dependenceof
the reservoir's surface on its level serves as a stabilizing factor on the
oscillations.

In this work it was established

that if the successive annual

oscillations of the Caspian Sea level were mutually independent, the

total amplitude of these oscillations would be equal to the sum of
random annual level oscillations. In this casethey would be an example
of a pure auto-oscillatory process.
Drozodov and Pokrosvkaya (1961) showed that the successiveannual oscillationsof the Caspian Sea level are not independent. A study
of their interaction lead to the conclusion that the main part of the
amplitude was caused by external factors (i.e. by general climatic
fluctuations), a significant part of this amplitude being a consequenceof
an auto-oscillatory process.

Far more complicatedauto-oscillatoryprocessescan developin the

atmosphere-ocean-polarice system. Models of such processeswere first

analyzed by Shuleikin (1941, 1953).
The possibility of explaining Quaternary climatic changes as autooscillatory processeshas been discussedin several papers including the
above works of Plass and of Wilson

(1964). The latter's

work was

elaborated by Flohn (1969) who explained the Quaternary ice agesby an

instability of the Antarctic glacier that causedauto-oscillatory processes
in the atmosphere.
Drozdov and Grigor'eva (1971) adopted the same approach to the
study of cyclic fluctuations of atmospheric precipitation over the territory of the U.S.S.R. They came to the conclusionthat these fluctuations
were of an auto-oscillatory nature. By using empirical relations established in this and preceding works, it was possible to predict the expected precipitation regime for the next decadefor various regions for
the U.S.S.R.

The Present Stage of the Problem. Aside from the propositions

outlined above, there are several other hypotheses which, for various
reasons, are less frequently referred to in climatic studies. However,
those we have gone into comprisean adequate list, illustrating the wide
range of opinions on the causesof climatic changes.
All contemporary reviews note the absenceof a generally

30

CLIMATIC

CHANGES

theory on this problem (Schwarzbach, 1950, 1961, 1968; Sellers, 1965;

Sinitsyn, 1967; Mitchell, 1968; Shell, 1971).
We think this is inevitable at a stage when many aspects of the
problem are dealt with mostly in qualitative terms, without any reliable
quantitative calculations.

It may be assumedthat a preciseestimate of the problemwill be

reached only after general theories have been developedon climate and
on the evolution of the atmosphere, ocean and upper layers of the
lithosphere, and that of the sun. Since it may take several decadesto
adequately develop these theories, it becomesurgent at the present time
to develop approximate methods for studying climatic changes which
will enable us to obtain quantitative estimates of the effects of these
factors that so obviously influence the climate. If these hypotheses
succeedin explaining the chief regularities observedin climate, there
will be no need to apply additional propositions for explaining the
processesunder study.
An urgent problem today is to work out a quantitative method not
only for explaining past climates but primarily to predict climatic
changes in the next decadeswhich may transpire as a result of man's
economic activity.
There are several ways of solving this problem. One of them is
based on data concerning the energy and water balance of the earth.
Such a method was developedby the author and his colleaguesduring
the past several years. The main results obtained by these specialists
are discussedin the following

2 The Genesis

of the Climate

2.1 The Energy Balance of the Earth

In Chapter 1 we noted that the climate is determined by the solar

radiation incident on the upper boundary of the atmosphere, by the
atmosphericcomposition,and by the structure of the earth's surface. In
order to study the processesthat go on under the influence of these
factors,or, in other words, in order to explain the genesisof the climate,
data on the energy balance of the earth are very important. These data
clarify the transformation of solar energy in the atmosphere, in the
ocean, and on the earth's surface.
Components of the Energy Balance. The problem of studying the
energy (or thermal) balanceof the earth was formulated in the 1880sby
A. I. Voeikov, the outstanding Russian climatologist. Later important

contributions
to thisstudyweremadebyW. Schmidt,
A. ngstrSm,H.

Sverdrup and F. Albrecht.

Since 1945 systematic studies of the earth's energy balance have
been conducted at the A. I. Voeikov Main Geophysical Observatory
(Leningrad), where two world atlases were compiled on the energy
balance (Atlas of Heat Balance, 1955; Atlas of Heat Balance of the
World, 1963), and other data obtained on the geographicdistribution of
the heat balance components(Budyko, 1956, 1971).
These efforts provided information on all the componentsof the
heat balance equation of the earth's surface:
R = LE + P + A,

(2.1)

where R is the radiation balance of the earth's surface, LE is the heat

expenditure for evaporation (L is the latent heat of evaporation, E is
evaporation);P is the turbulent heat flow from the underlying surfaceto
the atmosphere;A is the heat flow from the underlying surface to the
deeper layers.
Alongside with these data on the heat balance of the earth's surface, information was obtained on the componentsof the heat balance of
the earth-atmosphere system, i.e. the heat balance of a vertical column

CLIMATIC

32

CHANGES

of a unit cross-sectionthat passesthrough the atmosphereand the ocean

(or, on a continent, through the atmosphereand the upper layer of the
lithosphere).
The heat balance equation for the earth-atmospheresystem is
Rs= Fs + L(E-

r) + Bs,

(2.2)

where Rs is the radiation balance of the earth-atmosphere system, and

Fs is the difference between the input and lossof heat in this column, as
a result of horizontal motions in the stratosphereand hydrosphere.This
quantity is equal to the sum of the heat input to the atmosphere (Fa)
and to the hydrosphere (Fo); Lr is the heat input from condensation,
which is consideredproportional to the amount of precipitation r, and Bs
is the change of leat content in a vertical column passingthrough the
earth-atmosphere system, causedmainly by the heating and cooling of
the ocean'supper layers.
The radiation balance of the earth's surface (R) is equal to the
difference between the short-wave radiation absorbed by the earth's
surface, and its effective long-wave emission, while the radiation balance of the earth-atmosphere system (Rs) is equal to the difference
between the solar radiation absorbedby this system and the long-wave
radiation at the upper boundary of the atmospherebeing emitted into
outer space. The values of R and Rs in equations (2.1) and (2.2) are
consideredpositive when they correspondto heat gain, while all the
other terms in these equations are positive in the caseof heat loss.
When studying the energy balance, it is necessary to take into
consideration the water balance of the earth, the componentsof which
form the following equation:
r=E+f+

b,

(2.3)

where f is the run off for dry land, or the horizontal redistribution of
water for bodiesof water, and b is the change of moisture content in the
lithosphere's upper layers, or the change of water content in water
bodies due to level oscillations.

In this equation r is always positive, E and f positive when they

indicate water loss,b positive when the amount of water in the soil or in
bodies of water

increases.

We shall present the earth's heat balance componentscalculated

from the data offeredby the "Atlas of Heat Balance of the World" (1963).
On the basis of these maps, the mean latitudinal values of the heat
balance componentshave been calculated for the surfacesof continents,
oceansand the earth as a whole. These results are presentedin Table 3.
It follows from this table that only one element of the heat balance,
i.e. the radiation balance, has a similar distribution for land and

2 THE

GENESIS

OF

THE

CLIMATE

Table

Mean

Latitudinal

33

Values

of the Heat

Bal-

ance Components (kcal/cm2.yr)

Lati-

Oceans

Continents

Earth as a Whole

tude

Fo

70-60N

R
23

LE
33

16

-26

20

LE
14

P
6

R
21

LE
20

P
9

Fo
-8

60-50

29

39

16

-26

30

19

11

30

28

13

-11

50-40

51

53

14

-16

45

24

21

48

38

17

-7

40-30
30-20
20-10
10-0

83
113
119
115

86
105
99
80

13
9
6
4

-16
-1
14
31

60
69
71
72

23
20
29
48

37
49
42
24

73
96
106
105

59
73
81
72

23
24
15
9

-9
-1
10
24

0-10S
10-20

115
113

84
104

4
5

27
4

72
73

50
41

22
32

105
104

76
90

8
11

21
3

20-30
30-40

101
82

100
80

7
9

-6
-7

70
62

28
28

42
34

94
80

83
74

15
12

-4
-6

40-50
50-60

57
28

55
31

9
8

-7
-11

41
31

21
20

20
11

56
28

53
31

9
8

-11

Total

82

74

49

25

24

72

60

12

-6

The maximum net radiation balance, both for continents and oceans,
occursin the tropics, the average latitudinal values of the radiation
balance being almost constantwithin the tropical zone.
The difference

between

the radiation

balance values for land and

oceanvaries very markedly with latitude. This differenceis greatest in

the tropical zone, where it comprisesone-third of the ocean'sradiation
balance. The radiation balances for land and oceans converge with
latitude and at 50-70 becomepractically equal. One of the reasonsfor
this is the increase
latitude.

of the mean

albedo of the water

surface

with

The average latitudinal values of heat expenditureon evaporation

for the continents have a main maximum at the equator, while in the
latitudes of the high-pressure belt, the evaporation decreases.The
evaporationincreasesslightly with latitude in both hemispheresand
more noticeablyin the northern hemisphere.This is due to the greater
amount of precipitationin mid-latitudes than in the dry zonesof lower
latitudes. With further increase of latitude, the evaporation decreases
due to insufficient

heat.

In contrast to continental conditions,the maximum heat expenditure for evaporationfor oceans,averagedover the latitudes, takes place
in the high-pressurebelts. It shouldbe notedthat at the latitudes of

CLIMATIC

34

CHANGES

70, where the radiation balances of the continents and oceans are
approximately the same, the heat expenditure for evaporation over the
oceanis much higher than over the continents.This is partly explained
by the fact that evaporation from the ocean consumesmuch heat supplied by ocean currents.
The turbulent heat flow in the oceansincreasesmonotonicallywith
latitude. On the continentsit reachesits maximum in the high-pressure
zones;it is somewhatlower at the equator and dropsdrastically at high
latitudes. The totally different character of these changes on land and
on sea indicates

that

the transformation

of air masses over continents

and oceansis causedby different mechanisms.

The distribution of average latitudinal values of incoming and
outgoingheat in the oceansshowsthat the oceancurrents carry the heat
mainly from the zone between 20N and 20S,the maximum of the heat
loss being somewhat shifted to the north of the equator. This heat is
transferred to higher latitudes and is dissipatedmainly in the latitudes
of 50-70Nwhere particularly strong warm currents prevail.
The latitudinal distribution of the componentsof the heat balance
for the earth as a whole is characterizedby a regularity which depends
on the ratio

between

the

surface

areas

of continents

and oceans in

various zones. The average values of the heat balance componentsfor

continents and oceansare given in Table 4.
It is evident

from this table

that

on three

of the six continents

(Europe, North America and South America) most of the net heating is
consumedby evaporation. The other three continents (Asia, Africa, and
Australia) show the reverse, indicating the prevalence of dry climate.
The mean parameters of the heat balance differ very little on all three
oceans.It shouldbe noted that the sum of the heat expenditure and the
turbulent heat exchangefor each oceanis very closeto the net radiation
Table

Heat Balance Components for Continents

and Oceans (kcal/cm 2 yr)
Com-

Ocean

Continent

ponent

39

47

68

40

70

70

82

86

85

LE

24

22

26

23

45

22

72

78

77

15

25

42

17

25

48

2 THE

GENESIS

OF

THE

35

CLIMATE

balance value. This means that the heat exchange between the oceans
causedby ocean currents doesnot affect the heat balance of each ocean,
taken separately. A small excessof the radiation balance over the total
of the turbulent heat flux and the expenditure for evaporative heating
in the Atlantic correspondsto the transfer of a certain amount of heat
from the Atlantic

to the Arctic Ocean. Since this difference

is within

the

margin of error, further study of this problem is clearly necessary.

Table 4 does not include data on the heat balance of the Antarctic

continent or the Arctic Ocean, since this information is still unreliable.

On the basis of the foregoingdata and available estimates of heat

balance componentsfor the polar region, calculations have been made of
the heat balance componentsfor all continents, oceans,and the earth as
a whole. The results are given in the last line of Table 3.
From these data it follows that over 90% of the radiation

balance on

the ocean surface is consumedby evaporation and only 10% by direct

turbulent heating of the atmosphere.On the continentsthese expenditures are almost equal. For the entire earth the heat expenditure
comprises83% of the radiation balance while the turbulent heat exchange comprises17%.
The Atlas of Heat Balance of the World (1963) provides data on
average latitudinal values of the heat balance componentsfor the earthatmospheresystem and the atmosphere (Budyko, 1971). It can be seen
that the relation between the heat balance componentsfor the earthatmospheresystem in various latitudes, differ substantially for average
annual

conditions.

In addition to the large inflow of radiative energy into the equatorial zone, there is a significant heat influx in the earth-atmosphere
system as a result of water phase transformations, i.e. due to the
differencebetween heat of condensationand heat energy expenditure
for evaporation. These heat sourcesprovide a great amount of the
energy that is consumedby atmosphericand oceanadvection,for which
a comparativelynarrow near-equatorial zoneis an extremely important
energy

source.

In higher latitudes, such as about 30-40, there is a radiation

surplus (positive balance) of the earth-atmosphere system which decreaseswith latitude and the heat expenditure on water phasetransformations reachesa significant value. In the greater part of this zone, the
heat expendedon phase transitions is comparablewith the radiation
surplus; therefore the heat redistribution by atmospheric and ocean
currents is comparatively small.
The zoneof radiation deficit (negative balance), which increasesin
absolute value with latitude, is located at latitudes above 40. The
radiation deficit in this zone is compensatedby the inflow of

CLIMATIC

36

CHANGES

transferred by atmosphericand oceancurrents, the ratio of heat balance

componentsthat compensatefor the energy deficit being different at
various latitudinal zones.In the 40-60 zone, the main sourceof energy
is the energy released during water vapor condensationin excessof the
energy expenditure for surface evaporation.
At these latitudes the inflow of heat redistributed by oceancurrents
also plays an important role. The redistribution of heat by atmospheric
circulation becomesthe main sourceof thermal energy at higher latitudes, and especially in the polar region, where the heat inflow from
condensation

is small

and the effects of ocean currents

is either

absent

(south polar zone), or is weakened by a permanent ice cover (north pole

zone).

Calculations show that the zonally-averaged values of the atmosphere'sradiation balance changelessthan other componentsof the heat
balance. The large deficit observed at all latitudes is compensated
mainly by the heat inflow from condensation.
The heat inflow from the earth's surface causedby turbulent heat
exchange plays a smaller part, although its influence on the atmospheric heat balance is significant.
Mean latitudinal values of the heat balance componentsfor the
earth-atmospheresystemfor two half-years are shownin Table 5, where
Qa is the radiation absorbedin the earth-atmosphere system, and Is is
the outgoing long-wave radiation at the upper boundary of the atmosphere. The differencebetween these two values is equal to the radiation
balance Rs of the earth-atmosphere system. The total heat redistributions causedby horizontal motions in the atmosphereis equal to F0 +
L (E - r) and is designatedby Ca. The values ofF0 in Tables 3 and 5 were
determined by different methods and vary slightly.
The left-hand segmentof the upper half of the table and right-hand
segment of the lower half refer to the warm half-year, while the other
two segmentsrefer to the cold period.
As shownby Table 5, the absorbedenergy is not the only factor that
governsthe amount of outgoingradiation. In middle and high latitudes
during the cold period (and in high southern latitudes throughout the
year), the main sourceof heat is its transport from lower latitudes by
atmospheric circulation.
The chief component of the two heat-exchange processesin the
oceanconsistsof seasonalheat accumulation and dischargeby the mass
of ocean water. At some latitudes this componentof the heat balance
reaches 25-30% of the outgoing radiation. The heat redistribution by
ocean currents plays a smaller part, although in certain zonesit may
reach 15-20% of the radiation emitted into space.It shouldbe noted that
the values of these componentsof the ocean heat balance refer to

2 THE

GENESIS

OF

THE

CLIMATE

Table

Mean

Latitudinal

37

Values

of the Heat

Bal-

ance Components of the Earth-Atmosphere System for Two Half-Years (kcal/

cm . month)
First Half-Year

Lati-

tude

(IV-IX)

Second Half-Year

(X-III)

Qa

Ca

Fo

Bs

Is

Qa

Ca

Fo

Bs

Is

80-90N
70-80
60-70
50-60
40-50
30-40
20-30
10-20
0-10

7.8
8.2
11.5
14.6
16.9
19.2
20.0
19.7
18.4

-4.5
-4.4
-1.8
0.0
1.0
1.4
2.0
2.4
1.4

0
0
-0.4
-1.3
-2.0
-1.7
-0.4
0.9
2.2

0.8
0.8
1.2
2.8
3.9
4.3
2.9
1.4
-0.1

11.5
11.8
12.5
13.1
14.0
15.2
15.5
15.0
14.9

0.1
0.5
1.8
4.0
6.5
9.5
13.7
17.0
18.7

-9.1
-9.3
-6.7
-4.7
-2.9
-0.9
0.9
1.8
2.1

0
0
-1.5
-0.5
0.6
0.7
0.6
1.0
1.4

-0.8
-0.8
-1.2
-2.8
-3.9
-4.3
-2.9
-1.4
0.1

10.0
10.0
11.0
12.0
12.0
14.0
15.0
15.0
15.0

0-10S
10-20
20-30
30-40
40-50
50-60
60-70
70-80
80-90

18.0
16.2
13.0
8.9
5.9
3.3
1.0
0.2
0.0

2.3
2.6
1.4
-0.4
-1.9
-3.9
-9.5
-9.8
-8.8

2.2
0.9
0.3
-0.4
-1.6
-2.6
0
0
0

-1.5
-2.5
-3.4
-4.2
-3.5
-2.5
-0.8
0
0

15.0
15.2
14.7
13.9
12.9
12.3
11.3
10.0
8.8

19.7
20.7
20.6
18.9
15.9
12.8
8.1
4.4
3.4

2.4
3.5
3.5
2.0
0.0
-2.6
-4.3
-6.5
-6.9

0.8
-0.3
-1.2
-1.4
-0.6
0.6
0
0
0

1.5
2.5
3.4
4.2
3.5
2.5
0.8
0
0

15.0
15.0
14.0
14.0
13.0
12.0
11.0
10.0
10.0

total surface of the latitudinal

zones. In most cases this reduces their

valuessignificantlyas comparedto the heat balancecomponents

for the
surface areas of the ocean.

Table 5 indicatesthat in orderto determinethe outgoingradiation

from other componentsof the heat balance, it is necessaryto take into
considerationall the componentsincludedin the table. The values of the
heat balancecomponentsfor the earth as a whole are presentedschematically in Fig. I (Budyko, 1971).
The total flux of solarradiation reachingthe outer boundaryof the
troposphereis about 1000kcal/cm .yr. Due to the sphericalform of the
earth, a unit area of the upperboundaryof the tropospherereceives,on
the average, a quarter of this amount, i.e. about 250 kcal/cm.yr.
Assuming the earth's albedo as as = 0.33, we find that the short-wave
radiationabsorbedby the earth is equalto 167kcal/cm .yr. This quantity is depictedin Fig. I by the arrow Qs (1 - as).
Accordingto the foregoing data, the earth's surface receives 126

kcal/cm.yr. of short-wave radiation. The mean albedo of the earth's

surface a (consideringthe different amount of incident solar

CLIMATIC

38

CHANGES

Outer space

Atmosphere

Fig. 1. Heat balanceof the earth (components

are given in termsof kcal/
cm2.yr) .

for various regions)is equal to 0.14. Thus the earth's surfaceabsorbs108

kcal/cm2.yr, which is indicated in Fig. i by arrow Q (1 - a), while it
reflects 18 kcal/cm2.yr.
We see, therefore, that the atmosphere absorbs59 kcal/cm .yr, or a
great deal less than the surface.
The radiation balance of the earth's surface is thus equal to 72 kcal/
cm.yr, while the effective radiation at the earth's surface is, on the
average, equal to 36 kcal/cm2.yr (arrow/). The total amount of the longwave radiation of the earth is equal to the amount of absorbedradiation
and is about 167 kcal/cm2.yr (arrow Is).
It should be noted that the ratio of the effective radiation

from the

earth's surface to the total radiation of the earth (I/Is) is considerably

less than the correspondingratio of the absorbedradiations Q(1 - a)/
Qs(1 - as).This fact indicatesthe strong contributionof the "greenhouse
effect" to the thermal regime of the earth.
Due to the "greenhouseeffect"the earth's surfacereceivesabout 72
kcal/cm.yr of radiative energy (radiation balance), which is partly
consumedby water vapor (60 kcal/cm.yr, shown by circle LE), and
partly returned to the atmospherein the form of turbulent heat flow (12
kcal/cm2.yr, arrow P). As a result, the heat balanceof the atmosphereis
comprisedof the following elements:
1) incoming heat from the absorbedshort-wave radiation equal to
59 kcal/cm 2.yr,
2) incoming heat from water vapor condensationequal to 60 kcal/
cm.yr (in Fig. I circle LE);
3) incoming heat from the turbulent heat transfer from the earth's
surface equal to 12 kcal/cm2.yr;
4) expenditure of heat by effective radiation into outer spaceequal
to Is - I, i.e. 131

2 THE

GENESIS

OF

THE

CLIMATE

39

The last term is equal to the sum of the first three componentsof the
heat balance.
Water Balance.

Let us consider the water

balance

of the continents

first. The total precipitation over land has been measured over a long
period of time at numerous meteorologicalstations. Therefore there are
no difficulties in compiling precipitation maps based on observational
data.

The average annual amounts of precipitation on continents determined from the precipitation maps compiledby Kuznetsova and Sharova (1964) are given in Table 6. The table doesnot contain data for the
Antarctic, but these were taken into considerationin determining the
average precipitation for the entire land area.
It should be noted that meteorologicalinstruments often give an
inadequateestimate of the amount of precipitation, particularly of snow
which is often blown away by the wind.
Measurements of precipitation by various instruments contain different systematic errors. In compiling precipitation maps, attempts
have been made to correctmost of these errors. Generally, however, the
attempted correctionshave proved insufficient. Recent investigations
showthat in the regionssubjectto heavy snowfalls,the actual precipitation can exceedpreviously acceptedfigures by a high percentage. The
amount of rainfall is measured, as a rule, with a smaller margin of
systematic error. That is why the figures in Table 6 are somewhat
underestimated.

In the absenceof large-scaleobservations,the averageevaporation

from land surfacescan be determined solely by calculations.Table 6
gives the values estimated on the basis of the pertinent maps in the

Arias of Heat Balance of the World (1963).

It followsfrom the table that the ratio of evaporationto precipitation varies greatly for different continents. This ratio is closeto unity in
Table

Precipitation

and Evaporation

on Conti-

nents

Continent

Precipitation (cm/yr) Evaporation (cm/yr)

Europe

64

40

Asia

60

37

N. America

66

38

S. America
Africa
Australia

163
69
47

75
43
37

Entire

terrain

73

CLIMATIC

40

CHANGES

Australia, whereas in South America the evaporation comprisesless

than half of the precipitation.
The componentsof the oceanwater balancecan be calculatedon the
basisof the world precipitationmap compiledby Kuznetsovaand Sharova (1964), and the world evaporation map from the Atlas of Heat
Balance of the World (1963) (Table 7).

The difference between evaporation from the ocean surface and

precipitationis equal to the river runoff from the continentsinto the
oceans.For a given ocean,this differenceis equal to the sum of river
runoff and the horizontal water transfer between oceans due to circula-

tion processes.
It is difficult to determinethis quantity directly because
the difference between the inflow and outflow of water is comparatively
small, and each quantity is determinedwith a substantialerror. It is
simpler to evaluate the water exchangebetween the oceansas the
remainder term of the water balance of every ocean, although in this
case,there is very little precisionin the corresponding
valuesas well.
These values of river runoff for every oceanare given in Table 7
accordingto Zubenok (1956).
This table does not include data on the water balance of the Arctic

Ocean, which are less precisethan for the other oceans,but they are
taken into accountin determining the water balance of the earth as a
whole.

it followsfrom the table that the sum of precipitation and river

runoffis lessthan the evaporationin the Atlantic Ocean.This deficitis
compensated
by water flowingin from otheroceansincludingthe Arctic,
where evaporationis considerablysmaller than the sum of precipitation
and river runoff. In the Indian Oceanthis sam is slightly lower than the
evaporation,while in the Pacific it is higher, correspondingto the
transfer of surplus water to other oceans.

On the basisof precipitationand evaporationdata given abovefor

continentsand oceans,we find that for the earth as a whole, the annual
amountof precipitationcomesto 102cm,whichis equalto the amountof
evaporation.This figure is noticeablyhigher than similar valuesdeter-

Ocean

Table

Water

Balance

of the Oceans

PrecipitationEvaporation
River
Discharge
( cm/yr)

( cm/yr)

( cm/yr)

Atlantic

89

124

23

Pacific

133

132

indian

117

132

7
8

World ocean

114

126

2 THE

GENESIS

OF

THE

CLIMATE

41

mined in the course of most of the earlier studies (Wrist, 1936, M511er,
1951), but for the reasonsenumerated above, we must assume that it is
somewhat

underestimated.

The precipitation maps of Kuznetsovaand Sharova (1964)and the

evaporation maps from the Atlas of Heat Balance of the World enable us
to determine

the water balance of the earth's

latitude

zones.

The dependenceon latitude of the water balance componentsis

illustrated by Fig. 2. We find in various zonesthat the influx of water
vapor into the atmospherefrom evaporation can be either greater or
smaller than its consumptionby precipitation, the main sourceof water
vapor in the atmospherebeing chiefly the high pressurezones, where
evaporation predominates over precipitation. The expenditure of this
water vapor surplus occursin the near equatorial zone as well as in
middleand high latitudes,whereprecipitationexceedsevaporation.It is
obviousthat the quantity f, which is equal to the differencebetween
precipitation and evaporation, is at the same time equal to the differencebetweenthe input and outflowof water vapor in the atmosphere.
Reliability of Data on Components of the Energy Balance. The
total amount of data on the energybalancecomponentsobtainedduring
the last decadesis comparableto that of other meteorological
elements
observedat the basicnetworkof weatherstations.The precisionof these
data is of importancein the study of climate.
(cm/yr)

200 -

150 -

I00

5O

. f

-50
N 9()

60

,
()

30

,i(}

90' s

Fig. 2. Mean latitudinal distribution of water balance componentsof the

CLIMATIC

42

CHANGES

Most of the maps of the heat balance componentsare compiledby

means of calculationsas there are very few, if any, observationaldata
available.

In several cases,after these maps were published,direct observations of these componentswere collated. Thus the calculated maps
servedas a preliminary estimate of anticipated data from observations.
The first work of this kind was to determine the averagevalues of
the radiation balance of continental surfaces. The first maps of this
parameter were publishedin the early half of the 1950s,when long-term
observational

data of the radiation

balance did not exist.

In the secondhalf of the 1950s,during the International Geophysical Year, observationsof this type were initiated in many parts of the
world. Their results were first comparedwith previouscalculationsby
Robinson (1964), who found a good correlation between the measured
and calculated

data.

Later, studies of this kind were frequently made on the basis of a

much larger amount of data. Figure 3 gives a comparisonof the measured (Rm)and calculated(Re) averagemonthly values of the radiation

balance for 27 stations situated on different

continents.

The correlation

coefficientbetweenthesevalues in this caseis equal to 0.98.

The measurements

of the radiation

balance

on the ocean com-

mencedlater than on the continent, and the great amount of material

Re (K cal/cm? month)
12-

8-

,: ,.,,,',

.'

i- ::4:"
eee ee

-2...;2 4 0 8 10R
m
(K
cal/cm
?month)

Fig. 3. Comparison of the measured (R,,) and calculated (Rc) average

monthly values of the radiation balance of the

2 THE

GENESIS

OF

THE

CLIMATE

43

accumulated only by the secondhalf of the 1960s.Therefore the previously publishedmaps servedas a prognosisof what couldbe expectedin
the future. A comparisonof the measured and calculated values of the
radiation balance of oceansurfaceswas carried out by Strokina (1967),
and the results of this comparisonare presentedin Fig. 4. The correlation coefficientin this caseis equal to 0.96.
Another example of verification of the calculatedmaps is by means
of the radiation balance of the earth-atmospheresystem. It is known
that these maps were compiled before the first direct measurements
were carried out by means of artificial satellites. A comparisonshows
that although this correlation is worse than for similar maps of the
earth's surface,the main regularities establishedby direct observations
correspondto those of the calculated maps.
Note that the distribution of the heat balance componentsof the
earth-atmosphere system reveals an interesting peculiarity in desert
regions. It appears from the calculated map published in the Atlas of
Heat Balance of the World (1963), that the annual radiation balance of
the Sahara Desert is closeto zero, which means that there is no sourceof
heat advectionin this region for other regions.This paradoxicalresult
was later confirmed by direct measurements (Raschke, MSller, and
Bandeen, 1968).
The foregoingexamplescompriseonly a small portion of the availa-

ble material on heat balance components(Budyko, 1971); analysis

proves that these data are quite reliable for the study of climatic
genesis.

R,, (K cal/cm ? month)

15-

o-

..-.z'.'-

..

--5

10Rm (K cal/cm? month)

Fig. 4. Comparison of the measured (R,) and calculated (R) average

monthly values of the radiation balance of the

44

CLIMATIC

CHANGES

2.2 Semiempirical Theory of Climate

Contemporary Climate Theories. The theory of climate aims to

determine, by means of physical deduction,the average distribution of
meteorologicalelements in spaceand time and their variability under
the influence of given external factors. In order to developa theory of
climate, methods of fluid mechanics and thermodynamics are applied,
as well as the laws of energy transport and phase transformations of
water, along with other formulations of contemporaryphysical meteorology.In a review recently published (Inadvertent Climate Modification, 1971),F. Thompsondivided the existing theories into four groups:
1) modelsreferring to the planet as a whole, describingthe average
meteorologicalconditionsfor the entire atmosphere;
2) modelsfor zonally-averagedconditionsin which the longitudinal
variation of the fields is not considered;in these models large-scale
atmosphericmotions are statistically parameterized;
3) semiempirical models based on empirical regularities established by observations;
4) general modelstaking into accountall the details of large-scale
atmospheric motions.
This classificationis not strictly logical becausethese groups are
not uniform as regardstwo criteria: the range of applicationand the use
of empirical relations. Nevertheless,it reflectsthe main trends of modern attempts in the developmentof the theory of climate. One of the first
numerical modelsof this theory was developedby Milankovich (1920,
1930), who calculated the distribution of average latitudinal temperatures in the atmospherebased on the amount of radiation incident on
the upper boundary of the atmosphere.Milankovich believedthat the
thermal regime is determined by the heat exchangein each separate
latitude zone. He did not take into accountthe greenhouseeffect of the
atmosphere, nor the variations of albedo with latitude. Milankovich
also neglectedthe effect on the thermal regime of the meridional heat
transfer in the atmosphere,as well as the heat exchangecausedby the
phasetransformation of water and the interaction between the atmosphere and the ocean. Significant errors causedby these factors were
compensatedfor by the selectionof certain specialparameters, and as a
result Milankovich obtained a distribution of average latitudinal temperature that doesnot differ greatly from the observedone. This correlation, however, was highly provisional.
The first step in the application of methodsof dynamic meteorology
to the developmentof the theory of climate was taken by Kochin (1936).
Kochin calculated the average meridional profiles of pressureand wind
velocity in the lower atmosphere, on the basis of the given

2 THE

GENESIS

OF THE

CLIMATE

45

of air temperature and pressureat the earth's surface.Later, Dorodnitsyn, Izvekov and Shwets (1939)applied the same method to a model of
summer zonal circulation in the northern hemisphere. This trend in
research was further developedby Blinova (1947), who calculated the
averagedistributionof temperature,pressure,and wind velocityin the
northern hemisphere.

Of particular importancein the developmentof the theory of cli-

mate are the works of Shuleikin (1941), who establishedthat interaction

betweenthe oceanand the atmosphere

was a major and frequently

dominant factor in the formation of climatic conditions.

The treatment of the physicalmechanismof heat exchangeand

moistureexchangein the modelshas posedgreat difficulties. Several

worksby Rakipova (1967)were devotedto this problem;shecalculated
the distributionof averagelatitudinal valuesof air temperatureand
otherelementsby considering
all the main formsof heat input in the
atmosphere which contribute to the heat balance.

Later, many new modelswere developedmaking it possibleto

calculatethe averagefields of meteorological
elements.Progressin
computationaltechniquesfacilitatedthe developmentof quantitative
modelsrepresentingunaveragedfieldsof meteorological
elements.Calculationsof thesefieldsfor a sufficientlylongperiodof time resultedin
finding average fields that characterizeclimatic conditions.

Numerousstudiesin this directionwerecarriedout by Smagorinsky (1963),Smagorinsky,Manabe,and Holloway(1965),Mintz (1965),

ManabeandBryan (1969),Hollowayand Manabe(1971),Wetheraidand
Manabe (1975).

In the abovementioned
works, Manabe and Bryan developeda

numerical model of the theory of climate that takes into accountthe

effectof oceaniccirculationonclimaticconditions.
Hollowayand Manabewerethe first to compiletheoreticalworldmapsof the main componentsofthe heat andwaterbalanceof the earth'ssurface.Thesemaps

correlatedwell with similar mapsbasedon empirical data.

An originalapproach
to theproblemofthe generaltheoryofclimate
was suggestedby Shvets and his colleagues(1970). The improved
methodofparameterization
ofthe condensation
andradiationprocesses
introduced in this model allowed them to take into account the direct

andfeedback
relationsbetweentheseprocessed
andthe dynamicstateof
the atmosphere,
andthusall the links in the chainof a generalcirculation model were joined.

Suchmodelsmust meet a numberof requirementsin order to be

appliedto the studyof climaticchanges.Theserequirementsare not as
essentialfor modelsof the contemporaryclimatic regime. The first
requirementis that the modelmust not containempiricaldata on

CLIMATIC

46

CHANGES

distribution of climatic elements,especiallythosethat are highly sensitive to the effect of climatic changes.The secondrequirement is that the
modelmust realistically take into accountall types of heat inflow which
tangibly affect the temperature field and must conformwith the energy
conservationlaw. The third requirement is that the model must consider the main

direct

and feedback

relations

of various

climatic

ele-

ments.

The first two requirements are more or less obvious, but the third
one must be elaborated

on.

The stability of climate is achieved by those feedback relations

betweenits elementsthat can be called negative (Inadvertent Climate
Modification, 1971).This feedbackattenuates perturbations in meteorological fields and helps them to revert to their climatic norms. An
example of negative feedbackis the dependenceof long-wave radiation
on the earth's surfacetemperature. When the temperature rises, radiation increases, resulting in growth of heat energy expenditure and
causing a decreasein temperature use. Another example of negative
feedback is the dependenceof atmospheric heat transfer upon the air
temperature gradient. As a rule, the heat flow in the atmosphere is
directed from regions of higher temperature toward regions of lower
temperature, which leads to a leveling of the temperature distribution.
Positive feedback, on the other hand, plays an important part in
climatic changescontributing to an increasein perturbations of meteorologicalelements, thus decreasingclimate stability.
An example of positive feedback is the dependence of absolute
humidity upon its temperature mentionedin Chapter 1. As temperature
rises, evaporation from water or wet surfacesincreases,thus leading to
a comparatively stable relative humidity in most climatic regions (with
the exception of dry continental areas). Under these conditions, the
absolute air humidity increases with temperature, as confirmed by
numerous empirical data.
Sincethe effectivelong-waveradiation decreaseswith the growth of
absolutehumidity, the latter facilitates a further temperature rise. It is
obviously necessary to take this feedback into account in numerical
modelsof the thermal regime used for studying climatic changes.
Another positive feedback-the influence of snowand ice sheetson
the earth's surface albedo-plays an even greater role in the laws
governing the changesin the atmosphere'sthermal regime.
Influence of Snow and Ice Covers on Climate. The first to study
the influence

of the snow cover on climatic

conditions

was Voeikov

(1884), who proved that a snow cover aids in reduction of the air
temperature over its surface. Later, Brooks (1950) came to the conclusionthat the ice cover, due to its high albedo, significantly reduces

2 THE

GENESIS

OF

THE

CLIMATE

47

air temperature, as a result of which climatic fluctuations increase

greatly following the formation and melting of ice covers.
The values

of the albedo of the earth's

surface

and of the earth-

atmospheresystem in high latitudes can be determined on the basis of

data obtained during expeditionsto the Arctic and Antarctic as well as
from satellite observations;these figures may be compared with the
albedo of regions without any snow or ice cover.
Available data indicate that during the summer months the albedo
of the ice surface

in the Central

Arctic

is about 0.70 and is about 0.80-

0.85 in the Antarctic.

If we recall that the mean albedo of the earth's surface in regions

free of snow and ice is about 0.15, we come to the conclusion that the
snow and ice cover results in a manifold

reduction

of radiation

absorbed

by the earth's surface, other conditionsbeing equal.

Snowand ice also greatly affect the albedoof the earth-atmosphere
system. According to recent satellite data (Raschke, et al., 1968, 1973),
the system's summer albedo in the central Arctic and Antarctic is
approximately 0.60, which is about twice that of the earth's albedo as a
whole.

It is obvious that such great differences in albedo values must

strongly affect the thermal regime of the atmosphere.
If a snow or ice cover is formed on the surface as a result

of an air

temperature drop, this must facilitate a further temperature decreaseas

well as an expansion of the snow and ice covers. The reverse is true
when the temperature rises, if it leads to melting snow and ice.
That this feedback greatly affects the air temperature distribution
near the earth's surfaceis provedby taking this feedbackinto accountin
a numerical model of the atmosphere'sthermal regime (Budyko, 1968).
A simple example may be given in order to estimate this influence
indicating how the mean global air temperature will change if the
earth's surface under cloudlessconditions is completely covered with
snow and ice (Budyko, 1962, 1971).
Let us consider a hypothetical case in which the earth's surface is
completelycoveredby ice and snow, and cloudsare absent. Under these
conditionsthe earth's albedo is much higher than the actual value and
this shouldcausea change in the mean global temperature. The effective temperature of the earth correspondingto its long-wave radiation is
proportional to (1 - as)TM,where as is albedo. Therefore the absolute
effective temperature when the albedo changes from a.' to as" will
change as [(1 - as")/(1 - as')]TM.If we take the actual albedo of the earth
as equal to 0.33 and the albedoof dry snow surface as 0.80, we find that
for the earth coveredwith snow, the mean effective temperature will
drop by approximately

48

CLIMATIC

CHANGES

The mean air temperature near the ice-coveredsurfacewill probably drop even more. At present, the mean temperature of the lower air
layers increasesconsiderablyalmost over the entire surface,due to the
greenhouseeffect causedby the absorptionof long-wave radiation by
water vapor and carbon dioxide in the atmosphere.This effect is not
significant at very low temperatures,and the formation of denseclouds
that change the radiative flux is also impossible.Under these conditions, the atmospherebecomesmore or less transparent to both shortwave and long-wave radiation.
The mean temperature of the earth's surfacewhen the atmosphere
is transparent is determined by the simple formula [So(1 - as)/4cr]
TM,
where So is the solar constant and is the Stefan-Boltzman constant. It

follows from this formula that when as - 0.80, the earth's mean temperature is 186K, or -87C.

Thus, if the entire earth's surface were coveredwith snow and ice,
even for a short period, its mean temperature, which is now equal to
15Cwould drop by approximately 100C.This example illustrates what
a tremendous influence the snow cover has on the thermal regime.
We have made an attempt to evaluate the effect of polar sea ice on
the thermal regime in the Arctic (Budyko, 1961, 1962, 1971). By using
the data on the heat balance of central regions of the Arctic Ocean and
approximate formulas of the semiempirical theory of climate, it was
established that the polar ice coolsthe air in the Central Arctic by
several degreesin the summer and by 20Cin the winter.
We have come to the conclusionthat at present an ice-free regime
couldexist in the Arctic Ocean, although it would be very unstable, and
minor climatic

fluctuations

could lead to the restoration

of the ice cover.

Rakipova (1962, 1966), Donn and Shaw (1966) and Fletcher (1966)also
cameto the conclusionthat an ice-free regime in the Arctic was possible.
Thus, a permanent ice cover occupyingeven a comparatively small
fraction of the earth's surfacehas a strong influence on the atmosphere's
thermal regime, which must be taken into account in the study of
climatic

variations.

Semiempirical Model of the Thermal Regime. In spite of all the

outstanding achievements in the development of a general theory of
climate, it is still difficult to apply existing models to the study of
climate changes. This is due to the fact that these models do not, as a
rule, meet all the abovementionedrequirements. At the same time, as
pointed out by Smagorinsky (1974), computationsof climatic variations
using models of the general theory of climate require extensive computer time and are therefore usually far beyond the capabilities of
computer centers. In this connectionSmagorinsky noted the necessity

2 THE

GENESIS

OF

THE

CLIMATE

49

a parametric representation of the circulation processesin the atmosphere and ocean in order to statistically describe large-scale perturbations, as is done in the general study of small-scale turbulent processes.
This idea has been applied in several works in which heat and humidity
transfer processesin the atmosphere were treated as large-scale turbulent processes.
In view of the difficulties inherent in applying the general theory of
climate to the study of its variations, several attempts have been made
to use semiempirical models. These models can be made to meet the
requirements by means of extensive parameterization of large-scale
atmospheric processes,by limiting the given problems, and by using
empirical relations.
This trend includes a semiempirical model of the thermal regime of
the atmosphere (Budyko, 1968; Budyko and Vasishcheva, 1971), based
on a number

of considerations

which

we shall discuss below.

In the absence of an atmosphere, the mean temperature of the

earth's surface would be defined by the conditions of radiation equilibrium, i.e. the long-wave radiation of the earth's surface is equal to the
absorbedradiation. Let us expressthis condition in the following form:
(2.4)

o'T 4 - 8o(1

where a is a coefficient that characterizes the properties of the radiation

surface as distinct from that of a black body, (r the Stefan constant, T
the surface temperature, So the solar constant, and as the mean albedo
of the earth.

Assuming that = 0.95, (r = 8.14.10 - cal/cm2. min, and So = 1.95

cal/cm . min, it follows from (2.4) that for as = 0.33, the mean earth
temperature is 255K, or -18C. Since the mean air temperature near
the earth's surface, accordingto observations,is about 15C,we arrive
at the conclusion that with the same albedo, the atmosphere increases
the mean air temperature at the earth's surfaceby approximately 33C.
This temperature increaseis causedby the greenhouseeffect, i.e. by the
higher transparency of the atmosphereto the short-wave, as compared
to the long-wave radiation.
It must be noted that this evaluation

is true under certain

condi-

tions,namely, in the absenceof atmospherethe planetary albedocannot

be equal to the actual value of 0.33.
Under present conditions,the mean albedoof the earth's surface is
0.14. It may be assumedthat prior to the origin of the atmosphere,the
earth's albedo was lower and probably differed very little from the
moon's albedo, which is equal to 0.07. With this albedo the mean
temperature of the earth's surface couldbe about

50

CLIMATIC

CHANGES

In order to estimate the influence of incoming solar radiation and

the albedo on the mean temperature at the earth's surfaceunder existing conditions,it is necessaryto know the dependenceon temperature
distribution of long-waveradiation at the outer boundaryof the atmosphere. This dependencecan be establishedin two ways.
The first involves consideringa theoretical model of the vertical
distribution of radiation fluxes, temperature and air humidity. The
secondis basedon an empirical comparisonof data on long-waveradiation obtained from observations or calculations, and the factors that
affect it.

This method has been applied by the author (Budyko, 1968),who

utilized the averagemonthlyvaluesof the outgoingradiationcalculated
by Vinnikov in the courseof compiling maps for the Atlas of Heat
Balance of the World (1963).

These included data on outgoingradiation obtained at 260 stations

uniformly distributed on land and sea. We had at our disposal3120
average monthly values.
A comparisonof these data with various elements of the meteorological regime revealed that these average monthly values depended
mainly on the air temperature near the earth's surface and on cloudiness. This relation was expressedin the following form:
Is = a + bT-

(a + bT)n,

(2.5)

where Is is the outgoingradiation in kcal/cm2-month; T is air temperature in C;n is cloudinessin fractions of a unit, and the atmospheric
dimensional coefficients are: a = 14.0; b = 0.14; a = 3.0; b = 0.10.

The mean deviation of the outgoingradiation for each month, from

the results obtainedby (2.5), was very small (below5%), which confirms
the conclusionthat all other factors, including the peculiarities of the
vertical temperature distribution, exercisea comparativelysmall influence on the outgoing radiation.
It must be pointed out that this deduction applies to values of
outgoingradiation averagedover long periods,and that it refersto large
geographicalareas. The physical interpretation of the aboveconclusion
is that the outgoing radiation originates mainly in the troposphere
where the temperature deviations from its mean vertical distribution
are generally lower than its annual spatial variations.
Equation (2.5) can be comparedto the results of severaltheoretical
works including, in particular, that of Manabe and Wetheraid (1967).In
this work a relation was establishedbetween the outgoingradiation and
temperature and cloudinessunder radiational and convectionalequilibrium conditions in the atmosphere. This dependencewas presented
graphically, its curves couldbe expressedby the

2 THE

GENESIS

OF THE

Is = 14 + 0.14T-

CLIMATE

51

(2.6)

1.6n.

We can seethat the independently obtained formulas (2.5) and (2.6)

coincide in the case of a clear sky.
To verify the formulas showing the connectionbetween the outgoing radiation and meteorologicalfactors, we must apply the equilibrium
condition between the outgoing and absorbedradiation for the entire
planet:
Qsp(1- op) = Is,,,

(2.7)

where the values Qsp,-sp, and Is apply to the planet as a whole.

Assumingthat the average for the entire earth value of Qs is equal
to 20.8 kcal/cm2. month, the average temperature at the earth's surface
level is 15Cand the mean cloudinessis 0.50, we find from formula (2.5),

that condition (2.7) holds when "s = 0.33. This value of the planet's
albedo agrees with the estimates obtained by independent methods in
several modern investigations, thus confirming the accuracyof formula
(2.5).

It can be shown, however, that in accordancewith the above statements concerning the effect of snow and ice covers on the thermal
regime, equations (2.5) and (2.7) have at least one more solution.
Let us assume that at low negative air temperatures at the earth's
surfaceat all latitudes, it is entirely coveredwith snowand ice, and that
the global albedoof the earth-atmospheresystemin this caseis approximately equal to that of the central Antarctic region, i.e. 0.6-0.7. From
(2.5) and (2.7) it follows that in this casethe average planetary temperature at the earth's surface will vary from -47C (at the lower albedo
value) to -70C (at the higher albedovalue). At these average temperatures the total glaciation of the earth is inevitable, thus confirming the
above remark as to the ambiguous relation between contemporary
climatic conditions and external climatic factors. This problem will be
dealt with in greater detail further on.
Let us find from (2.5) and (2.7) the expressionfor the average earth
temperature'

Tp =
This formula

b - b,n

[Qsp(1- o) -a

is correct within

the limits

+ a,n].

(2.8)

of the actual fluctuations

of

mean monthly temperatures at the earth's surface,its precisiondepending on that of expression(2.5). We must bear in mind that in accordance
with the structure of formula (2.8), the precisionof temperature calculations for heavy cloudiness(i.e. for n closeto unity) is considerably lower
than for low or medium cloudiness.Although this limits the possibilities
of applying formula (2.8), certain conclusionsmay be drawn as to

52

CLIMATIC

CHANGES

effectof cloudinesson the thermai regime of the lower atmospheric

levels.

For this purposewe must take into considerationthe dependenceof

albedo on cloudiness, which has the form

Os
= Os.n
+ So(1- n),

(2.9)

whereas,andasoarethe albedovaluesofthe earth-atmosphere

system
for overcast cloudinessand a clear sky, respectively.
From (2.8) and (2.9) it follows that
1

T,= b- blQs,[1
- o,n
- o%(1
- n)]- a+ aln). (2.10)
Assuming, in accordancewith the method applied in compilingthe
maps for the Atlas of Heat Balance of the World (1963), that

Sop
= 0.66 + 0.10

(2.11)

where is the albedo of the earth's surface, we find for the northern

hemisphere
an averagevalue of o = 0.20. Then from (2.9)we infer
that for sp- 0.33 and n = 0.50, the averagevalue o = 0.46.
Takingintoaccount
the obtainedvaluesof o ando, from(2.10)

we find that for an averageglobal value of Q the effectof cloudinesson

temperature is comparatively small and probably lies within the limits
of precision, which is not high in this case.
A more detailed analysis of the relation between cloudinessand the
air temperature at the earth's surfaceis given by Schneider(1972),from
which it follows that this influence strongly dependson the altitude of
the upper boundary of the clouds.Sincethe relation betweenthe clouds'
altitude and the temperature has not been sufficiently studied, it was
impossibleto determine the sign of this relation under average global
conditions.Subsequentlythe questionof the effect of cloudinesschanges
on the mean air surfacetemperature was studied in the author's works
(Budyko, 1975)and of Cess(1976),basedon the analysis of satellite data
on outgoing long-wave radiation. The conclusionwas derived in these
works that this effect is rather

small.

Formula (2.10) makes it possibleto draw certain conclusionsas to

the effect of climatic factors on the average temperature at the earth's
surface.

A 1% variation in the solar radiation changesthe average temperature by approximately 1.5Cfor cloudinessequal to 0.50 and the actual
albedo value. This estimate is about twice as high as that of temperature variations in the absence of atmosphere. Thus the radiational
properties of the atmosphere increase the effect of radiational fluctuations on the termperature distribution of the earth's

2 THE

GENESIS

OF

THE

CLIMATE

53

The variation of albedo by 0.01 changesthe temperature by 2.3C.

Hence the thermal regime greatly dependson the albedo if its variation
is not causedby cloudiness.As stated above, the effect of cloudinesson
the thermal regime connectedwith albedo variations is compensatedto
a great extent by correspondingvariations of the outgoing long-wave
radiation.

Model of an Average Latitudinal

Temperature Distribution. Horizontal redistribution of heat in the atmosphere and hydrosphere strongly influences the thermal regime at certain latitudinal
zones.

This problem is very difficult to deal with since all the forms of
horizontal heat transfer in the atmosphereand hydrospherewhich are
comparable with the amount of absorbed solar radiation have to be
taken into consideration.They include heat transfer by mean and largescale turbulent motions in the atmosphere and hydrosphere as well as
heat redistribution due to phase transformations of water. Since existing theoretical models are not yet sufficiently developed,it is difficult to
determine precisely the parameters of horizontal heat distribution, in
order to take into account all these forms of heat transfer.

Data

on the

heat balance componentsmay be utilized in the solution of this problem.

The heat balance equation of the earth-atmosphere system has the
form

Qs(1 - o) - Is = C + Bs,

(2.12)

where C = Fs + L(E - r), i.e. the sum of heat inflow due to horizontal
motions in the atmosphere and hydrosphere.
For annual mean conditions,the term B s correspondingto the gain
or lossof heat during a given period is equal to zero, and the heat influx
C is equal to the radiation balance of the earth-atmosphere system.
Since the values of this balance can be determined either by observations or calculations, we can simultaneously determine the value of the
horizontal

heat redistribution.

It may be assumed that the value of C is in some way connected

with the horizontal distribution of the mean temperature in the troposphere. Considering that the deviation of the earth's temperature from
the mean vertical distribution in the troposphereis small comparedto
geographictemperature fluctuations, we find that the mean air temperature in the troposphereis closelyconnectedwith the temperature at
the earth's surface. This assumptionis fully confirmed by a comparison
of average monthly temperatures at the earth's surfacewith those at the
altitude of the 500 mb level for various geographicalregions and seasons
(Kagan and Vinnikov, 1970).
Thus there exists a dependencebetween horizontal heat transfer
and temperature distribution near the earth's

CLIMATIC

54

CHANGES

Let us analyze the relation between these parameters for latitudinal zones of the northern hemisphere under mean annual conditions.
For this purposelet us calculatethe radiation balanceusing the equa-

tion

Rs = Qs(1 - as) - Is.

(2.13)

For latitudes 0-60we shall determine the albedoaccordingto formulas

(2.9) and (2.11), while for polar latitude, where formula (2.11) is not
sufficiently precise, we shall use albedo values obtained from satellite
observations (Raschke, MSller and Bandeen, 1968).
It is possibleto determine the meridional heat flux in the earthatmosphere system from the values thus obtained for the radiation
balance of the earth-atmospheresystemequal to the term C. A comparison of these values to relevant air temperature gradient at the earth's
surface shows that the relation between the parameters differs at various latitudes. Thus using the assumptionthat the meridional heat flow
is proportional to the average latitudinal temperature gradient, we
must assumethat the proportionality factor, i.e. the coefficientof largescale turbulent heat exchange, is a function of latitude. Although,
under present conditions,this function can be determined empirically,
its behavior under changing climatic conditions remains unknown.
Therefore it is necessaryto find other empirical relations between the
parameter C and the temperature distribution.
Consideringthat meridional heat transfer is directedfrom warmer
to colderregions, we may say that C dependson T - Tp, where T is the
mean temperature at a given latitude and Tp is the mean global temperature of the lower atmosphericlevel.

A similar assumption
was usedby Opik (1953)and Sawyer(1963,

1966).

To investigate this dependencelet us comparethe obtainedvalues

ofR with correspondingvalues of T - Tp (seeFig. 5). We find that there
is a definite relation betweenthe parametersthat can be expressedby
the empirical formula
Qs(1- as) -Is = fi(T- Tp),
(2.14)
where fi = 0.235 kcal/crn. month. deg.
This dependencegreatly simplifies the estimate of the meridional
heat redistributionin the modelof the thermal regime.
From (2.5) and (2.!4) we find that

T = Qs(1
- as)- a + a,n+ fiTp

B+ b- b,n

'

(2.15)

from which it is possibleto calculatethe mean temperatureat various

latitudes, these results being presented in Fig. 6 by the To

2 THE

GENESIS

OF

THE

CLIMATE

55

Rs (K cal/cm? month)
4-

-2

-4

-6

-8
-30

I,

-20

-10

10

r-Tp
Fig. $. Dependenceomeridional
heat transer
on temperaturedifference.

together with the T curve for measuredtemperatures. Obviously the

calculated and observedtemperatures correlate well.
It should be borne in mind that this correlation was obtained by
using a model containing only a single empirical parameter (/g) which
dependson the temperature distribution but not on the latitude.
Note that in the caseof a polar ice coverboundaryconcentricto the
T(C)
J0
20 -

10-

-10
-20,

..,

20
.t0
60
80 o
Fig. 6. Mean latitudinal distribution of air temperature;To, calculated
values; T, measured

56

CLIMATIC

CHANGES

pole, we obtain spasmodicvariations of the zonal-meanair temperature

at the ice field boundary;these variations correspondto variations in the
albedo of the earth-atmosphere system. But since the actual ice boundary is not concentric,it is advisableto calculate the average latitudinal
air temperatureby interpolatingit in the zonefrom the lowestto the
highest latitude of the ice boundary thus eliminating its spasmodic
variation within t,his latitudinal belt.
Observationaldata confirm that the temperaturejump on the ice
boundary,impliedby calculationsaccordingto this model,actuallytake
placein varioussectorsof the Arctic seasduring the coldseason.It may
be assumedthat in the caseof a concentricice distribution, this temperature jump would have been pronouncedin the distribution of winter
temperatures derived from observational data.
Model for Various Seasons. In order to study the atmosphere's
thermal regime during various seasons,it is necessaryto modify this
model by taking several additional factorsinto account.
Sucha more general modelwas suggestedby Budyko and Vasishcheva (1971) with the heat balance equation of the earth-atmosphere
system having the form
Qsw(1- Ow)-i,w = C,.,,+
Qc(1 - asc)- Ic = Cc - B,

(2.16)
(2.!7)

where B is the heat gain or lossdue to coolingor heating of the e arthatmospheresystemwhich dependsmainly on the processof coolingor
heating of the ocean.Here and further on, the subscripts"w" and "c"
correspondto the warm and coldhalf-years.
We shall make use of equations (2.5), (2.12), and (2.14) in order to
determine the I and C terms.

The useof the last equationis madepossibleby a comparisonof the

values of meridionalheat redistribution for various seasons,determined
fromdata on the heat balanceof the earth-atmospheresystem,usingthe
appropriate air temperature differences at the earth's surface. As a
result of this comparison,it was establishedthat relation (2.14)holdsfor
various seasons,the fi proportionalityfactor increasingslightly for the
cold, and droppingfor the warm half-year. The value of this factor for
the warm period in the southern hemisphere equaled that for the
northern hemisphere, while for the cold period it was higher than that
for the northern hemisphere.
The value of B is found from
Bs = s'(Tww- T,c),

(2.18)

where Twwand T are average latitudinal temperatures of the ocean

surface waters for the warm and cold half-years, respectively;s is

2 THE

GENESIS

OF

THE

57

CLIMATE

fraction of the total area of a given latitudinal zone occupiedby the

ocean, and 7 a dimensional factor.
In order to determine the ocean surface temperature we use its heat
balance equation in the form
Bs
Rww = LEw + Pw + --,

(2.19)

Rwc
+Pc- Bs
= LEe
-- + Fo,

(2.20)

where Rw is the radiation balance of the ocean surface; LE the heat

expenditure for evaporation; P the turbulent heat flux between the
ocean surface and the atmosphere;Fo is the heat transfer by ocean
currents which, accordingto empirical data, is of the utmost importance
during the coldtime of the year, when its absolutevalue is comparable
to that

of the radiation

balance.

To determineLE, P and Fo, we apply the approximaterelations

LB - fI'w,
P = c(Tw - T),
Fo = fi'(Tc- Tp),

(2.21)
(2.22)
(2.23)

where Tw is the surface temperature of oceanwater in C, and Tc is air

temperature for the cold half-year, f, c and fi' dimensionlessfactors.
The approximate relations (2.18), (2.21) and (2.22) were obtained
through simplifying the equationsdescribingthe heat exchangeprocessesin the upper oceanlevels and on its surface.Formula (2.23) was
obtained by a method similar to the one used for deriving the formula C
= fi(T- Tp).
In obtaining expression(2.21) we have simplified the well-known
evaporation formula
E = A(V)(qw - q),

whereqwis the specificair humidity equalto its saturationvalue at the

temperature of the ocean surface, q is the specificair humidity, and

A(V) is a coefficient depending on wind velocity. If we neglect the

dependenceof coefficientA on the wind velocity V and the deviations of
q/qw from the mean value, then, by using the relation between the
saturation humidity values and temperatures, we arrive at the approximate relation (2.21).

Our calculations involve only average latitudinal and average sea-

sonalvalues of the pertinent terms of the heat balance, and, as can be

seenfrom numerical experiments,the resultsof calculationsof temperature distribution dependvery little on the errors of the heat balance
components

CLIMATIC

58

CHANGES

This makes it possibleto neglect the influence on the evaporation

from the ocean surface, of changes in the relative humidity over the
oceanand in the wind velocity in various latitude zones,thus simplifying the determination of heat expenditure for evaporation and calculations of air temperature, without substantially affecting the precisionof
obtained

results.

The numerical factors in these equations (when the heat flows are
determined in kcal/(cm2 month) have the following values:7 = 3.0; f =
0.4; fiw = 0.22; fie -- 0.27 for the northern hemisphere south of the mean
boundary of Arctic ice; fie = 0.40 for the southernhemispherenorth of
the mean boundary of Antarctic ice; fie = 0.22 for zones with an ice
cover;fi' = 0.14 for the northern hemisphere;fi' = 0.20 for the southern
hemisphere; c = 0.84.
From this relation we may obtain the following formulas for determining the average latitudinal temperatures in the northern hemisphere:
TW

Qsw(1- osw)- a + alnw + fiwT,l

'3

f+c+27

/s(' - c)

(Rw
-Rwc
- fi'T)-f +c+27Qs(1
- )

(2.24)

- 2a + alnw -]- anc + f]wTpl -]- f]cTp2

+ Qsc(1- asc)
b - blnc-- 13c

b - blnw + 13w+

'y sc

f+c+27

7s(fi' -c)(b - blnw -- 13w)

(f+c+27(b-b,nc+fic)

Tc

Qsc(1- Osc)
-a

+ alrtc -- ficT, +

(Rww- Rwc+ fi'T,) +

'ysc

f+c+27

'3

f+c+27

Qsw(1- asw)

(2.25)

- 2a + aln w -]- alnc -]- wTpl + cTp2

+ Q sc(1- Oc)
b - bnw + fiw
b - bl nc + tic+

7s(c - fi') + 7sc(b - blnc -- 13c)

f+c+27(f+c+27)(b-bnw+fiw)

In this and the following formulas subscript"p" of values T, Qs, n,

as, R, s refers to the planet as a whole, subscript"1" to the warm

2 THE

GENESIS

OF THE

CLIMATE

59

year in the northernhemisphereand to the coldhalf-year in the southern hemisphere,and subscript"2" to the coldhalf-year in the northern
hemisphereand the warm half-year in the southernhemisphere.
Thus formulas (2.24) and (2.25) can also be used for calculating

averagelatitudinal temperaturesin the southernhemisphereas well,

assumingin this casethat T, refersto the coldhalf-yearin the southern
hemisphereand T2 to the warm half-year in the samehemisphere.
For an ice-covered zone formulas (2.24) and (2.25) are somewhat

simplified.Sincein this zonethe annual changeof heat contentin the

earth-atmosphere
systemis small and the meridionalheat transfer by
currentsis negligible, if it existsat all we find that we may use the
equations

T=
Tc--

Qsw(1- asw)- a + a,nw + fiwT - lh

b- bnw
+ w

Qc(1 - ac) -a

+ anc + ficTp2+ lh

b- bn
+ tic

'

(2.26)
(2.27)

wherelh is the gain (or loss)of heat due to cooling(or warming)of the
ice cover and the accumulation (or melting) of ice.
Sincethe value of lh is comparativelysmall, we shall considerits

average value for all latitudes and zones as equal to 0.8

kcal/cm . month.

To determineaverageplanetary temperatureswe shall use formu-

las which can be derived from (2.24) and (2.25), assuming that for the

planet as a whole the meridionalheat exchangeis equal to zero;

/Sp

f+ c + 2/
(Rw - Rw) +

Qp(1 - ,)
Tp1 z

+ Q(1

/SpC

f+ c + 2/

-- s) --2a + anp + an2

b - bnp
(2.28)

/spc
/spc(bb - bnp +
+
f + c + 2/ (f + c + 2/)(b - bnp)
/Sp

f+ c + 2/
(R,

- R2) +

/SpC

CLIMATIC

60

CHANGES

Qsp(1- sp)(1 - p2) - 2a + an, +

b - bn,

Tp2 = -

(2.29)
b-

/s
/s,c(b- bnp2)
blzp2qqf + c + 2,/
(f + c + 2,/)(b -

By applying formulas (2.24)-(2.29), we can calculate the distribution of average latitudinal temperatures for every half-year.
In this calculation it was assumed that the boundary of polar ice
correspondsto the average latitudinal temperature of the warm halfyear, which is equal to -IC. For the polar-icezone the albedoat the
earth-atmospheresystemwas taken to be 0.62 and 0.72 for the northern
and southern hemispheres respectively. The albedo of latitude zones
without any permanent snow or ice cover was determined accordingto
Table 8. The values given in this table were obtained from satellite
observations (Raschke, M511er, and Bandeen, 1968).
In determining the total radiation incident on the outer boundaryof
the troposphere,the radiation flux incident on a unit surfaceperpendicular to the solarbeam (meteorologicalsolar constant)was assumedto be
1.92cal/cm2. min. The average latitudinal values of the radiation balance for the ocean surface were taken from the Atlas of Heat Balance of

the World (1963).

These results are presentedin Fig. 7 together with the distribution

of average latitudinal temperatures obtained by observations.As we
see,the discrepancybetween the measuredand calculatedtemperatures
at various

latitudes

is at most 1-2 .

The describednumerical model of the atmospheric thermal regime

for various seasonsis not the only one. In a number of studies several
other numerical models have been suggested, in many cases more
detailed than this one. These models, however, did not as a rule meet all
the requirements listed at the beginning of this section.
Table

Albedo

Latitude
(N)

Half-Year

First

of Latitude

Zones

Latitude
(S)

Second

Half-Year

First

Second

65

0.49

0.54

0.25

0.27

55

0.40

0.48

15

0.25

0.26

45

0.35

0.45

25

0.29

0.28

35

0.29

0.40

35

0.40

0.32

25

0.27

0.30

45

0.46

0.41

15

0.27

0.25

55

0.53

0.49

0.29

2 THE

GENESIS

OF THE

CLIMATE

61

T(C)

3o(a)

"'\

(b)

20

-1o

-20

,,

-30
80 N

40

40

80c S

t,, , , , , , , O,

80c.m. 40

40

Fig. 7. Distribution of average latitudinal air temperatures.(a) First halfyear; (b) second half-year. Solid line, observational data; dashed line, calculated data.

Our model of the thermal regime satisfies all these requirements to

a first approximation, so that it is possible to use it in calculating
climatic changes.
Unambiguity and Stability of Climate. We may analyze the unambiguity and stability of climate by using these models of the atmospheric thermal regime.
Let us assume, in accordancewith empirical data obtained both for
polar sea ice and for continental glaciations on a plain and in mountains
outside of the tropical zone, that the mean latitudinal boundary of the
permanent snow and ice cover correspondsto an average annual temperature of -10C. Under this condition we will calculate (by means of
the thermal regime model for average annual conditions) the dependence of the mean boundary of the polar ice cover in the northern
hemisphere as a function of the solar radiation at the upper boundary of
the atmosphere.
This dependenceis depictedin Fig. 8, where AQs,
/ Qs,is the relative
variation of the solar radiation at the outer boundary of the

62

CLIMATIC

CHANGES

80 '3
_

-10

10

20

30

40

(%)

Fig. 8. Dependenceof the mean latitude of the polar-ice boundary in the

northern hemisphere on the radiation influx.

in per cent, and is the mean latitude of the polar ice boundary in the
northern hemisphere.
The dependenceof on AQsp/Qsp
is presentedby a set of heavy lines
showingthat this relation is ambiguousand that it differs substantially
in the casesof increasing and decreasinginflux of solar radiation.
First let us consider the case when heat influx

increases from an

initial low value. When the increaseis small, we have completeglaciation of the earth ( = 0), which is maintained until the heat influx
reachesits presentvalue (point 1) and up to a value increasingthe given
value by several tens of per cent (point A ). The glaciation regime
correspondingto point A is unstable and when the heat influx slightly
increases,it changesto a regime where glaciation is completelyabsent
(point A'). The ice-free regime is preserved with a further increase of the
heat

influx.

If the heat influx decreasesbelow the initial value which greatly

exceedsthe solar constant, then we have, at first, conditions for an ice-

free regime ( = 90).When the curve reachespoint E, which is near the

current value of the heat influx, polar glaciation occurs,which grows
rapidly as the heat influx decreases.After reaching point 3, corresponding to the current climatic regime, the decreaseof the heat influx by
about 2% causes the ice cover to spread southward to latitude 50N
(point B ). The glaciation regime correspondingto this point is unstable,
and with a small decreaseof heat influx, changesto completeglaciation
(point B'), which remains as the heat influx further decreases.

By means of this model we can also obtain the relation between

and AQp/Q depicted by the dashed line AB. Since the curve correspondsto growth of as AQp/Qv decreases,and vice versa, it also
correspondsto unstable glaciation regimes, which are replaced by

2 THE

GENESIS

OF

THE

CLIMATE

63

total glaciation regime or an ice-free regime, with small fluctuations of

heat influx. Consequently, point 2 on curve AB representing the actual
value of the heat influx, indicates an unstable regime that cannot exist
for an extended period of time.
Thus the relation between the polar glaciation and the heat influx
bears the form of a hysteresis loop, whose sectionsAA' and BB' correspondto the transition from one solution of the given equations to the
other, as shown in Fig. 8 by arrows. Other sectionsof the hysteresisloop
correspondingto stable regimes represent the dependence of on
AQsp/Qsp,both for increase and decreaseof the heat influx.
In studying the relation given in Fig. 8, we can also use the thermal
regime model for various seasons.The use of this model yields results
that quantitatively differ somewhat from the diagram in Fig. 8 but
entirely confirm the main qualitative regularities. This could be expected,sincethese regularities can be establishedon the basis of general
considerations that must be taken into account in compiling various
models of the thermal regime.
As shown above, the present value of the solar constant does not
excludethe possibility of a stable and completeglaciation of the planet
with very low temperatures at all latitudes (white earth). The stability
of this regime is due to the very high albedo of a surface covered with
snow and ice. In this case climatic conditions typical for the Antarctic
might prevail over the entire surface.
The possibility of a stable white earth regime apparently follows
from any realistic theory of climate.
The mean air temperature at the surfaceof a white earth can vary
greatly, dependingon the albedoof the earth's atmospheresystemover
a planet coveredwith snow and ice. Since the albedocan be calculated
only approximately, existing estimates of the mean temperature vary
widely. However, in all casesit appears to be much below the waterfreezing point, which indicates a high stability for a thermal regime
under these conditions.

Thus the current influx of solar radiation permits a completeglaciation of the earth; this regime is designatedin Fig. 8 by point 1. It is also
possibleif the level of incoming radiation dropsbelow the current level
(as shownin Fig. 8 by a line running along the horizontal axis to the left
of point 1), and also if the incoming radiation increases above the
presentlevel up to a value correspondingto the ice-melting temperature
in the warmest regions of the earth's surface. Once this temperature is
reached,part of the surfacemay get clear of ice, reducingthe albedoand
increasing the absorbed radiation. The point correspondingto the
boundaryof the completeglaciation regime with increasingsolar radiation is indicated in Fig. 8 by

CLIMATIC

64

CHANGES

it is natural to assume that, between the conditionsoutlined at

pointsA and 3, there exist regimesof partial glaciationwhich can be
shownon this diagram by a line connectingthese points.The path of
this line may be determined in the following manner.
Observationaldata as well as general physical considerationssuggest that under present climatic conditions,an increaseof the heat
influx to the earth's surface would lead to a rise in the mean tempera-

ture at the surfaceand to a recessionof polar ice, while a decreaseof the

heat influx would causea drop in the mean temperature and an expansion of the ice sheet.

In accordancewith this rule, the line running from A to 3 must

approachthis point from the left, which is possibleif this line intersects
the vertical axis at least at one more point (point 2).
Thus we arrive at the conclusionthat a third climatic regime may

exist under presentconditions,that is a secondversionof the partial

g!aciationregime of the earth, the surfaceof the ice sheetbeing larger
than one we have now.

Ater intersectingthe vertical axis at point 3, the line corresponding to the regime of partial glaciation,at a certain value of incoming
radiation exceedingits presentamount, will reach the horizontal line
correspondingto the ice-free regime (point E). Further on, the line
corresponding
to a continuingradiation increase,becomeshorizontal
and exits at the right.
In several of the mentioned works, calculations have been made
proving that with the current radiation influx, an ice-free regime is

probablein the Arctic.If we agreewith this contention,then this regime

canbe representedon our diagram by a point on the vertical axis at =
90, and in this case the line for the partial glaciation regime will
approachthe horizontal line correspondingto the ice-free regime, not
from the right of this point, but from the left. This position of the
intersectionis causedby the fact that if an ice-free regime is possible
under presentconditions,it is alsopossibleunder slightly lower incoming radiation.

Thus, in this casethe line representinga possibleglaciationregime

passesthrough the given point as a horizontal line that must crossthe
vertical axis at least once more. Therefore, with conditionswhen an ice-

free regime can exist in the Arctic at the presentlevel of incomingsolar

radiation, there must exist another regime of partial glaciation with a
smaller surface of ice cover than is the casefor present climatic conditions.

In addition to the relation presentedin Fig. 8, we showin Fig. 9 a

relation betweenthe average planetary air temperature and the

2 THE

GENESIS

OF

THE

65

CLIMATE

Tp(C
)

A'

60

4O

2O

I0

!0

20

30

--(%)

40

'o

Fig. 9. Dependence
o
theaverageplanetaryair temperature
on theradiation influx.

inflow, calculatedaccordingto the samemodel.All designationsin this

figure are the same as in Fig. 8.

It followsfrom Fig. 9 that this relation has much in commonwith

that for the boundaryof polar ice, and also has the form of a hysteresis
loop. The result presentedin Figs. 8 and 9 stronglydependson the
albedo difference of the earth-atmospheresystem with and without a
snow and ice cover. The albedo values used in these calculations were
determined from satellite observations. In this connection it should be

noted that the new reduction of these observations (Raschke, et. al.,

1973)producedsomewhatlower albedovaluesfor the polar regionsthan

the earlier works which were used in our calculations. The estimate,

based on the new data, somewhat modified the quantitative results

presentedin Figs. 8 and 9, but did not affecttheir main character.

Our conclusionthat modern climate is ambiguousis based on the

use of a thermal regime model for the stationary condition of the
atmosphere-ocean-ice
cover system. This condition correspondsto a
stable or slowly changing climate.
For a climate whosechangesare comparativelyrapid, it may not be
appropriateto assumethat the systemis stationary. Since the continental ice coversas comparedto other parts of the system are

66

CLIMATIC

CHANGES

stable, a greater number of different climatic versions can exist under

nonstationary conditions as compared to the few solutions corresponding to stationary conditions.
Let us assume that during the postglacial warming there still
existed a continental glaciation with a boundary not correspondingto
the changed climatic conditions. In this case a new type of thermal
regime appears, which can be calculated by means of a numerical model
for a given ice-cover boundary. This thermal regime changes together
with the ice-coverboundary, producing an unlimited number of climatic
versions corresponding to the given values of the external climatic
factors.

However, despite the fact that in our epoch there exist such vast
glaciations as the Arctic and Antarctic, the polar ice can be approximately consideredstationary, since their mean boundaries in the northern and southern hemispheres correspond to the same values of the
climatic

factors

that

determine

these boundaries.

The stability of the current mean boundary of the polar ice in both
hemispheresis apparently causedby the fact that the main part of this
boundary consistsof not very thick sea ice.
The actual temperature variations during the main part of the
earth's history were limited to short sectionsof the lines in Figs. 8 and 9,
from just above point E to just above point B.
It is interesting to comparethis part of the line in Fig. 9 to the dotand-dashline passingthrough point 3 and correspondingto variations of
the average planetary temperature, when the polar ice doesnot affect
the thermal regime. This comparisonshowsto what extent the polar ice
intensifies the temperature variations causedby fluctuations in incoming heat.
Thus the present climate apparently is not the only possibleone
under the existing external climatic factors, i.e. it is not transitive in
the sense of the definition given by Lorentz. It is possible that in
addition

to the

considered

mechanism

there

are

also other

forms

of

climate intransitivity that will be found in further studies.

In order to estimate the stability of the current climate, we may
make use of the relation presented in Figs. 8 and 9. As we see, the
climatic regimes correspondingto sections BE are highly sensitive to
small variations of the heat influx. It follows therefore that comparatively small variations in external climatic factors can considerably
affect the climate of today.
This conclusionis of great importance for understanding the laws of
climatic changes.
Results of Other Investigations. After the first work on the stability of current climate had been published (Budyko, 1968)this

2 THE

GENESIS

OF THE

CLIMATE

67

was studiedin several other investigations,the results of which we shall

now analyze.The first of thesestudieswas carriedout independentlyof
our work by Sellers (1969).Sellerscompileda semiempiricalmodelof
the atmosphere'sthermal regime which was basedon the energybalanceequationof the earth-atmospheresystemand took into consideration the feedback between the thermal regime and polar ice. From this
model Sellers determined the dependenceof the average latitudinal
temperatureat the earth'ssurfaceuponthe solarconstant,the albedoof
the earth-atmospheresystem,and other climatic factors.Then he used
these relations to estimate the climatic effectsof the polar ice and solar
radiation.

From Sellers'

work

it followed

that

if the albedo of the earth-

atmospheresystem in the Central Arctic drops from 0.62 to 0.32 as a

result of melting polar ice, the average annual air temperature at the
earth's surfaceincreasesin this region by 13C,while at low latitudes it
increasesby 2C. In calculating the decreasein the solar constant
sufficientto causethe completeglaciation of the earth, Sellers found, on
the basis of the albedo-temperaturerelations obtained from observational data, that it amounted to 2%. For a weaker albedo-temperature

dependence,this value increasedto 5%. However, Sellers considered

that this relation did not agree with observations.
It is noteworthy that Sellers' results are closely correlated with
those of our paper, which likewise state that the existing ice cover
reducesthe average annual temperature in the Central Arctic by 14C
and at low latitudes by 2C.
This correlation of the main results is very interesting, because,
although the main principle of the climate models was the same, the
calculationswere based on many different assumptions(Budyko, 1970;
Sellers, 1970).

In his next work, Sellers (1973) developeda more detailed model of

the atmosphere'sthermal regime for different seasons,the main results
of which were closeto those of his previous works.
Another semiempirical model of the atmosphericthermal regime
that takes into account the feedback between the temperature and the
polar ice was developed by Faegre (1972). Faegre confirmed the main
deductions

of the above works and called attention

to the fact that his

model indicated the ambiguity of the current climate. Faegre obtained

five different solutions of the equation for the average latitudinal temperature distribution at the earth's surface. Out of these he selected
three as having physical sense:1) for existing climatic conditions,2) for
the white earth, and 3) for an intermediate one.
Berger (1973) also proved that Sellers' model likewise implies that
during the present epoch,three climatic regimes may prevail. The

68

CLIMATIC

CHANGES

correspondsto actual observedclimatic conditions;the secondto a colder

climate with temperature 10-15Cbelow the present one, and the third,
to very low negative temperatures at all latitudes (white earth).

Schneider and Gal-Chen (1973) studied climatic stability on the

basisof semiempiricalmodelsof the thermal regime, supplemented
by a
considerationof the instability of these processes.For this purpose
another term was added to the heat balance equation of the earthatmospheresystemthat describesthe instability of the thermal regime:

rn

0T
0t

= M,

(2.30)

where 3T/3 t is the variation of the average temperature at the earth's

surface with time, rn the coefficient of the therma] inertia of the earthatmospheresystem, and M the algebraic sum of the heat balance
componentsfor the earth-atmosphere system.
This work has confirmed the conclusionon the high sensitivity of
the atmospheric therma] regime to small variations of the solar constant, and at the same time provedthat when the inflow of solar energy
is steady, the thermal regime returns to its initial condition even after
very big drops in the average global temperature, probably reaching
-18C. Total glaciation of the Earth is possibleonly when the average
temperature drops even more. Thus, the thermal regime of the earthatmosphere system is highly stable relative to the fluctuations of the
mean global temperature that are not causedby long-term variations of
the heat inflow.

An analysis of semiempirical models of the atmosphere'sthermal

regime was carried out by Held and Suarez (1974). They used empirical
parameters determined from satellite observationsof the outgoinglongwave radiation, their results being similar to ours. Held and Suarez
have confirmed the finding that the feedback between the thermal
regime and the surface area of polar ice strongly enhancesthe sensitivity of the ice-coverboundary to variations of the solar constantand other
climatic factors. They have also confirmed the existence of "critical
latitudes" beyond which the polar ice expands equatorward under constant

external

climatic

factors.

Held and Suarez analyzed the effect on climate stability of the

additional term corresponding to the horizontal heat redistribution
causedby macroturbulent processes.Thus it was possibleto obtain a
continuous distribution of the average latitudinal air temperature for
concentricdistribution of the solar ice relative to the poles.
A semiempirical model of the atmospheric thermal regime was
developedby Gandin, II'in and Rukhovets (1973). They constructeda
model of air temperature distribution along latitude and longitude and
averaged over the

2 THE

GENESIS

OF

THE

CLIMATE

69

By using the heat balanceequationof the earth-atmospheresystem

these authors have consideredthe horizontal heat transfer, assuming
that it was determined by microturbulent exchange. It was also assumedthat the coefficientof turbulent heat exchangedid not dependon
the coordinatesand that the outgoing long-wave radiation was determined as a function of temperature accordingto the above-givenempirical formula. The effect of polar ice on the atmosphericregime was taken
into account in a manner

similar

to that used in our model.

On the basis of empirical albedo values, the authors found the

distribution of the mean annual air temperature over the northern
hemisphere to be closeto observational data.
Later on, the variations of the air temperature and the ice boundary
were studied for a changing solar constant. It was found that when it
increasesby several tenths of a per cent, the polar ice melts completely,
and when it decreasesby the same amount, a total glaciation of the
earth takes place, accompaniedby a drastic drop in the average global
temperature (white earth).
These authors have pointed out that the high sensitivity of the
results to small variations of the solar constant, dependedpartly on the
choice of the albedo values of the earth-atmosphere system with or
without ice. By narrowing the gap between these values, it is possibleto
obtain the critical value of the solar constantcorrespondingto the total
glaciation of the earth, which is closeto the results given above.
Since the empirical values of the albedo are determined with a
certain error that can not noticeably affect the results of calculations of
the critical values of the solar constant, correspondingto the total
glaciation of the earth and to the completemelting of the polar ice, it is
obviousthat all the calculationsof these critical values are approximate
and will be made more exact as the albedo is determined

with

more

precisionand the models of the thermal regime improved.

Nevertheless, the agreement of the results obtained from all semiempirical modelsthat take into considerationthe feedbackbetween the
temperature field and the polar ice, should be noted. It follows from all
the models that

a decrease of the solar constant

of at most several

percent can causecompleteglaciation of the earth. As mentioned above,

this conclusiontestifies to the very high sensitivity of the current
climate to changesin climatic factors.
Besidesthe aboveworks, semiempiricaltheoriesof the atmospheric
thermal regime were discussedin many studies recently published
(Dwayr, Peterson 1973; Gordon and Davis, 1974;Hantel, 1974; Suarez,
Held, 1975; North, 1975; 1975a; Cess et al. 1976). Interest of many
authors in these modelsshowsthat they appear to be useful means for
studying the climate

CLIMATIC

70

CHANGES

Let us comparethe sensitivity of the atmosphericthermal regime to

the change of solar constantobtained from semiempirical modelswith
the results of similar calculations based on more general theories of
climate.

Of particular interest is the study by Wetheraid and Manabe (1975)

of modern climate sensitivity to a change in incoming radiation.
In contrast to the abovementioned works that usedthe semiempirical models of the average temperature distribution, Wetheraid and
Manabe used a three-dimensional model of the general theory of climate. This model took into account all the details of the dynamic
processesin the atmosphere, as well as the effect of the water phase
transitions on the thermal regime and the feedback between the snow
sheets, polar ice, and air temperature.
Wetheraid

and

Manabe

found

that

a 4% decrease

in the

solar

constant causesthe expansion of the snow and ice coversover the entire
earth.

A 2% increase of the solar constant leads to a recession of the ice

sheet and to a 2Cincrease of the mean annual air temperature at the

earth's surface at low latitudes, and to as much as a 10Cincrease at
high latitudes (80N).
For comparison we give the temperature variations curve calculated accordingto our semiempirical model for a 2% increasein the solar
constant (Fig. 10).
In these calculations (as well as in those on the basis of which Fig.
27 is plotted) the albedo values of the earth-atmosphere system were
obtained from satellite data.

It is noteworthy that Wetheral and Manabe's results agree with the

main conclusionsimplied by semiempirical models on the high sensitivity of the current climate to small variations of the heat influx at the
AT

86

10

20

30

40

50

60

70

80

90 N

Fig. 10. Effect of a 2% increase in the solar constant on average latitudinal air temperature. Solid line, results obtained from the semiempirical model;
broken line, Wetherald-Manabe

2 THE

GENESIS

OF

THE

CLIMATE

71

outer boundary of the atmosphere.Note also the agreement of conclusions on the much higher temperature increase at high latitudes as
comparedto low latitudes with increasing heat inflow.
The results obtainedin these works have considerablychangedour
conceptionof the unambiguity and high stability of present-dayclimate.
As we shall seein the following chapters,it is impossibleto explain the
chief regularities of climatic changeswithout taking these results into

3 Contemporary Climatic
Changes

3.1 The Warming During the First

Half of the Twentieth Century

Secular Temperature Trend. The general climate changesthat

took place in the first half of the twentieth century were describedin
Chapter 1. Accordingto investigationscarried out by Rubinshtein (1946,
1973), Rubinshtein and Polozova (1966), Mitchell (!963), Lamb (1966)
and others, these climatic changesvaried in different regions and from
season to season.

In particular, it o1!ows
romthe works of Rubinshtein and Polozova
that the mean anomalies of monthly air temperature at a given point
vary great!y during successivemonths, these variations often being of

an irregular nature. The relation betweentemperatureanomalies'at

various points weakens rapidly with distance.

It is possiblethat the time and spacestructure of temperature

anomalies is caused to a certain extent by random fluctuations of

atmospheric circulation. Averaging temperature anomalies over a period of time for certain stations or limited regions does not entirely
exclude the effect of suchfluctuations, thus making it difficult to reveal

the regularities in climatic changes.

For a study of global fluctuationsof climatic conditions,the largescaleaveraging of the air temperature anomalies is of great importance.
This approach is based on a number of physical considerations.
As we know from Chapter 2, there is a linear relation between the
!ong-waveradiation emitted into outer spaceand the air temperature at
the earth's surface. Therefore the averaging otemperature anomalies
is equivalent to averaging the outgoing radiation anomalies.
As a rule, the anomalies of the outgoingradiation in limited regions
are generally compensatedto a great extent by averaging over the
surface area, and their effect on the global climate is limited. Climatic

3 CONTEMPORARY

CLIMATIC

CHANGES

73

conditions are strongly affected by radiation anomalies over large regions and especiallyby global anomalies.
The algebraic sum of this anomaly and that of the short-wave
radiation absorbedby the earth-atmospheresystemdeterminesthe sign
of the air temperature variations over the planet as a whole.
Average air temperature anomaliesfor large areas were calculated
by Mitchell (1961, 1963) on the basis of air temperature observations
carried out by meteorologicalstations in various geographical regions.
For a more precisedetermination of mean anomalies we have used
maps of air temperature anomalies and not the observational data of
separate stations. Such maps compiledat the Main GeophysicalObservatory give the distribution of the averagemonthly temperature anomalies for each month beginning with 1881through 1960 in the northern
hemisphere,with the exceptionof the equatorial zone,for which there is
not enough information referring to the first part of this period.
In Fig. 11, the secular variations of the air temperature anomalies
are presentedas calculated by Spirina, for the extraequatorial zone of
the northern hemisphereand for the 70-85Nlatitude zone.All the data
were averaged over running five-year periods.
From this figure it followsthat at the end of the nineteenth century

-to

1890

1900

1910

1920

1930 1940 1950

1960 1970

Fig. 11. Secular variations of air temperature anomalies over running fiveyear periods. (1) Anomalies of the average annual temperature of the northern
hemisphere;(2) temperature anomalies of the 70-85N latitude zone during the
warm half-year; (3) the same during the cold

74

CLIMATIC

CHANGES

a warming began in the northern hemisphere which reached its first

and weak maximum during the last years beforethe turn of the century.
It was followedby a slight temperature drop which soontook an upward
leap. This leap greatly acceleratedduring the coldperiodat the end of
the 1910s and the beginning of the 1920s. The positive temperature
anomaly reached its maximum at the end of the 1930s,while in the
1940sa coolingprocessbegan that acceleratedin the 1960s.In the 1960s
the mean temperature in the northern hemisphere reachedthe level of
the 1910s. Thus the cooling of the last decadeshas compensatedfor
approximately half of the temperature increase that had been taking
place since the end of the nineteenth century.
It appears that the secular temperature trend in the extraequatorial zone of the northern hemisphere follows that of the air temperature
at the earth's surface for the entire planet. The data available for the
southern hemisphere are more limited than for the extratropical latitudes of the northern hemisphere and show that in the equatorial zone
and the extratropical latitudes of the southern hemisphere, the variations of the average air temperature were weaker than in the extratropical zone of the northern hemisphere. Nevertheless, the nature of these
variations in most of the zones for which data exists apparently coincided with the variations in the zonescoveredby numerous observations
(Mitchell, 1963).
Figure 11 showsthat secular temperature variations increase with

latitude, the air temperature during the cold period changing more
drastically than during the warm period, especiallyat higher latitudes.
The general picture of the thermal regime variations presentedby Fig.
11 is similar to that obtained by Mitchell.
It is noteworthy that in addition to regular temperature variations
over long periods, Fig. 11 reveals numerous short-term temperature
fluctuations, which reflect the effect of unstable circulation processes
that have not been entirely excludedby averaging over spaceand time.
Due to the fact that temperature anomalies of various seasonsdiffer
comparatively little at low latitudes, the average meridional temperature gradient changedlessduring the warm seasonthan during the cold
seasonof the year.
Secular variations of anomalies of the mean meridional temperature gradient are presentedin Fig. 12 (Budyko and Vinnikov, 1973).The
anomalies were determined by the root-mean-square method on the
basis of the average latitudinal temperatures for every 5-zonein the
interval from 25 to 70Nand subjectedto five-year running averaging.
It follows from the figure that from the 1880s to the 1930s the
meridional temperature gradient kept dropping. During this period
there were two comparatively short intervals when the

3 CONTEMPORARY

CLIMATIC

75

CHANGES

Ot--
(C/10
lat)

07

-0,
2

(b)

01

-0,2

1890

1900

1910

1920

1930

1940

1950

1960

1970

Fig. 12. Secular variations of meridiona! air temperature anomalies in

latitude zone, 70-85N. (a) mean annual anomalies; (b) mean anomalies for the

warm half-year; (c) mean anomaliesfor the cold half-year.

gradient rose-during the first years of the twentieth century and

during the secondhalf of the 1910s.
Beginningwith the latter half of the 1930s,the meridional gradient
began to climb again, its anomaliesby the end of the 1960sreaching the
same level as at the turn of the century.
Secular Variations of Precipitation. Along with temperature
variationsduring the warming period,notablechangesin the amount of
precipitation took place in several regions.
The secular variations of precipitation during the cold part of the
year (from Novemberto March), in the steppeand forest-steppezoneof
the U.S.S.R. shown in Fig. 13, are based on data compiled by Grigor'eva. The precipitationis calculatedfor running five-year periodsand
is based on observational

material

obtained

at 21 stations.

From

comparisonof Figs. 11-13 we can see that when the average temperature rises and the meridional temperature gradients drop, there is a
tendencytoward a decreasein precipitation in the regions of unstable
moisture.

Drozdovand Grigor'eva (1963)have noted that during the epochof

maximum warming (1930s), widespread droughts in the regions of
insufficient moisture over the territory of the U.S.S.R. and North America were more frequent than in the previous and following decades.

76

CLIMATIC

CHANGES

R(mm)
--

160
140
120

--

.,

1900

1910

1920

1930

1940

1950

1960

Fig. 13. Secular variations of total precipitation during cold half-year in

steppeand forest-steppezones of the U.S.S.R.

we know, during these years, the level of the Caspian Sea dropped
drastically by about 170cm owing to the decreaseof precipitation in the
Volga River basin.
Shnitnikov

(1969) and others have noted that the fluctuations of the

Caspian Sea level are similar in many respects to those of several

European lakes, and therefore reflect the large-scale anomalies in the
regime of atmosphericprecipitation.
It may be assumedthat to a large extent the secular variations of
precipitation are causedby the variation of the meridional temperature
gradient which affectsthe nature of the atmosphericcirculation.
As the meridional temperature gradient decreases, so does the
intensity of the water-vapor transfer from the oceanto inland regions,
and conversely.
This conclusionis well confirmedby observationaldata. Figure 14
showsthe dependenceof the precipitation during the cold period of the
year in the steppeand forest-steppezonesof the U.S.S.R. on the meridional temperature gradient for running five-year periods from 1891 to
1960. It follows from this figure that the correlation coefficientis 0.78.
The precipitation over this territory varied by 50% from its mean value
during various five-year periods.It is obviousthat suchvariations in the
amount of precipitation over a vast territory have a great influence on
streamflow parameters and that agricultural cropyields dependconsiderably on precipitation in the regions of insufficient moistening.
The fluctuations of the Caspian Sea level are a goodillustration of
the dependencebetween the amount of precipitation and the river
discharge.The drasticdrop of its level in the 1930swas the consequence
of a precipitation decrease over most of the European part of the
U.S.S.R. During the same period the meteorological conditions for
agriculture over the greater portion of Eurasia and North America also

worsened.

It is interesting to comparethe changesin humidity conditions

3 CONTEMPORARY

CLIMATIC

CHANGES

77

r(mm)
170"
160

150

140

130

ee

ee

120

!1o

,I . :

-0.4 -0.3 -0.2

-0.1

0.1

0.2

' Ot--'(C/10
lat)

0.3

Fig. 14. Dependenceof precipitation during cold half-year in steppeand

forest-steppezonesof the U.S.S.R. on meridional temperaturegradient in 2575N latitude

zone.

the middle latitudes

of the continents in the eastern and western hemi-

spheres.In orderto studythesevariationsan analysiswas madeof the

wheat crop deviations from its mean trend in the U.S.A. for separate
five-year periods (1910-1971).It follows from these data that during
individual intervals the cropchangedby as much as 16%as comparedto
the norm, and that this was causedmainly by the fluctuation of humidity.

Since such data are not available for vast territories of Eurasia, we

had to rely on the informationconcerningthe CaspianSealevel in order

to study humidity conditions.

In Fig. 15 the wheat cropanomaliesin the U.S.A. are comparedto

the fluctuations AH of the Caspian Sea level during the same period.
These two parameters seem to correlate comparatively well with a
coefficient of 0.74. Here we find confirmation

that this mechanism

of

precipitation change exercisesa similar influence on the humidity

regime of middle latitudeson different continentsas well.
This conclusionis likewise confirmed by direct measurements of
precipitationin North America. Landsberg(1960)comparedthe air
temperature and precipitationin the U.S.A. for two 25-year periods:
1906-1930and 1931-1955.During the secondperiod the average annual
air temperature at the earth's surfacewas 0.5higher than during the
first.

For a comparisonof precipitationamountsduring theseperiods,the

choiceof the years was not a fortuitousone, as a sharp changein the
meridionalgradienttookplacenot at the boundaryof the given

78

CLIMATIC

CHANGES

(%)
_

$_

0-

e
ee

ee

eee

I f f

i I I,

-100

I I Ifil

-50

! I i lili

50 A H (em)

Fig. 15. Correlation betweenhumidity conditionsin Eastern Europe (level

fluctuations of the Caspian Sea) and in North America (relative wheat crop
anomalies in the U.S_4.).

but around 1920. Nevertheless, Fig. 12 showsclearly that the meridional temperature gradient of the secondperiodwas lower than that of the
first.

As noted by Landsberg, the annual total precipitation over the

greater part of United States territory was lower in the secondperiod,
which agrees with the above conclusion.
A detailed investigation of the dependenceof precipitation on the
variations of the meridional temperature gradient was carried out by
Drozdov and Grigor'eva (1963, 1971), who found that though at high
latitudes the general pattern of precipitation during the warming and
cooling periods is comparatively complex, in regions of insufficient
humidity in middle latitudes, a tendency neverthelessprevails toward
an increase of precipitation as the temperature drops in the Arctic. A
similar conclusion could be drawn from the study by Lamb (1974).
Drozdov and Grigor'eva explained this effect by the intensified transfer
of water vapor to intracontinental regionswhen the temperature difference between low and high latitudes increases.
This mechanism requires special study. It seems that with the
growth of the meridional temperature gradient, the subtropical high
pressurebelt shifts to a lower latitude. This augments the amount of
water vapor being transferred from the ocean to the intracontinental
regions in middle latitudes.
Methods of studying atmospheric circulation variations based on
the standardizationof synopticprocesseshave led to diverse and

3 CONTEMPORARY

CLIMATIC

CHANGES

79

contradictory results. This is apparently due to the differencesin the

indicesof circulation intensity applied in various works and also to the
fact that the regions under study did not coincide.
It follows from general considerationsthat the intensity of the
westward transfer and other forms of circulation can be expected to
increase when the temperature difference between high and low latitudes increases,becausethis differenceis the main sourceof energy for
large-scaleatmosphericmotions.
In order to study the circulation variations during the warming
period of the twentieth century, it is necessary to apply the methods
either of the general theory of climate, or of the empirical analysis of the
circulation regime on the planet as a whole (or at least over one of the
hemispheres).
Polar Ice. As shownby observations,the distribution of polar sea
ice has a great infieunce on the atmosphericthermal regime. At high
latitudes the air temperature during the coldtime of the year over the
ice-free ocean surface drops, as a rule, to only several degreesbelow

zero, because the ocean releases much heat into the air. But in the

presenceof an ice cover, which significantly reducesthe heat flow from

the oceanto the atmosphere,the temperature of the lower air layer can
drop to several tens of degreesbelow zero.
Thus the migration of the polar sea ice boundary greatly affectsthe
air temperature in related regions,especiallyduring the coldtime of the
year. Unfortunately, observationaldata of the sea ice boundary are
greatly limited for the current period of climatic change.
Data on the ice boundary for the last century mainly refer to the
Atlantic sectorof the Arctic and the adjacentseas;similar information
on other regionsof the Arctic Oceanand Antarctica is available only for
the last 20-30 years.
Many attempts have been made to reveal the relation between
climatic fluctuations and ice cover in the North Atlantic
ents and Kara

and the Bar-

Seas.

Rubinshtein and Polozova (1966) have found that the amount of ice

in the Atlantic sectorof the Arctic began to decreasein the 1920s.This

processdid not stopwith the end of the warming period, but continued
in the Barents Sea up to the middle of the 1950s,when the amount of ice
began to increase gradually. Off the coast of Iceland, the ice had been
decreasingsincethe end of the last century and through the 1940s,then
it began to expand. It is obviousthat the polar ice which is closely
related to the thermal conditionsof the upperlayer of seawatersat high
latitudes,possesses
a certaininertia that causesthe ice-coverchangesto
lag behind the fluctuations of the global climate.
An interesting result was obtainedby Flohn (Inadvertent

8O

CLIMATIC

CHANGES

Modification, 1971),in describingthe variations of the ice boundaryin

the Arctic. He calculatedthe differencein mean latitudinal temperature
averagedover the year in the northern hemisphere,betweenthe periods
1931-1960and 1961-1970.Flohn'sdiagram implies that this differenceis
closeto zero from the equatorto 50N;it increasessharply in the polar
latitudes, reaching its maximum value of 1 between 70 and 80N, and
then dropsagain. Such a distribution may be related to the shift of the
mean boundary of polar ice toward lower latitudes in the 1960s.
3.2 Causes of Climate Warming

Radiation Factors of Temperature Variations. At the beginning

of the twentieth century it was already established that the average
amount of direct solarradiation that reachesthe earth's surfacethrough
a cloudlesssky can vary noticeablyfrom year to year. Thesefluctuations
are clearly discernible on the curves of secular variations of direct
radiation

based on actinometric

observations.

The

curves also show

considerablechangesover longer periodsof time, such as several decades (Budyko and Pivovarova, 1967).
It is interesting to compare the secular temperature variations in
the northern hemisphere with those of the radiation reaching the
earth's surface. For this purposeactinometric observationscarried out
at several European and American stations between 1880and 1965,were
reduced and an average curve obtained of the secular variations of
direct radiation under cloudlessconditions(Pivovarova, 1968).In Fig. 16
curve b presents the solar radiation smoothed over running 10-year
periods.As we see, the solar radiation had two maxima; a brief one, at
the end of the nineteenth century, and a longer one, in the 1930s.Let us
compare curve b with curve a representing the secular temperature
variations, also smoothed over decades. We find that there is a resemblance between them. Thus, both curvespossesstwo maxima, one at the
end of the nineteenth century, and the secondand principal one, in the
1930s. At the same time there are certain differences also. For instance,
the first maximum is more pronouncedon the radiation curve than on
the temperature curve. The similarity of curves a and b leads us to
believe that radiation fluctuations causedby irregularities in the atmospheric transparencyare an important factor of climatic change.In order
to clarify this problem it is necessary to calculate temperature variations as a function of the atmospheric transparency for short-wave
radiation.

In his studiesmentioned in Chapter 1, Humphreys establishedthat

the main factor affecting the atmosphere'stransparency on a

3 CONTEMPORARY

CLIMATIC

CHANGES

81

(C)

'!f
(a)

O,

1,

102 (b)
100

98

1880

1890

1900

1910

1920

1930

1940

1950

1960

Fig. 16. Secular variations of (a) temperature and (b) direct radiation
anomalies.

scale is the presenceof comparatively small aerosol particles that remain in the lower level of the stratospherefor a long time.
Humphreys and Wexler assumed that the smallest particles can
remain in the atmosphere for several years. These have a very small
effect on the long-wave radiation but significantly augment the scattering of short-wave radiation, as a result of which the planetary albedo
increases, while the amount of radiation absorbed by the earth as a
whole decreases.It shouldbe noted that due to the preferred scattering
of the radiation in the direction of the incident beam (Mie effect), the
direct radiation decreasesmuch more because of the scattering than
does the total solar radiation. Since the earth's thermal regime is
affected by the fluctuations of the total radiation, it is necessary to
determine the relation between this parameter and the presence of
aerosol particles in the stratosphere. For this purpose we can use the
methodsapplied by Shifrin and his colleaguesin their studieson atmospheric optics (Shifrin and Minin, 1957;Shifrin and Pyatovskaya, 1959).
This study resulted in a determination of the ratio of the total
radiation decrease to the direct radiation decreasefor average conditions at various latitudes, when a layer of aerosolparticles is present in
the stratosphere. The results are given in Table 9.
Shifrin's studies have shown that the values given in this table
dependcomparatively little on the sizeof the particles if their prevailing
diameters are between (2-3). 10-2 and (2-3). 10- tm.
When determining the effects of a change in the amount of direct
radiation on the average temperature at the earth's surface, it is

CLIMATIC

82

Table

CHANGES

Effect of Statospheric Aerosol on the Ratio

of Total
Latitude
Ratio

to Direct

Radiation

Decrease

90

80

70

60

50

40

30

20

10

0.24

0.23

0.22

0.21

0.19

0.18

0.16

0.14

0.13

0.13

sary to considerthe dependenceof mean temperature on the incoming

solar

radiation.

These

calculations

have

been

carried

out time

and

again. During recent years it has been found that a 1% change in

radiation causes a 1.2-1.5C change in the temperature for a constant
albedo of the earth-atmosphere system (see Section 2.2).
Let us compare the radiation and thermal regimes of the earth
during two 30-year periods: 1888-1917 and 1918-1947. From the data
shown in Fig. 16, it follows that during the secondperiod, the direct
radiation was 2% higher than during the first one. Since, accordingto
Table 9, the mean weighted ratio of the total radiation changesto the
direct radiation changesis 0.16, we find that the total radiation during
the secondperiod was higher in the northern hemisphereby 0.3%. This
correspondsto the increase in the average temperature by approximately 0.4C. The actual temperature difference for these periods, as
determined from the data shown in Fig. 16, is equal to 0.33C,which

agreeswell with the results of calculations(Budyko, 1969).

For a more detailed study of radiational changesand their effecton
the air temperature, we shall fall back on the model of the thermal
regime for various seasonsdealt with in Chapter 2.
Figure 17 shows secular variations of the smoothed anomalies of
direct radiation for a cloudlesssky in the years 1910-1950.This diagram
is basedon the data obtainedat severalactinometricstationsin Europe
and North

America

scattered

over the latitude

zone of 40-60N.

In

accordancewith the above stated hypothesis,let us assumethat the

variations of direct radiation are causedmainly by the changingtransparency of the lower stratospheric layer, due to fluctuations in the
concentrationsof aerosolparticles.
By using the data of Fig. 17, consideringthe relative optical depth
of the aerosollayer, and assumingthat the aerosolis homogeneously
distributedin the stratosphereover the entire northern hemisphere,we

can determine

the secular variations

of the direct radiation

at various

latitudes both for the average annual conditionsand for separate seasons.

As previously noted, the contribution of aerosol concentrationto the

changesin total radiation representsa small portion of the variations of
direct radiation. By taking into accountthe dependenceof the ratio

3 CONTEMPORARY

CLIMATIC

CHANGES

83

total and direct radiations on the altitude of the sun above the horizon,

it is possibleto calculate the time variations of the total radiation in

different

latitude

zones.

Figure 18representsthe ratio of the variation of total radiation over

the northern hemisphereto that of direct radiation at 50Nfor the warm
and cold half-years, as a function of latitude.
In order to determine the temperature variations correspondingto
the abovegiven data on radiation, it is necessaryto take into account
the relation between the thermal regime and the ice cover.
It is obviousthat we cannot rely on the assumptionthat the oceanpolar ice-atmospheresystem is stationary. Due to the large thermal
inertia of the ocean and the continental ice cover, this system can be
consideredstationary only over long periods of time. Therefore, the
assumptionof its being stationary is not applicable to the study of
(%)
2-

o
-1
-2

1910

1920

1930

1940

1950

Fig. 17. Secular variationsof the direct radiation anomaliesduring 19101950.

/S5o
1.0-

0.5

30

60

Fig. 18. Ratio of variation of total radiation over northern hemisphereto

that of direct radiation at 50N. (1) Warm half-year; (2) cold

CLIMATIC

84

CHANGES

current climatic fluctuations that have been going on for only a few
decades.

At the same time, we cannot ignore the changesof polar ice regime
which are connected with climatic variations. For instance, observa-

tional data show that the warming of the Arctic has led to an approximately 10% shrinkage of the surface area of sea ice (Ahlman, 1953;
Chizhov and Tareeva, 1969).

It is very difficult to carry out precisecalculationsof nonstationary

processesin the ocean-polarice-atmospheresystem, especiallysincethe
heat exchangemechanismbetweenthe surfaceand deeperlayers of the
oceanis not very well known. Therefore, it seemshelpful to assumethat
for a given time period, there is a certain relation between the surface
area of polar ice and the climatic factorsinvolved. By determining these
parameters from empirical data, we can use these relations as a supplementary equation in our model of the thermal regime, instead of the
expressionrelating the surface area of the ice to the elements of the
thermal regime under stationary conditions.
It follows from the semiempirical theory of the thermal regime that
the relative change in the ice surface is approximately proportionalto
the radiation change. This relation is valid for comparatively small
surface variations and can be presented in the following form:
Ap
AQ
=
,
P
Qp

(3.01)

where Ap/p is the relative changeof the polar ice surfacearea; AQp/Qis
the relative variation of the total planetary radiation, and p a dimensionless coefficient.

For current climatic changes the coefficient p can be determined

from empirical data on the variation of the total radiation and the ice
cover during the Arctic warming period. Figures 16 and 17 imply that
during this epoch the direct radiation at 50N exceeded that of the
previous period by about 2%.
Assuming that the surface of polar ice shrank during the same
epochby 10%, we find that p = 40. This is much less than the similar
coefficientwhich can be derived on the basisof the semiempirical theory
of the thermal regime for stationary conditions.
It must be pointed out that this coefficient depends on the time
periodto which it refers and that its precisionis not too high. Numerical
experiments have shown, however, that the error in the determination
of coefficientp doesnot strongly affect the calculated air temperature
distribution, which justifies the idea of using an approximate value of
this

3 CONTEMPORARY

CLIMATIC

CHANGES

85

By using Fig. 17 and the describednumerical model of the thermal

regime, where the relation (3.1) replaces the dependence of the ice
boundary on the air temperature, we can calculate temperature variations at different latitudes during a given period. These results are
presented in Fig. 19 by broken curves which appear to be quite closeto
the observed temperature variations smoothed over 10-year periods
(continuous curves).

It may be noticed that the calculated values somewhat precedethe

observed values. The lag of observed curves is apparently due to the
inertia of the ocean-polarice-atmospheresystem, though in this caseit
is not too great. In the thermal regime model applied here the derived
empirical parameters do not take into considerationtemperature variations, and therefore its results do not dependon the applied experimental data.

From an analysis of temperature variations at different latitudes, it

follows that in the northern hemisphere the main reason for temperature increase

in the 1920s and 1930s was the increase

of total

radiation

reaching the earth's surface. The temperature increases during the

warm half-years that were revealed in the calculations and determined
from observations, differed very little at different latitudes.
In the cold half-year temperature fluctuations at low and middle
AT

CLIMATIC

86

CHANGES

latitudes differ very little from thoseof the warm half-year. But at high
latitudes, mainly in the 70-80 zone, temperature fluctuations are
greatly augmented.
In this case temperature variations dependedvery little on the
fluctuations of solar radiation in the same season,becauseduring the
coldhalf-year the radiation at high latitudes is very low and scarcely
affectsthe atmosphericthermal regime. The primary causeof temperature variations in this casewas the changing area of the polar sea ice
that resulted in a considerablerise of air temperature during the cold
half-year.
This effect on the temperature in the warm half-year is comparatively small; it alsodiminishesrapidly in the coldperiodas the distance
from the 70-80N latitude zone grows.
This conclusionagreeswell with the earlier conceptionof the influenceof polar ice on the air temperature, accordingto which the presence
of ice causesa sharp decline in the air temperature at high latitudes
during the winter, exercisesa smallereffectin the samezoneduringthe
summer, and a still smaller influence on the air temperature in middle
and low latitudes (Budyko, 1971).
Role of Volcanic Eruptions. Savinov (1913), Kimball (1918), and
Kalitin (1920) as well as other authors have established that explosive
volcaniceruptionsresult in a drasticreductionof solar radiation reaching the earth's surface.
In such cases the average amount of direct radiation for vast
territoriesmay drop10-20%in the courseof severalmonthsor years.An
example is given in Fig. 20 which illustrates the average monthly
values of direct radiation under a clear sky, relative to normal value
(%)

11o[

IOO

90

IV

I,I

VI

VIII

1912

A'11 I1

1['

_1

'1 l'll/

.11

1913

Fig. 20. Arerage monthly ralues of direct radiation after rolcaniceruption

relatire

to normal

3 CONTEMPORARY

CLIMATIC

CHANGES

87

after the eruption of the Katmai volcanoin Alaska. This curve, basedon
observationscarried out at several actinometric stations in Europe and
America, shows that the atmospheric aerosol caused a more than 20%
drop in the direct radiation during specificmonths.
In certains areas the radiation drop was even greater. For instance,
in Pavlovsk, a suburb of St. Petersburg (Russia), situated at an enormous distance from Alaska, the solar radiation for half a year was 35%
below the norm. Similar effects were producedby the eruption of the
Krakatao volcano in Indonesia in 1883. In both cases, after the eruptions, anomalous optical effects in the atmosphere were observedover
great territory, thus confirming the global character of changes in
radiation regime resulting from the spread of the aerosolin the stratosphere.
The establishment, mainly in the 1950s,in the course of preparations for the International GeophysicalYear, of a world-wide network of
actinometric stations greatly facilitated studies of the effectsof volcanic
eruptions on the solar radiation regime. Since then, the first major
explosiveeruption occurredin March, 1963 (the Agung volcano on the
island of Bali, in Indonesia). The effectsof this eruption on the radiation
regime were studied with far greater thoroughness than was ever
possiblein the past.
Shortly after this eruption, the effect on incoming radiation was
detectedin various regions of the planet (Burdecki, 1964;Flowers and
Viebrock, 1965; Dyer and Hicks, 1965, 1968; Budyko and Pivovarova,
1967).

Budyko and Pivovarova reduced observations of direct noontime

radiation, carried out at several actinometric stations of the Soviet
Union during 1957 to 1966.
They established that the average monthly values obtained at 22
stations situated between 40 and 68N, registered very little change
from 1957to November, 1963. However, in December, 1963, these values
sharply declined (Table 10). A systematic change in these values since
Table

10

Deviation

of Direct

Radiation

from

Aver-

age over U.S.S.R. (in %)

II

Year
1963

-1

1964

-10

1965

1966

III

IV

-2

-3

-6

-7

-15

-7

-11

-5

-10

-9

-7

-4

VI

-2

VII
2

VIII
2

-1

-3

-4

-3

-3

-2

-1

IX
i

-2
-2

-3
0

An-

XI

XII

-1

-2

-10

-6

-8

-16

-5
-6

-5

-2

-9

-12

-2

-4

-7

-7

nual
0

88

CLIMATIC

CHANGES

the end of 1963 can be seen along with certain variations of direct
radiation from month to month causedapparently by the instability in
the atmospheric circulation.
There was a big drop in the intensity of direct radiation in the
winter as comparedwith the summer months, which means that it was
causedby changes in the atmospherictransparency and not by the
fluctuations

of the solar constant.

Similar calculationswere carried out for other parts of the globe.

The differencebetweenthe intensity of direct radiation at noonin 1964
and 1958 was determined

on the basis of observations

conducted

at

several stations outside of the Soviet Union while the transparencyof

the atmospherein 1958was assumedto be closeto the average value for
the end of the 1950sand the beginning of the 1960s.
These results are given in Table 11. They clearly show the same
attentuation of direct radiation in 1954 over Eastern Europe, North
America, and the Central Pacific.

The absence in this case of a noticeable

annual

variation

in the

intensity difference can be explained by the fact that all the stations
listed in Table 11 are located at comparatively low latitudes where the
average midday altitudes of the sun are not very different all year

round.

Observations of scattered radiation show that after 1963, the decreasein direct radiation over the U.S.S.R. was accompaniedby a sharp
increase of scattered radiation. This conclusionfollows, for example,
from Table 12 which presents the relative variations of the scattered
radiation observedat noon under clear sky conditions, at 22 Soviet
stations.

We find by comparing the absolute values of the decreaseddirect

radiation and the increasedscatteredradiation over the territory of the
Table

11

Deviation of Direct Radiation

that of 1958 (in %)
Station

Porto, Lisbon, Faro

(Portugal)
Blue
Hill,
Albuquerque
(N.
America)
Mauna
Loa
(Hawaii)

in 1964 from

III

iV

VI

VII

VIII

IX

XI

XII

AD-

nual

-7

-10

-6

-11

-9

-9

-10

-7

-12

-10

-12

-14

-10

-12

-5

-5

-3

-4

-4

-4

-4

-6

-5

-9

-4

-6

-7

-6

-6

-4

-6

-6

-6

-4

-4

3 CONTEMPORARY

CLIMATIC

Table

CHANGES

89

12

Deviation

of Scattered

Radiation

From

Av-

erage over U.S.S.R. (in %)

Year

II

III

IV

VI

1964
1965

10
20

7
14

12
24

22
17

17
22

6
25

VII

VIII

IX

XI

19
12

6
6

3
8

9
9

11
11

XII

Annual

25
12

14
14

Soviet Union, that the total radiation changed very little during these
years. Although it is difficult to determine the amount of this variation
from empirical data, it can be calculated by taking into account the
variations

of direct

radiation.

Dyer and Hicks (1968)studiedthe effect of the Agung eruption on

the direct

radiation

observed

at several

dozen actinometric

stations

situated in the northern and southern hemispheres.They found that the

aerosolparticles spread over the entire globe, from the south pole to
high latitudes of the northern hemisphere.This did not take more than
several months and their concentration remained high for about two
years; for the Soviet Union this period was longer. It is possiblethat
only comparatively small particles reachedthe middle latitudes of the
northern hemisphere and that it took them longer to settle. Dyer and
Hicks describethe propertiesof the aerosolcloudwhich formed after the
Agung eruption. The average altitude of this cloud was 15-20 km, the
mean radius of the aerosolparticles varied from 0.5-1.0/m during the
first month of its existence, to 0.1-0.15 pm one year later. Dyer and
Hicks assumedthat the cloud began to form at an altitude of 22-23 km
and that its particles later settled at a speedcorrespondingto their sizes.
Considering the great influence of volcanic eruptions on the radiation regime, we can assumethat the radiational increase at the end of
the nineteenth century was the result of atmospheric clearing of the
aerosolcausedby the Krakatau eruption. The succeedingdrop in radiation was the consequenceof the eruption of the Mont Pele and other
volcanoes(anomalies began to decreaseon the curves of secular variations of radiation in Figs. 16 and 17 before the time of these eruptions,
but this is due to the 10-year smoothingeffects).The increasein radiation during 1915-1920was apparently due to the transparency increase
caused by the settling of aerosol after Katmai eruption, after which
there were no intensive eruptions for a long time.
It is interesting to analyze the radiational drop that began in the
1940s.A number of scientistshave expressedthe opinion that this was
the result of industrial pollution (Davitaya, 1965) as well as nuclear
tests. It couldalso be due to eruptionsfrom the Spur volcano (Alaska)

9O

CLIMATIC

CHANGES

1953, the Bezimyannaya Sopka (Kamchatka) in 1956, and others. This

problem is dealt with in greater detail in Chapter 6.
Several empirical works show that after major eruptions the average temperature at the earth's surfacewas lowered by several tenths of
a degree during periods varying from several months to several years
(Humphreys, 1929; Mitchell, 1961, 1963).
Let us considerthe effectof these radiation changeson the temperature. We can make a theoretical estimate by taking into account the
heat conductivity and the specificheat of ocean water. However, since
we are interested in only an approximate estimate of this effect,we shall
use a simple relation between radiation and temperature variations
that can be established empirically from the annual variations of the
parameters.

We shall assumethat the rate of the temperature variations at the

earth's surface is proportional to the difference between the present
temperature (T) and the long-term average temperature (Tr), i.e.,
dT
dt

k (T - T),

(3.2)

where k is the proportionality factor.

If the temperature at some initial time is designatedby T,, then
from (3.2) we obtain

T - T = (T, - T) e-t.

(3.3)

For an approximateestimate of let us use the data on the annual

variations of solar radiation and temperature in the northern hemisphere neglecting in these calculations the interaction between the
climatic factors in the northern and southern hemispheres.
Assuming that the ratio of the solar radiation at the outer limits of
the atmosphere in the northern hemisphere during the warm half-year
(April-September), to the mean annual radiation is equal to 1.29, we
find from the above formula, relating the temperature with the radiation changes,that this changein radiation could lead to a temperature
increase of approximately 40 in the case of zero heat inertia. The
observeddifference between the average temperature in the northern
hemispherefor the warm half-year and that for the entire year is equal
to 3.5C.Assuming here that T - T = 3.5Cand Tr - T = 40Cfor t =
year, we find from (3.3) that k = 0.4 yr -.
With this in mind, let us find the change in average temperature
one year after a volcanic eruption that led to a 10% drop in direction
radiation. In this casethe drop in the total radiation would be 1.5% and
the decreasein T about 2C. From (3.3) we find that the

3 CONTEMPORARY

CLIMATIC

CHANGES

91

variation after the eruption is equal to several tenths of a degree. This

result agrees well with the annual average temperature anomalies
registered after large eruptions of an explosivenature.
A more detailed calculation of temperature changesafter volcanic
eruptions was carried out by Borzenkova (1974) who also applied the
semiempiricalmodel of the atmosphericthermal regime and the given
formula for finding the average temperature variations at the earth's
surface.Borzenkovafound that after a volcaniceruption the temperature drop in the northern hemisphere increaseswith latitude both in the
warm and cold half-years.
Oliver (1975) and Sagan with co-workers (Pollack et al., 1975)
studied the volcanic eruption effect on climate. The results of their
calculationare in goodagreementwith the empirical data on the mean
temperature variations. These studiescorroboratethe assumptionthat
contemporary climatic changes depend considerably on the level of
volcanic activity.
Detailed empirical studiesof the effectof volcaniceruptionson the
air temperature, carried out by Pokrovskaya (1971) and Spirina (1971),
have revealed regularities in the effect of a single eruption on the
thermal regime. It has been establishedby these works that for several
years after powerful eruptions, there is a reduction in air temperature

during the warm part of the year, this reductionbeing the greatestin
the northern part of the middle latitudes. During the cold seasons
temperature variations after eruptions are more complicated:the temperature usually drops in the polar zone and often increases at middle

latitudes.As a result, the averageannual temperaturetakes a bigger

drop in high latitudes as comparedwith middle latitudes.

The abovementionedregularities, accordingto Pokrovskaya, are
the result of the predominant effect of radiation factors on climatic
conditionsduring the warm seasons,when the increasein the optical
depthsof the aerosollayer enhancesthe temperature drop in the polar
regions. During the cold season,the main factorsof temperature distribution are the fluctuations in atmosphericcirculation causedby variations of the heat inflow to various latitudes.
Tables 10 and 11 show that at middle latitudes

the relative

effect of

aerosolparticles on radiation is much smaller during the warm than

duringthe coldseason,the reasonfor this probablybeingthe noonsolar

altitude.

Since the radiation in middle latitudes during the cold seasonis

much lower than during the warm season,the annual variations of the
absoluteradiation are significantly less than the relative values given
in Tables 10 and 11; nevertheless,a comparatively large decrease

92

CLIMATIC

CHANGES

radiation during the cold period cannot but affect other componentsof
the heat balance of the earth-atmosphere system. It is possiblethat it
influencesmainly the heat exchangein the upper layers of the ocean,
whosegreat thermal inertia causesvariations of the summer temperatures discoveredby Pokrovskaya and Spirina.
These scientistshave shown that the variations in the atmospheric
thermal regime causedby a single volcaniceruption are not stationary
and that the temperature drop in this caseis only a part of the changes
that would have taken place under stationary conditions.
The effect of volcanic activity on temperature changesoccurring
over longer intervals lasting several decades was studied by Lamb
(1970),who comparedtemperature anomaliesfor various time intervals
with the index representingthe averagedrop in atmospherictransparency after eruptions.
Lamb has expressedthis index in relative units which describethe
effect of volcaniceruptionson the atmospherictransparencyin terms of
the effect of the Krakatau eruption of 1884.
A highly negative correlation coefficient of -0.94 was obtained
between the temperature anomalies in the northern hemisphere(from
the equator to 60N)for the period of 1870-1959,averagedover 10-year
intervals, and the describedindex, which proves the closecorrelation
between the volcanic activity and temperature fluctuations. Lamb
pointedout that by the end of this periodthe relation was disturbed.The
reasonsfor this will be analyzed in Chapter 6.
For certain regionsthis correlation appearsto be lessclose,and this
seemsonly natural consideringthe great influenceof atmosphericcirculation changeabilityon temperatureanomaliesin limited regions.From
the point of view of our theory on the mechanism of modern climatic
changes,the relation discoveredby Lamb betweenthe amount of seaice
off the coastof Iceland and his index of atmosphericopacity,is of great
interest. The correlation coefficientbetween these parametersfor 10year intervals, from 1780to 1959, was found to be 0.61, which is rather
high in view of the fact that Lamb was comparingvery approximate
estimates, especiallyfor the first part of the consideredperiod.
Aerosol Layer in the Stratosphere. In order to study the mechanism by which volcaniceruptions influence the climate, it is necessary
to make a more thorough study of the dependencebetween climatic
conditionsand atmosphericaerosol.The atmosphereincludesnot only
water droplets and ice particles of clouds and fog, but also a large
amount of solid and liquid suspendedparticles of varying chemical
composition.The size of these particles ranges from 10-2 to 10-7 cm.
The greater portion of the total aerosolmassis comprisedof the

3 CONTEMPORARY

CLIMATIC

CHANGES

93

called large particles with radiuses ranging from 10-5 to 10-4 cm and
gigantic particles measuring over 10-4 cm.
Available

estimates

show that

from

800 to 2200 million

tons of

matter enter the atmosphere annually, as a result of natural processes,

from which aerosolparticles are formed. Another 200-400 million tons of
matter is added as a result of human activity. Among the aerosol
particles are productsof rock and soil weathering, sea salt from the
oceansurface, soot and ashesfrom forest fires and combustiongases,as
well as products of chemical reactions of sulphurous gas, hydrogen
sulphide, ammonia and other gasesentering the atmospherefrom the
earth's surface (Inadvertent Climate Modification, 1971).

Atmospheric aerosolis concentratedmainly in the lower layers of

the trophosphere and its residence time is, as a rule, comparatively
short. The largest particles settle down rapidly under the force of
gravity, while the smaller particles are carried to the earth's surfaceby
the descendingair currents and by precipitation playing an important
part in cleansing the air of aerosol.
According to available data, the average residence time of an
aerosol particle in the troposphereis about 10 days. Therefore we find
that the atmosphere contains approximately 30-70 million tons of aerosol.

In the stratosphere the amount of aerosol is considerably less. It

remains there much longer, however, than in the troposphere.
The main mass of stratosphericparticles ranges in size from I to 0.1
pm (large particles) whoserate of settling under the force of gravity is
comparatively small. Since the vertical air motions in the stratosphere
are very poorly developedand there is no precipitation, the residence
time of aerosol in the stratosphere ranges from several months to
several years.
Direct observationsof the compositionof the stratosphere aerosol
initiated by Junge (1963) have proved that this aerosol is made up of
droplets of sulphuric acid and also (to a lesser degree) of sulphuric acid
salts, basically compoundedwith ammonia. Most of the stratospheric
aerosolis generally concentratedin a layer several kilometers thick, its
middle level being at an altitude of 18-20 km. This is known as the
sulphate layer, or Junge's layer.
Accordingto Junge, the aerosolparticles of the sulphate layer are
productsof sulphurous gas that enters the stratosphere from lower
levels. After photochemicalreactionswith atomic oxygen in the stratosphere, the sulphurous gas forms sulphur dioxide, from which, as a
result of interaction with water vapor, droplets of sulphuric acid are

CLIMATIC

94

CHANGES

Recent investigations show that if the sourceof stratospheric aerosol is outside the tropical latitudes, the aerosol spreadscomparatively
fast over the pertinent hemisphere but very slowly over the other
hemisphere. If the source of aerosol is close to the equator, it spreads
over both hemispheres (Karol', 1972).
The particles of stratospheric aerosol settle gradually under the
influence of gravitation force and as a result of large-scale air motions

whichcarrythemto the troposphere,

wherethey arerapidlywashedout

by precipitation. The latter mechanism is apparently the main one for

large particles, while the former is more effective for gigantic particles,
the amount of which in the stratosphere is low due to their rapid
descent.

Karol' (1973) assertsthat large particles remain in the stratosphere

at an altitude of 20-30 km on an average for 20-40 months, while at the
tropopausethis period is reduced to 6-20 months. These periods vary
greatly with the intensity of air exchangebetween the stratosphereand
the troposphere.
In studying the climatic effects of atmospheric aerosol, it is very
important to establish the relation between the vertical radiation flows
in the atmosphereand the aerosolconcentration.This problem has been
studied by many researchersbeginning with Humphreys.
Humphreys' contention that the atmospheric aerosolgenerally has
comparatively little effect on long-wave radiation has been confirmed by
later research.

At the same time it was established that atmospheric aerosol can

noticeably affect the flow of short-wave radiation due to the backscattering and absorptionof radiation by aerosolparticles.
The scattering index for particles within the size range of 10--10
m is strongly prolate in the direction of the incident beam and as a

result the direct radiation

is more attenuated

than the total radiation.

Due to a shortage of experimental data, the absorption of shortwave radiation by aerosolparticles is lessclear to sciencethan backscattering. The available data showthat as a result of absorptionby aerosol
particles, the depletion of direct radiation is apparently less than that
causedby backscattering,although both effectsare comparablein magnitude (Gaevskaya, 1972).
A number of estimates couldbe listed of an influence of atmospheric
aerosol mass upon attenuation of short-wave radiation flux. The simplest way to handle this problem is to comparethe average climatological value

of short-wave

radiation

flux attenuation

due to aerosol with

the average mass of atmosphericaerosol.

Let us assume, in agreement with Pivovarova (1968) and

3 CONTEMPORARY

CLIMATIC

95

CHANGES

authors, that this decrease in direct radiation due to aerosol comes to

about 10%. From the above data it follows that the average mass of
atmosphericaerosolis 50.106 t, or 10-5 g/cm 2. About half of this amount
is made up of gigantic particles; their number density is not high and
therefore their influence on the radiation processesis small. Thus a
mass of aerosol particles equal to 0.5.10 -6 g/cm2 reduces the direct
radiation by 1%.
If this attenuation was due only to backscattering of radiation by
aerosol particles, the correspondingdrop of the total radiation under
average conditions would have been about 0.15%. If this attenuation
was due entirely to absorption by aerosol particles, it would have been
equal to 1%. Let us assume that both effects have comparable influence
on the direct radiation

attenuation.

In this case the amount

of aerosol

causing a 1% drop in the total radiation is equal to about 10-6 g/cm .

This estimate is of coursevery approximate.
In several works the effects of the aerosol on the attenuation

of total

radiation was attributed to backscattering.In the first work of this kind

Humphreys (1929) found that the radiation dropped by 1% when the
aerosol mass M was equal to 0.6.10 -6 g/cm . Similar calculations were
recently carried out at the Main Geophysical Observatory by Novosel'tsev and Vinnikov, who respectively followed the schemeof multiple
scattering suggestedby Soboliev, and the schemeof a single scattering
developedby Shifrin. The values of the mass M obtained by them are
0.4.10 -6 and 0.6.10 -6 g/cm , respectively. A somewhat larger value of M
= 10-6 g/cm was obtained by Barrett (1971).
The value of M can also be found by calculating the backscattering
accordingto Yamamoto and Tanaka's model (Inadvertent Climate Modification, 1971). These data contend that M = 1.3.10 -6 g/cm .
The relation

between the aerosol mass and the short-wave

radiation

can alsobe estimatedby .ngstrSm's

formula(1962)whichshowsthe

connectionbetween the albedo of the earth-atmosphere system and the

coefficientof optical opacity for an average mass of aerosol;here M =
0.6.10 -6 g/cm .
Hence these and similar calculationsimply that M is on the average
between 0.4.10 -6 and 1.3.10 -6 g/cm .
Since the value of M is much less than aerosol mass variations

we

cometo the conclusionthat fluctuations of atmosphericaerosolin space

and in time can noticeably affect the flow of short-wave radiation
reaching the surface of the earth. These fluctuations apparently have a
strong effect on the thermal atmosphericregime.
It is instructive to study the connectionbetween the fluctuations of
the atmospheric aerosol concentration and the volcanic activity.

CLIMATIC

96

CHANGES

cumstantial evidence here is provided by the established decreaseof

direct solar radiation after major volcanic eruptions as a result of the
considerable

increase

of the aerosol concentration.

There is also direct evidence. Measurements carried out by highaltitude

aircraft

have shown that the aerosol concentration

in the at-

mosphereover the northern hemisphereincreasedgreatly by the end, as

comparedto the beginning, of the 1960s.This variation was apparently
due to the eruption of Mt. Agung in 1963 and subsequenteruptions of
other volcanoes (Inadvertent Climate Modification, 1971).
Causes of Climatic Changes. We may now draw a conclusionin
the light of the data given in this chapter about the causesof climatic
changesthat took place in the first half of our century. The warming
which reached its peak in the 1930s apparently resulted from an increasein the transparency of the stratospherewhich led to an increase
in the solar radiation flow to the troposphere(the meteorologicalsolar
constant).This, in turn, led to an increasein the average globaltemperature

at the earth's

surface.

The air temperature variations of different latitudes and seasons

dependedon the optical depth of the stratosphere'saerosol and on the
shift of the polar sea ice boundary. The subsequentrecessionof the
Arctic ice has led to an additional and noticeableincreasein air temperature during the cold seasonin high latitudes of the northern hemisphere.
These conclusionsare confirmed by calculationsbasedon a model of
the atmosphericthermal regime, whoseresults agree well with observational

data.

It seemspossiblethat the changesin the stratosphere'stransparency that took place during the first half of the twentieth century were
related to the volcanic activity, when the products of the eruptions,
including sulphurousgas, were ejectedinto the stratosphere.This conclusion, although supported by abundant observational material, is
neverthelessnot as obvious,in our opinion, as the previous one.
It should be noted that the given explanation concernsonly the
main pattern of the climatic fluctuations that took place in the first half
of the twentieth century. Parallel with the above-describedgeneral
regularities, this processexhibited many specific features of climatic
variations during shorter time intervals and in certain geographical
regions.
It seemsto us that these fluctuations were causedto a large extent
by changesin the circulation of the atmosphere and hydrospherethat
were sometimes of a random nature and other times, the result of auto-

3 CONTEMPORARY

CLIMATIC

CHANGES

97

There are groundsfor believing that the climatic variations of the

last 20-30 years began to dependto someextent on the activity of man.
This problem is discussedin Chapter 6.
Although the warming in the first half of the twentieth century had
a certain effect on human economicactivity and provedto be the largest
climatic fluctuation that took place during the epoch of instrumental
observations, its scale was negligible as compared to the climatic
changesthat occurredduring the Holocene,to say nothing of the Pleistocene, when vast glaciations evolved.
This warming, however, is of great importance for revealing the
mechanism of climatic changes, since it is the only one that has been
comprehensivelyand exhaustively studiedby means of reliable instruments at many research centers.
Therefore any quantitative theory of climatic fluctuations must be
backedby data referring to the warming of the first half of the twentieth

4 Climates

of the Past

4.1 The Quaternary Period

Climate of the Quaternary Period. In Chapter I it was pointed

out that the Quaternary-the last geologicalperiod-was marked by a
great variability in climatic conditions,especiallyat middle and high
latitudes.

Natural conditionsduring this period have been studiedmuch more

extensivelythan any earlier ones.Particularly important contributions
to the study of these conditionswere made by the Soviet scientists
Gerasimov and Markov (1939), Markov (1955, 1960), Saks (1953), Rukhin (1962),Velichko (1973),and others. Valuable data on the climate of
the Quaternary can be found in the works of Butzer (1964), Emiliani
(1955),Erikson (1968),Ewing and Donn (1956),Fairbridge (1967),Flint
(1957), Flohn (1963, 1964), Mitchell (1965a), Schwarzbach (1950, 1961,
1968), Zeuner (1959), and others.
Though numerous outstandingachievementshave been made in
the study of the Pleistocene,many important natural processes
of that
time are still insufficiently clear; among these is the dating of cooling
epochsassociatedwith the spreadingof ice coversover land and sea.
Thus the questionof the total duration of the Pleistoceneremains open.

Of signalimportancefor workingout the absolutechronology

of the
Quaternary period are the methodsof isotopeanalysis, which include
the radiocarbonand potassium-argonmethods.The first of these gives
us more or less reliable results only for the last 40-50 thousand years,
i.e. for the final phaseof the Quaternary period. The latter methodis
applicableto periodsof longerduration but its precisionis much lower
than that of the former method.

In view of the major discrepancies

in the opinionsof variousauthors
on many important climatic features of that time, we shall limit ourselvesto a generaldescriptionofthe Pleistoceneclimate, supplementing
the brief information already given on it in Chapter 1.
The Pleistocenewas precededby a long coolingprocesswhich was
mostpronouncedin the middleand high latitudes.This processacceler-

CLIMATES

OF

THE

PAST

99

ated during the Pliocene, the last epochof the Tertiary period, when
large ice covers originated in the polar zones of the northern and
southern hemispheres.
Paleogeographicdata reveal that the glaciation of the Antarctic and
the Arctic lasted over several million years. At the beginning the ice
covers were comparatively small but they spread gradually to lower
latitudes and later receded,sothat ice agesalternated with interglacial
periods.
It is difficult for many reasonsto determine the point at which the
repeated migration of the ice cover boundaries commenced,but it is
generally believed to have been about 700 thousand years ago.
Moreover, researchers have often claimed that this epoch of the
active development of widespread glaciation was preceded by a still
longer transitional period, the Eopleistocene,which would add more
than a million years to the duration of the Pleistocene.
The total number of glaciations must have been very large because
the main ice ages, already recognizedby scientists in the last century,
seemed to have consisted of a series of alternating warm and cool
subperiods,the latter of which may be consideredice ages in their own
right.
As mentioned in Chapter 1, the total duration of all the ice ages
constitutedless than half of the Pleistoceneepoch,i.e. they were shorter
than the comparatively warm interglacial periods.
The scaleof glaciation differed greatly in the various ice ages. Some
authors are of the opinion that by the end of the Pleistocene the
glaciations were more intensive than the earlier ones had been.
The last one-The Wfirm, or Valdai, glaciation-that took place
several tens of thousands of years ago has been studied most intensively. During that epoch,along with the developmentof continental ice
covers,the climate becamevery dry over vast areas in middle latitudes.
This was probably due to the sharp drop in evaporation from the oceans
as the sea ice expanded to lower latitudes. As a result, the rate of the
hydrological cycle was reduced, leading to a decreasein precipitation
over the continents, which in turn limited the further growth of continental glaciation. In addition, the enlargement of the terrain due to the
lowering of the ocean level could also have influenced the precipitation
regime, as a great quantity of water was trapped in the ice cover. This
processalso enhanced the continental nature of the climate and led to a
reduction in the amount of precipitation over land. There can be no
doubt that the permafrost area increasedenormouslyduring the last ice
age.

Although climatic conditionsin the ice-coveredzone at that time

were in many respectssimilar to those of today, it is quite possible

100

CLIMATIC

CHANGES

the climate in the ice-freezonesof the middlelatitudeswas then quite

different from all currently existing types of climate.

This climate wasmarkedby a combinationof comparativelylow air

temperature all year round, relatively high influx of solar radiation,
and a small amount of precipitation over a considerablepart of this
zone. These climates were conduciveto the formation of specificlandscapesresembling both the tundra and steppeof our times.
During the Wfirm glaciation, the tropical climate changedlittle in
comparisonwith the climate of middle latitudes, although there was a
certain drop in the average air temperature and a redistribution of the
amount of precipitation.

The last glaciation ended 10-15 thousand years ago and this is
generallyconsideredthe end of the Pleistoceneand the beginningof the
Holocene,the epochin which natural conditionsbeganto be affectedby

the activity of man.

Astronomical

Causes of Climatic

Variations.

It is difficult

to

point out the causesof past glaciationsbecausepresumablyno great

changesin climatic factorsoccurredduring the Quaternary period.Only
two factors can be definitely identified.
The first is the variation in the amount of radiation reachingthe
earth's surface at various latitudes throughout the year, due to the

changingpositionof the earth relative to the sun. This positiondepends

on the eccentricityof the earth's orbit, the obliquity of the eclipticplane
and the precessionof the equinoxes.All these orbital elementschange

periodically and cause variations in the amount of radiation that can be

calculatedquite preciselyfor the pasttensof thousandsof years,though

with mounting uncertainty for more remote times.
The secondfactor is the variation in atmospherictransparencydue
to the amount of aerosolproducedmainly by volcanic activity. These
fluctuations somewhat affect the amount of radiation reaching the
earth's surface, which can be estimatedon the basis of contemporary
actinometricobservations.Someidea of the atmospherictransparency
in the past is providedby geologicaldata on volcanicactivity during
various epochs.
It should be noted that

the effect of these two factors on the

glaciation regime was not clear until recent times. Since these factors
exert only a limited effecton the radiation regime of the earth, many
authors have expresseddoubt as to their possible influence on the
development of glaciations.
In order to clarify this problem, it is necessaryto apply a numerical

modelthat allowsus to calculatethe positionof the ice sheetboundary

as a function
astronomical

of external

climatic

conditions. Let us first consider the

CLIMATES

OF

THE

PAST

101

For this purpose we have applied the numerical model of the

thermal regime for average annual conditions described in Chapter 2
(Budyko, 1968). This model implies that the variations of the radiation
regime during the last Wiirm glaciation couldhave led to a 1southward
spread of the ice cover in the northern hemisphere, but this is much
below the actual

value.

In analyzing this result, we have come to the conclusionthat the

determination of average annual temperatures is not sufficient for
estimating the effect of orbital changeson glaciation, becauseit depends
mainly on the thermal conditionsduring the warm half-year.
In a later study Budyko and Vasishcheva (1971) used the second
model describedin Chapter 2, which gives the distribution of the average latitudinal temperature for various seasons.
On the basis of this model the approximate position of the average
ice-coverboundary was determined by the method of successiveapproximations for the periodswhen the incoming radiation in high latitudes
during the warm half-year was reduced due to astronomical factors.
Data on the radiation regime for suchperiodswere taken from the work
of Milankovich

(1941).

In this calculation the dependenceof the planetary albedo on the

polar ice boundary was taken into accountby the equation

Aas,--

Qs ALi
Qi ALis
si - a v
sif)+
(a v
(a sSi
- a i),
QspLo
QspLo

(4.1)

where Aas is the variation of the planetary albedo as comparedto its

present value; AL is the variation of the zonal area coveredby ice in one
of the hemispheres,as comparedto the present regime; Lo is the earth's
surface;asis the albedoof the polar ice zone;as is the albedoof the icefree zone which freezes as a result of radiation decrease;Qsi is the
radiation at the outer boundary of the atmospherein the ice-freezone,
whose surface is equal to L; Qs is the average planetary value of
radiation. The superscripts "N" and "S" indicate the northern and
southern hemispheres.
In this calculation the fluctuations
and cloudiness were not considered.

in the ocean radiation

balance

The results of these calculations are given in Table 13. This table
showsthat the fluctuations causedby astronomicalfactors can lead to
significant climatic variations. Calculations show, however, that although the average planetary temperature changes vary slightly, the
shift of the ice-coverboundary is substantial.
According to these calculations which give the current position of
the average ice boundary as closeto 72Nand 63S,the maximum

102

CLIMATIC

CHANGES

Table

13

Climatic Changes During Glaciation Periods

Time (103yr before 1800)

22.1
71.9
116.1
187.5
232.4

(Wiirm III)
(Wiirm II)
(Wiirm I)
(Riss II)
(Riss I)

Atbs

Atbs

8
10
11
11
12

5
3
2
0
-4

AT
-5.2
-5.9
-6.5
-6.4
-7.1

Note: Av and Atbs

, decreasein mean latitude of the polar ice boundaryin
the northern and southern hemispheres, respectively, as compared to their
present value; AT, variation of average temperature of warm half-year at
latitude

65N.

of the ice boundary during the studied period was equal to 12 in the
northern and 5 in the southern hemispheres.
The zone into which the ice cover spreadsexperiencesa considerable drop in temperature. Thus, at 65Nthe average temperature of the
warm half-year decreasesby 5-7C.These figures apply to temperatures
at sealevel; it is possiblethat during the formation of massivecontinental glaciations, the temperature drop at the ice level can be much
greater.

It is interesting to comparethe results of Table 13 with paleogeographic data on natural conditionsduring the ice ages. Such a comparison, however, meets with certain difficulties resulting from the schematic nature of our calculations and our insufficient empirical knowledge of past climatic conditions.
One of the numerousassumptionsmade in the courseof the calculations is that the positionsof the ice coverand the thermal regime were
stationary during the periodsunder consideration.It is known, however, that the developmentof ice coverswas actually a gradual process
which lagged behind the maximum radiation decreasein high latitudes.
Neglecting this effectmay causea certain error in the calculationsof the
surface occupiedby the ice sheet.
A comparisonof these results with empirical data is also difficult
becauseof the absenceof exact dating of the Quaternary glaciations.
The correlation between the data given in the table and the two main
epochsof the last two glaciations correspondsto Milankovich's views,
which, however, are not sharedby all scientistsengagedin the studyof
Quaternary glaciations. Another difficulty of such a comparisonis the
absenceof reliable data on the averagelatitudinal boundariesof the ice
coverduring variousice ages.It is thereforepossibleto compareonly the
main results of our calculations with the empirical data on the

CLIMATES

OF

THE

PAST

103

conditionsduring the ice ages, and this particularly applies to the ice
cover.

The maximum calculated migration of the average ice boundary in

the northern hemisphere accordswell with the empirical data. For
example, Lamb (1964)found that during the stage of greatest glaciation
the average ice boundary in the northern hemisphereexpandedas far as
57N,which correspondsto a 15shift of this boundary as comparedto its
present position. In our calculations this shift is 12, which is close
enough to Lamb's result.
It is difficult to carry out this comparisonfor all ice agesbut it may
be done for those regions of the earth for which data are available. For
example, Zeuner (1959) gives information about the expansion of glaciers during various epochs in Central and Northern Europe. This
spread, expressedby Zeuner in percentagesof the distancetravelled by
the glaciers during the Mindel II epoch,are comparedin Fig. 21 to the
calculated variations of the boundary latitudes of the north polar ice
during the Riss I, Riss II, Wiirm I, Wiirm II, and Wiirm III epochs.Since
Zeuner gives a range of values for each of these epochs,the results are
shown on the diagram in the form of line segments.As we can seefrom
the diagram, there is a certain correlation between the quantities,
which proves that it is possibleto estimate the relative parameters of
different glaciations by a quantitative method.
The conclusion that the development of Quaternary glaciations
could be caused by astronomical factors does not exclude the effect of
volcanic activity on glaciations. According to contemporary observations, fluctuations in the transparency of the stratosphere due to the
changing amount of aerosolparticles causevariations in the total radia-

10

90

100

110(%)

Fig. 21. Expansion of glaciers during various

CLIMATIC

104

CHANGES

tion, averaged over 10-yearperiods,which amount to several tenths of a

percent.

Calculations carried out accordingto this schemeshow that if such

variations in the radiation had lasted for several thousand years, they
could have caused a migration of the ice boundary of up to several
hundred kilometers. Although it is known from geology that volcanic
activity in the past varied enormously, it is at present difficult to
estimatethe long-termvariations in the amountof atmosphericaerosol.
This limits any opportunity to make a more detailed study of the
relation between volcanism and the Quaternary glaciations.
We note that this agreementbetweenthe calculateddistributionof
the polar ice and the paleogeographicdata could be reached only by
taking into accountthe feedbackbetween the polar ice and the atmospheric thermal regime. An interesting example of a study which ignored
this feedback is given in the work of Saltzman and Vernekar (1971).
They used a numerical model of the average latitudinal distribution of
temperature, winds, evaporation and precipitation, based on the integration of the equationsof atmosphericdynamics. By this model they
calculated the variations of the temperature distribution at the surface
in the northern hemisphere that took place 10 and 25 thousand years
ago, as compared with that of today. All the climatic factors were
assumedto be constant, with the exception of the radiation at the outer
boundary of the atmosphere,which differed from the present value due
to astronomical

factors.

Saltzman and Vernekar came to the conclusionthat during those

periods,the largest temperature differenceat certain latitudes for the
warm half-year did not exceed 1.0. They concludedthat such minor
variations were inadequate for the development of glaciations.
Since Saltzman

and Vernekar

did not take into account the feed-

back between the ice cover and the temperature field, we can expect
their temperature values to be underestimates. It is interesting to
comparetheir results to thoseof a similar calculation doneby meansof
our model without feedback, i.e. considering the albedo of the earthatmospheresystemconstantas the radiation varies. The results of this
calculation are presented in Fig. 22, which gives the distribution of
average latitudinal temperature differencesfor the warm half-year 10
and 25 thousandsof years ago, comparedto modernconditions.
It follows from the figure that the obtained results correlate quite
well, and this is rather remarkable since the Saltzman-Vernekar model
differs greatly from our more schematicmodel. The satisfactoryagreement

of these

calculations

based on different

models

testifies

reliability of these modelsand to the possibilityof making many

to the

CLIMATES

OF

THE

PAST

105

AT

1- (a)
.

30
60
90 0
30
60
90
Fig. 22. Distribution of average latitudinal temperature differences for
the warm half-year caused by astronomical factors. (a) 10 thousand years ago;
(b) 25 thousand years ago. Solid line, calculations by means of the model
describedin Chapter 2; broken line, calculations by Saltzman and Vernekar.

fying assumptionsin developingthe semiempiricaltheoriesof the thermal regime.

One of the reasonsfor the agreement between the results of the
thermal regime variations calculated from the above semiempirical
model and those obtained using more general climatic theories presumably is the fact that the basic empirical hypothesis used in the indicated
model, and expressed as (2.5), is fulfilled not only for mean annual
conditionsbut also for different seasons,and is confirmed by observa.
tion.

It is known that the average temperature differencebetween the

pole and the equator changesin the courseof the year by several tens of
degrees;this correspondsto extremely great variations in the incoming
energy, producing large-scale atmospheric motions. At the same time
the chief patterns of the summer and winter global atmospheric circulation in eachhemisphereremain virtually unchanged,making it possible
to describe the meridional heat transfer during various seasonsby
similar empirical relations.
Of interest from this standpoint is the work of Williams, Barry and
Washington (1973), in which a numerical model of the atmospheric
circulation during the last Wfirm glaciation was constructed.
Flohn and other authors have voicedthe opinion that during the ice
agessummer climatic conditionsin middle and high latitudes resembled
modernwinter conditions.Williams, Barry and Washingtonarrived at
a similar conclusionby physical deduction.
It follows from their work that during the ice ages the summer
westward transfer in the middle latitudes of the northern hemisphere
was approximately the same as during contemporary winter conditions.
Under the influence of glaciation, the stable pressure systems were
shifted somewhat and there was someredistribution of

CLIMATIC

106

CHANGES

There were no drastic changes,however, in the atmosphericcirculation

during that period. Similar results were obtained by Alyea (1972).
Thus it is apparent that the simplified numerical models that
satisfactorily describe modern climatic conditions for various seasons,
may be applied to the Quaternary period as well.
Lessclear, however, is the questionof whether it is possibleto apply
the thermal regime models for average annual conditions in order to
discern the mechanismby which Quaternary glaciations developed.
Fluctuations

of astronomical

factors do not affect the total annual

radiation as heavily as the total radiation for the warm and cold halfyears separately. Since the ice cover regime is controlled mainly by the
thermal conditions during the warm season, it is obvious that the
application of average annual data can underestimate the effect of
astronomicalfactorson glaciation. This is confirmedby our calculations
and by Sellers (1969, 1970). Sellers used his model of the atmospheric
thermal regime for studying the relation between astronomical factors
and the Quaternary glaciations, and came to the conclusionthat the
dependencewas negligible. On the other hand, Berger (1973) applied
Sellers' model for average annual conditionsand obtained the ice boundary migration during the Quaternary period, a result quite consistent
with paleogeographicdata.
Calculations based on the semiempirical model describedin Chapter 2 involving the feedback between the ice cover and the thermal
regime have shown that in this case the influence of the radiational
factors on the air temperature in high latitudes is tremendously enhanced, making it possible for the glaciations to expand over large
areas. It may be assumedthat if this feedback is taken into consideration, a similar result can be reached with the Saltzman-Vernekar model
as well as other available modelsof the atmosphericthermal regime.
It follows from the foregoing that astronomical factors began to
significantly influence climatic conditionsonly after the appearanceof
large polar glaciations, since it is the latter that made the Quaternary
climatic

conditions

so sensitive

to the minute

fluctuations

of climatic

factors. During earlier epochs,before the formation of the polar ice

sheets, the astronomicalfactors varied in a similar manner as during
the Quaternary, but their effect on the climate was much smaller. The
extent of this influence seems to correspondto that calculated on the
basis of our model and the Saltzman-Vernekar model, as illustrated by
Fig. 22.
This figure implies that fluctuations of astronomical factors in the
absence of ice covers change the air temperature by at most

CLIMATES

OF

THE

PAST

107

Therefore, during the Pre-Quaternary times, astronomical factors could

not have affected the climate to any great extent.
Thus, in order to explain the origin of Quaternary ice ages, it is
important to find the mechanism of temperature decreasein the polar
zonesthat took place during the Pliocene.
The Role of Atmospheric Carbon Dioxide. This problem is analyzed in Section 4.2, where it is established that fluctuations of carbon
dioxide concentration in the atmosphere did not affect the thermal
regime while it remained above 0.10%. Under those conditions the air
temperature was about 5Chigher than at present.
When its concentrationdroppedbelow this level, the air temperature began to decrease,slowly at first, and then more rapidly.
Taking into account the estimate given below of change in CO
concentration during the Cenozoicera it appears that the lowering of
temperature caused by a decreasein CO concentration began in the
Oligoceneand acceleratedduring the Pliocene.
It follows from the formulas of the semiempirical theory of the

thermal regime that the first polar glaciationmust have developedin

the southern hemisphere, where the major part of the polar zone was
occupiedby the Antarctic continent.
This processwas accelerated by the influence of mountains with
even lower surfacetemperaturesand by the absenceof oceancurrents
that supply an influx of heat in the north polar zone. According to the
semiempiricaltheory, the contemporarysize of this glacier providesfor

a 2Cdrop in the global air temperature at the earth's surface.It is

therefore possiblethat its development facilitated the origin of the ice

cover in the northern hemisphere.
Thus, the development of the polar glaciations could have been
inducedby the drop in the carbon dioxide-concentration,assumingthat
the continentswere situated favorably for the temperature decreaseat
the poles.
It is interesting to ascertain whether fluctuations in the carbon
dioxide concentrationduring the Quaternary period were adequate for
the expansion and recessionof the ice cover. The next section will show
that the boundariesof the polar glaciationcanmigrate by 5-20latitude,
when the concentration changes by 0 005-0.010% i.e. by i of its
present amount. Although such fluctuations are comparatively small,
there are no indicationswhatsoeverthat they actually occurredduring
that time.

On the other hand, there can be no doubt that the orientation of the
earth's surface relative to the sun changed periodically during

108

CLIMATIC

CHANGES

Quaternary. Since the calculationscited aboveprove that the variations

in the radiation regime caused by this effect are sufficient for the
developmentof glaciations,we need not countthe fluctuationsof carbon
dioxide concentration among the phenomena causing the migration of
the ice cover boundaries.

The Role of Volcanic Activity. It was mentioned in Chapter i that

several authors point to volcanic activity as a major factor of climatic
variations during the Quaternary.
It follows from general considerationsthat fluctuations in volcanic
activity can affect climatic conditionsin two ways: by increasing the
amount of carbon dioxide in the atmosphere and by reducing atmospheric transparency.
The first of these mechanisms led to a rise in air temperature
during the Quarternary, when the level of carbondioxidewas low, while
the secondmechanismhas reducedthe temperature throughout geologic
time.

We can assume that during the Quaternary the secondeffect prevailed over the first one. If a volcanic eruption ejects into the atmosphere similar amountsof carbondioxideand sulphurousgas, their effect
on the thermal regime is quite different. A comparatively small amount
of sulphurous gas (say, several million tons) can reduce the atmosphere'stransparencyconsiderablyby producingaerosolparticles in the
lower stratosphere, and this can causea sensibledrop of air temperature
at the earth's surface (seeChapters 3 and 7). At the same time an equal
or even a thousandfoldaddition of carbon dioxide changesthe mass of
atmosphericcarbon dioxide very little and, consequently,cannot possibly affect the climate.
The conclusionis confirmedby the noticeableincreasein the mean
global temperature during a quiet volcanic period (1920s-1930s).
In accordancewith the data given in Section 4.2, we may conclude
that in the courseof the last hundreds of millions of years there was a
tendency toward a reduction in the amount of carbon dioxide in the
atmosphere, but there is no proof of any noticeable increase of its
concentrationduring the Quaternary.
Since the volcanic activity in the Quaternary period led to a decreaseof air temperature, it is important to estimate its possiblevalue.
In the next sectionwe go on to analyze the mechanism of drastic
short-term air temperature changes over the planet caused by the
statistical regularities of the simultaneous eruption of a number of
volcanoes.In the courseof hundreds of thousandsof years these events
could have resulted in numerous air temperature decreasesamounting
to several degreesover several decades.It might be thought that

CLIMATES

OF

THE

PAST

109

temperature changes had no considerableinfluence upon the development of glaciation.

It is possible that the fluctuation in the number and power of
volcaniceruptionsfor time intervals of hundredsand thousandsof years
could have played a great part in the formation of climatic conditions
during the Quaternary.
From Chapter 3 it followsthat the rise in global temperature in the
first half of the twentieth century was limited to about 0.5C chiefly
becausethe period of increasedatmospherictransparency was so brief.
Had this period lasted several centuries instead of several decades,the
polar sea ice would have melted considerably, and this would have
meant an additional considerabletemperature increase. A similar result could be reached by a long-term reduction of the air transparency
which attenuates the total radiation flux by severaltenths of a per cent,
i.e. by an amount correspondingto the current climatic fluctuation.
Such a reduction would be sufficient to producea migration of the polar
ice boundary into lower latitudes by hundreds of kilometers, which is on
the same scale as the ice expansion during the major glaciations.
Thus, if changesin volcanicactivity which occurredin the courseof
the last century had continued for a sufficiently long time, say, thousandsor tens of thousandsof years, they could have brought about the
ice ages.
Unfortunately, the available material on the frequencyof volcanic
eruptions is not sufficient for calculating these fluctuations over prolonged periods of time. Statistical estimates based on the recurrence of
eruptions as a random processare more reliable for brief periods of time,
when the number of eruptions remains considerably lower than the
number of active volcanoes.From materials on epochsof higher volcanic
activity that undoubtedly took place, it is difficult to estimate the
transparency changes in the atmosphere and the duration of these
changesin the courseof the given epochs.
In estimating the role of volcanic activity in the development of
Quaternary glaciations, therefore, the deductionthat changesin astronomical factors are sufficient for the developmentand melting of glaciations is highly significant.
From this it follows that the volcanic activity has not been the
major factor in the formation of the ice sheets, although its fluctuations
might have had someeffect on the Quaternary climatic conditions.This
effectwas probablynot as strongas that of the astronomicalfactorsand
apparently manifested itself mainly during briefer time periods.
It is interesting to analyze in this connection the mechanism of
climatic variations during the last several thousand years, a short
descriptionof which has been given in Chapter

110

CLIMATIC

CHANGES

These fluctuations lasted only about several hundreds or thousands

of years, thus excluding the possibility of any relation to the astronomical climatic factors which are of considerably greater duration.
Auer (1956) and other scientists have established the following
epochsof increased volcanic activity: ca. 7000 B.C.; 3500-3000 B.C.; 500200 B.C., and A.D. 1500-1900. Lamb (1970) contends that the volcanic

activity could have affected the climatic conditions during these epochs
but it was not the only cause of these climatic variations.
The climatic fluctuations during the Holocene were characterized
by comparatively small changes in the air temperature. They could
have been produced by fluctuations in the atmospheric transparency
due to volcanic activity. They could also have been caused by autooscillatory processesoccurring in the ocean-polar ice-atmosphere system.

Causes of Climatic Variations. The specificclimatic conditionsof

the Quaternary period were apparently due to a drop in the carbon
dioxide concentration in the atmosphere and to continental drift and
orogeny, which will be analyzed in the next section. This has led to a
partial isolation of the Arctic Ocean and to the shift of the Antarctic
continent to the polar zone of the southern hemisphere.
The Quaternary period was precededby a climatic evolution of long
duration leading to the formation of distinctive thermal zones. This
processwas causedby changes on the earth's surface and found expression mainly in the air temperature decrease in the middle and high
latitudes.

During the Pliocene climatic conditionsbegan to be affectedby the

lower concentration of carbon dioxide in the atmosphere, which led to a
drop in the global air temperature of 2-3C (and 3-5Cin high latitudes).
After this the polar ice coversbegan to developand this led to a further
reduction in the mean global temperature, especiallyin high latitudes.
The existence of polar ice caps sharply increased the sensitivity of
the thermal regime to small fluctuations in climatic factors. Thus,
enormous boundary migrations of the snow and ice covers on land and
sea became possible due to the changes in the position of the earth's
surface with respect to the sun, which previously had not exerted a
noticeable

effect on the climate.

The continued decrease of the carbon

dioxide

concentration

in the

atmosphere facilitated the intensification of later glaciations as compared to earlier ones, although the scale of glaciation dependedprimarily on the combination of astronomical factors.
It seemsthat astronomicalfactorsexerted the greatest influence on
climatic variations during the Quaternary, though it is very

CLIMATES

OF

THE

PAST

111

that the changesin volcanicactivity couldalso play an important role in

climatic variations during time intervals of several hundreds or thousands of years.
4.2 Pre-Quaternary

Variations

As pointed out in Chapter 1, the farther back we go into the ages,

the less amount of evidence do we find on climatic conditions, and the
more difficult it becomesto interpret these data.
The most reliable, though schematic,information on the climates of

the past stems from the very fact of the continuousexistenceof living
organismson our planet. The ecologyof organisms of the past may of
coursehave differed greatly from that of contemporary related forms.
Nevertheless it is highly improbable that they could have subsisted
outside the limits of a comparatively narrow temperature range, say,
from 0 to 50C,within which the overwhelming majority of present-day
animals and plants live.
It may therefore be assumedthat the surface temperatures, as well
as that of the lower atmosphericand upper water layers, did not exceed
these limits over the greater part of the earth during the past 2-3 billion
years, i.e. since living organisms first came into being.
The actual fluctuations of the mean surfacetemperature over long
time intervals were smaller than this temperature range, and in most
cases,did not exceed several degrees over millions of years.
Hence it is extremely difficult to study past variations in the earth's
thermal regime on the basis of empirical data, because the errors in
temperature determined either by means of radioisotopic analysis of
organic remains or by any other known method, are generally not less
than several degrees.
Another difficulty in studying the climates of the remote past
mentioned in Chapter I concernsthe location of different regions relative to the poles, causedby continental drift and possiblythe motion of
the poles themselves.
We shall therefore limit ourselves to an analysis of the secular
variations of air temperature that occurredduring the secondhalf of the
Mesozoicera and the Tertiary period. We will be using paleotemperature determinations obtained by Emiliani (1966) by means of radioisotopic analysis.
Curve I in Fig. 23 gives the secular distribution of the mean
temperature of the ocean surface level in middle latitudes, according to
Emiliani. We see from this figure that the air temperature near the
earth's surface during the last 130 million years had a tendency to
decrease.This tendency was interrupted several times by increases

CLIMATIC

112

CHANGES

T(C)

25-

20-2
15-

140

120

100

80

60

40

20

Time, m/n. years

Fig. 23. Secular temperature variations in the middle latitudes of northern hemisphere. (1) Observational data; (2) calculated curve.

the air temperature, which, however,gave way to coolingand did not

changethe general trend of climatic evolution.
In order to explain these climatic fluctuations,it is necessaryto
analyze the influenceof the atmosphere'scompositionand the structure
of the earth's surfaceon the thermal regime.
Atmospheric Carbon Dioxide. In studyingthe Pre-Quaternary
climatic conditionsit is important to deal with the effectof atmospheric
carbondioxidedecreaseon the thermal regime.
Carbondioxideis one of the comparativelysmall constituentsof the
atmosphere.At the presenttime the atmospherecontainsabout 2.3.10 2

t of carbon dioxide which amounts to 0.032% of the entire volume of the

atmosphere.

A much larger amount of carbondioxide is present in the hydrosphere,where 130.10 t of carbondioxideis dissolved,primarily in the
oceans.A constant exchange of carbon dioxide goes on between the
atmosphereand the hydrosphereby means of molecular and turbulent
diffusion.

Measurementsof carbondioxideconcentrationin the atmosphere

show that it changesvery little with location and also with altitude
within the troposphere.The relatively constantamount of carbondioxide in the atmosphere,as comparedto water vapor,is explainedby the
more stable nature of the sources and sinks of carbon dioxide on

CLIMATES

OF

THE

PAST

113

earth's surface and by its low sensitivity to temperature variations.

Observational

data show that

the carbon dioxide concentration

in-

creasesslightly in the equatorial region and dropsby approximately

0.005%in high latitudes. This effectis explainedby the higher solubility
of carbondioxidein the coldoceanwaters in high latitudes, as compared
to warm tropical waters. As a result, the atmosphere in high latitudes
loses its carbon dioxide, which is then dissolved in the oceans. Its

surplus is transported by the deep, cold currents to the lower latitudes

where it is releasedback into the atmosphere.The flux of carbondioxide
between the equator and the north pole producedby this mechanismis
about 2.1020t/yr (Bolin and Keeling, 1963).
A continuous carbon dioxide exchange exists between the atmosphere and the hydrosphere,on the one hand, and living organismsand
the lithosphere, on the other. This cycle plays an extremely important
part in the maintenance of life on our planet.
The main ingredients of this cycle are determined by biological
processes.Carbon dioxide is absorbed by autotrophic plants during
photosynthesis,which is the sourceof the direct or indirect energy of the
overwhelming majority of living organisms. In the course of their biologic activity (particularly, during the breathing process)carbondioxide
is released and returned to the atmosphere.
A certain amount of carbon dioxide is ejected into the atmosphere
from the deeplayers of the earth's crust during volcaniceruptions,from
mineral water sources,etc. Carbon dioxide is consumedby silicate rocks
during weathering and in the processof formation of various carbonaceous combinations.

Thus, two cycles, biological and geological, are involved in the

balance of atmospheric carbon dioxide, and each of these has its sources
and sinks of carbon dioxide. Available data show that the biological
componentsof the annual carbon dioxide cycle considerablyexceedthe
geologicalcomponentsof this process.
The annual carbon dioxide expenditure for photosynthesisat the
present time is about 10 t. The same amount of carbon dioxide is
producedin the processof breathing and the decompositionof living
organisms. A much smaller amount of carbon dioxide, about 10st, comes
from volcanic eruptions. An amount of carbon dioxide of the same order
of magnitude is used up in various geologicalprocesses(Plass, 1956;
Mfiller, 1960; Lieth, 1963).
Living organisms, products of their metabolic activity, and the
lithosphere contain large amounts of carbon drawn from the atmosphere
and hydrospherein the form of carbon dioxide. The amount of carbon in
living organisms apparently correspondsto about 10' t of carbon

CLIMATIC

114

CHANGES

ide. The estimatesof different researchersvary considerablyas to this

value (Takahashi, 1967; Man's Impact on the Global Environment,
1970).

The amount of carbonin the lithospherecorresponds

to about 2.10 7
t of carbondioxide, the major part of which is trapped in carbonaceous

rocks (Vinogradov, 1972).

Although the estimatesgiven of both the componentsof the carbon

dioxide balance, and of the carbon stored in various natural environments, are very approximate, someconclusionsmay be drawn about the

rate of the carbon dioxide cycle.

For example, the living organismshave a comparativelylow carbon dioxide capacity"; the amount of carbon in them is renewed in a
rather brief period of about 10 years, while in the atmospherethis
processtakes about 20 years.
The period during which carbon has been accumulating in the
lithosphereis very long and is commensuratewith the period of existence of the biosphere.
There are groundsfor believing that the carbondioxide concentration in the atmosphereduring the geologicalpast differed greatly from
the present. It follows therefore that the carbon dioxide balance in the
atmospherehas not always been equal to zero.
However, this differencebetween the gain and loss of carbondioxide comprisesa very small fraction of the absolute values of the main
componentscontributing to the carbon dioxide budget of the atmosphere. For instance, if we assumethat over the last million years the
carbondioxideconcentrationin the atmospheredroppedto onehalf, its
annual variation would correspondto only 0.002%of the annual amount
of photosynthesis.
Geochemicalstudiesindicatethat duringthe last part ofthe geological history of the earth, the carbon dioxide concentration in the atmosphere has tended to decrease.

According to Vinogradov (1957), the atmospheric carbon dioxide

originatedtogetherwith othergasesin the courseof the zonalmolting of
the earth's mantle, heated in the processof radioactive decay of the
elements. During the Proterozoicera, the atmospherewas anaerobic
and contained a large amount of carbon dioxide. By the end of that time
the compositionof the atmospherebegan to changeunder the influence
of photosynthesis,
the amount of oxygengradually increasingand the
atmosphere acquiring an oxidizing property. The increase in oxygen
content was accompanied by a decrease of carbon dioxide.
Ronov (1959, 1964, 1972) studied the regularities of changing the

amount of carbondioxide in the atmosphereduring the geologicalpast

and established that throughout the end of the Paleozoic and

CLIMATES

OF

THE

PAST

115

eras the carbon dioxide concentrationdroppedgradually, chiefly as a

result of its consumption in the formation of carbonaceousrocks. The
amount of carbon dioxide in the atmosphere increased somewhat in the
period of higher volcanic activity.
Estimates of the atmosphericcarbondioxideduring various geological periodsare highly approximate.For example, Hoffman (1951)suggeststhat its concentration during the Permian period (about 250 million years ago), was 7.5%. On the basis of this figure we come to the
conclusionthat during the Mesozoicand Cenozoiceras, the average rate
of its decreasecame to 0.03% per million years.
There are also other opinionson the rate of carbon dioxide changes
during the geologicalpast. For example, Rutten (1971)contendsthat its
maximum concentration during the Phanerozoic exceededits modern
level by not more than a factor of ten. This agrees with the assumption
that the main part of the free carbon dioxide was removed from the
atmosphere during the Proterozoic era, after which the rate of its
decreasedroppeddrastically.
In the study carried out by the author jointly with Ronov and
Yanshin, using data on the carboncontent in sediments,the change in
atmosphericCO2 mass during the last 570M yr was computed. Fig. 29
(curve CO2)showsthe results of this calculation.From this figure it is
inferred that during this time interval the CO concentration changed
approximately by an order of magnitude. This conclusionfits the view
point of Rutten. From the end of the Mesozoicera the tendency for CO
mass to decreaseprevailed, resulting in lowering its concentration from
a few tenths to three hundredths per cent.
Let us consider the influence exercisedby the carbon dioxide concentration on living things.
The vegetative cover of the land consists mainly of autotrophic
plants producing organic matter from atmosphericcarbon dioxide, water, and minerals from the soil. Together with the atmospheric carbon
dioxide, plants can utilize the carbon dioxide from the soil; its amount,
however, is negligible as compared to that in the atmosphere. The
autotrophic plants of the hydrosphereutilize the carbon dioxide dissolvedin the water for their photosynthesis.
It has long been established that under sufficient insolation and
other favorable conditions,the rate of photosynthesisis strongly limited
by the low concentrationof carbondioxide in the atmosphereand water
reservoirs.

It has been proved by numerous experiments that an increase in

carbon dioxide concentrationfacilitates the increase of the photosynthetic rate, other conditions being equal. For most plants this rate is
higher when the carbon dioxide concentration increasesup to

CLIMATIC

116

CHANGES

tenths of one per cent. After this the photosynthesisrate stopsrising,

and when the concentrationexceedsseveral per cent, the photosynthesis
begins to drop.
The considerableenhancement of photosynthesisunder increased
carbon dioxide concentration is often interpreted as evidence of the
adaptability of autotrophic plants in the courseof their evolution to
higher concentrationsin the past.
Since the geochemicaldata indicates a gradual decreasein the
amount of atmosphericcarbondioxide during the last hundred of millions of years, this assumptioncan be consideredplausible.
The questionarises in this connectionas to whether, during the
time of the existence of autotrophic vegetation, the carbon dioxide
concentration could exceedthe value at which the photosynthesisrate
begins to drop.
It is rather difficult to answer this questionbecausethe dependence
of photosynthesis
on the amount of carbondioxidefor the plants of the
remotepastdoesnot necessarilycoincidewith a similar dependence
that
exists for present-day plants.
Let us note that the variation

in the amount of carbon dioxide in the

atmospherehas an effect on the total mass of living organismson our

planet. The major portionof this biomassconsistsof autotrophicplants,
the amount of which is controlledby their photosynthesis.This dependencehas a very simple explanation. The consumptionof organic matter
by autotrophicplants through breathing, through the dying of certain
parts of plants, etc. can be considered proportional to plant mass.
Inasmuch as the total amount of photosynthesisis equal to the expenditure of organic matter and the rate of photosynthesisis approximately
proportionalto the concentrationof carbondioxide, we find that the
mass of vegetation is also proportionalto the concentrationof carbon
dioxide.

We may assumethat in somecasesthe dependenceof the vegetative

mass on the carbon dioxide concentration is somewhat weaker, since the

rate of photosynthesiscan increasewith the surfacearea of the plant.

However, for a compact vegetation cover with an optimal structure
providing the maximum photosynthesis,the dependenceof the total
photosynthesison the size of the photosynthesizingsurfaceis not great
(Budyko, 1971).
Thus the mass of autotrophic vegetation depends greatly on the
carbondioxide concentrationand sometimesthe mass is directly proportional

to the concentration.

Obviously,this conclusionpertains mainly to those casesin which

the amount of water, mineral nourishment and other factors are in
sufficientsupply, that is, they do not thwart the photosynthesis

CLIMATES

OF THE

117

PAST

It is possible,however, that even if there were unfavorable factors, some

relation would exist between the vegetation mass and carbon dioxide
concentration, one that would be particularly noticeable at low concentrations.

From this it follows that during the geological past, when the
carbondioxideconcentrationwas considerablyhigher than it is now, the
mass of plant cover on land was probably greater than at present. This
refers chiefly to vegetation in zoneswith sufficient moisture. The total
amount of the biomassof autotrophicvegetation in the oceanduring the
time of high carbondioxidecontentcouldalso be higher than its present
level.

Thus, the processof the carbon dioxide decreasein the atmosphere

and hydrosphere was probably accompaniedby a gradual decrease in
the massof autotrophic plants and consequentlyin the massof all living
organisms on our planet.
This law mplies that for every plant of a certain size there exists a
lower bounr.ary of carbon dioxide concentrationbelow which the given
plant cannot go on living. Under these circumstances,autotrophicmicroorganismslast longer than any others, due to the fact that the ratio
of their massto the area of the photosynthesizingsurfaceis minimal.
The question arises as to the limit of the diminishing global biomass if the tendency of carbon dioxide to decreasecontinues in the
future. As noted above, over the last 100 million years CO concentration decreasedby approximately 0.2%. Since its present value is only
about 0.03%, it seems obviousthat autotrophic plants will disappear
within a few million years, which is a short period compared to the
duration

of their

existence.

There are, however, insufficient groundsfor this forecast. During

the most recent

times

the decrease in the amount

of carbon dioxide

ceasedand gave way to an increasedue to the economicactivity of man.

There are reasons to believe that, had this processof carbon dioxide
decreasecontinued further, there would have been a drastic change in
the climatic

conditions that would have made the continuation

of life

impossibleon this planet, and this would have happenedlongbeforethe

dying-out of autotrophic vegetation causedby an insufficiencyof carbon
dioxide.

Climatic Effects of Atmospheric Carbon Dioxide. Let us consider the dependenceof air temperature on the carbon dioxide concentration in the atmosphere and on the albedo of the earth-atmosphere
system (Budyko, 1973). For this purpose we shall make use of the
semiempirical model of the atmospheric thermal regime described in
Chapter 2.
Let us assume

that

in the absence

of feedback

between

the

CLIMATIC

118

CHANGES

temperature and the polar ice, the temperature at the earth's surface
varies with carbon dioxide concentration, in accordancewith the dependenceestablishedby Rakipova and Vishnyakova (1973).
To estimate the effect of carbondioxide on the thermal regime by
consideringthe feedback, we shall use the relation
Is' = '),'Is,

(4.2)

whereis is the long-waveradiation lost into outer spacefor the present

valueof carbondioxideconcentration
in the atmosphere;
Is' the outgoing
radiationfor variousvaluesof carbondioxideconcentration, a dimensionlesscoefficientdependingon the carbondioxide concentration.This
relation can be consideredsufficientlyexact when the coefficient is
closeto unity.
By using the empirical relation given in Chapter 2 for determining
Is we find
Is'= '[a+ bT-

(a, + b,T)n],

(4.3)

where T is the air temperature at the earth's surface in C, n the

cloudinessin fractions of unity; a, b, a, b, the dimension factors.
Considering that for the earth as a whole the amount of absorbed
radiation is equal to the outgoing radiation we find the relation

Qs',(1- as',)= [a + bT',- (a, + b,T,)n',],

(4.4)

where Qs,,is the solar radiation at the outer boundary of the atmosphere, Os',the albedo of the earth-atmosphere system, T', the mean
planetary temperatureat the earth'ssurface,n', the averagecloudiness.
For determining the average planetary temperature we obtain the
formula

T', =

Qs,(1- Os',)- '),,a+ '),,a,n',

.
'),(b- b,n',)

(4.5)

Assumingthat the coefficientT varies only slightly with latitude,

and applying the semiempirical theory of the atmosphere'sthermal
regime for determining the mean annual temperature of various latitude zones, we obtain the formula
T =

Qs(1 - o's)-'),a + '),,a,n+ fiT,,

,
fi + '),,
b - '),,
b, n

(4.6)

where Qs, as,n, and T refer to the latitude zoneunder consideration;fi is

a dimension factor describing the intensity of the meridional heat
exchange.
The T coefficientfor various concentrationsof carbondioxidecan

CLIMATES

OF

THE

PAST

119

A Tv
-

0 //
-

-2

-4

-6
-8
-10

-50

O,05

0,15 C(%)

0,10

Fig. 24. Effect of carbon dioxide on air temperature. Solid line, albedo
depending on air temperature; broken line, constant albedo.

found from formula (4.4) by using it together with the relation established by Rakipova and Vishnyakova and presentedin Fig. 24 by curve
1. The determined values of this coefficientare given in Table 14.
By taking into account these values and the relation between the
boundaries of the polar ice cover and the albedo, on the one hand, and
the air temperature, on the other, we can calculate the variations of the
mean planetary temperature for the changing concentrationof carbon
dioxidewith an existingfeedbackbetweenthe polar ice and the thermal
regime. The result of this calculationis illustrated in Fig. 24 by curve 2.
As we see from Fig. 24, fluctuations in the surfacearea of the polar
ice greatly enhance the effect of the carbon dioxide fluctuations on the
mean planetary temperature. When the carbon dioxide concentration

increases30% above its present level, the temperature rises significantly, and when the amount of carbon dioxide exceeds 0.05%, the
temperature rises several degrees above its present level.
When the amount of carbon dioxide drops, so doesthe average
Table

14

Dependenceof Coefficient /on Carbon

Dioxide

c (%)

Concentration

0.000

0.005

0.010

0.020

0.032

0.050

0.070

1.038

1.030

1.030

1.011

1.000

0.994

0.990

0.100

CLIMATIC

120

CHANGES

planetary temperature, and when the carbon dioxide concentration is

equal to about half of its present amount, the average temperature at

the earth's surfacedropsby several tens of degreeswhich correspondsto

conditions of total glaciation of the earth.
The dependenceof the polar ice boundary in the northern hemisphere on the amount of carbon dioxide calculated accordingto this
scheme is presented in Fig. 25. The increase of the carbon dioxide
concentrationby 30% above its present level (designatedby a vertical
line) appearsto be sufficientfor completemelting of the polar ice, while
a reduction to half its present value leads, as pointed out above, to the
total glaciation of our planet.
It shouldbe notedthat the generalrelation betweenthe averageair
temperature and the polar ice boundary, on the one hand, and the
carbondioxideconcentration,on the other, is morecomplicatedthan the
curvesshownin Figs. 24 and 25. This relation is similar to the dependenceof these parameters on the solar constantwhich is ambiguous,as
proved in Chapter 2.
Sincein this casewe are concernedsolelywith the climatic effectsof
variationsin the carbondioxideconcentrationfrom its presentlevel, the
pertinent relations becomesimplified and obtain the form given above.
When analyzing the relations presentedin Figs. 24 and 25, it is
necessaryto remember that they characterize stationary conditionsof
the atmosphere-ocean-polar
ice system that can be reached only over
long periods, such as several thousand years. If the changesin the
carbon dioxide concentrationoccur at a much higher rae (as, for instance, is the case at the present time), then the variations of the
thermal regime and the polar ice boundarywill be considerablysmaller
than thosedeterminedby the relations presentedin Figs. 24 and 25.
90

.....

60

30

O.025

0,050

C(%)

Fig. 25. Dependence of polar ice boundary in northern hemisphere on

carbon

dioxide

CLIMATES

OF THE

PAST

121

By means of formulas (4.5) and (4.6), we can calculate the distribution of mean latitudinal temperatures at the earth's surface for two
cases:1) an ice-free regime correspondingto a carbondioxide concentra-

tion of 0.042%,and 2) the total glaciationof the earth corresponding

to a
concentration

of 0.015%.

In the first instance, calculations were carried out under the as-

sumptionthat the albedo of the earth-atmospheresystem in high latitudes is equal to 0.40 in the absenceof ice. The effect on the thermal
regime of the possiblechangesin cloudinesswas neglected.
In the second instance, it was assumed that the albedo at all

latitudeswas equal to 0.64 and that the effectof cloudinesswas negligi-

ble.

The results of calculationsfor the northern hemisphereare presentedin Fig. 26. We may seefrom this figure that when the concentration of carbon dioxide is equal to 0.042% (curve 1), the mean annual
temperaturein low latitudes risesapproximately2Cand in high latitudes as much as 10Ccomparedto present conditions(curve 2).
In the case of total glaciation (white earth) the mean annual
temperature drops approximately 70Cat the equator and 45Cat the
pole (curve 3). It must be noted that these temperature values for the
case of total glaciation differ somewhat from the values obtained in
Chapter 2. This discrepancyis due to the fact that in the present
T(C)
40-

2O

-2O

-4O

30
60
90 c?
o
Fig. 26. Distribution of mean latitudinal temperature as a function of
carbon dioxide concentration. (1) Ice-free regime; (2) present conditions; (3)
total

CLIMATIC

122

CHANGES

calculation we neglectedthe effect of cloudinesson the thermal regime

at very low air temperatures, and we also assumeda different value for
the albedo of the white earth. Since the mean air temperature at the
earth's surface in the case of total glaciation is far below zero under
various assumptionsabout the values of albedoand cloudiness,it can be
deducedthat the white earth regime is highly stable.
Due to the enormouseffectof the feedbackbetween the temperature
of the atmosphereand the polar ice, a comparatively small variation in
the carbon dioxide concentration (from 0.015 to 0.042%) changes the
mean annual temperatures of the atmosphereat various latitude zones
by several tens of degrees.
The precision of our calculation is limited by the fact that in our
model of the thermal regime we have used several simplifying assumptions. This is why our estimates of the effect of carbondioxide concentration on the thermal regime are very rough and must be checkedand
made more precise in future researches.
Thus, Manabe's and Wetherald's (1975)results are of interest. They
calculated the changesin the atmosphere'sthermal regime for a doubling of the amount of carbon dioxide. For this purposea three-dimensional model of the general climatic theory was used that takes into
account the effect of water phase transitions and feedbackbetween the
thermal regime and the snow-and-icecover on the air temperature.
The obtained variations of the mean latitudinal air temperature at
the earth's surfaceobtainedby Manabe and Wetheraid are presentedin
Fig. 27 (curve M). For comparisonthe same figure gives the variations
of the air temperature calculatedaccordingto the semiempiricalmodel
of Chapter 2, for a case when the local temperature effect of carbon
AT

M
l

IO

10

20

$0

40

50

60

70

80

90 N

Fig. 27. Effect of carbon dioxide concentration on mean latitudinal air

CLIMATES

OF

THE

PAST

123

dioxide is taken into accountaccordingto the Manabe-Wetherald (1967)

scheme (curve 1), and to the schemesuggestedby Rakipova and Vishnyakova (curve 2).

The figure showsthat temperature variations at different latitudes

determined by completely different models are in agreement.
It was mentioned earlier that there has been a long-term tendency
for the carbon dioxide concentration in the atmosphere to decrease. It
may be assumed that the fluctuations in the amount of free carbon
dioxide in the atmosphere and hydrosphereover protracted time periods
cannot depend on the increase or decrease of the total mass of living
organisms. For example, if we assume that this mass is equal to the
values given above then its variation by 100% can be compensatedfor by
the inflow of carbon dioxide from the lithosphere (or by its expenditure
on geologicalprocesses)in about 10,000 years. If we take the previous
amount of biomass to be 10 times as high, then the relative time
estimate will increase only to 100,000years, which is still small according to the time-scale of geological processes.
It is obviousthat the changein the massof living organismshas an
insignificant influence on the balance of carbon dioxide, as compared
with that of the gas exchange between the atmosphere and the lithosphere during long time intervals commensurate with geological periods.

Unlike living organisms, the lithosphere can accumulate a very

large amount of carbon and can also release a large amount of carbon
dioxide over a long period of time. Therefore the geologicalcomponents
of the carbondioxide balance, which are much smaller than its biological components,play the major part in the fluctuations of the amount of
free carbon dioxide.

In contrast to the biological cycle of carbon dioxide, where the gain

and loss are practically equal due to the comparatively small mass of
living organisms, the gain and loss of carbon dioxide in the geological
cycle are poorly related. Therefore there are casespossiblein which the
input of carbon dioxide from the lithosphere greatly exceedsits absorption by the lithosphere, and vice versa.

Consideringthat, under modern conditions,there is a difference of

about 10s t per year between the lossof carbondioxide due to geological
processes,and the input of carbon dioxide from the lithosphere, we find
that the amount of free carbon dioxide in the atmosphereand hydrospherecausedby this differencechangesat the rate of 10TM
t in a million
years.

If we assume that the amount of atmospheric carbon dioxide is

equal to a few per cent of that in the hydrosphere, we find that

CLIMATIC

124

CHANGES

amount of atmospheric carbon dioxide probably decreasesby 102 t, or

several hundreths of a per cent in a million years. This estimation is not
at variance with the above-given rate of decreasein the concentrationof
atmospheric carbon dioxide over the last million years.
Thus geologicalprocessescan cause variations in the atmospheric
carbon dioxide concentration. It is difficult to carry out a quantitative
analysis of the effect of lithospheric processeson the carbon dioxide
balance

because the available

data on the carbon balance

in the litho-

sphere is far from precise. In fact, from present-day estimates, we can

determine only the order of magnitude of pertinent componentsin the
carbon

dioxide

balance.

Although we know very little about the changes in the carbon

dioxide concentrationin the geologicalpast, we can still give a comparatively reliable estimate of the influence of carbon dioxide on the thermal
regime in the atmosphere during the Pre-Quaternary time.
From

the aforesaid

data

it follows

that

when

the carbon

dioxide

concentrationexceeds0.10%, the mean global air temperature increases

about 5C over its present value. The carbon dioxide concentration
apparently reached this level by the secondhalf of the Tertiary period,
and therefore, during the Mesozoicera and part of the Tertiary period,
the mean global air temperature at the earth's surface was 5 higher,
while the fluctuation in the carbon dioxide concentration during the
Pre-Quaternary time, when its mean level was rather high, had no
effects on climatic

conditions.

The secular temperature variations in the Pre-Quaternary, presentedin Fig. 23, demonstratethat, although the temperature then was
appreciablyhigher than at present, it was not constant.This indicates
that there were other factors then which exercised an effect on the
climate.
Climatic Effects of the Structure of the Earth's Surface.
In the

courseof the secondhalf of the Mesozoicera and the first part of the
Tertiary period (the Paleogene), the mean continental level was cornparatively low, so that a considerablepart of the continental platform
was coveredby shallow seas. It was during that time that the various
continental regions repeatedly rose and fell. This, however, did not
prevent the continents from being criss-crossedby wider or narrower
straits

and seas.

During the latter part of the Tertiary period (the Neogene), intensive tectonic activity led to the elevation of continentsand to a gradual
disappearanceof many seas. For example, a vast body of water on the
territory of what is now western Siberia, that formerly joined the
tropical oceansto the polar basin, then ceasedto exist. Thus the

CLIMATES

OF

THE

PAST

125

Oceanbecamea more or less isolatedbody of water which mergesonly

with the Atlantic Ocean and is connectedwith the Pacific only by the
narrow Bering Strait.
It is possiblethat the movementof the Antarctic continent to higher
latitudes was completedduring the Tertiary period and that it has since
then been situated concentricallyrelative to the south pole.
Such radical changes in the structure of the earth's surface must
have had a great effecton the meridional heat exchangein the ocean.
In Chapter 2 it was assertedthat the present-daymeridional heat
transfer in the ocean comprisesapproximately one-half of that in the
atmosphere.It is probable that the increase in the surface area of bodies
of water connectinghigh and low latitudes leads to an increase in heat
transfer. The relation between the meridional heat exchange and the
distribution of mean latitudinal temperatures can be found from the
semiempiricaltheory of the atmosphericthermal regime.
The result of suchcalculationsfor the northern hemisphere is given
in Fig. 28 where curve I correspondsto the distribution of the mean
annual temperature under ice-free conditionsin high latitudes for the
present ratio of meridional heat transfer in the atmosphereand in the
hydrosphere. Curve 2 gives the temperature distribution when these
transfers are equal, and curve 3 correspondsto a casewhen the transfer
in the hydrosphereis twice as large as that in the atmosphere.
In the formulas of the semiempirical theory, the ratio of the transfers is determined by the parameter
Figure 28 implies that the fluctuations in the meridional heat
exchangedo not affect the averageplanetary temperature. Accordingly,
the air temperature at latitude 35 remains constant for various conditions of meridional heat exchange.
Figure 28 also demonstrates that the relative increase of the meridional heat exchange in the oceansleads to a small decreasein the mean
air temperature near the equator, and to a temperature increase in the
high and some of the middle latitudes.
We may assume that the decrease of the meridional heat flow
caused by the elevation of continents played a certain role in the
temperature decreasein middle and high latitudes during the past 100150 million years.
Assuming that at the beginning of this period the meridional heat
transfer in the oceanswas considerablyhigher than in the atmosphere
but was closeto its current value at the end of that period, and taking
into account the drop in the global temperature at the end of the
Tertiary period due to the decreaseof carbon dioxide concentration, we
obtain the secular temperature variations in middle latitudes

CLIMATIC

126

CHANGES

T(C)
30

2O

I0
3
2

-I0,

20

40

60

80 oo

Fig. 28. Effect of heat transfer in oceanson distribution of mean latitudinal air temperature. (1) Ice-free regime; (2) meridional heat transfers in atmosphere and hydrosphere are equal; (3) heat transfer in hydrosphere is twice as
large as in atmosphere.

spondingto curve 2 in Fig. 23. This curve agrees reasonably well with
the empirical curve 1, though they do not coincide.
It is interesting to analyze the reasons of temperature fluctuations
(seecurve I in Fig. 23) during time intervals of tens of millions of years.
Owing to the differences of opinion in the interpretation of data on
paleotemperatures of the Mesozoic era and the Tertiary period, the
reliability of these temperature variations and their global character is
not entirely clear (Teis and Naidin, 1973). Probably these temperature
fluctuations

were

connected

with

local

variations

in the

circulation

processesof the atmosphere and ocean.

A possible effect of the earth's surface structure on climate is the
developmentof continental glaciations. At certain heights in the mountains, a constant snow cover as well as glaciation often form depending
on the air temperature at sea level and on precipitation.
The lack of any traces of mountain glaciationsduring the Mesozoic
era and the greater part of the Tertiary period is an indication not only
of a warm

climate

but also of the low level of the continents.

From this viewpoint the nature of the last paleoglaciation-the

Permo-Carboniferous- that lasted tens of millions of years, is of special
interest.

As we pointed out in Chapter 1, there are traces of this glaciation in

several continental regions in tropical latitudes. Paleogeographicdata
show that the vast regions of cold climate that appeared during the
Quaternary period did not yet exist during this

CLIMATES

OF

THE

PAST

127

According to one hypothesis, the Permo-Carboniferousglaciation

covered a continent that was then situated in high latitudes of the
southern hemisphere and which later broke up.
At the same time, simple calculations indicate that due to their
high albedo, large glaciations could have existed at any latitude (Budyko, 1961). This possibility follows particularly from the fact that the
absorbedradiation in the region of a vast glaciation near the equator
can be equal to the absorbedradiation at the pole under ice-free conditions.

If we take the albedoof the earth-atmospheresystemfor the ice-free

polar basin as equal to 0.30, we find that the mean annual amount of
absorbed radiation

at a latitude of 80 is 7.7 kcal/cm . month. The same

amount of radiation will be absorbedat the equator if the albedois 0.70,

which is a realistic figure for a large continental glaciation.
By means of semiempirical theory of climate, we can calculate the
mean latitudinal temperature distribution in the case of an equatorial
glaciation (Budyko, 1971).Let us assumethat this glaciation is situated
in a latitude zone near the equator occupying5% of the total area of the
hemisphere. If the albedo of this zone is equal to 0.70, the mean
weighted value of the planetary albedo will drop by 0.023, as compared
to ice-free

conditions.

The temperature distribution, calculatedaccordingto formula (4.6),

where fi = 0.40 kcal/cm. month .deg, showsthat it is nearly 0Cin the
vicinity of the glacier. This provesthat a glacier can exist on a plain at
sea level in the equatorial region. Obviously, the parts of the glacier
abovethis level will exist under negative temperatures.
It is interesting to note that the temperature at the pole in this case
will remain positive, so that the developmentof the polar glaciation is
not inevitable.

The possibility of existence of large stable glaciations at low latitudes doesnot contradict the hypothesesof continental drift and polar
motion during these epochs. This conclusion only implies that the
centersof large glaciationsdid not necessarilycoincidewith the poles.
Calculations of the average latitudinal temperature distribution,
the results of which are presented above, were carried out without
taking into accountthe global temperature increaseunder the influence
of high carbon dioxide content in the atmosphere.By taking this into
consideration,we see that if the Permo-Carboniferousglaciation had
developedat low latitudes, it would have been mainly situated at an
altitude of several hundred meters. As there is no evidenceof any
influenceof this glaciation on the climate of other geographicalregions,
it is possiblethat this was a mountain

128

CLIMATIC

CLIMATES

OF THE

PAST

129

Climatic Effect of Volcanic Activity. The intensity of volcanic

activity in the past varied to a large degree, as witnessedby the amount
of eruption productscontainedin the layers of different geologicalages.
We have plotted the diagram characterizing the change in volcanic
activity during the Phanerozoiceon (Fig. 29, curve V).
Figure 29 showsthat the intensity of volcanic activity during different epochsunderwent manifold variations. The carbon dioxide concentration changedin a more or less similar way, which is understandable
as the eruptions are one of the main sources of carbon dioxide in the

atmosphere. The diagram reveals a tendency toward a decrease of

carbon dioxide from the second half of the Mesozoic.

As already mentioned, the climatic conditionsduring the main part

of Phanerozoiceon did not dependon the variations of the atmospheric
carbon dioxide. Therefore the only factor of volcanic activity that could
affect the climate was an increase in the amount of aerosolparticles in
the stratosphere. As shown above, this processled to a decreasein the
air temperature at the earth's surface.
In the previous sectionwe arrived at the conclusionthat the longterm climatic effectsof volcanic eruptions were not great. At the same
time, we believe that the short-term fluctuations could have had a
perceptibleinfluence on the atmosphericthermal regime.
Several geologicalstudiesnote that the possibledeviations of many
natural phenomena from their normal intensity increasewith time. For
example, disastrousearthquakes that are hardly probable during shortterm intervals becomevery likely over sufficiently long periods. Therefore, from general considerations, it seems natural to assume that
throughout the long geologicalhistory of the earth, great anomalies in
the natural processescould have taken place that have not been observed during the comparatively short time of man's existence, and
particularly in the very brief period in which environmental studies
have been conducted.

Let us apply this principle in estimating the volcanic activity of the

past.

Since the effect of a single eruption on the temperature is comparatively low due to the limited amount of aerosol ejected into the atmosphere, it is obvious that the earth's temperature would be much more
affectedif a seriesof eruptionstook place within a short time interval.
Certain difficulties arise in calculating the greatest amount of
aerosolthat can be ejectedinto the atmosphereby volcanoesover a short
period of time.
An approximate estimate can be made from the table of explosive
eruptions in the eighteenth-twentieth centuries, compiled by

130

CLIMATIC

CHANGES

(1969). This table also gives the amount of aerosol ejected during each
eruption in relative units as compared to that of the Krakatao 1883
eruption, assumedto equal 1000 conventional units.
An analysis of Lamb's table shows that the amount of aerosol
ejectedinto the atmosphereeach decadeis subjectto great fluctuations,
its maximum value increasing with the total duration of time, i.e. with
the nmnber of considereddecades.This relation can be expressedby the
simple empirical formula
T

N = a logio--,
t

(4.7)

where N is the maximum amount of dust ejected into the atmosphere

during the time interval t of the period T, and the factor a is equal to
3000.

This formula is applicable to values of T much greater than t. It

followsfrom this formula that for T = 300 the amount of dust ejectedin a
decade (t = 10) is equal to 4400, which is in full accord with Lamb's
maximum

value.

If we use this formula to extrapolate over great time intervals, we

find that N = 9000 for T = 104,and N = 18,000for T = 107.Thus, during
a long time span, several eruptions within a single decadeare possible,
their combined effect from the ejected aerosol being equal to nine or
even 18 Krakatao eruptions.
The reliability of this result obtainedby extrapolation shouldnot be
overestimated.At the same time it must be emphasizedthat the above
amounts of aerosol are more likely than not to be understated.
Although in Lamb's opinion the volcanic activity of the recent
centurieshas been rather high, there were long periodsof much higher
activity in the geological past. Therefore, for such periods, the aerosol
amounts

based on Lamb's

data are not overestimated.

Let us consider the earth's temperature in the case of multiple

eruptions.
If 10 eruptions comparableto the Krakatau eruption occurwithin a
decade,the average decreasein direct radiation will be at least 10-20%
(according to Chapter 3), because the rate of aerosol ejected into the
atmosphereduring this time is not less than its rate of settling. In this
case the total radiation will drop by 2-3%, while the average temperature of the earth's surface will decreaseby 3-5 .
If 20 eruptions take place within a decade, the radiation decrease
will be more drastic. From atmosphericopticsit followsthat in this case
the total radiation will drop by at least 5%, and this would lead to

CLIMATES

OF THE

PAST

131

temperature drop of 5-10 at the surfaceas this coolingspreadsto both

the atmosphereand the upper level of the oceanwaters.
Although all these results are highly approximate, they nevertheless showthat from time to time, volcanic eruptions lead to short-term
but considerabletemperature dropsthat affect the entire planet.
Under stable climatic conditionswith a relatively small temperature differencebetween the polesand the equator that existed during
the Mesozoicera and almost the entire Tertiary period, even suchgreat
short-termtemperature dropscouldnot lead to glaciation on accountof
the high initial temperatures in the polar latitudes.
Causes of Climatic

Variations.

Climatic

conditions

of the Meso-

zoicera and the Tertiary periodwere characterizedby two main properties:

1) The mean temperature at the earth's surface was much higher

than now, especiallyin high latitudes. Consequently,the temperature
differencebetween the equator and the poles was smaller.
2) The air temperature had a tendency to decrease,especiallyin
high latitudes, and this tendency grew during the last half of the
Tertiary period.
These two regularities were caused by changes in two climatic
factors-the amount of carbondioxide in the atmosphereand the elevation of the continents. The higher concentrationof carbon dioxide provided for an increase of 5 in the air temperature as comparedwith its
present value. This temperature persisted through the Pliocene when,
owing to a drop in the carbondioxideconcentration,the average temperature began to fall.
The

low level

of the

continents

facilitated

the meridional

heat

exchangein the oceans,thus increasing the air temperature in middle

and high latitudes.
The heat exchange in the oceanswas weakened by the elevation of
the continentsand led to a gradual temperature drop in middle and high
latitudes. The rate of coolingwas acceleratedduring the Pliocenewhen
the thermal regime of the atmosphere appreciably changed under the
influence of CO2 concentrationvariation. This cooling was most pronounced in high latitudes and made the increase of polar glaciation
possible.
Although the thermal regime during the Mesozoic era and the
Tertiary period was in general highly stable due to the lack of large
polar glaciation, nevertheless short-term temperature drops of the air
and upper water level could occur. They were causedby the frequent
occurrenceof a series of explosive volcanic

5 Climate

and the Evolution

of

Living Organisms

5.1 Ecological Systems

The Climatic Zone of Life. There are many ways in which climatic conditions determine the structure of living organisms and their
spread and interaction with the environment. These conditionsexert a
particularly strong influence on the energy processesthat sustain plants
and animals.

Nearly all the primary productivity of living organisms in the

contemporary epoch is determined by photosynthesiscarried out by
autotrophic plants (chemosynthesissuppliesan insignificantly smaller
quantity of organic matter than doesphotosynthesis).
The general pattern of energy transformations in living organisms
can be represented in the form of an energy pyramid. Energy assimilated by autotrophic plants in the courseof photosynthesisis at the base
of this pyramid. This energy is found within the organic matter of
plants. A significant part of it is consumedin the vital activity of the
plants themselves,while someof it is consumedby the organismswhich
feed directly on the autotrophic plants.
The secondtier in the energy pyramid characterizesenergy influx
to plants and animals which use the energy foundin autotrophicplants.
Usually, no more than 10%of the original energy consumedis transmitted to the living organisms which utilize the energy of autotrophic
plants. These organisms also lose much energy in the courseof their
vital activity, in breathing, and in other functions.
The third stage of the pyramid involves the energy influx to living
organisms, which exist not by means of the primary energy of autotrophic plants, but by converting energy consumedby heterotrophic
organisms.This stageof the pyramid includes, in particular, carnivores.
Many living organismsare simultaneously found at the secondand
third stagesof the energy pyramid. This is the casewith man, who uses
some of the energy directly created by plants and some of the energy
created by animals.

5 CLIMATE

AND

THE

EVOLUTION

OF

LIVING

ORGANISMS

133

The energy pyramid reflects the closerelation between animals and

plants. This relation, basedon the fact that plants are the only sourceof
the energy that sustainsthe vital activity of all types of animals, exerts
a profound influence on the body structure of animals and on their
ecology.
Energy resourcescreated by the activity of autotrophic plants define the energy base animals may exploit in different natural regions, so
that the geographiclocation of animals and the size of their populations
are functions of the energy resources.
Since the productivity of plants is conditioned to a significant
degreeby climatic conditions,these conditionsare also a decisivefactor
in many zoogeographiclaws.
All living organisms are highly dependent on climatic resources.
The question therefore naturally arises whether organisms are able to
vary climatic conditions and make them more favorable for their existence.

A comparisonof the energycharacteristicsof climatic and biological

processespersuades us that this is not so (Budyko, 1971). Such a
comparisondemonstratesthat living organisms exploit a small part of
the energy involved in climatic processes.The base of the energy pyramid characterizingthe part of the influx of solar energy usedby plants is
much smaller than the energy involved in atmospheric processes.Typically, not more than tenths of a percent-in rare cases, several percent-of the energy of solar radiation is used in the biological cycle.
Consequently, living organisms cannot substantially vary the energy conditionsof climatic processes.Animals and especially plants
nevertheless affect the climate in a particular way. Such effects are
basically limited to local climate and do not extend to the climate of
large areas. Thus climate exerts a vast influence on the activity of
animals and plants, whereas the inverse relations between living organismsand climate are comparativelyweak. This feedbackhowever,is of
someimportance for explaining contemporaryclimatic changesand for
estimating possibleclimatic changesin the future; we will discusssuch
changesin detail in succeedingchapters.
The influence of climate on living organismscan be also illustrated
by consideringthe range of conditionswithin which life exists on the
earth.

The zone of propagation of living organisms was consideredby

Vernadskii, who studiedthe biosphere,including the atmospheric,hydrospheric,and lithosphericlayers within which biologicalprocesses
have developed(Vernadskii, 1926).ThoughVernadskii'sconceptionwas
highly fruitful, we shouldnot forget that exaggeratedconceptions
as

134

CLIMATIC

CHANGES

the dimensions of the climatic zone within which life exists on earth and

related conceptionsas to the dimensions of the biosphere are often

encounteredin particular in popular literature. These conceptionsare
based on the idea that life pervadesall regions on the surfaceof the
globe, in the atmosphere,and in the hydrosphere.
In speaking of the zone within which living organismsare found,
we shouldnote that organismshave existedpermanently in more or less

significantnumberswithin a rather restrictedspace.Life is basically

concentratednear the surfaceof the earth, though even here extensive

high-latitude regions as well as a number of desertregionscan be found
where the amount of living organic matter per unit area is either
negligible or simply zero. Living organisms are found there only by
accident, and the basis for sustained life does not occur.
One Irther feature in the spread of living organismsis usually
neglected.Life doesnot constitutean independentsystemin all areas of
the globe, i.e. living organic matter cannot be created in all regionsin
which living organismsare found. It is well known that organicmatter
evolves only where autotrophic plants are found (disregardingthose
modesof organicmatter productionwithout any significantrole in the
general balance of the biomass).
However, there are extensiveareas as well as a number of regions
with long seasonsin which the photosynthesisprocessis absent. Autotrophic plants cannot exist in many continental regions due to inadequate heat or moisture. The photosynthesizingactivity of plants is
restricted on a significant part of the ocean surface because of the
absenceof minerals, so that these zones are nearly barren.
Thus the energy pyramid of life clearly cannotbe sustainedover the
entire surface of the globe. Even after billions of years of evolution,
living organismsare still not able to adapt to many of the climatesof our
planet.
The contemporaryclimate is less favorableto biologicalprocesses
than those climates that have prevailed over the past hundreds of
millions of years.
As we noted in the precedingchapter, the decreasein the amount of
carbondioxide in the atmospherehas led to the result that its concentra-

tion is nowsomewhatlessthan that whichwouldguaranteethe highest

productivity of autotrophicplants. In addition, a significant part of the

earth's surface is either constantlyor temporarily coveredby snow and
ice at the present time, and photosynthesisunder these conditionsis
either absent or significantly diminished.
Thus the base of the energy pyramid, which representsthe total
mass of living organisms,is noticeably reducedin the contemporary
epochin comparisonwith earlier periodsin the history of the

5 CLIMATE

AND

THE

EVOLUTION

OF

LIVING

ORGANISMS

135

Although living organisms possesssufficient plasticity to adapt

somewhat to less favorable climatic conditions, the fact that extensive
zonescan be found on the surface of the earth where biological processes

occur slowly indicates that plants and animals are limited in their
ability to adapt to changesin atmosphericconditions.
It is therefore obvious that the laws of climatic change are of
enormousvalue for explaining the conditionsunder which the biosphere
has developedin the past and for estimating how it will evolve in the
future.

Ecological Energy Systems. Similar trophic relations unify animal and plant populations into elementary energy systemsthat are part
of a global biological energy system. Any elementary energy systemis
characterizedby a set of relations among the organisms comprising it
and between these organisms and the environmental conditions.
Let us considera highly schematizedmodel of an ecologicalsystem
containing four groups of living organisms:
a) autotrophic plants; b) herbivorousanimals; c) predators;d) parasites utilizing the biomassof the first three groups of organisms.
Following Lotka and Volterra, we will compile balance equations
for living organic matter virtually coincidingwith energy balance equations.

We let B denote the biomassof a particular group of organismsper

unit area occupiedand P their productivity (i.e. the gain of biomass per
unit time, lessthat consumedby their life activities); D will refer to the
consumptionof biomassby parasites. The conceptof productivity used
here correspondsto the algebraic sum of all forms of gain and loss of
living energy matter, except for those forms of loss expressedby separate terms in equations (5.1)-(5.3).
The biomassbalance equationsfor autotrophicplants have the form
dBa
dt

= Pa - bB - Da;

(5.1)

- P-

(5.2)

for herbivores

dB
dt

bc-

D;

and for predators

dBc
dt

- Pc- Dc;

(5.3)

where b and bc are coefficientsrepresenting the consumption of biomass by organisms that herbivores and predators subsist on per

136

CLIMATIC

CHANGES

mass of these groups of animals. The derivatives in the left sides of

equations (5.1)-(5.3) define the rate at which biomass varies over time.

Sincebband bccanbeconsidered
constantto a first approximation,
nine variablesoccurin equations(5.1)-(5.3).Consequently
six addi-

tional relations are required to solve them. These relations can be

compiledfrom empiricaldata for differentecological

systems.
We may usethe followingsimplifiedrelationsfor a widerangeof
systems:

Pa =
P, =
Pc =
Da =
Dr, =
D =

Pa(Ba,B,),
P,(B,,Ba, Be),
Pc(Bc,B,),
Da(B.),
D,(Bt,),
D(Bc).

(5.4)
(5.5)
(5.6)
(5.7)
(5.8)
(5.9)

The nature of theserelationshas beenstudiedin manyempirical

investigations.

The productivityof an autotrophicplant coverP, at first incre,es

slowlyas its biomassB, grows(the surfaceof leavesis still small),then
increasesmore rapidly, and again slowlywhen the biomassreachesa
high densityas a result of the decreasein photosynthesis
that occurs
whenthe lowerstrataof leavesare significantlyshaded.The P,(B,)
dependence
is represented
by an S-shapedcurvewithin a particular
rangeof biomassincrease.ProductivityP beginsto decrease
at high
biomass,
whenthe additionallossforbreathingis notentirelycompensatedby photosynthesis.
The dependence
P(B,)is determinedby the
factthat herbivores
oftenincreasethe productivityof plantsby accelerating the growthof plants,someof this incrementbeingconsumed
by
animals,

The productivityP of herbivoresis usually connectedto their

biomassBbby a dependence
qualitativelyanalogous
to Pa(Ba),sincethe
ratio of productivityto biomassdecreases
at very low and very high
animal populationper unit of area, chieflydue to a decrease
in birth
rate. Evidently,P dependsonplant biomass
B,, increasingasthe latter
increases.
It is affectedin a similarway for a givenB by biomass
Bcof
predators,who mainly eliminatesick and old animals,facilitatingan
increase in the birth rate.

The productivity
Pcof predators
varieswith increasing
biomass
B
in a way similarto P(B),Pcincreasingwith increasingbiomass
P.
Theuseofplantandanimalbiomass
byparasitesis usuallyproportional to biomass,thoughin a numberof casesthis dependence
is
strongerthan direct proportionality,sincean increasein population
densitytendsto spreadparasitesfromonelivingorganismto

5 CLIMATE

AND

THE

EVOLUTION

OF LIVING

ORGANISMS

137

If we represent equations (5.4)-(5.9) quantitatively, we may study

the laws of population dynamics for organisms within an ecological
system.

A number of conclusionsregarding the conditionsof many ecological systemsmay be obtained as a result of such an analysis.
1. There is often more than one solution of these equationsfor timeindependentvalues of the biomassof each group of organisms (stationary state of ecological system), which indicates that there may be
different biomass values for the componentsof an ecological system.
However, not all these solutionscorrespondto stable ecologicalsystems,
sothat small deviationsin the biomassof one groupof organismsfrom
its value for the steady state results in a return to the steady state.
2. An ecologicalsystemhas somestability relative to deviationsin
the biomassof each of its componentsfrom the biomassin the caseof the

steadystate. This stability is determinedby the interval within which

the biomass of the system componentsvaries and within which the
steadystate of an ecologicalsystemis restored.Restorationof the steady
state doesnot occurfor biomassvariations exceedingthis interval, and
one or several componentsof the system are destroyed.
3. The stability of ecologicalsystemsthat include all these components usually exceedsthe stability of systemsin which some groups of
animals are absent.

Since trophic interrelations (as well as nonenergy ecologicalrelations) between the componentsof ecologicalsystems confer a definite
tabtlity on these systems, such systemsexhibit features of completeness(a similar point of view has been previouslyexpressedby Elton in
1930).Ecologicalsystemsmay thus exert a substantial influence on the
evolutionary process.
Evolution of Ecological Systems. The population of each group
of living organisms within an ecologicalsystem constantly varies.
The variation in the number n of individuals of animals of a given
speciesover time t per unit area occupiedby them is determined by the
equation
dn/dt = an - fin,

(5.10)

where a and/] are the birth and death rates. If et and fl are constant,
letting n - N at the initial moment of time, we find from (5.10) that
n = Ne t("-).
(5.11)

When a > fl, the population of the animals will increase, while
when a </], it will decrease.

If a populationdecreases,
the animalsin it will becomeextinctsoon

after n attains the critical value n at which the low number of

CLIMATIC

138

CHANGES

no longer insurestheir further reproductionby multiplication.If the

populationincreases,an ecologicalcrisisis reachedwhen n attains the
critical value n2 correspondingto the population that can no longer
subsiston the foodresourcesof the given locality. In this case,destruction by the animals of their foodsourcesis possible,which leads to a
breakdownin the ecologicalsystemand threatens the existenceof the
speciesoccurring in it. 1
Bearing in mind these concepts,we find from equation (5.11) the
duration for the existence of an animal population in the first and
second cases:

t=--ln

,-3

'

t=--ln

,-3

n2

(5.12)

These equations can be rewritten in the form

gl = ,/3

In'n. g= I - I,/ InN

n-'

(5'13)

where g = 3t and g = 3tcorrespondto the number of generationsof

animals, renewed from the initial moment of time to the moment the
animals

die out.

Obviously, variations in the birth and death rates are unavoidable

under natural conditions, as a result of which their mean values over
long intervals of time will not exactly coincide. A difference in the
values of these coefficients

will

also occur due to variations

in external

conditions,whether favorable or unfavorable for a given animal species.

Equation (5.13) implies that any population may disappearcomparatively rapidly, dying out oncethe birth and death rates do not greatly
differ.

Let us bear in mind that ln(N/n)

and ln(n/N)

are as a rule less

than 10 under natural conditions (this assumption correspondsto the

possibiliWthat the number of animals per unit area occupiedby them
may vary nearly by a factor of 10"from the minimum to the maximum
values). Under these conditions a difference in the birth and death rates

of only 1% will lead to the extinction of the populationwithin a period of

time less than the lifetime of several hundred generations. Since such
' A breakdownin an ecologicalsystemin a given caseis to be understoodas
the destructionof oneor more principal componentsof this systemas a result of
external factors or variations in the relations between the principal components
of the system. The groupsof organismspreservedfollowing a breakdown in a
previousecologicalsystemcan form a new ecologicalsystemeither on their own
or in combination with organismsfrom other

5 CLIMATE

AND

THE

EVOLUTION

OF LIVING

ORGANISMS

139

an interval of time is somewhat less than the mean duration of existence

of speciesas determinedby the rate of evolution,it is obviousthat the

birth and death rates are regulated under actual conditions.Populations couldnot endurefor a long periodof time without this regulation.
Populationsare able to last for extendedperiodsof time becauseof
this dependenceof the birth and death rates on populationsize and
becauseof the interaction between organismsoccurring in ecological
systems.

The most important of suchrelations are presentedabovein the

form of equations(5.4)-(5.9). Theserelationscharacterizethe decrease
in productivityof excessively
large populations
due to foodshortages,
the increasein the death rate of animals of thesepopulationsas a result
of an increasein the number of parasitesand predators,etc. Inverse
laws affect the dynamicsof small populations.
As a result of the influenceof ecologicalfactors,the success
of living
organismsin the strugglefor existenceis governedto a significant
degreeby their ability to adaptto conditions
that sustainthe stabilityof
ecologicalsystems.

An animal speciesthat possesses

a number of advantagesin com-

parisonwith a species
that hasadaptedto a givenecological
systemcan
upsetthe equilibriumof this system,destroyit, and as a result,be in
danger of dying out.

For example, an increasein the size of predatorsthat pursue

herbivoresand an increasein the aggressiveness
of thesepredatorscan
leadto a destructionof their foodobjects,andthus to the extinctionof

the predatorsthemselves.
If herbivores
developeffectivemethodsof

defenseagainstpredators,their populationmaygrowandasa resultthe

plant coverof their habitat may be destroyed.
There is somebasisfor supposing
that the criterionthat determines
natural selectiondoesnot necessarilyimply an increasein the popula-

tion of a given speciesif this increasereducesthe stability of the

ecologicalsystem.

As we notedabove,all ecologicalsystemsare subjectto regular

fluctuationsof their plant and animalpopulations.
Thesefluctuations
can occurwhen environmentalfactorsvary or when theseconditionsare

constant(self-oscillatory
variationof population).As a result, unstable

ecological
systems
will havea reduced
life spanandtheir breakdown
will placethe populations
that compose
them in dangerof dying out,
Highly stableecological
systemshave a greaterchanceof maintaining
their populationsover a long periodof time.
Natural selectionthus tendsto preserveorganismswhose

140

CLIMATIC

CHANGES

increasesthe stability of ecologicalsystems.In addition, it supportsthe

existence of more stable systems and eliminates the less stable ones.
Numerical simulation of energy relations in ecological systems
leads us to concludethat many systemsattain a high stability at early
stages of their existence, after which the impetus for further development became highly restricted. An unusually large number of these
systems include representatives of ancient forms of organisms in the
open sea where environmental conditionshave always been more constant.

The rapid courseof the evolutionary processwhich can induce great

fluctuations in living organisms is less likely within stable ecological
systemsin more or less constant environmental conditions. Variations
of the organisms of these systems should correlate somewhat, which
significantly decreasesthe rate of evolution for each organism separately.
Nevertheless, even the most stable ecologicalsystemshave evolved
both as a result of fluctuations in the inorganic medium (for example,
due to fluctuations in the gaseouscompositionof the earth's atmosphere) and as a consequenceof purely biological factors. The rate of
evolution of such systemsis apparently minimal.
The evolutionary process in stable ecological systems is microevolutionary, i.e. it is primarily restricted to the readaptation of low
taxonomic groups of living organisms.
This process can also lead to a fluctuation of higher taxonomic
groupsover sufficiently long periodsof time; we are in agreement here
with the previously expressedpoint of view that no sharp boundary
exists between micro- and macro-evolution (Mayr, 1963). The rate at
which new taxonomic groups spring up, however, is usually comparatively low.
Fluctuations in the environment that disturb the stability of ecological systems have played a substantial role in accelerating the evolutionary process.If these disturbances are sufficiently great, they can
bring about the extinction of many groupsof living organismswhose
ecologicalniches then remain empty, sometimesfor a long time. Less
significant disturbancesin the stability of ecologicalsystemshave led to
the extinction of other organismswhoseecologicalniches were immediately occupiedto someextent by new organisms.In these cases,the rate
of evolution has increasedmore rapidly with greater fluctuations of the
environmental

conditions.

Let us consider an example that illustrates the influence of a

fluctuation

of external

conditions

on the succession of fauna.

South America became separated from the other continents

5 CLIMATE

AND THE EVOLUTION

OF LIVING

ORGANISMS

141

after the start of the Tertiary period, so that its fauna developedin
isolation for a long period of time. Members of two orders of herbivores,
the Litoptera and the Notoungulata, occupieda significant position in
this fauna. South America wasjoined to North America at the end of the
Pliocene, and as a result many mammals that significantly differed
from the animals in the ecologicalsystemsof South America were able
to reach the continent. This led to the rapid extinction of many speciesof

fauna of SouthAmerica,includingthe overwhelmingmajorityof species

of these orders. These specieswere replacedby tapirs, horses,deer,
mastodons,and other animals, which formedstable ecologicalsystems

with the predators that had arrived with them from North America
(puma, jaguar, bears, and others).
Over nearly the entire Tertiary period the herbivores of South
America comprisedecologicalsystemsthat includedpredatorsfrom the

orderof Marsupialia,whichwere apparentlylessactivethan the placental predatorsof North America.It wasonlynatural that the predatorsof
SouthAmericabecameextinctat the sametime as the animalsthey had

pursued.

The extinction of many speciesof widespreadanimals in South

America at the end of the Plioceneis generallyexplainedin terms of the
displacementof these animals by more advancedanimals from North
America (Colbert, 1958). This explanation of the extinction of these
animals appearsoversimplifiedfrom our point of view. A more comprehensiveexplanationcan be seen in the breakdownof the ecological
system of South America once its isolation was disturbed.

Using energy balance equations for ecological systems, it is in

principle possibleto quantitatively calculate the rate of extinction of
species and to determine what were the factors that resulted in the
extinction of the herbivores of South America, whether it was their
pursuit by new speciesof predators,trophiccompetitionwith herbivores

from North America, infectionby parasitesbroughtby the new groups

of animals, or a combination of all these factors.

Let us discussthe role that variationsin the sizeof living organisms

play in the courseof their evolution.We notedabovethat the energy
flux directedfrom autotrophicplants to different heterotrophicorganisms is the basis for the energy systemthat includesall living organisms of the biosphere.The energy flux gradually narrows as it travels
to new organismsdue to great energy lossesin each link of the trophic
chains. It is completelyexhaustedwhen the organic matter of living
creatures that complete the food chain is mineralized.
The structure of the energy flux is notoriously nonuniform. A

number of links of the trophic chains characterize energy of

142

CLIMATIC

CHANGES

matter concentrated in large organisms, while in other links it is

distributedin many small living bodies;energymay be transmittedin

bothdirections(for example,herbivorous
whalesfeedon planktonand
whale biomassis utilized by different microorganisms).
A highly importantevolutionaryfactoris that of the sizeof living
organisms.This determinestheir energylevel and the amountof energy

required to sustain their vital activity. It has been noted for sometime

that an increasein size can be seen in progressivelines of living

organisms (Cope'sLaw).

The role of sizein the ecologyof the groupplant coveris governed

by the preferenceof largeplantsfor the greatestamountof solarenergy
in comparisonwith small plants, which are often extensivelyshaded.
Large plants with long root systemsthat are part of the noncompact
plant coverof arid regionscanoftenobtainwater from deeperlayersof
the soil, while great succulentscollect a greater amount of water in
their tissuesin comparisonwith the surfacefrom which evaporation

occurs.Large plants usually live longer than small plants.

An increasein the bodysize of animals increasestheir resistanceto
changesin thermal conditions,decreases
the threat posedby encounters
with other animals, and in most casessubstantiallyprolongstheir
lifetime. In addition, in the caseof both plants and animals, an increase
in size means an increasein the energy required for their activity,
whichplacesa limit on the maximally possiblesizeof living organisms.
It shouldbe notedthat the trophicenergyobtainedby many plants
andanimalsfromthe environmentis proportional(to a first approximation) to their surfacearea, whereasthe energyconsumedin activity is
proportionalto their mass. Sincethese energy fluxes are equal under
averageconditions,it is clearthat the rate at whichenergyarrivesfrom

the environmentperunit surfaceof an organismwill increasein propor-

tion to the size of the organism.

A variation in the energyinflux requiredby living organismsnot
only limits the maximal size of organisms,but also establishesthe
optimalsizeof eachorganismfor the ecological
nicheit occupies,
sothat

the difficultiesin maintainingthe energybalanceis compensated

by the
advantagescreatedby an increasein the sizeof the organism.

We should note that a group of organismswith identical biomass

per unit area, but including organismsof different sizes,is characterized
by differingsensitivitiesof eachorganismto a changein environmental
conditions.

Let us considerthis questionfor two animal populationsprovided

with the same foodenergy base, but significantlydiffering in size and
lifetime. The productivity of the populationof smaller animals will

5 ClAMATE

AND

THE

EVOLUTION

OF

LIVING

ORGANISMS

143

somewhatgreater than the productivity of the population of the larger

animals under such conditions,due, in particular, to the higher rate of
metabolism of the first group of animals. Calculation using population
dynamics equations demonstrate that an increase in the death rate
correspondingto the same loss in biomassfor small and large animals
leads to much more rapid extinction of larger animals than of smaller
animals (cf. Chapter 7).
Thus larger animals with a low biomass-reproductioncoefficient
that possessa number of advantagesover small animals under stable
environmental conditionsare in greater danger when their death rate
increasesdue to changing environmental conditions.
For this reason many large speciesof organisms in our example
vanished,these speciesbeing the mostvulnerable componentsof ecological systemswhosestability has been disturbed. This probably explains
why ecologicalniches for larger organismsare more often occupiedby
new, more advancedforms of organisms,whereasmany small organisms maintain their ecologicalniches even when environmental conditions significantly change.
5.2 Climatic

Factors

in Evolution

Climatic Changes. A change in the composition of the atmospherehas been an important factor in the evolutionof living organisms.
The enrichment of the atmosphereby biogenicoxygen that occurredin
the Proterozoicled to an increasein the rate of metabolic processesand
made possiblethe developmentof more complexforms of living beings.
A subsequentchange in the content of carbon dioxide in the air decisively affected the evolution of plants.
As we noted in the precedingchapter, the decreasedconcentration
of atmospheric carbon dioxide that prevailed at the end of the Phanerozoic must have led, other conditionsbeing equal, to a decreasein the
mass of autotrophic plants, and consequently,to a decreasein the mass
of all living organisms. Under these conditionsthe evolution of photosynthesizing plants would have accelerated the rate of assimilation of
carbon dioxide per unit weight of the plants. Such a change in autotrophic plants can retard the processby which the mass of living
organismsdecreasesas the concentrationof atmospheric carbon dioxide
decreases.

We should, however, note that this is difficult to prove using data

from observations on photosynthesis in contemporary plants from
groupsof more and lessancient origin. Available data apparently do not
indicate a significant differencein the rate of photosynthesis

144

CLIMATIC

CHANGES

these groups. However, since these observationsrefer to contemporary

plant forms, they cannot characterize the laws of photosynthesisin the
distant past.
Fluctuations in thermal conditions exerted a significant influence
on the evolutionary processin the Cenozoic and particularly in the
Pleistocene.These fluctuations, as indicated above, led to a temperature

dropin all regionsbf the globeand to a particularly noticeabledecrease

in the middle and high latitudes. The coolingprocessacceleratedtoward
the end of the Tertiary and resulted in a sharp temperature drop during
the ice ages of the Pleistocene.
The influence of climatic changes on the evolution of terrestrial
vertebrates was treated by Mathew (1915). By analyzing paleontological
data and data on the contemporary geographicdistribution of animals,
Mathew was led to concludethat climatic changes exert a substantial
influence on the evolution and dispersionof animals.
Mathew assumed that new animal forms appeared in the most
recent geologicalperiodsprincipally in the holarctic region, from which
they spread to all the continents. He concluded that the widespread
theory in zoogeographyregarding the previous existence of different
land "bridges" between the continents was unnecessary and that the
processby which animals have spread can be explained by retaining the
contemporaryshape of the continents and assuming that they may have
been uplifted and sunk within limits correspondingto the depth of the
continental

shelf zone.

Mathew acceptedChamberlain's hypothesisregarding the mechanism of climatic changes,which states that the increase and decreasein
the level of the continents

that has occurred from time to time has led to

changesin climatic conditions,and that coolingbegan in high latitudes

as the continents rose, and subsequentlyspread to low latitudes.
Such climatic changes, in the opinion of Mathew, accelerated the
evolution of animals and facilitated their spread from the holarctic
region into the tropic zone and into the continents of the southern
hemisphere.
It should be noted that Mathew's work is now chiefly of historical
interest.

Mathew was one of the first to opposethe previously popular theory

in biogeographyof "continentalbridges"which, in the opinionof investigators of the geographical spread of plants and animals, repeatedly
arose between different continents and then disappeared. While Mathew's views on this question have been subsequently accepted, his
claim that continental shifts did not occurin the past is not acceptedby
most geologistsat the present time. Chamberlain's theory (which

5 CLIMATE

AND

THE

EVOLUTION

OF LIVING

ORGANISMS

145

thew accepted)regarding the laws of climatic changes,is clearly outdated. Finally, the theory that the holarctic region was the chief center
for the genesisof speciesin the most recent geologicalperiodsis also
oversimplified in the light of currently available data.
Works by Axelrod and Baily (Axelrod, 1967;Axelrod and Baily,
1968) are some of the most recent studies of the influence of climatic

changeson the evolutionof living organisms.Axelrod and Baily in the

latter work consideredwhy the dinosaursbecameextinct toward the end

of the Cretaceous.

In their opinion,continentaluplifting occurredin this period,which

led to the drying up of many shallow inland seas. As a result, the air
temperature in middle and high latitudes significantly decreasedand
the temperature changesintensified over the courseof the year.
Axelrod and Baily surmised that the increase in the range of
environmental temperature fluctuations to which the dinosaurs were
unable to adapt due to their previousevolutionary development,was the
chief cause of their

extinction.

In his first work, Axelrod proposeda similar explanation for the

well-knownextinctionof many large mammalsduring the Pleistocene.
Axelrod assumedthat the influence of coldwinters in high and middle
latitudes on the survival of animal calveswas particularly substantial
and led to the extinction of a number of mammalian speciesin the ice
ages.

Without enumerating other studies of the influence of climatic

changeson the evolutionof living organisms,let us note that different
authors often disagree as to the mechanism of this influence. For
example,there exist numeroushypothesesconcerningthe causesfor the
extinction of the dinosaurs. Some authors have proposedthat an increase in cosmicradiation occurred simultaneously with the extinction
of the dinosaurs,or that the air temperaturerosebeyondthe limits that
the dinosaurscould endure, which stimulated the activity of mammals
who destroyeddinosaureggs.To a greater or lesserextent, noneof these
hypothesesare truly justified, and the problem remains unsolved. A
number of suggestionsregardinga possiblemethodfor dealing with this
problem are presented below.
Just as many difficulties arise in explaining why many large animals became extinct

in the Pleistocene.

It is well known that particularly significant changesin the fauna

occurred toward the end of the Pleistocene with the end of the Wfirm

glaciation. At this time the compositionof the animal world outside the
tropicregionssubstantiallychanged.Many previouslywidespreadlarge
herbivoresdied out in Europe: these included the mammoth,

146

CLIMATIC

CHANGES

rhinoceros, European steppe bison, and gigantic deer as well as a

number of large predators, including the cave lion and cave bear.
Substantial fluctuations in the spreadof a number of animals that have
been preserved also occurred;for example the large flocks of reindeer
that had inhabited Western and Central Europe disappeared.
Still more significant fluctuations in the animal world occurredat
this time in America, where about 30 speciesof large animals became
extinct.

Other hypotheses besides the decisive influences of climatic

changeshave been given as to the possiblecausesof the extinction of
large mammals toward the end of the Quaternary.
Different authors even in the nineteenth century expressedthe
opinion that primitive man may have exterminated the mammoth and a
number of other speciesof animals. This point of view has been maintained in the contemporary literature by Pidoplichko (1963), who connects it, however, to the debatable conceptionof antiglacialism.
Much data confirming the significant role of man's activity in the
extinction

of animals

at the

end of the

Pleistocene

can be found

in

studiesby Martin (1958, 1966, 1967),who considersthat primitive man

decisivelycontributedto the disappearanceof many animals not only in
the middle, but also in the tropic latitudes.
In one of our works (Budyko, 1967) a numerical model for the
biomassbalance of animals pursued by Upper Paleolithic hunters was
used to explain why many large mammals became extinct toward the
end of the Pleistocene.

The results of these calculations (which are set forth in Chapter 7)

demonstrate that hunting for large animals could have made these
animals extinct within a period of time not exceedingthe duration of the
Upper Paleolithic.
A similar calculation has sincebeen carried out by Martin (1973)to
explain why many large mammals becameextinct toward the end of the
Pleistoceneon the North and South American continents. Using simplified balance equations to describethe biomassof pursued animals, he
came to the conclusionthat the migration of a comparatively small
number of hunting tribes (not more than a thousand men) from Northeast Asia into North America in the Upper Paleolithic couldhave led to
the rapid extermination of all large animals in the region where these
tribes originally settled, after which the hunters would begin to migrate
to new, still unused territories. In the course of this migration, the
hunting tribes pursuing the herds of large herbivores that still remained would move along a wide front from north to south, increasing
their numbers. Martin found that Upper Paleolithic hunters could

5 CLIMATE

AND

THE

EVOLUTION

OF

LIVING

ORGANISMS

147

reached the southern extremity of South America within a comparatively short period of time (about a thousandyears), destroyingtens of
animal speciesin their wake.
Much attention must be paid to the influenceof climatic changeson
ecologicalsystemsfrom the point of view that the rate of the evolutionary processdependson the stability of ecologicalsystems.
It is well known that plants and animals are distributed zonally
particular geographiczonescomprisingecologicalsystemsthat are similar in many respects.
A closerelation between geographiczonesand climatic conditions
had been established far back by V. V. Dokuchaev. This relation was
subsequentlystudied by Grigor'ev and the author, who compileda table
of geographiczonality (Grigor'ev and Budyko, 1956).
Let us present one version of this table (Table 15). The columnsof
the table correspond to different gradations in moisture conditions,
which are characterized by the aridity index (i.e. the ratio of the
average annual radiation balance of the surface of the earth to the heat
necessaryfor evaporating the annual sum of precipitation), while the
rows correspondto gradations in the annual values of the radiation
balance of the earth's

surface. The values of the radiation

balance used

in constructing this table are for moist surface conditions (Budyko,

1971).

Similar types of plant covers correspondto each column in the

table. Forest vegetation predominates at all latitudes under excess
moisture conditions, but not under extremely excessivemoisture conditions (aridity index below 2/5). In this caseforest vegetation is replaced
by tundra at higher latitudes and by swampsat low latitudes. (Since an
aridity index below 2/5 is observedover more or less extended areas only
a comparatively high latitudes, the swamp regions of low latitudes
cannot be consideredas independent geographiczones.)
Each gradation in moisture is characterized not only by the type,
but also by the productivity of the plant cover. In these works, Grigor'ev
and the author proposedthat the productivity of the natural plant cover
growsas the moisture conditionsapproachthe optimal state (for a given
radiation balance) and that productivity grows as the radiation balance
increases

under

fixed moisture

conditions.

A particular sequenceof changes in soil type with substantially

similar properties correspondsto each column of Table 15. In particular,
the following sequenceof soil types is characteristic for excessmoisture
conditions:tundra, ash-gray and brown forest soil, subtropical red and
yellow soil, tropic ash-gray red soil and laterite. The sequence for
moderately insufficient moisture is black and dark brown soil, black

148

5 CLIMATE

AND

THE

EVOLUTION

OF

LIVING

ORGANISMS

149

brown weakly leached subtropical soil, reddish brown subtropical soil,

and so on.

A special place in the table is occupiedby the region of perpetual

snow. This region is characterized by negative or near-zero radiation
balance, negative aridity index, and practical absenceof plant and soil
cover, and also by a subzoneof arctic desert with low annual radiation
balance and high moisture.
The geographiczonality regularities represented in the table can be
explained using the following considerations.
Since the earth is spherical, its surface can be divided into several
latitude zonesthat differ with respectto the radiant energy influx to the
earth's surface. Within each zone (other than the perpetual snow zone)
various moisture conditions can be observed, from excess to highly
insufficient moisture [i.e. different zonally-averaged values of evaporation minus precipitation, Editor]. Zones with similar moisture conditions in different

latitude

bands can be seen to have

similar

features

(which combine into distinct features due to the change in the radiant
energy influx). If we successivelycompareregions in two latitude bands
with increasing moisture (or increasing aridity), these similar features
will periodically repeat.
A comparison of the annual values of the radiation balance to the
mean air temperatures over the growing period showsthat these variables are closely related (Budyko, 1971).
Let us consider how geographic zonality conditions vary with climatic changes, for example, with those that occurred in the course of
Quaternary glacial development.
Though glaciers developedas a result of particular changes in
radiation conditions, as is evident from the data of Chapter 4, simple
calculations

demonstrate

that the radiation

balance of the earth's sur-

face outside the ice zone hardly varied as the ice cover shifted to low
latitudes. Since the air temperature in middle latitudes noticeably
decreasedduring the ice ages, it is evident that the present-day relation
betweenthe radiation balanceand the mean temperature of the growing
period was disturbed during these epochs.
Many geographic zones may be thought of as having gradually
spread to lower latitudes as air temperature decreased. This zone
spreading could also have occurred as a result of a variation in the
amount of precipitation, which increasedin someregions and decreased
in other regions as glaciers developed.
Climatic changesin the Quarternary period were slow enough to
allow not only animals, but also most plants to shift their geographic
ranges into regions with favorable climatic conditions;this is

CLIMATIC

150

CHANGES

larly evident from the noticeable change in the plant cover in high
latitudes that occurredas a result of the relatively very brief warming of
the 1920sand 1930s(Grigor'ev, 1956).
We should, however, bear in mind that the change in geographic
zonesthat accompaniedthe developmentof the glaciers was not limited
to their spatial shifting. A decreasein air temperature and more or less
constant radiation created new types of climatic conditions and, consequently, led to a change of the plant cover and of other componentsof
the landscapein these geographiczones.
Many examples are available in the contemporary epochof different landscapesin the same geographiczonesfor different combinations
of thermal

and radiation

conditions.

These differencesexist, for example, in regionswhere the air temperature in the warm part of the year is significantly lower than its
mean latitudinal values. In particular, a relatively low summer air
temperature in Sakhalin combineswith a high radiant energy influx.
As a result, the plant cover on Sakhalin is characterized by a number of
distinctivefeatures, suchas gigantic forms of many plants. Theseforms
are assumed to have appeared as a result of the combination of a
relatively high assimilationproductivitywith high radiation influx and
a comparatively low expenditure of organic matter in breathing due to
the low temperatures.
This example demonstratesthat the shifting of geographiczonesas
a result of climatic changeshas led to fluctuations in many natural
processesthat form intrazonal ecologicalsystems. In particular, new
types of ecologicalsystemsapparently aroseduring the last glaciation
over wide middle latitude areas of the northern hemispherethat were
not coveredby ice.
Limited fluctuations in ecologicalsystemsdue to the migration and
partial rearrangement of the structure of geographiczonesmay have
been accompaniedby sharp fluctuations in these systems,when climatic
changeswould have led to the total disappearanceof certain isolated
regions of these zones.This may also have occurredon islands, where
climatic changes may have resulted in the replacement of forest by
tundra or tundra by forest, on coastlines,in regionsadjacentto extended
mountain areas, etc. Intrazonal ecologicalsystemsmay have been disturbed under these conditions,and different types of plants and animals
whose areas were limited to the correspondingparts of the geographic
zone became extinct.

It is, however, likely that the influence of comparatively slow

climatic changes on ecologicalsystems led chiefly to their gradual
rearrangement, thus accelerating the evolution of living

5 CLIMATE

AND

THE

EVOLUTION

OF LIVING

ORGANISMS

151

This conclusionhas been confirmed by Kurten (1965), who established that the rate of evolution of mammals in the Pleistocenesignificantly exceededtheir rate of evolution in the Tertiary.
Short-Term Climatic Changes. The environmental temperature
range within which different plants and animals can carry on their vital
activity is an important factor in limiting the climatic zone of life.
All plants exist within definite temperature ranges. At low temperatures plants are harmed by ice crystalsforming in their tissues.These
crystals break down cell walls both directly and by forcing water into
tissues as they freeze.
The size of the temperature drop that plants can tolerate varies
within a wide limit for different plants. It also dependson the rate at
which temperature falls (plants tolerate slow coolingmore easily than
rapid cooling), and on the degree of hardiness of the given plant, i.e. on
the coolingit has previously experienced.
Temperature increasesabove certain limits also harm plants, for
example by destroying chlorophyll and blighting leaves, which lead to
the death of the plant if the temperature increasesfurther. The breakdownof different protein compoundsin plant tissuesapparently plays a
substantial role in the mechanism through which overheating harms
plants.
Resistanceto the effects of overheating differs for various plants;
succulentdesert plants endure the highest temperatures without being
harmed.

Many studies have establishedthat the rate of photosynthesisand

the productivity of autotrophic plants substantially dependon temperature, productivity usually approaching zero at high and low temperatures.

Thus, thermal factors exert a profound and decisive influence on

the vital activity of plants. The geographicdistribution of plants, their
seasonal changes, and the speciescompositionand productivity of the
plant cover thus depend on the thermal conditions.
The influence of temperature on the vital activity of animals is no
less substantial than on plants. All animals can be divided into two
groupsdependingon their thermal conditions,the cold-bloodedanimals,
whose body temperature varies within wide limits, and the warmbloodedanimals, whose body temperature is maintained at an approximately constant level. The intermediate group of heterothermal animals, which can regulate the temperature of their body to a limited
extent, is sometimesadded to these two groups.
The body temperature of cold-bloodedanimals usually does not
coincidewith the temperature of the environment becauseof

CLIMATIC

152

CHANGES

processes
resulting in the release of heat. However, the heat production
of theseanimals is, as a rule, insignificant,sothat thesedifferencesare
usually comparativelylow. Exceptionsto this rule related to momentsof
high activity by the animal are possible.For example, the bodytemperature of many insectssharply increasesduring flight.
There exist other factorswhoseeffectcan facilitate a changein the
temperature of cold-bloodedanimals relative to the environment. This
temperature can decreasedue to hat loss from evaporation from the
body surface of the animal and due to long-wave radiation into the
surrounding space. The absorptionof solar radiation can lead to an
increase in body temperature.
Fluctuations in body temperature for large cold-bloodedanimals
(for example, somereptiles) can lag noticeablybehind fluctuationsin
the temperature of the environment becauseof the thermal inertia of
the animal's body.
The warm-bloodedanimals include two higher classes(birds and
mammals) that have developedan effectivemechanismin the courseof
evolutionfor maintaining a constantbodytemperature independentof
external

conditions.

This mechanismis controlledby a heat regulation center, the

hypothalamus, found in one sectionof the brain. As the external factors

of the thermal conditionsvary, this center correspondinglystimulates

the physicaland chemicalheat regulation systems,which allows the

animal to maintain a constantbodytemperature.The wide variability
of metabolism,which increasessharplyas temperaturedecreases,is the
principalmeansof temperatureregulation.Fluctuationsin evaporation

from the body surface, the state of the hair coat, etc. are other means of
temperature regulation in many animals.
Body temperature differsfor different warm-bloodedanimals. It is
usually between 40C and 42C for birds and often between 36C and
39Cfor mammals. The constancyof the body temperature of warmbloodedanimals is relative, and it oftenfluctuatessomewhatduring the
day with variations in muscular tension, as a function of feedingconditions and other factors. An animal's body temperature can fluctuate
noticeably with external environmental conditions that are anomalous
for the particular warm-bloodedanimal.
Both cold and warm-bloodedanimals carry out their vital activity
within a given range of bodytemperature, limited by what is called the

"lethal" temperature.This temperaturemay differ significantlyfor the

same speciesof animal as a function of the animal's acclimatization, i.e.

on its prior exposureto increased or decreasedtemperatures.

The lethal temperature of cold-bloodedanimals is highly varied.
Some animals die when their body temperature fluctuates

5 CLIMATE

AND

THE

EVOLUTION

OF LIVING

ORGANISMS

153

tively slightly, while otherstolerate temperaturefluctuationsof up to

tens of degrees without harm.
The differencebetween the upper and lower lethal temperatures for
warm-blooded animals often is 15-25C. The lower lethal temperature

for a human being with normal body temperature of about 37C is

between 24-25C, and the upper lethal temperature is 43-44C.
The biological functions of both cold-bloodedand warm-blooded
animals vary significantly as their body temperature fluctuates. The
vital activity of cold-bloodedanimals is highly susceptibleto the influenceof temperature,suchactivity beingpossibleonly within a fixed and
sometimesquite narrow temperature range. Thus the behavior of coldblooded animals is often determined by the need of maintaining an
optimal body temperature. For example, many insects, reptiles, and
other animalswill moveto areasilluminated by the sun and assumea
positionproviding the greatest radiant heating of their body in the case
of insufficient heat. These animals will leave the surface of the earth

when it becomesextraordinarily heated by the sun, either burrowing or

climbing up the branchesof plants where the air temperature is below
the temperature of the earth's surface.We shouldparticularly note the

important effectsof thermal conditionson the breedingprocessof coldbloodedanimals, which is governed in many ways by their ecologyin
the correspondingperiod.
Though the ecologyof warm-bloodedanimals dependsto a lesser
extent on the temperature conditions, under unfavorable conditions
these animals will also use the microclimatological features of the
landscapein order to bring their bodytemperature to an optimal level.
These actions are particularly necessaryin the case of small animals whosebodiespossessinsignificanttemperature inertia. In addition, they coolvery easily at low environmentaltemperaturesdue to the
high ratio of body surface, through which heat is lost, to body mass,
which determines heat production as a result of metabolism.
The substantial influence of temperature factors on the vital activity of animals makes the thermal conditions of the environment an
important factor of the geographicdistribution of animals and their
ecology.Obviously, the geographicdistribution of different plants and
animals dependssubstantially not only on the mean climatic conditions,
but also on deviations

of weather

conditions from their norm.

Data from meteorologicalobservationsdemonstrate that the air

temperature at the surface of the earth may differ by tens of degrees
from their mean monthly value for a given region at different moments
of time. Great temperature anomalies(particularly negative anomalies)
often lead to systematic destruction of plants and animals.
The influenceof great weather anomalieson the number of

CLIMATIC

154

CHANGES

was noted as far back as 1859by Charles Darwin, who describeda sharp
decreasein the number of birds following a severewinter in England. In
his book,The Origin of Species,"Darwin wrote (pp. 68-69), The action
of climate seemsat first sight to be quite independentof the strugglefor
existence;but in sofar as climate chiefly acts in reducing food,it brings
on the most severe struggle between the individuals, whether of the
same or of distinct species,which subsiston the samekind of food.Even
when climate, for exampleextreme cold,actsdirectly, it will be the least

vigorousindividuals,or thosewhich have got least foodthrough the

advancingwinter, which will suffermost."It is clear from this remark
of Darwin that systematicdestructionof animals during highly anomalous weather can occurlong beforethe meteorologicalelementsreach
their

lethal

values.

It has been established in climatological studies that great anoma-

lies in meteorologicalelements often encompassareas extending over

thousandsof squarekilometers. In many casesthe areas of the anomalies exceedthe areas occupiedby different plants and animals. Thus
sufficientlygreat anomaliesin weather conditionscan lead to the extinction of several species.

Since the size of a probable anomaly in air temperature, like a

number of other meteorological elements, increases with increasing
periodof time, it is obviousthat fluctuationsin climaticconditionscould
have influencedliving organismsin a very substantialway over the
long epochsof the earth's geologicalhistory.
The influence of short-term climatic changeson the evolution of

living organismshas been consideredin one of our previousworks

(Budyko, 1971), where it was noted that the evolutionary processmay
have dependednot only on slow climatic changes,but also on shortterm, sharp fluctuations in climatic conditions.
This possibilityis a consequence
of the enormoustime scalesof
nature's developmentin the geologicalpast, scalesbeyondthose that
can be directly observed by man.
In the precedingchapterwe consideredshort-termclimatic changes
causedby volcanic activity. These changesvery probably exerted a
definite influence on the progressionof fauna.
It is well known from paleontologicalstudies that the processby

which higher taxonomic groups of animals (orders, suborders,and

families) became extinct is characterized by the fact that biologically
distinct groupsof animals that occupiedentirely different ecological

niches often all became extinct within short (from the standpoint of

geologicalhistory) time intervals. Statistical conceptsimply that this

phenomenoncannot have been a result of the struggle for existence
under

invariant

external

5 CLIMATE

AND

THE

EVOLUTION

OF LIVING

ORGANISMS

155

As an example let us considerthe well-known simultaneous extinction of the large groups of animals towards the end of the Cretaceous
period. At this time, five of the 10 existing orders of reptiles vanished,
including the dry-land dinosaurs (two orders), the pterosaur, and also
ichthyosaur and Plesiosaurus.These five orders included 35 families
containing a great number of species(Colbert, 1965).
Let us assume that the mean duration of existence of the reptile
orders that became extinct in the Cretaceous was roughly 100 million
years, and let us also assumethat the time at which an order disappeared can be determined to within 5 million years.
The probability that five out of 10 orders will randomly die out
within 5 million years is roughly one in 10,000 as calculated using
Poisson'sformula. Such an extremely small probability indicates that
the extinction of these groupsof reptiles was not a random coincidence,
but was due to definite

fluctuations

in nature.

In view of the paleontologicaldata, it is easy to verify that the

higher taxonomic subdivisionsof the animal world in most casesdid not
disappear one by one, but in comparatively large groups.In particular,
mostof the ordersand subordersof reptiles and amphibiansthat became
extinct vanished within three critical epochsof geologicalhistory, toward the end of the Permian, the Triassic, and Cretaceous. Calculation

shows that such a coincidencewould have been highly unlikely for

constant or slowly varying external conditions, even if these critical
epochslasted several million years.
It is as well impossibleto explain the disappearanceof many large
groupsof animalsby the strugglefor existenceunder invariant external
conditionsbecauseof a number of other qualitative conceptsbesidesthe
above calculations.

it has been repeatedly noted in the paleontologicalliterature that

mammals occupiedthe ecologicalnichesof the reptiles of the Mesozoic,
not competing with these reptiles, but only following their extinction,
when the correspondingniches were free.
We may recall from a number of other examples of this type the
interesting case of the extinction of the phytosaurs at the end of the
Triassic. These rptiles suddenly disappeared,freeing the corresponding
ecologicalniche, which was appropriated after sometime by the crocodiles who had sprungfrom the sameinitial forms as the phytosaursand
who were extremely similar to them. It would be quite difficult to
explain such examples without taking into account fluctuations in
natural

conditions.

Thus to understandthe causesfor great fluctuationsin fauna in the

geologicalpast, it is necessaryto take into accountthe effective

156

CLIMATIC

CHANGES

associatedwith sharp fluctuations in environmental conditions as well

as the influence of purely biological factors.
The enormous effect that may have been exerted by short-term
temperature drops on different forms of organic life cannot be doubted.
As we noted above, all living organisms without exception are characterized by a definite temperature zone within which they exist, which in
many casesis very narrow. A sharp decreasein temperature lasting for
several years or tens of years will lead to the destruction of many species
of plants and animals, primarily those habituated to climatic conditions
near the lower boundary of their temperature zone. While a decreasein
temperature over the entire planet would allow various species to
continue living in the warmest regions, many plants and animals would
re-establish their previous geographicranges oncea temperature reduction ceased.But if cooling of the warmest regions was sufficiently great,
a given speciesor group of specieswould completely vanish.
We should note that the actual methods by which sharp temperature dropsaffect the vital activity of different organisms vary widely.
For every organism we may establish the size of a temperature drop
that may (a) lead directly to its destruction; (b) decrease its activity
down to the limits at which it perishes in the course of its struggle for
existence; (c) decrease the resistance of the organism to infectious
illnesses, as a result of which it becomes their inevitable victim; (d)

disturb the breeding process.

The temperature drops correspondingto these conditions may differ, though even the least of them will be sufficient to cause the
extinction of some species of animals.
The extinction of the reptiles and amphibians toward the end of the
Triassic, which was outstanding in its scale, and the still greater
extinction of reptiles toward the end of the Cretaceous, are the most
interesting simultaneous disappearances of large groups of animals
from the geological past.
Before passingto a discussionof these two milestonesin geological
history, we must aboveall emphasizethat both occurredduring a period
of intensive orogenesis.This has been repeatedly noted in the geological
literature (Strakhov, 1937).

The probability that a large number of volcanoeswill simultaneously erupt is extremely high during a period of active orogenesis.We
should bear in mind that the significant general intensification of
volcanism was accompaniednot only by an increase in the probability of
a particularly great temperature drop, but also by an increase in the
probability of a seriesof such drops, which could have been grouped over
hundreds of thousandsor millions of

5 CLIMATE

AND

THE

EVOLUTION

OF LIVING

ORGANISMS

157

It is well known that one important feature of the reptiles and

amphibians was their lack of temperature regulation, which made these
animals particularly susceptible to thermal conditions.
The climate of the Mesozoic was characterized by a weakly expressed thermal zonality, a higher temperature in the low latitudes
than is the casein the contemporary epoch, and a much higher temperature outside the tropic regions, which created favorable conditions for
the existence of cold-bloodedanimals lacking temperature regulation
over the entire

earth.

A different situation held towards the end of the Paleozoic, when

the air temperature over extensive areas of the earth's surface decreased
from time to time, which apparently created difficulties for cold-blooded
animals. Peculiar features developed by some reptiles of the Permian
(dimetrodon, edaphosaur) illustrate this in an interesting way. These
reptiles carried a large crown on their spine, which, in the opinion of
some paleontologists(Romer, 1945), was used to heat their body in the
case of low temperatures. By positioning this crown perpendicular to
solar rays, these reptiles were able to receive an additional amount of
heat sufficient for further physical activity.
The absence of such an adaptation in Mesozoic reptiles is one
indication of a comparatively warm climate at all latitudes during this
epoch. Furthermore, sudden and significant temperature drops must
have exerted a pronouncedeffect on cold-bloodedanimals of that time.
Reptiles that still exist are the vestiges of a rich and varied fauna
from the Mesozoic

era which

have survived

after

the extinction

of the

overwhelming majority of the large (and many small) forms of this

fauna. There are groundsfor assumingthat contemporaryreptiles have
a greater resistanceto temperature drops than those which have died
out, sothat their ecologymay provide someidea of the upper limit of the
resistance of reptiles to cold.
The comparatively small number of reptiles whose habitat is poleward of the subtropicalzone are active only in the warm time of the year
and protect themselves in the cold periods by different methods. Many
reptiles have typical daily rhythms of activity which usually cease at
the coldesthours of the day.
Works by Bogert (1959) and his colleagues,who have studied the
ecologyof reptiles, established that reptile activity is enormously dependent on the necessity of maintaining a fixed body temperature.
When heat is insufficient, reptiles will attempt to get warm using solar
rays or the heat from the earth's surface. When cooled, some lizards
change their color in order to absorbmore solar radiation.
We may reliably conclude, using data from laboratory and

158

CLIMATIC

CHANGES

observations,that a decreasein the mean temperatures in the habitat of

many reptiles of about 10will lead to their destruction.The ecologyof
one animal belonging to the most ancient of the still existing orders of
reptiles (Rhynchocephalia),is thus of particular interest. The only
species occurring in this order is the tuatara, which now inhabits
several islands in New Zealand. The existence of this animal under the
cold climatic conditions of New Zealand is sometimes considered evi-

dencefor the comparativelyhigh resistanceof ancientreptilesto low

temperatures (Bellairs and Garrington, 1966).

However,sucha conclusion
is basedon an incorrectunderstanding
of the climate of New Zealand. It is a pronouncedmaritime climate
distinguishedchieflyby the absenceof any significantchillsduringthe
year, an extremely favorable factor for cold-bloodedanimals.

Recent studieshave establishedthat the embryo in the tuatara's

egg matures very slowly and that this processcontinuesfor 15 months;
the embryoceasesdevelopmentduringthe coldesttime ofthe year. Such
a long period greatly increasesthe risk of destruction of the tuatara's
eggsby other animals, which demonstrateshow difficult it is for ancient
types of reptiles to survive at insufficiently high temperatures.Evi-

dently, evena rather smalltemperaturedropwill make it impossible

for
the tuatara's eggs to develop and will lead to the extinction of this
species.

It may be that a sharp temperature drop could have affectedthe

cold-bloodedanimals of the Mesozoic.However, it is now difficult to
establishthe precisecriteria governingthe sensitivity of different, now
fossilizedreptiles to cold and causingsome forms to becomeextinct
while preservingothers.We can only assumethat a temperaturedrop
will have been particularly destructiveto large animals and also to
oceanic species,which could not have sheltered themselves in warmer
spotsat moments of the greatest chills, as is done by contemporary

reptiles.

This hypothesisreasonablyexplainsthe nearly total extinction of

the large land reptilesas well as bothordersof marine reptilestoward

the end of the Upper Cretaceous.

Though the comparatively small forms of land reptiles had a
greater chanceof survivingbecauseof their ability to find shelter,only
in limited, very warm regions could several groups of reptiles have
survived following sharp chills.
It is entirely obviousthat sharp chills were far lessdangerousto
animals with temperature regulation, i.e. mammals and birds. These
animals are known to have successfully
enduredthe transition from the
Cretaceousto the Tertiary and have sinceoccupiedall the ecological
niches of the reptiles that had meanwhile become

5 CLIMATE

AND

THE

EVOLUTION

OF

LIVING

ORGANISMS

159

The degree to which the extinction of the reptiles at critical points

in their history was in fact simultaneousis of someinterest. Though the
analysis of available paleontologicaldata encountersgreat difficulties
due to the unreliability of dating different deposits,it appearsthat the
different speciesof reptiles becameextinct toward the end of the Upper
Cretaceous within comparatively close but not coinciding periods of
time. This conclusioncan be made to agree with theories of the influence
of sharp, short-term cooling on the extinction of reptiles if we bear in
mind the probability of a seriesof short-term temperature dropsduring
periods of increased volcanism. The first drops would destroy animals
least adapted to cold, while succeedingdrops would lead to the extinction of other groupsweakened by the earlier coolings.
Let us now discussthe degree to which this theory allows us to
overcome the difficulties mentioned in some studies in explaining the
extinction of Mesozoicreptiles. We will considerhere an idea of Colbert,
who set forth the modern viewpoint on this question in th concluding
sectionof his book, The Age of Reptiles" (Colbert, 1965).
In discussingthe possibility for the influence of catastrophic fluctuations in external conditionson the extinction of reptiles, Colbert (pp.
203-204) wrote, Did some great event take place that wiped out these
reptiles? If so, why should a great catastrophe be so selective-why
shouldit have causedthe extinction of five ordersof reptiles and allowed
the continuation of five other orders? Why should all families and
genera and species belonging to these five orders have disappeared
completely?Moreover, if it was a physical catastrophe of somesort, why
is it not recorded in the rocks?The continuity of sediments to be seen in
certain localities as Cretaceous beds are succeededby Tertiary beds
indicates an orderly world in those distant days, like the orderly world
we know. And indeed this is an expressionof the principle of uniformitarianism, so important to the scienceof geology. From all of the records
of the rocks, the world around, there is ample reasonto think that earth
processesand life processesin the past were as they are today; that
events of past ages must be explained by the forceswhich we see acting
in Nature at the present time. There is no place for world-wide catastrophes in the world of the past or of the present if th principle of
uniformitarianism has any validity."
Colbert

further

noted that the Cretaceous

ended in the Laramide

Revolution," a term referring to the formation of a number of contemporary mountain systems. He assumedthat this change in natural conditions was quite slow and that reptiles could have adapted to it in the
courseof biologicalevolution. Colbert proposedthat the difference in the
climate of the Upper and Lower Cretaceousperiodswas extremely small
and could not have causedthe extinction of the reptiles. In

CLIMATIC

160

CHANGES

the possibleimportanceof a temperature changein the transition from

the Cretaceousto the Tertiary, Colbert was chiefly drawn to the hypothesisthat temperaturesharply increasedat this time. He indicatedthat
it was difficult to make this hypothesisagree with the fact that some
groupsof reptilesbecameextinct while otherswere preserved.Colbert
particularly doubtedthat a temperature changecouldhave influenced
the extinction of marine animals, since the ocean possessesa high
thermal

inertia.

Colbertalsorejectedsuchhypothesesas competitionwith mammals

ableto destroythe reptile eggs,the exposureof the reptilesto radiation,
etc. He deducedthat the extinction of the reptiles was causedby a
combination of complexphenomena whose true nature will probably
never be explained.
From our point of view, sucha pessimisticconlcusioncan be considered justified only on the basis of paleontologicaldata and without
resorting to methodsand data from related disciplines.Recourseto data
of paleoclimatologyallows us to overcomethe difficulties indicated by
Colbert and to answer the questionshe posed.
In particular, this theory implies that sharp temperature drops
whose destructive

influence on cold-blooded animals

is doubtless must

have occurredduring a period of time comparableto the duration of the

Mesozoic era. These drops were most likely in the period of active
volcanism, when the extinction of the reptiles also occurred. Calculations showthat the upper layers of the oceanwere significantly chilled
during these coolings,and this led to the destructionof marine speciesof
reptiles.
Becauseof their short-term nature such coolingscould not leave
direct traces in geologicaldeposits,though their effect on the organic
world would have been catastrophicin the full senseof the word.
As we noted above, the qualitative laws governingthe extinction
process,accordingto which only land and chiefly small forms of reptiles
were preserved,reasonablyagree with the likely consequences
of shortterm cooling. The extinction of all members of some orders and the
preservation of some members of other orders could reflect to a certain
degreethe different resistanceof membersof different ordersto cooling.
The direct relation between the evolutionary origin of reptiles in different orders and the influence of the temperature conditionson them is
quite natural.
Thus the largest fluctuationsin the stockof living organismsapparently occurredas environmental conditionschangedsharply. The example of suchchangesgiven abovefar from exhauststheir possiblemechanisms, which require further

5 CLIMATE

AND

THE

EVOLUTION

OF

LIVING

ORGANISMS

161

Fluctuations in environmental conditions, and particularly great

changesin these conditions,could have significantly increased the rate
of evolution.

This conclusionagrees with Simpson'sviewpoint (Simpson, 1944),

who noted that the extinction of ancient groups of animals often coincided with periods of mountain formation. He wrote, ... among reptiles and mammals the proportion of orders that seem to have arisen
during times of pronouncedemergenceand orogeny is greater than can
reasonably be ascribed to chance. This is probably true of birds,
also..." (p. 113). Simpsonalso tried to draw attention to the decreased
rate of evolution in tropical latitudes, where external conditionswere
always more stable.
In his work, Simpson isolated a specific type of the evolutionary
processhe called quantum evolution."
In quantum evolution, the higher taxonomic units, families, orders
and classesare replaced, and the rate of variation of living organismsin
this case attains the highest value. We must assume that the mechanism of quantum evolution is of great importance when ecological
systemsbecomeunstable, that is when their organisms, in the courseof
experiencing rapid changes, partly die out and partly become components of new ecological systems.
Thus climatic changes have exerted a definite influence on the
evolutionary process,this influence apparently being particularly significant in epochswhen short-term sharp climatic changes intensified.
5.3 Climatic

Factors

in Human

Evolution

Human Evolution. The hypothesis that man is descendedfrom

ape-like ancestors was expressedeven in the eighteenth century. I.
Kant advocatedthis hypothesisand it was supportedby J. B. Lamarck
at the beginning of the nineteenth century.
The works by C. Darwin and A. Wallace, publishedat the end of the
1850sand containing a theory of the origin of the speciesby means of
natural selection of the most adapted organisms, were of great importance for the explanation of the origin of man. Though the problem of
human evolution was not considered in these works (Darwin

limited

himself to the remark that his theory will shed new light upon human
evolution), it immediately became obvious that the Darwin-Wallace
theory could be used to explain human evolution.
As early as 1860, a famousdisputebroke out between T. Huxley and
S. Wilberforce in which Huxley defended the hypothesis that

162

CLIMATIC

CHANGES

originated from ape-like animals as a result of the processof natural

selection. Works by T. Huxley, E. Haeckel, and K. Vogt and other
authors that developedthis point of view were published in the 1860s.
A somewhat different position was taken by Wallace. Though he
expressedviewpointssimilar to thoseof Huxley in his first work dealing
with human origin (Wallace, 1864), he came to the conclusion after
several years that natural selection could not have been the only basis
for the origin of man (Wallace, 1869). In the opinion of Wallace, the
intellectual abilities of man would only slightly surpassthe abilities of
the higher apes as a result of natural selection. He noted that the
development of the human brain significantly exceededthe requirements of human existenceat earlier stagesof human society.In order to
account for this discrepancy,Wallace allowed for the intervention of a
"higher intelligent will" in the processin which man originated.
It is worth noting that, though Wallace was interested in the
spiritualism much in style during the secondhalf of the nineteenth
century, the hypothesis that nonmaterialistic factors influenced the
origin of man was the only such assumption taken in all his many
biological studies (George, 1964).
Darwin's opinion regarding human evolution was expressedin his
book, "The Descent of Man and Selection in Relation to Sex," published
in 1871. This book contained voluminous data demonstrating that man
arosefrom ape-like ancestorssimilar in several features to a number of
contemporary higher apes. Darwin suggestedthat the likely spot at
which man appeared was Africa, where there now exist two apes most
similar to man, the chimpanzee and the gorilla (the same opinion was
expressedby Wallace).
In his book Darwin only fieetingly referred to Wallace's concepts
regarding the difficulty of explaining the development of the human
brain and indicated that he disagreedwith these theories.
This lack of agreement was apparently based on a comparison
Darwin carried out between the intellectual activity of man and animals. In fact, he found much in common between them. As a result of
this comparison,Darwin was led to concludethat there was no qualitative difference between animal and human intelligence. Such a conclusion can hardly be consideredjustified from the standpoint of contemporary concepts.

It is worth noting the conclusionwith which Darwin ended his

book. "We must, however, acknowledge,as it seemsto me, that man,
with all his noble qualities ... with his god-like intellect which has
penetrated into the movements and constitution of the solar systemwith all these exalted powers-Man still bears in his bodily frame the
indelible stamp of his lowly origin" (Darwin,

5 CLIMATE

AND

THE

EVOLUTION

OF

LIVING

ORGANISMS

163

It is clear from this sentencethat Darwin recognizedto somedegree

the fundamental

differences

between

man and the most advanced

ani-

mal forms.

Though Wallace outlived Darwin by several decadesand witnessed

many new advancesin the study of the origin of man, he nevertheless
preservedhis opinions on this question to the end of his life.
The difference of opinion between Wallace and Darwin on such an
important question is of some significance in view of the scientific
authority Wallace held and the fact that he highly valued Darwin's
studiesand only in rare casesexpressedhimself against his conclusions.
Nevertheless it was quickly forgotten, probably as a result of Wallace's
clearly unacceptable hypothesis that nonmaterialistic factors affected
the evolutionary process.
Only very little paleontologicaldata illuminating the origin of man
were available in the middle of the nineteenth century. Many outstanding discoverieshave been made in this field over the last hundred years,
the most important of them being the discovery of the remains of
Australopithecus in southern and eastern Africa as well as the remains
of Pithecanthropus and beings related to him (Homo erectus) in Java,
China, and Africa.
Data from paleontological studies have confirmed the DarwinWallace viewpoint that man arose in Africa, though many features of
Australopithecus, who was similar to the ancient ancestors of man,
differ from contemporaryhigher apes.Australopithecuswas to a significant degree a carnivore, and hunted different animals using the simplest tools. Thus the use of animal meat, which is characteristic of
modern man, has a very ancient origin.
Australopithecus existed toward the end of the Pliocene and
throughout the Quaternary, i.e. within an interval of time from several
million years to several hundredsof thousandsof years beforeour era.
Members of the speciesHomo erectus, who occupiedan intermediate
positionbetweenAustralopithecusand man in many respects,lived for
several hundred thousand years in the Pleistocene.
Modern man (Homo sapienssapiens)appeared about 40,000 years
ago, though it is assumedthat modern man is descendedfrom Neanderthal man (Homo sapiens neanderthalesis), who existed for a longer
period of time and vanished approximately when modern man appeared. Since later forms of Neanderthal man differ from modernman
to a greater extent than do earlier forms, the origin of modern man from
later Neanderthal man has often been subject to doubt.
The greater part of time during which modern man has existedis
encompassedby the first culture he created, the culture of the Upper
Paleolithic. During this period hunting for large animals was the

164

CLIMATIC

CHANGES

nomic foundation of life in human society. It was concludedin the

studies whose results are set forth in Chapter 7 that the culture of the
Upper Paleolithic ended in an ecological crisis encompassinggreat
areas, due to the destruction by Upper Paleolithic hunters of the animals they pursued.
As we noted above, this crisis was a highly unusual phenomenonin
the history of the biosphere,since ecologicalsystemsas a rule regulate
the population of organisms that are part of particular food chains,
making rapid extinction of speciesof plants and animals impossible.
Disturbance of this regulation indicates that the emergenceof modern
man constituted a definite boundary, and in passingthis boundary man
was no longer governedby the biologicallaws that determine the size of
animal populations.
The extermination of many large animals apparently led to a
reduction in the population of several extensive regions up to the time
when hunting and gathering of useful plants began to be replacedby
cattle breeding and agriculture, the latter occurring in the Neolithic
epoch, about 10,000 years ago. Thus the period during which man has
producedthe foodhe uses amounts to only about one-fourth of the total
duration

of his existence.

Industrial civilization developed within an even briefer period of

time-a fraction of a per cent of all human history-while the scientific
and technical revolution takes up only somehundredths of a per cent of
human history.
Though available paleontological and archaeological data illuminate many stages in human evolution, the mechanism of this evolution
requires further study.
One difficulty in explaining this mechanism was understoodby
Darwin, who wrote in his conclusionregarding the origin of man from
ape-like ancestors, that %he high standard of our intellectual powers
and moral disposition is the greatest difficulty which presents itself,
after we have been driven to this conclusion on the origin of man"
(Darwin, 1871, p. 609). To overcome this difficulty it is necessaryto
clarify the question posedby Wallace over one hundred years ago.
It is well known that natural selectiongenerally preserveschanges
in organisms that are useful for their vital activity and which are
employed in the courseof their existence. However, it appears that the
vast abilities of the human brain have not been utilized to any great
extent throughout most of the entire history of modern man.
There are groundsfor believing that the central nervous system of
modern man has scarcely changed since the Upper Paleolithic.

5 CLIMATE

AND

THE

EVOLUTION

OF LIVING

ORGANISMS

165

paintings of animals in caves of Western Europe, which are often

consideredthe greatest achievements of animal painting over all of
human history, is one proof of this fact.
The impression that an enormousgap exists between the abilities of
modern man in the pursuit of intellectual activity and the use of these
abilities at previous stagesof cultural development arose from observations on transmitting the achievements of contemporary civilization to
members of the most backward tribes, some at the cultural level of the

StoneAge. Since they easily enter modern industrial civilization if they

receive the proper education from childhood, it follows that their intellectual abilities were not usedto any great extent under the previously
existing conditions.
It is highly likely that the greatest achievementsof moderncivilization are still much below the level modern man could reach were he to

completelyuse his intellectual abilities. This possibility further widens

the gap between man's potential abilities and their use throughout his
history.
An explanation of this gap within the framework of the ordinary
theory of evolution by means of natural selectionof the most adapted
organisms encounters a number of difficulties. Though many animals
master

certain

forms of behavior

as a result

of domestication

that

are

not natural to them in the wild state, there exist no analogies for this
gap in the animal world. The hypothesisthat socomplexand perfect an
organas the brain of modernman couldhave appearedas a result of preadaptation, i.e. to somedegreeby accident, is clearly unlikely.
Thus an explanation of the origin of man requires that we take into
account the effect of particular factors that led to man's qualitative
differencefrom all animals. The nature of thesefactorswarrants special
attention.

The Role of Climatic Changes. We may suppose,following the

viewpoint of a number of researchers (Gerasimov, 1970), that climatic

changesthat occurredin the Plioceneand Pleistoceneexerted a particular influenceon human evolution.As we notedin the precedingchapter,
in this time coolingdevelopedin middle and high latitudes, which led to
significant climatic changesover all the continents.It is highly likely
that these climatic

fluctuations

accelerated

human

evolution

as well as

the evolution of many other organisms.

Any quantitative estimate of the rate of human evolution meets
with special difficulties. According to some data (Kurten, 1965), the
mean duration of existence of European speciesof mammals in the
Pleistoceneamountedto aboutonemillion years. This value is

CLIMATIC

166

CHANGES

ble with the duration of human-like speciesin the Pleistocene.Thus the

rate at which speciesthat were ancestorsof man were replaced doesnot
appear to be very high.
However, we should assume that such an estimate of the rate of

evolution is quite tentative.

The acceptedsystem of classificationof living organismsdoesnot
reflect their ecologicalfeatures to any great extent. It is well known that
this system is basedmainly on morphologicaland phylogeneticprinciples. However, there are many examples of ecologicallysimilar organisms that are highly distinct in origin. In addition, closely related
organisms often possessentirely different interrelations with the environment.

The assertion that the higher apes are morphologicallymore similar to man than to the lower apes was presented as one argument
favoring this theory during the first studiesof the origin of man based
on the idea of natural

selection.

It seems that the difference between modern man and Neanderthal

man was greater from the ecologicalpoint of view than that between
Neanderthal man and many mammalsthat donot belongto the orderof
primates. The basis for this difference is man's vast influence on the
environment, which manifesteditself even during the Upper Paleolithic

(when man exterminated tens of speciesof large animals) and has

reached an enormousscale in recent times. Though Neanderthal man
lasted roughly ten times as long as has modernman, he did not disturb
the balanced ecologyof the biosphereand in this respectwas far more
similar

to other mammals

than to modern man.

There is now a tendency to reduce the rank of taxonomic groups

that divide man from the primates related to him. This tendency,which
is entirely justified from the standpoint of the principles by means of
which the classificationof living organismsis constructed,amplifies to
an even greater extent the lack of correspondencebetween this classification and the differencesbetween the ecologicalproperties of man and
his ancestors.

We must conclude,bearing in mind the scaleof this difference,that

changesin human ecologyoccurredmuch more rapidly than the mean
rate of change of the ecologyof different animals.
This processmay have been due to the developmentof the human
central nervous system, whose elements are not taken into account to
any great extent within the classificationof man and beingsrelated to

him.

Available

data demonstrate

that the size of the brain of man's

ancestorsnearly tripled over two million years (Kurten, 1971). At the

same time, the brain structurebecamemore

5 CLIMATE

AND

THE

EVOLUTION

OF LIVING

ORGANISMS

167

One cause for such a rapid development of the central nervous

systemmay possiblybe the critical circumstancesof man's predecessors
in the tropics during the Pliocene and Pleistocene, when noticeable
climatic changesoccurred.These changeswere a consequenceof cooling
at high latitudes, which led to an increase in the mean meridional
gradient of the air temperature. As a result, the atmosphericcirculation
system changed and the high pressure belt shifted to lower latitude.
Since little atmosphericprecipitation falls within this belt, the moisture
conditionsvaried over large tropical regions, which led to the replacement of tropical forests by savannas and semidesertsin a number of
regions.

There exist groundsfor supposingthat prior to this climatic change

man's ancestorsinhabited forests and hardly differed from other higher
primates exceptfor a more developederect posture.It is highly probable
that, like all primates, they were basically herbivores.
The direct ancestorsof Australopithecus (who may be provisionally
called pre-Australopithecus)were morphologicallysimilar to Australopithecus and were comparatively small (the body weight of ancient
Australopithecus was at most 30 kg). The rate of travel of erect preAustralopithecuswas much less than for most four-legged animals of
that size, and they also lacked suchorgansof attack and defenseas large
canine teeth and claws.

Pre-Australopithecuswas almost defenselessin opencountry conditions, where the possibility of concealment in trees was often absent,
even when attacked by small members of the many predators of the
African savannas. In addition, the disappearance of many edible plants
he fed on in the tropical forestssignificantly limited his supply of plant
food.

The low birth rate of pre-Australopithecus, which is characteristic

for the higher primates, and the great length of time for the development of young, who were a particularly easy catch for predators pursuing pre-Australopithecus, made it considerablydifficult to maintain preAustralopithecus populations.
The chief factors which could have insured the survival of preAustralopithecusincludedtheir comparativelyhighly developedbrain
and their erect posture, which completelyfreed their hands.
Pre-Australopithecuswas exposedto conditionsof exceptionally
intense natural selection, which led to a high rate of evolution of this
small population.Such an evolutionaryprocess,called quantum evolution by Simpson, doesnot usually leave traces in the paleontological
recordsdue to the small number of animals involved (Simpson, 1944).
Paleontologicalstudieshave shownthat Australopithecus

CLIMATIC

168

CHANGES

the technique of applying the simplest tools for defense and attack.
Using these tools, Australopithecussucceededin attacking such large
apes as the baboon and, possibly, different hoofed animals. As a result,
they were able to catch a significant amount of animal food.
A number of discoveriesof Australopithecus remains demonstrates
that their populationswere not too small and that, consequently,they
were able to overcome the crisis situation

discussed above.

It is worth noting that groupsof comparativelysmall Australopithecus were in a difficult situation when attacked by large predators
even if they used such tools as stones, sticks, and bones of hoofed
animals, It was no lessdifficult for the slowly moving Australopithecus
to catch any rapidly moving animal. Their successin the struggle for
existence could have come about only as a result of the significant
advantage providedby their intellectual abilities in comparisonwith all
the animals who attacked them or which could have been their prey.
Natural selectionin the courseof the evolutionof Australopithecuswas
therefore enormously favorable to the development of a brain, which
made possiblea gradually increasing successagainst predatorsand in
hunting for different animals.
An increase in the intellectual abilities of Australopithecusmade
him more capable of using different objectsas tools and made possiblea
transition to the productionof tools.It was of enormousimportancethat
Australopithecusfound it necessaryto engage in collective actions for
self-defense
and when attackingother animals.Elementsof socialorganization subsequently arose out of these collective actions, which led to
the developmentof moral conceptsto define the behavior of the individual in a collective.The importanceof the labor processand the development

of social relations

for human

evolution

was clarified

in a well-

known work by F. Engels. (F. Engels, The role of labor in the processby
which apesbecame man, The Dialectics of Nature (in Russian), Moscow,
1950,pp. 132-144).Engels also drew attention to the important role of
the processin which man's ancestors switched to animal food, which
significantly acceleratedman's physical development.
Paleontologicaldata indicate that some lines of Australopithecus
attained enormoussize in the courseof their evolution. This significantly decreasedthe pressureof natural selectionon the evolutionof the
nervous system and led to a loss of the habits of tool use. These lines
came to an evolutionary dead end and their members died out.
Lines of man's ancestorscharacterizedby progressivedevelopment
of the central nervous system existed alongsidesuch regressivelines.
This developmentwas probably not uniform. It acceleratedwhen

5 CLIMATE

AND

THE

EVOLUTION

OF

LIVING

ORGANISMS

169

pressure of natural selection on the populations of man's ancestors

intensified. It is highly likely that such epochscoincidedwith fluctuations in natural circumstancesand with the developmentof Quaternary
glaciations.
Since climatic fluctuations encompassedenormous areas in the
course of the glaciers' advance and recession,these fluctuations could
have significantly varied the conditions under which man's ancestors
lived, which set new problemsbefore them and acceleratedtheir evolution.

The last great stagein human developmentapparently took place

during the Wfirm, when Homo sapienssapiens appeared.
These conceptionsdescribethe role of climatic changesin the origin
of man, though they are not sufficient for decidingwhy the human brain
developedbeyond the limits determined by the needs of primitive society.

From our point of view, this problem can be explained if we take

ino account the discrepancy between the crude means available to
man's ancestorsand those tasks it was necessaryfor them to solve.
Until recently hunting for large animals and protection against
predators involved significant difficulties and risk even for civilized
peopleswith firearms. These difficulties were incomparably great for
man's ancestors,who used primitive toolsfor hunting, and particularly
for distant ancestors, who were significantly smaller and weaker than
modern

man.

Nevertheless, man's ancestorsinevitably encounteredlarge predators. Dangerous circumstancesarose also when man's ancestorsgathered animal food, that is, when only large animals could supply a
sufficient

amount

of meat for more or less numerous

tribes.

Undoubtedly, man's ancestors as a rule were the victors in such

encounters. This is confirmed, in particular, by discoveriesof the remains of such enormousanimals as elephants, mammoths, and cave
bears at campsites of ancient man.
Man's ancestors, when equipped with primitive tools, were able to
successfullyfight large animals only becauseof highly developedspecial
habits far exceedingthe habits that highly equipped modern man finds
necessaryfor successfulhunting.
A paradoxical situation arises. The lessperfectedwere the hunting
tools available to man's ancestors,the greater the degree to which they
compensatedfor the disadvantages of these tools by a high level of
intellectual activity, which permitted them to get the better of encounters with large animals. Therefore, enormousdemands on the

170

CLIMATIC

CHANGES

tual developmentof man's predecessorswere apparent in the distant

past, which led to the crystallizing of progressivechangesin the development of the brain by means of natural selection.
On the basis of this theory, it appearsthat human evolution accelerated in those regionswhere hunting for large animals was the chief
source of food. Such conditions prevailed, in particular, in the vast
forestlessregionsadjacentto ice coversin the ice ages.It is possiblethat
modern man emergedin the middle latitude region.
Modern man couldhave appearedunder ecologicalconditionsthat
placed very stringent demands on his intellectual activity both for
successfullycombatinglarge and dangerousanimals, and, possibly,in
order to comeout the victor in competitionwith related primates.
Following the emergenceof the subspeciesHomo sapienssapiens,
man's material culture progressedmore rapidly in regions where the
natural conditionsdid not provide him with plant food, but required
hunting for large animals.
Members of this subspecies,which populated all the continents,
appearedin a number of casesin zoneswhere foodgathering was aided
by the availability of easily accessibleedible plants (humid tropics and
subtropics)or of animals that could not strongly resist hunters. The
latter caseprobably occurredwhen Australia was colonizedby modern
man. Marsupials were an easy prey for the experiencedhunters of the
Upper Paleolithic. Nearly the same thing may have happenedwhen
America was first colonized,the various animals there never having
comeinto into conflictwith man and lacking protective instincts against
hunters (Martin, 1973). Under these conditions,cultural development
was retarded and previously attained advances may have sometimes
been entirely lost.
These theories imply that enormousreservesof intellectual activity
were freed as material culture, which insured man's struggle for existence, developed.Such reserves, when stimulated by complexenvironmental conditions, could be used to further increase the level of material
culture. Thus culturespontaneously
developed,and at somestagemade
possiblethe emergenceof the first elementsof civilization. The development of social relations in human society was of fundamental importance in this process.
This progresswas highly nonuniformin spaceand time. Whenever
environmental conditionsgave rise to great stresseson human society,
the useof the potential intellectual abilities of man reacheda high level.
This level significantly decreasedunder the comparatively favorable
conditionswhich were far more frequent.
These thoughts can explain the difficulties involved in the explanation of the origin of man and pointed out by Darwin and

6 Man's

Influence

on Climate

6.1 Changes in Local Climate

Influence on Plant Cover. The influenceof man on climate began

several thousandyears ago as a result of the developmentof agriculture.
Forest vegetation was destroyed in many regions in order to cultivate
the land, which led to an increase in the wind speed at the earth's
surface, a change in the temperature and moisture conditions in the
lower layer of the atmosphere, and also to a change in the soil moisture,
evaporation, and river discharge. Destruction of forests in comparatively arid regions was often accompaniedby an increase in dust storms
and destruction of the soil cover, which significantly changed the natural conditions

in these areas.

In addition, the destruction of forests even over extended areas

apparently exerted a limited influence on large-scale meteorological
processes.

A decreasein the irregularity of the earth's surface and a fluctuation in evaporation on areas free of forest somewhat varied the precipitation regime, though this variation was comparatively low if the
forests were replaced by other types of plant cover (Drozdov and Grigor'eva, 1963).
Complete destruction of the plant cover in an area, which repeatedly occurred in the past as a result of man's agricultural activity,
exerted a more substantial influence on precipitation. These caseswere
particularly frequent following the felling of forestsin mountain regions
with a weakly developed soil cover. Under such conditions erosion
rapidly destroyedsoil not protectedby the forest, as a result of which the
further existence of a mature plant cover became impossible. Similar
circumstancesarose in some regions of dry steppes,where the natural
plant cover, destroyed as a result of unlimited grazing of agricultural
animals, was not restored; these regions became deserts.
Since areas of the earth's surface lacking a plant cover are intensively heated by solar radiation, the relative humidity over these areas
falls, which increases the level of condensation and decreases the

172

CLIMATIC

CHANGES

amount of precipitation. It is possiblethat such a mechanismis of

importancefor explainingcasesin which the natural vegetationin arid
regionswas not renewedafter being destroyedby man.
Planting of forests is also accompaniedby changes in weather
conditionswhich, however, are basicallylimited to the lowestlayers of
the atmosphere.Protectiveforeststrips exert the greatestinfluenceof
all the differenttypesof forestplantingson climate and are widely used
as means of reclamation.

Forest strips are usually in the form of plantationsfrom several

meters to several tens of meters wide, which border square or rectangu-

lar fields measuringfrom severalhundred metersto betweenone and

two kilometers.

Forest strips are often grown in drought-strickenregionswhere

they are usedto maintain more favorablemoistureconditionsin farm
lands.

The wind-breaking effect of forest strips, resulting from the decreaseof the mean wind speedin fields betweenthe strips and from the
decreasein the rate of turbulent exchangenear the earth's surface, is
the principalway in which they influencethe meteorological
regime of
the lowest air layer.
The attenuation of turbulent exchangein fields between the strips
is explainedby the fact that vorticesmovingnear the earth'ssurfaceare
brokenup to a significantextent after passingthrough the foreststrip.
As a consequence,
the vorticeswill lack large eddies,which substantially decreasesthe intensity of eddymotionsin this flow (Yudin, 1950).
It should be noted that such an effect is observedonly if the vortex
has passedthrough the foreststrip comparativelyunimpeded.A thick,
impenetrableforest strip acts in a completelydifferent manner. A
comparativelysmall calm zone is createdbehind such a strip, beyond
whichthe wind rapidly picksup speedand approaches
the conditionsof
the open steppe.In this casedecreasesin the size of the eddiesare not
observedin the lowest layer.
Thesephenomenaare explainedby the fact that an air stream rises
somewhatas it approachesan impenetrable forest strip, archesover it,
and then immediately descendsand returns roughly to its original state.
A decreasein the intensity of the turbulence in the lower air layer
over fieldsbetweenthe strips is of great practical importance.Data from
recent studiesshow that turbulence directly affectsthe developmentof
two meteorologicalphenomena,the blowing of snowfrom fields and the
formation of dust storms. A decrease in turbulent exchange near the
earth's surfacesafeguardsagainst or weakens dust stormsand helps to
maintain snow on farm lands between the

6 MAN'S

INFLUENCE

ON

CLIMATE

173

A decreasein turbulent exchange is also of substantial importance

for preservingmoisture in the soil during the warm part of the year.
The amount of possibleevaporation (evaporative power), as well as
a number of other meteorologicalfactors, dependson the rate of turbulent exchangein the lowest air layer. Previous calculations (Budyko,
1971) have demonstrated that evaporative power decreasesby roughly
10% when the mean values of the exchange coefficient in the lowest air
layer decreaseby 20%.
Besides decreasing evaporative power, protective strips help to
increase snow reserves on fields and increase the amount of precipitation. The influence of all these factors leads to a significant increase in
soil moisture in fields protectedby strips.
As noted above, numerous observationshave shown that the run-off
of snow waters significantly decreasesin fields with protective strips.
This decreaseis chiefly explained by the change in the distribution of
the snow cover on protected fields from that on unprotected fields.
Abatement of the wind speedand of turbulent exchangein the lowest air
layer in fields within the strips create conditionsfor a comparatively
uniform distribution of the snow cover, whereas a significant part of the
snow in open fields is blown away in gullies and other topographic
depressions,and after thawing simply drains away. Moreover the increased permeability coefficient of the soil under strips impedes to a
greater extent the snow melt in comparisonwith openfields, which also
decreasesthe spring run-off of snow water.
A fluctuation in the amount of precipitation due to a variation of
vertical velocities in the atmosphereand of evaporation may also exert
some influence on the water balance of soil protectedby strips.
We may use heat and water balance equations, taking into account
the values of the integral diffusion coefficients,run-off, and precipitation correspondingto the conditions of protective strips in order to
estimate the total influence of these hydrometeorologicalconditionson
the water balance of the soil (Budyko, 1956).
Calculations have demonstrated that soil moisture significantly
increases and evaporation somewhat increases over protected fields.
The increasein soil moisture dependson the seasonand the conditions
of turbulent exchange, run-off, and precipitation.
Since a system of afforestation strips, besidesimpeding run-off of
snow melt, decreasesturbulent exchange in summer and simultaneously increases precipitation, soil moisture increases not only at the
beginning of the planting season,but also during its secondhalf.
But if (as often happens) the influence of forest strips is chiefly
manifestedin an increaseof snowreservesand a decreaseof spring

174

CLIMATIC

CHANGES

off (winter effect"), soil moisture will increaseonly in spring and at the
beginning of summer.
The amount of productive moisture attained in the soil under these
conditionscan be several tens of per cent greater than in unprotected
fields, other conditionsbeing equal.
These,conclusionsare satisfactorily confirmed by data from observations in regions with mature penetrable forest strips.
A significant increase in the amount of free moisture in the soil and
an increase in total evaporation can substantially improve the yield of
agricultural crops under average climatic conditions. Yields may improve as a result of an increase in the productivity of plant transpiration
(which is aided also by a decreasein exchangeand wind speedin the
lowestair layer). In addition, the ratio of the amount of water consumed
in transpiration to total evaporation noticeably increaseswith increasing soil moisture, which also enhancesthe productivity of agricultural
cropsas total evaporation grows.
Thus the use of forest strips can significantly vary the water balance in the soil and noticeably increase productivity, which, as is well
known, is confirmedby data from numerousexperimental studies.
Forest strips, in addition to influencing the water balance in the
soil, also play an important role in abating dust storms which drasti-

cally damage the soil cover in arid regionsin some years. This was
clearly manifested in the 1968-69 winter when powerful dust storms
passedthrough the southernpart of the Europeanterritory of the USSR.
Inspection showed that the winter crops on protected fields suffered
significantly less damage than on other fields.
Forest strips are also used in regions of sufficient humidity, where
they increasethe mean temperature of the earth's surfaceduring the
warm part of the year by attenuating turbulent mixing. Under such
conditionsthey exert a favorable influence on the developmentof thermophilic crops and accelerate the maturation of many agricultural
plants (Gol'tsberg,1962).Without dwelling on other methodsof changing the meteorologicalregime, we should note that the effect of man on
the plant cover changes climatic conditions in the lowest air layer
within

certain

limits.

The Modification of the Water Regime. One way by which man

affects climate involves his use of artificial irrigation. Irrigation has
been utilized in arid regionsfor thousandsof years, beginning with the
most ancient civilizations in the Nile valley and the Tigris-Euphrates

interfiuvial

area.

Irrigation sharply varies the microclimate of irrigated fields. A

significant increasein heat lossby evaporation leads to a decrease

MAN'S

INFLUENCE

ON

CLIMATE

175

temperature and an increase in the relative humidity of the lower air

layer. This variation in the meteorological regime, however, rapidly
weakens beyondthe irrigated fields. Irrigation leads only to fluctuations
in the local climate and doesnot really affect large-scale meteorological
processes.

Let us discussin detail the physical mechanisms by which irrigation affects the meteorological regime (Budyko, 1956). The radiation
balance substantially increaseswhen irrigation is used in arid steppes,
semideserts, and deserts; this increase may reach several tens of per
cent of the initial

balance.

This change in the radiation balance can be explained, on the one

hand, by the increase in the amount of absorbedshort-wave radiation as
a consequenceof a decrease in albedo, since the albedo of moist soil
coveredby more or less abundant vegetation is significantly less than
that of a desert or semidesert. On the other hand, a decrease in the
temperature of the underlying surface and an increase in the humidity
of the lower air layer with irrigation decreasesthe effective radiation,
which

also increases

the radiation

balance.

Irrigation under arid climatic conditionsleads to a sharp increase in

heat lossesin evaporation, whose magnitude is governedprincipally by
the irrigation rates. An increase in the loss of heat in evaporation as a
rule exceedsthe increase in the radiation balance under ordinary irrigation rates. As a consequence,turbulent heat transfer sharply decreases
and becomesnegative, correspondingto the direction of the mean turbulent heat flux from the atmosphereto the underlying surface in the case
of sufficiently high irrigation rates. This is manifested in the appearance of daily temperature inversions.
Thus irrigation under arid climatic conditions significantly decreasesboth the turbulent heat flux (which can even change its direction), and the heat flux transmitted by long-wave radiation. This may
lead to sharp fluctuations in air mass transformations over a given
region when irrigating sufficiently large areas.
Fluctuations in the componentsof the heat balance in the case of
irrigation can be estimated from the observations at Pakhta-Aral (central Asia). Results of these observationsdepicted in Fig. 30 allow us to
compare the heat balance components of an irrigated oasis and its
surrounding semidesert for summer conditions.
The data in Fig. 30 demonstrate the radiation balance R noticeably
increased in the oasis in comparison with the semidesert regions and
that a large heat loss in evaporation from irrigated fields occurred
(evaporation over the investigated period of time was practically zero in
the semidesert region). Accordingly, the turbulent heat flux P in

CLIMATIC

176

Rcal/cm'

CHANGES

. min

LE
(b)

(a)

12

16

20 (hrs)

Fig. 30. Daily variationsof the heatbalancecomponents

in an irrigated

oasis (solid line) and semidesert(broken line). (a) Radiation balance; (b) heat

losson evaporation
(LE), turbulentheatflux (P), and heatexchange
in thesoil
(A).

desertregioncanbe seenfrom Fig. 30bto be muchgreaterthan the flux

in the oasisduring the day and has an oppositesign. The flux of heat in

the soil for these conditionschangescomparatively little with irrigation.

We noted abovethat irrigation exerts a substantial influenceupon

the thermalregime.Due to low heat lossesin evaporationin desertsand

steppes,solar radiation absorbedby the earth's surfaceis basically
consumedin atmosphericheating by meansof turbulent heat transfer
and long-wave radiation. Under these conditions,extremely high
groundsurfacetemperaturesare observed.To calculatethesetemperatures, we may use the heat balance equation in the form

R o- 4&rT 3(T,,-

T) =pcpD

(T,.-

T) +LE +A,

(6.1)

where Ro is the radiation balanceof the earth's surfacedeterminedby

calculating the effectiveradiation with respectto air temperature;4 &r
T 3 (T - T) is the differencebetweenthe effectiveradiationdetermined
from the temperature Tw of the earth's surfaceand the air temperature
T; a is a coefficientthat characterizesthe differencein the propertiesof a
radiating surfacefrom the propertiesof a black body,tr is the Stefan-

Boltzmannconstant,p cpD (Tw - T) is the turbulentheatflux fromthe

earth'ssurfaceto the atmosphere,p is air density,c is the specificheat
of the air for constantpressure,D is the integral coefficientof turbulent

diffusion, LE is heat consumptionby evaporation,L is latent evaporation heat, E is evaporation,and A is the heat influx to the

6 MAN'S

INFLUENCE

ON

CLIMATE

177

Equation (6.1) implies that

Tu,- T =

Ro -LE

pC

-A

(6.2)

+ 4,50-T 3'

Since LE and A are much less than R ounder these conditions, Tu, T can reach high values in daytime. Calculations using (6.2) have
shown that it may be as high as 10-20C.
Large values of Tw - T correspondto strong heat fluxes from the
earth's surface to the atmosphere. The temperature in the lower air
layer increases as a result of heating and the relative humidity falls. A
decrease in the relative humidity in turn decreases the amount of
precipitation.
The water balance of the soil changes considerably when droughtstricken areas are irrigated. Evaporation sharply increasesas a result of
irrigation, this increase being equal to the irrigation norm minus the
loss of irrigation water due to infiltration. Heat consumption in evaporation correspondinglyincreases, which leads to a substantial decrease
in the temperature of the earth's surface.
Since the air temperature fluctuations observed in irrigation are
significantly less than the temperature fluctuations of the earth's surface, we may write a formula for the temperature fluctuations of the
earth's surface in the course of irrigation:
Tw-

Tw'=

(Re - Re') - (LE - LE')

pC D + 4 ,50.T

'

(6.3)

where the values referring to irrigation are indicated by a prime. In

derivingequation(6.3)it was assumedthat IA - A' I < ILE' - LE I.

We indicated above that though the radiation balance of the earth's

surface strongly grows in the course of irrigation due to a decrease in
albedo, the increase in heat consumption in the course of evaporation
significantly exceedsthe increase in the radiation balance for ordinary
irrigation rates.
As a consequence,the difference T,, - Tw' is quite high for sufficiently high irrigation rates, the temperature of the earth's surface over
the irrigated section approaching the air temperature under daytime
conditions;it is less than the air temperature in the case of abundant
irrigation. We noted above that this leads to large changes in the
conditions of air masses transformation when sufficiently great areas
are irrigated, as a result of which the temperature and humidity conditions of the lower air layers change over irrigated areas. Dry warm air
entering from outside is humidified and cooled as it moves over an
irrigated

178

CLIMATIC

CHANGES

Temperature and humidity fluctuations of the air over irrigated

sections can be considered

as results

of the transformation

of the air

arriving from outside in different directions. The magnitude of these

fluctuations dependson distance from the border of the irrigated oasis.
The further is the given sectionfrom the border of the oasis, the greater
will be the fluctuations. In addition, temperature and humidity fluctuations of the air with irrigation dependon a number of factors, including:
1) Irrigation rates and time intervals between irrigations. The
greater the amount of water an irrigated field receives and evaporates
over a given interval of time, the greater will be the temperature and
humidity differences between irrigated and nonirrigated fields.
2) Wind speeds and coefficient of turbulent exchange. These two
closely related factors govern the thickness of the air layer in which
temperature and humidity fluctuations occur. The temperature and
humidity distribution with altitude within the lower air layer chiefly
depend on these factors.
3) Radiation properties of the underlying surface, chiefly its reflectivity (albedo); we have spoken of this factor above.
Any empirical study of the influence of all the factors taken separately is extremely difficult so that it is more suitable to calculate
theoretically the temperature and humidity fluctuations and to verify
the results by means of data of stationary and field experimental studies
(Budyko, et al., 1952).
On the basis of theories that have been developed,we may calculate
temperature and humidity fluctuations at different altitudes with a
satisfactoryprecision as a function of water lossin evaporation, fluctuations of albedo, rate of turbulent exchange, and distance from the oasis
edge. These calculations can be used to plan irrigation systems over
previously nonirrigated areas.
Let us pass to a consideration of empirical data.
Table 16 presentsdata generalized by S. A. Sapozhnikova (Budyko
et al., 1952) of observations carried out at weather stations located in a

desert and in irrigated oases.The data are reducedto a single latitude

(42) and altitude over sea level (100 m). Oases whose width did not

exceed 3 km were called small, while the other oaseswere called large.
It is evident from the table that fluctuations in temperature T and
humidity e were maximal in absolute value during summer months
when irrigation water dischargewas greatest. At this time the temperature drop at the level of the weather station due to an increase in
evaporation and transpiration was 2.5-3C,while the absolute humidity at the center of a large oasiswas roughly 5 mbar greater than in the
desert. At lesser altitudes, the absolute values of the differences

6 MAN'S

INFLUENCE

Table

ON

CLIMATE

16

Effect of Irrigation
Regime

Apr.

179

May

June

July

Aug.

0.0

-0.5

-1.6

-2.4

-0.6

-1.1

-2.2

-3.1

on the Meteorological

Sept.

Oct.

Nov.

-2.5

-1.7

-1.4

-0.4

-2.8

-2.3

-1.7

-0.8

Temperature difference:
small

oasis-

desert

big oasis-desert
Humidity difference (mb)
small

oasis-

desert

1.1

1.8

3.4

3.6

3.7

2.5

1.2

0.4

0.4

1.8

4.2

5.4

5.4

3.6

1.6

0.8

big oasis-desert

The influence of irrigation on air temperature and humidity is

somewhatless in a steppezone, mainly due to a lower additional water
discharge through evaporation.
It shouldbe noted that irrigation of drought-strickenareas, which
decreasesthe temperature over irrigated fields, increasesthe mean
atmospherictemperature.A temperaturedropin irrigated areasresults
from an increasein heat consumptionfor evaporation, but this growth
for the earth as a whole is compensatedfor by an identical increase in
the heat influx due to condensation;this heat is released in the atmosphere over other regions where water vapor created by irrigation is
condensed.

In addition, the albedoof the earth's surfacesignificantly decreases

when desertand arid steppesare irrigated, which increasesthe amount
of radiation absorbedin the earth-atmosphere system.
Drainage of marshland usually has an effect on the climatic conditions that is oppositeto that of irrigation. The soil temperature increasesand evaporation decreasesdue to the decreasein soil moisture,
tillage can begin earlier, harvestingin autumn is facilitated, etc. (Budyko, Drozdov, and Yidin, 1966).
Furrow ploughing,which canbe usedin the north when there is not
enoughheat, is one method of drainage and thereby of warming waterloggedsoil. The soil in the ploughhorizonbecomes1-1.5warmer than
in level fields as a result of this form of tillage.
A natural peatbog exhibits a number of properties that differ from
those of mineral soils. In the moist state peat contains a great quantity
of water and therefore possesses
a high heat capacity with significant
heat conductivity.In contrast,dry peat and the mossyvegetationcovering it is characterizedby low heat conductivity.The dry surfacelayer of
peat, since it is strongly heated by day and cooledat night,

180

CLIMATIC

CHANGES

little heat to the soilunderlayers.Thereforeundersuchpeat layersfrost

is retained for a long time at the beginning of summer, sometimes
disappearingonly after rain moistensthe soil.In summer,the peatysoil
of a bogis relatively cold,while in winter, on the other hand, the bog
hardly freezes,and often thaws out under the snowcoverin the courseof
the winter. Evaporationfrom undrained swampsvaries within wide
limits. If the peak is thin (as is usually the casein low-level grassy
bogs),the bog will be coveredby thick grassyvegetationin summer,
from which a quantity of water that is near the maximally possiblefor
these energy resourceswill evaporate (given moisture reservesare
sufficient).Grass-landis usuallyreplacedby mosswhen a thick layer of
peat is present (usually high bogs). Moss easily absorbswater, but
evaporatesit comparatively weakly, particularly if the bog is covered
with shrubvegetation.The total evaporationfromsuchbogsdependson
the moisturesaturationof the bog.Accordingto Romanov(1961),bogs
in the northern European territory of the USSR are saturated with
moistureand evaporateseveraltens of percentmore than dry valleys.
The mean yearly evaporationfrom bogs in the central part of this
territory is not greater than evaporationfrom dry valleys, while in the
most arid regions it may be even less.
The bog climate also varies depending on the manner and purposes
of its reclamation. Thus bogsare thoroughly drained in order to extract
peat. Due to the low capillary conductivityof peat, the surfaceof the bog
easily dries up, vegetation on it dies, and evaporation substantially
decreaseswhen it is thoroughly drained. Daily temperature fluctuations
increasesharply over a dry peat surface.
Drainage of marshland is carried out in an entirely different way
for the purposeof convertingit into agricultural land, the microclimatological conditions varying correspondinglywith reclamation. In this
case reclamation encompassesonly the surface layers. Under these
conditionsevaporation from vegetation is determined by the energy
resources as well as by the state and phase of development of the
vegetation. In general, evaporation approachesevaporative power during vegetation periods.
No such increase in daily temperature ranges noted over parched
marshland is observedover a moist peat surface. At the same time the
soil temperature in drained marshland under thick swards is less than
the temperature of arid valley loamy soil by 3-6 and is less than the
temperature of dry valley sandy soil by 4-8 .
Large water storage basins on rivers created in the course of constructing hydroelectric power stations constitute large but comparatively shallow reservoirs.Therefore their influence on a climatic

6 MAN'S

is similar

INFLUENCE

to that of small reservoirs.

ON

181

CLIMATE

Their main effect is to decrease the

irregularity of the earth's surface and correspondinglyincrease wind.

The wind speedover water reservoirs is several tens of per cent greater
than over openlevel country. This increaseis greatest in autumn, when

the water

is warmer

than

the air and when

intensive

turbulent

ex-

changedevelopsover reservoirs,and is comparativelyweakly expressed

in spring both when the reservoirsare coveredby ice and immediately
after it has melted when the water is cold. At this time, the wind speed
over water basins is about the same as over open level country. Daily
fluctuations of the air temperature decreasefollowing the construction
of a water basin, the radiation balance increases(as a consequenceof a
decrease in the albedo of the area), and the mean yearly evaporation

increases,while it is distributed differently than over dry land in the

courseof a year.
The fluctuation in environmental climatic conditions is not great
when a water storagebasin is built in a locationwith excessor sufficient
moisture. For example, it is difficult to note any systematictemperature
fluctuations in the vicinity of the Rybinsk water storage basin at any
point along its presentshores.
Substantially larger climatic changes arise when reservoirs are
locatedin regions characterizedby insufficient humidity (Tsimlyanskoe, Volgogradskoe,Bukhtarminskoe reservoirs, etc.). The temperature alongthe coastlinesof thesebasinsis noticeablyless(by 2-3)in the
warm time of the year than in regionsfar from the water, due to the
greater evaporationfrom a reservoir than from the surroundingdry
land (where the rate of evaporationis limited by low soil moisture and
from which dry air arrives at the coastalparts of the basin). A daytime
air temperature drop promotesthe developmentof rather strong (up to
3-4 m/sec)breezesextendingvertically to severalhundred meters.
Artificial reservoirs,like irrigation, decreasethe earth-atmosphere
system albedo and, consequently, increase the amount of absorbed
radiation. A water basin therebyincreasesthe mean temperature of the
atmosphere. This increase is obviously less than the temperature
change resulting from irrigation.
Urban Climate. Climatic conditions in cities are usually more
variable than in the surroundingregions;these variations are greater,
other conditionsbeing equal, the larger is the area of the city.
Available data indicate that climatic changesin large cities appeared hundreds of years ago. Thus, Landsberg (1956) presentedeyewitnessevidenceregarding the severeair pollution in Londonin the
seventeenthcentury; such air pollution significantly attenuated solar
radiation in the city as comparedwith agricultural

CLIMATIC

182

CHANGES

The chief factors influencing the meteorologicalregime of a city

include:

1) a fluctuation in the albedoof the earth's surface,which is usually

less in built-up areas than in a rural area;
2) a fluctuation in mean evaporationfrom the earth's surface,which
is noticeably less within the boundaries of a city (though immediately
after rainfall evaporation from roofsand pavements may exceedevaporation in a rural area);

3) heat release createdby different types of man's economicactivity,

whose amount may be comparable to the amount of solar radiation
incident upon the area of a city;
4) an increase in the irregularity of the earth's surfacewithin a city
as comparedwith a suburban area;
5) pollution of the atmosphereby different solid, liquid, and gaseous
impurities created in the courseof economicactivity.
One of the chief features of the urban climate is the appearancein a
city of a "heat island," which is characterized by greater air temperatures than a rural area. This effect has been investigated in many
experimental works, and several numerical modelshave been proposed
for its study.
Results of previous investigations (Inadvertent Climate Modification, 1971;Landsberg, 1970;Berlyand and Kondrat'ev, 1972)have demonstratedthat a heat island is usually of a complexstructure, each city
block of an urban developmentbeing a sourceof heat for surrounding
undevelopedsections.The mean air temperature in a large city is often
1-2 greater than the temperature of surrounding regions, though the
temperature differencemay reach 6-8 at night with a light wind. This
difference usually decreaseswith strong winds.
Interesting observationscharacterizing the growth of a heat island
in the processof urban developmentwere carried out by Landsbergand
Maisel (1972). Constructionwas begun in 1967in the USA, near Washington, D.C., of the city of Columbia. In 1968,when only a small part of
the city had been constructed, the air temperature in its center was
about 0.5 greater than the suburban area at nighttime. In 1970, after
significant progress was made in its development, the temperature
increaseat the center of the city reached4.5,and over a large part of its
area, 2. The temperature increasewas noticeably less in the daytime,
when the intensity of air movement was greater.
The relative humidity decreasedby severalpercent over the area of
Columbia at the same time as the temperature increased,this decrease
apparentlybeing due to both the temperature increaseand a decreasein
evaporation throughout the

6 MAN'S

INFLUENCE

ON

CLIMATE

183

Evidently, a heat island appearschiefly as a result of the influence

of the first three factors given above, which govern urban climatic
conditions. The relative role of each of these factors may strongly vary
in different

cities and in different

seasons.

It should be noted that not only the relative, but also the absolute
air humidity usually decreasesin heat islands due to the decrease in
evaporation in built-up sections.
Additional heating of the air over cities creates local circulation
systemsthat resemble sea breezes,and also amplifies ascendingconvective motions over cities. In addition, an increase in the surface irregularity leads to a noticeable decreasein wind speedin cities in comparison with

suburban

areas.

Air pollution from different impurities, which in many cities

reacheshigh levels, has the greatest practical importance of all climatic
features. Dischargesfrom industrial plants, heating systems,and traffic
are the source of these impurities.
Some of the anthropogenic aerosol in cities is formed from dischargesof solid and liquid particles, while the rest of it is due to gases
mixed with the atmosphere.
An increase in aerosol concentration over cities sharply decreases
the solar radiation arriving at the earth's surface. According to Landsberg (1970),direct solar radiation in large cities often decreasesby about
15%, ultraviolet radiation, on the average, by 30% (it may completely
vanish during winter months), and the period of solar irradiation, by 515%. The thin atmospheric boundary layer, which contains the greatest
amount of aerosol particles, plays the chief role in the attentuation of
solar radiation in cities. Horizontal visibility is usually sharply diminished within this layer, sometimesto 10-20% of its value in rural
regions.
The high concentration of aerosol particles in urban air increases
the frequency of fog and, particularly, of smog-a type of stable fog
whose droplets contain a significant amount of impurities that pollute
the atmosphere. In some cities (Los Angeles is a particularly wellknown example) where local conditionsreduce atmospheric circulation,
smog may remain for many days, greatly harming the health of the
population.
Urban fog plays an important role in attenuating solar radiation as
well as in decreasing the range of visibility in cities.
A decreasein solar radiation in cities and industrial regions with a
very high level of air pollution can lead to a decreasein air temperature
in the daytime. This decrease sometimes completely compensatesthe
increasein temperature due to the heat island (Berlyand, et al.,

CLIMATIC

184

CHANGES

The increasedquantity of condensationnuclei in the air over cities

and the intensification of ascendingair movements lead to an increase
in cloudinessand precipitation. Available data indicate that there exists
a week-long cycle in the amount of precipitation in some industrial
centers,which can be explainedby a decreasein precipitation on weekends, when the industrial plants are not in operation.
Someincreasein precipitationand a significantdecreasein evaporation lead to an increasein run-off, facilitated in many cities by surface
water sewer systems.

Snow removal and its transportation outsidethe city are often used
in cities with great amounts of solid precipitation, which acceleratesthe
increasein the temperature in the city in the spring in comparisonwith
the suburbs.

Urban climate may be significantly improved by rationally situating residencesand industrial plants and by creating green belts and
reducing air pollution.
There are many examples that illustrate how a change in the
heating system (conversionfrom solid fuel to gas or to electrical energy)
sharply decreasesthe pollution of urban air and fluctuations of related
climatic elements. Of no less importance is the removal of industrial
plants from the city, the use of an effective method of air purification in
smokestacksand other sourcesof pollution, the creation of large parks
within a city, and landscapingstreets and various sectorsthat have not
been developed.
At the contemporary level of development of industry, energetics,
and transport, urban construction cannot be carried out without a
detailed study of the influenceof man's economicactivity on the meteorological regime. Calculations of possible changes in local climate are
necessaryfor proper city planning in order to prevent any climatic
conditionsthat couldpresenta hazard to the health of the population.
It is worth noting that a number of features of urban climate have
spread over extensive areas reaching hundreds of kilometers as urbanization has progressedin regions with high population density. Under
theseconditions,many closelysituatedand denselypopulatedareas act
as a sourceof heat and air pollution, and their effect is cumulative. This
situation occursin a number of regionsin the USA, Japan, and Western
Europe, where local climatic changesencompassever-greater areas.
6.2 Changes in the Global Climate

Atmospheric Carbon Dioxide.

Data have been obtained over re-

cent years indicating that man's contemporaryeconomicactivity

6 MAN'S

INFLUENCE

ON

CLIMATE

185

encesnot only the local climatic conditionsof different regions, but also
the climate of our planet as a whole. Variations in the amount of carbon
dioxidein the atmosphere is one of the factors involved in this influence.
The hypothesis that the mass of carbon dioxide in the atmosphere
began to increase as a result of burning large amounts of coal, petroleum, and other types of fuel was expressedas early as the first half of
our century (Callender, 1938). However, only in the 1950s, in conjunction with the advent of the International Geophysical Year, were systematic observationsbegun of atmospheric carbon dioxide at a number
of stations, which permitted a quantitative estimate to be made of the
growth in its concentration.
These observations demonstrated that significant annual fluctuation of carbon dioxide at the earth's surface (decrease in concentration in

summer as a result of the intensification of photosynthesis)is accompanied by a stubborntendencytoward a growth in concentrationfrom year
to year. According to observational data at Mauna Lea (Hawaii), the
yearly increase in the concentration of carbon dioxide amounted to
0.64.10-4% between 1958 and 1968. This magnitude correspondsto approximately 0.2% of the amount of carbon dioxide contained in the
atmosphere (Inadvertent Climate Modification, 1971).
Results of observations at Mauna Lea agree with data from similar
observations in Alaska, Sweden, and Antarctica (Belin and Bischof,
1970). Since these observations were carried out at stations situated at

great distancesfrom each other, there is no doubt that they indicate a

real tendency in the variations of the atmospheric carbon dioxide concentration.

Data of these observationsimply that approximately half the carbon dioxide created in the course of human activity is retained in the
atmosphere;the other half is apparently absorbedby the ocean, and, to
a lesser degree, by living organisms.
The mechanism by which the additional carbon dioxide is absorbed
by water reservoirs and living organisms has not been adequately
studied. Though the oceanshave a potentially great capacity and can
absorban enormous amount of carbon dioxide, the actual role of absorption of carbon dioxide by sea water is significantly reduced due to the
slow rate of exchangebetween the surfaceand deeplayers of the ocean.
The rate of photosynthesisincreases as carbon dioxide concentration increases,though the additional amount of organic matter created
here is mineralized within a limited time, releasing the carbon dioxide
consumedin the process.
The constructionof a completequantitative theory that would take
into accountthe effect of buffer processesin the ocean and biosphere

CLIMATIC

186

CHANGES

variations in carbon dioxide concentrationin the atmosphereis a task

for the future. At the present time we may use empirical schemesfor
this purpose, one example of which is the simple numerical model
proposedby Machta (Inadvertent Climate Modification, 1971).
Calculations using this model imply that the amount of carbon

dioxide
intheatmosphere
hasincreased
by0.003%
overthelastcentury,

i.e., by roughly 10%of its contemporaryconcentration;the chief part of

this increasein the content of carbon dioxide occurredduring the most
recent

decades.

Using the results of recent calculations of the effectsof CO2 concentration on the air surfacetemperature, we find that its increasecaused
by man's economicactivity couldhave increasedthe mean global temperature of the earth's surfaceby 0.3-0.4.
This value is comparableto the temperature fluctuationsthat have
occurredover the past century.Thus it is highly likely that the burning
of different types of fuel exerts a definite influence on contemporary
climatic

conditions.

Atmospheric Aerosol. We noted above that a large number of

particles, noticeably increasingthe concentrationof atmosphericaerosol, enters the atmosphereas a result of man's economicactivity.
Interesting observationson the fluctuation of dust content in the
atmospherewere carried out by Davitaya (1965), who used data on the
vertical

distribution

of dust concentration

in the snow cover of Cauca-

sian glaciers.
These data indicated

that the amount

of dust in a unit volume of the

upper snow layers significantly exceededthat in the deeper layers. In

the opinion of Davitaya, this differencecorresponded
to the sharp increasein dust concentrationin the atmospherethat had occurredover
the preceding decades.
McCormick and Ludwig (1967)as well as other authorshave presenteddata indicating a decreasein the transparencyof a clear atmosphere over the last several decades,apparently due to the increase in
the concentrationof atmosphericaerosol.
Available estimates (Inadvertent Climate Modification, 1971) indicate that about 200 million to 400 million tons of anthropogenicaerosol
is now ejectedinto the atmosphere,which comprises10-20% of the total
amount of aerosol particles that enter the atmosphere.Only a small
portion of this .quantityis ejectedinto the atmospherein the form of
solid or liquid particles; their chief sourcesare the gaseousmixtures
createdby man, such as sulphur dioxide, nitric oxide and others, from
which aerosol particles are formed as a result of various chemical

6 MAN'S

INFLUENCE

ON

CLIMATE

187

These observational data indicate that anthropogenic aerosol noticeably increases the total concentration of aerosol particles not only
locally in cities and separate industrial regions, but also over large
areas, which confers a global nature on the contemporary processof
atmosphericpollution.
The influence of stratosphericaerosolon the air temperature conditions at the earth's surfacewas consideredin Chapter 3. Humphreys had
already established that this influence is basically determined by a
decreasein the flux of short-waveradiation entering the troposphere.
Aerosol particles in the stratosphereplay the role of a screen that
changesthe meteorological solar constant to a greater or lesser degree..
Absorptionof radiation by these particles can causelocal heating of the
stratosphere,as occurredin particular, after the eruption of the Mount
Ageingvolcano(Newell, 1971),though this heating scarcelyaffectsthe
thermal regime at the earth's surfacedue to the insignificant air density
in the stratosphere and weak heat exchangebetween the stratosphere
and the trophosphere.Thus, as has been demonstratedin a number of
studies, an increase in aerosolconcentrationin the stratosphereinvariably leads to a temperature drop at the earth's surface.
Aerosolparticles in the troposphereexert a more complexinfluence
on thermal regime. These particles attenuate the flux of short-wave
radiation arriving at the earth's surface as a result of back-scattering
and absorption of radiation by aerosol particles.
While the first processincreasesthe albedoof the earth-atmosphere
system, the secondmay decreaseit.
Vertical redistribution of absorbed radiation within the trophosphere, where vertical heat exchange is intensive, influences the mean
temperature of the trophosphereand the mean air temperature of the
earth's surfaceto a comparatively small extent. Therefore the influence
of troposphericaerosolon the thermal regime is basically determined by
the dependenceof the albedoof the earth-atmospheresystem on aerosol
concentration, an increase in the former leading to a temperature drop
at the earth's surface, while its decrease, to a temperature increase.
Obviously, a change in the albedoof the earth-atmospheresystemdue to
aerosolwill dependon the albedoof the earth's surface.The lower is this
albedo, the more likely is it that atmospheric aerosol will increase the
system albedo. The probability that the earth-atmosphere system'salbedowill decreasebecauseof aerosolwill grow for higher surface albedo
(snow, ice).
Numerical

models have been constructed in several studies to de-

scribe the influence of aerosolon the radiation and thermal regimes of

the atmosphere(McCormickand Ludwig, 1967;Charlsonand

188

CLIMATIC

CHANGES

1969;Atwater, 1970;Barrett, 1971;Yamamoto and Tanaka, 1971;Rasool

and Schneider, 1971;Mitchell, 1971;Ensor, et al., 1972;Newman and
Cohen, 1972; Kondrat'ev, et al., 1973). Some of these studies took into
accountonly the influence of aerosolon back scattering of short-wave
radiation, a circumstancewhich leads to an increasein the albedoof the
earth-atmospheresystem,and, consequently,to a decreasein air temperature at the earth's surface.
Fluctuations in the radiation and thermal regimes of the atmos-

phere due to both back-scatteringand absorptionby aerosolparticles

has been considered in other studies. The second of these mechanisms

can decreasethe albedoof the earth-atmospheresystemunder certain

conditions, which tends to increase the temperature at the earth's
surface.

Quantitative criteria dependingchiefly on the albedoof the earth's

surface and determining the direction in which temperature changes
under the influence of aerosol have been proposedin a number of
studies. The values of these criteria found by different authors differ
noticeably,which is explainedboth by particular featuresof the models
used and by the fact that the values of the empirical coefficientscharacterizing back scatteringand absorptionof radiation by aerosolparticles,
did not coincide.

In some cases, causes for differences between these criteria are

easily understood.For example, Mitchell suggestedthat the influence of
aerosolon the thermal regime of the atmosphereis associatedwith a
decreasein heat expenditure on evaporation from the earth's surface,
due to a decreasein the influx of radiant energy. Mitchell apparently
obtainedthe resultsby referring mainly to the weather conditionsunder
which the condensationof water vapor is insignificant.
Let us considerresults of one study along these lines (Kondrat'ev, et
al., 1973), which concludedthat the direction in which air temperature
changesin the troposphereis determined by the ratio of the effective
albedoof the aerosolcomponentand the albedo of the earth's surface. If
the first albedois greater than the second,air temperature will decrease
if aerosol is present, and vice-versa. A number of conclusionsmay be
drawn from this work regarding the influence of troposphericaerosolon
the thermal regime for mean global conditions.
Calculations carried out in this study imply that the effective
albedoof a "mean aerosolparticle" varies from 0.20 to 0.60, dependingon
the coefficient of radiation absorption by aerosol particles and on the
distribution of aerosol particles by dimensions. The mean effective
albedo for the group of values given in this work was 0.42.
By comparingthesevalues with typical values of the earth's

6 MAN'S

INFLUENCE

ON

CLIMATE

189

albedo, we may concludethat under cloudlessconditions atmospheric

aerosol as a rule decreasesair temperature over surfacesfree of snow
and ice, whosealbedo is usually at most 0.20, and increasesair temperature over a snow cover, whosealbedo is usually greater than 0.60. Since
the snowcoveroccupiesa small part of the total surfaceof the globe,and
since solar radiation in this zone is on the average less than its mean
global value, it can be seen that a tendency predominatestoward a
decreasein mean air temperature in the presenceof aerosolfor that part
of the earth's atmospherefree of clouds.
The influence of aerosolon mean air temperature when cloudsare
present is of significant interest. Kondrat'ev set the albedo of clouds
equal to 0.70, which is a possiblevalue for separate, thick cloudformations, but cannot be consideredthe mean global albedo of clouds.
The mean global albedoof clouds,accordingto data given in Chapter 2, is 0.46, which nearly coincideswith the mean value of the effective
aerosol albedo.

The differencebetween these two values is indisputably less than

the probableerror of each, which implies that the influenceof aerosolon
the thermal regime of the atmosphereis scarcelyof substantial importance if cloudsare present.
The sameconclusionmay be drawn by bearing in mind the fact that
the bulk of troposphericaerosol is situated below the upper cloud
boundary,as a result of which the influenceof aerosolon the radiation
and thermal regime is substantially weakened if there are clouds.
We find that the combinedinfluence of troposphericaerosolon the
mean air temperature is to decreaseit by roughly half the amount that
wouldbe the casein a clear atmosphere,sinceon the averageabouthalf
the earth's sky is coveredwith clouds.
It is difficult to quantitatively calculate the influence of tropospheric aerosol on the thermal regime of the atmosphere, since not
enough reliable data are available for mean climatological parameters
in the modelsthat are usedas, for example, the coefficientsof radiation
absorptionby aerosolparticles. Thus it is of substantial importance to
analyzeempiricaldata in orderto estimatethe influenceof tropospheric
aerosol on the radiation and thermal regimes.
Observationaldata from a groupof actinometricstationssituatedin
Europe,Asia, and America were processed
by Z. I. Pivovarovaand used
by Budyko and Vinnikov (1973)to estimate the influence of anthropogenic aerosolon climate. A graph describingthe secular variations of
direct solarradiation over the past severaldecadesin the caseof a clear
sky was constructedfrom these data.
Figure 31 presentsvaluesof directradiation anomaliesexpressed

190

CLIMATIC

CHANGES

.S(%)
S
4-

1890

1900

1910

1920

1930

1940

1950

1960

1970

Fig. 31. Secular variations of direct radiation.

percentagesof its norm from 1885to 1970.Curve A in Fig. 31 character-

iZsvaluesaveraged
over5-yearrunningperiods.
(Theuseof 5-year
averaging
leadstoa numberofdifferences
between
thiscurveandthose

presentedabove,whichwere averagedover 10-yearperiods.)It should

be noted that the number

of stations at the end of the nineteenth

and

beginning of the twentieth centuries was less than during the recent
period. Therefore values for that interval of time are less reliable.

It is evident from Fig. 31 that several sharp decreasesin the

magnitude of direct radiation occurredfrom the end of the nineteenth
century to the middle of the 1910sas a result of powerful volcanic
eruptions, such as Krakatau, Mont Pele, Katmai, and others, after
which the aerosolconcentrationin the lower layers of the stratosphere
significantly increased. Direct radiation attained maximal values and
varied comparativelylittle from the end of the 1910sto the beginningof

the!940s,afterwhichit beganto gradually

decrease.

A number of studies(McCormickand Ludwig, 1967;Bryson, 1968;

BudYko,
1969;
Man'sImpactontheGlobal
Environment,
1970;
Inadvert-

entClimateModification,
1971)haveclaimedthat the gradualdecrease

in direct radiation beginning with the 1940scan be explained by the

increasein the massof atmosphericaerosolcreatedby human activity.
We noted above that anthropogenic aerosol is basically concentrated in the lower layers of the troposphere.Its influence on the
radiation that reachesthe earth's surfaceis thereforesignificantlyless

undercloudy
conditions
thanwitha clearsky.Wealsosaidthatdefinite
relationships exist between the variations in direct radiation as a result

of scatteringon aerosolparticlesand the corresponding

changein

6 MAN'S

INFLUENCE

ON

CLIMATE

191

radiation. Using theserelationships,we may calculatethe variations in

total radiation from data on fluctuations of direct radiation.
If these fluctuations result from radiation that has scattered

on

aerosolparticlesin the stratosphere,the correspondingcalculationcan

be performedby usingdata on variations of direct radiation represented
by curve A in Fig. 31. But if part of these fluctuations is due to
attenuation of radiation in the lower layers of the troposphere, the
influence of clouds on the radiation regime of the atmosphere must be
taken into account to calculate the mean global variations in total
radiation.

Let us assumethat sincethe beginning of the 1940sdirect radiation

has varied becauseof two processes,namely the growth in the mass of
anthropogenicaerosol,which would monotonicallydecreasethe amount
of radiation, and a fluctuation in the transparency of the stratosphere,
which

would decrease

direct radiation

as aerosol concentration

in the

stratosphereincreased,or increasethe direct radiation to a limit correspondingto total or nearly total absenceof aerosolin the stratosphere.
We will also assume that the variation

in direct radiation

due to an

increase in anthropogenic aerosol is characterized by a straight line

passingthrough points of the secular variation curve correspondingto
atmospherictransparency maxima in different years. We may suppose
that the stratospherewas either completelyor nearly completelyfree of
aerosolparticles with increasedtransparencyof the atmosphere,which
is confirmed by data of direct observations of stratospheric aerosol
obtainedby Cadle (1972)and by other investigators.The estimate found
here for the decreasein radiation by anthropogenicaerosolcorresponded
to clear sky conditions.The mean cloudcoverfor the globewas taken to
be 0.5 and the influence of anthropogenicaerosolon the total radiation
in the case of a cloudy sky is consideredsmall in comparison with its
influence if no clouds are present. We then find that values of direct
radiation must be increasedby one-half of the variation in direct radiation for a clear sky causedby anthropogenicaerosol, in order to calculate the influence of radiation changeson air temperature. The secular
variation of direct radiation, constructed by taking into account this
correction, is depicted in Fig. 31 in the form of the broken line B.
It is possibleto approximate the massof anthropogenicaerosolfrom
variations in direct radiation resulting from an increase in this mass. It
is evident from Fig. 31 that anthropogenicaerosol has recently decreasedthe direct radiation in clear sky by about 6%. This corresponds
to a decreasein total radiation by roughly 1% for average conditionsin
the caseof a clear sky. Accordingto the data given in Chapter 3, roughly
106g/cm2 of aerosolwill change the total radiation by 1% as a result

192

CLIMATIC

CHANGES

scattering by aerosolparticles. Thus the amount of anthropogenicaerosol is about 5 million tons for the entire globe.
This amount correspondsto the massof optically active aerosolwith

particledimensionsof tenthsof a micrometer.It is usuallyassumedthat

approximately
one-halfthemassofaerosol
in thetroposphere
consists
of

so-calledgigantic particles that are insignificantly smaller in number

than particles with dimensionsof tenths of a micrometer. As a result the
gigantic particlesdo not substantially induceany attenuation of radiation (Junge, 1963).Taking this into account,we find that the total mass
of anthropogenicaerosol can reach about 10 million tons.
Let us compare this value to results obtained by direct methods.
According to modern data (Man's Impact on the Global Environment,
1970), between 1,000 and 2,600 million tons of matter forming aerosol
particles annually enter the troposphere, 20% of this amount being
created by human activity. If we assume that the mean lifetime of
particles in the troposphere is roughly 10 days (Inadvertent Climate
Modification, 1971), we find that the total aerosol mass in the troposphere is equal to 30-70 million tons, while the mass of anthropogenic
aerosol is about 3-14 million

tons. The above estimate

for the amount

of

anthropogenic aerosol reasonably agrees with these values. This confirms that the influence of human activity on direct radiation is calculated with sufficient accuracy.
Fluctuations in the mean latitude air temperatures from 1910 to
1950 were calculated in Chapter 3 from data on fluctuations in solar
radiation. It is necessaryto take into account the influence of human
activity on climate in order to explain the causesof more recent changes
in air temperature.
We demonstrated

above

that

over recent

decades variations

in

direct radiation in the case of a clear sky have depended not only on
fluctuations in aerosol concentration in the stratosphere, but also on
variations in the amount of aerosol in the troposphereresulting from
human activity. It can be seen from the data presented in Fig. 31 that
variations in the amount of troposphericaerosol have had a gradually

increasing influence on radiation and now exert a greater influence

than variations of the aerosol mass in the stratosphere.
Increase in the concentration of carbon dioxide in the atmosphere
has, in addition, influenced the air temperature of the earth's surface
over recent decades. For this reason, the mean temperature of the
earth's surfaceshouldhave grown over recent decadesby approximately
0.3-0.4. Since, accordingto observational data, the mean air temperature in the northern hemisphere has decreasedby roughly 0.3 within
this period of time, we may question whether an increase in the mass

6 MAN'S

INFLUENCE

ON

CLIMATE

193

anthropogenicaerosol could decreasethe mean temperature of the

earth's surfaceby a value consistentwith these measurements.
This questioncan be answeredonly by estimating the influence of
variations in the aerosolconcentrationin the troposphereon air temperature at the earth's surface, which is more complicatedthan the alalogousinfluenceof variationsin aerosolconcentrationin the stratosphere
(Budyko, 1974a).

The data presentedaboveimply that an increasein the concentration of anthropogenicaerosoldecreaseddirect radiation in the northern
hemisphereby roughly 6%. In the case of a clear sky, the level of
radiation would then decreaseby approximately 1%. Since tropospheric
aerosoldoesnot strongly affect the total radiation if cloudsare present,
and alsosincethe mean planetary cloudinessis roughly 0.5, we find that
the mean decreasein total radiation is 0.5%. By using the semiempirical
theory of the atmosphericthermal regime, we find that sucha decrease
in total radiation would lead to a decreaseof 0.75in the mean temperature of the earth's surface. This result correspondsto the case in which
aerosolparticlesonly scatterradiation, but do not absorbit. A number
of studies have indicated that absorption of short-wave radiation by
troposphericaerosolparticles can compensateto somedegree the temperature drop resulting from back scattering of radiation by aerosol
particles.
It is difficult to estimate the influence of radiation absorption by
aerosol due to the absenceof sufficiently general studies of the mean
values of absorption coefficients. It is, nevertheless, possible to use
results of a study by Yamamoto and Tanaka (1972),in which the drop in
the mean global temperature as a result of radiation scattering by
aerosol with and without radiation absorption was calculated using a
semiempirical theory of the thermal regime. Yamamoto and Tanaka
obtained a ratio of 0.68 between the temperature fluctuation in the first
case and the temperature fluctuation in the secondcase with a mean
global cloud cover by assuming that atmospheric opacity was not too
high. If we multiply this value by the temperature drop determined
above, we find that this drop is nearly 0.5for actual conditions.
Though our calculation is highly approximate, it neverthelessgives
us groundsfor assumingthat changesin the mean global air temperature over recent decadesmay partly be explained in terms of man's
economicactivity.
This confirms previous conclusionsobtained in a number of the
studies mentioned above, as well as by Bryson (1972).
Figure 31 shows that after the mid-1960sdirect radiation stopped
decreasing.This fact is worthy of notice becauseit possiblypoints to

194

CLIMATIC

CHANGES

stabilization of the aerosolmass that is of considerablesignificancefor

estimating the climatic conditions of the last decade.
The influence of aerosolparticles on direct radiation determined by
actinometricobservationsin Europe, Asia, and North America was used
in this discussionas more or less correspondingto the mean conditions
of the entire northern hemisphere.
This hypothesis is strongly justified in the case of stratospheric
aerosol whose concentrationvaries comparatively slightly within the
hemisphere.It is lessobviousfor troposphericaerosoldue to its limited
lifetime in the troposphere.Observational data indicate that extensive
regionscan be found in the lower layer of the atmospherewith significantly increased aerosol particle content.
Thus the extent to which actinometric data usedin this analysis are
representative enough for the hemisphere as a whole requires further
study.
Other Factors of Climatic Changes. It is likely that influence of
all other anthropogenicfactorsthan the quantity of atmosphericcarbon
dioxide and aerosol, on the contemporary global climatic conditionsis
comparatively small.
One of these factors involves the productionof energy in the course
of economicactivity, which leads to additional heating of the atmosphere and of the earth's surface. All energy used by man is in the end
converted into heat, the chief part of it being an additional source of
energy for the earth that increasesits temperature.
Hydroelectric power and the energy in woodand agricultural products are the only substantial componentsin man's modern use of energy
that involve transformation of solar energy annually absorbedby the
earth. Consumptionof these types of energy doesnot vary the earth's
heat balance and doesnot lead to additional heating.
However, these types of energy constitute a small part of all the
energy usedby man. Other types of energy, suchas the energy of coal,
petroleum, natural gas, as well as atomic energy, are new sourcesof
heat that do not depend on transformation of solar radiation in the
contemporaw epoch.
In previous works (Budyko, 1962)we estimated the amount of heat

generatedin the courseof human economicactivity. This amount is not

great, being about 0.01 kcal/cm2-yr for the earth as a whole. It is about
100 times greater, and reaches 1-2 kcal/cm2.yr, over the most highly
developedindustrial regions that extend over tens and hundreds of
thousandsof square kilometers. It grows to tens and hundreds of kcal/
cm2.yr over large cities (tens of square kilometers). Let us estimate how
this additional productionof heat influencesthe mean temperature of
the

6 MAN'S

INFLUENCE

ON

CLIMATE

195

We noted above that a 1% change in the influx of solar energy will

change the mean temperature of the earth's surface by 1.5. Since the

productionofheat asa resultofhumanactivityconstitutes

about0.006%
of the total amount of radiation absorbed by the earth-atmosphere
system,the increasein mean temperature correspondingto this amount
of heat is roughly 0.01. This value is comparatively small, though the
increase in temperature can be significantly greater in different regions
sinceheat sourcescreated by man are distributed highly nonuniformly
on the earth's

surface.

It was also noted in this work that temperature could increase

about 1 in the most developed industrial regions if no atmospheric
circulation is present, and by tens of degreesin large cities, which would
obviously make life there impossible. The influence of atmospheric
circulation noticeably weakens correspondingtemperature increases,
and this weakening is greater, the less is the area over which,the
productionof the additional thermal energy is concentrated.
A detailed analysis of the energy consumptionthat represents an
additional heat source for the atmosphere was carried out by Flohn
(Inadvertent Climate Modification, 1971). He found that the heat influx
created by man in the central part of New York and Moscow is several
times that of the amount of energy arriving from the sun. The influx of
additional heat is between 10 and 100%of the influx of solar energy in a
number of smaller cities as well as in the most highly developedindustrial regions tens of thousandsof square kilometers in area.
The additional influx of heat is 1% of the energy of solar radiation
received in many countries measuring up to hundreds of thousands of
square kilometers in area.
Flohn's data confirm our previous conclusionthat man's production
of energy is a substantial climate-forming factor not only in large cities,
but also over significant areas of industrial regions.
In the precedingsectionit was noted that irrigation of desert areas
exerts a definite influence on the mean temperature at the earth's
surface. Let us determine how a currently existing irrigation system
affects the mean temperature of the earth's surface. The albedo of the
earth's surfacemay decreaseby about 0.10 as a result of irrigating arid
regions. By taking into accountthe relation between the albedo of the
earth's surface and the albedo of the earth-atmosphere system (cf.
Chapter 2), we find that with scarce cloudiness this decrease in the
albedoof the earth's surfacecorrespondsto a 0.07 decreasein the albedo
of the earth-atmosphere system.
There are now roughly two million square kilometers under irrigation, which constitute about 0.4% of the total surface of the earth. Thus

irrigation decreasesthe earth's albedo by roughly

196

CLIMATIC

CHANGES

It was establishedin Chapter 2 that a variation in the earth's

albedoby 1%will vary the meantemperature

of the earth'ssurfaceby

2.3. With this in mind, we find that irrigation increasesthe mean

temperatureofthe earth'ssurfaceby approximately

0.07. Thistemperaturefluctuationis notnegligiblysmallandmayplayan importantrole
if the areas under irrigation are increased.

Theconstruction
ofreservoirs
canexert,in conjunction
with irriga-

tion, a specialinfluenceon the mean temperature of the earth'ssurface.

The mean value of the albedo of the earth's surface decreaseswhen

reservoirsare constructed
in plantedregionsby approximatelythe same
amountaswhendesertregionsare irrigated.However,sincethe largest
artificial reservoirshave beenbuilt in regionswith a comparatively
moistclimate where a moreor lessdensecloudcoverexists,the albedo
of the earth-atmospheresystem in this casevaries less than over irrigated regions,where cloudinessis low. In addition, sincethe total area
of artificialreservoirsis appreciablylessthan the area ofirrigatedland,
their influenceon the mean temperature of the earth's surfaceis comparatively slight.
Releaseof heat by peoplein the courseof their activity is onemore
process
that may exert a stronginfluenceon air temperature.This heat

influx is significantonly over limited spaces,where large crowdsare

concentrated,and its role is slight in the caseof larger areas.
The data of this sectiondemonstratethat changesin the global
climateoverthe last severaldecades
haveapparentlydepended
heavily
on man's economicactivity.

This may explain why the relations between fluctuationsin mean

global temperature and volcanic activity (cf. Chapter 3) established
from earlier observationaldata have been upsetin this era.
The further evolutionof man's economicactivity will lead to even
more changesin the global climate than have occurredtill

7.1

The Climate

The Outlook

of the Future

for Climatic

Changes

Natural Climatic Changes. In the study of future climatic

changes,we must first discussthose changesthat may result from
natural

causes.

Chapters 3 and 4 consideredthe influence of a number of factorson

climatic changes,each of thosefactorsbeing characterizedby different
time scales.These include the following fundamental factors.
1. Volcanic activity. A study of contemporary climatic changes
showsthat fluctuations in volcanic activity can affect climatic conditions for years and decades. Volcanism can also influence climatic
changesover periodson the order of centuriesand over longer intervals
of time.

2. Astronomical factors. A change in the position of the earth's

surfacerelative to the sun will create climatic changeswith time scales
of tens of thousands of years.
3. Composition of the atmosphere. A decrease in the content of
carbon dioxide in the atmosphere particularly influenced the climate
towards the end of the Tertiary and Quaternary periods. Bearing in
mind the rate of this decrease and the air temperature changes correspondingto it, we may concludethat the influence of natural variations
in carbondioxide content is of importance for intervals of time greater
than 100,000 years.
4. Structure of the earth's surface. A variation in topography and
related variations in the positionof the shoresof seasand oceanschange
climatic conditions over large spacesfor periods of time of at least
hundreds of thousands and millions of years.
5. Solar constant. We should bear in mind that slow fluctuations of

solarradiation are possibleas a result of the evolutionof the sun, setting

aside the questionof the existenceof short-periodfluctuationsin the
solar constant. These fluctuations can substantially influence climatic
conditionsover periodsof time lasting at least 100 million years.

CLIMATIC

198

CHANGES

Climatic conditionscan vary as a result of self-oscillatingprocesses

in the atmosphere-ocean-polarice system appearing alongsideexternal
factors. These variations have periods of years and decadesand, possibly, hundreds and even thousands of years.
The time

scales in this

list for the action

of different

factors

on

climatic changesbasically agree with similar estimates made by Mitchell (1968) antl others.
Let us consider how these factors can influence
tions of the future.

the climatic

condi-

No reliable methodsfor predicting the level of volcanic activity are

available at the present time. Thus, in studying the influence of this
factor on climate, we are forced to limit ourselves to estimating its
probable scale. It was indicated in Chapter 3 that the rise in temperature in the first half of the twentieth century, due basically to fluctuations in volcanic activity, increased the mean global temperature by
roughly 0.5. It is evident from an earlier study (Lamb, 1970)that the
abating in volcanic activity in the 1920sand 1930sconstituteda large
and comparatively rare anomaly in volcanic activity over recent centuries.

A quantitative analysis of data from Lamb's report on volcanic

activity as well as data on the secularvariations of mean global temperature lead us to concludethat the mean-decadeglobal temperature will
probably change toward the end of the current century due to fluctuationsin volcanicactivity by about 0.2, i.e., by an amount smaller than
the below estimate of the probable change in air temperature due to
man's economicactivity.
i Unlike predictionsof the level of volcanicactivity, future changes
in the position of the earth's surface relative to the sun can be calculated
with reliable accuracy. Such calculations (Sharaf and Budnikova, 1969;
Vernekar, 1972) allow us to estimate fluctuations in the influx of solar
energy over the warm part of the year, the positionof the boundariesof
the polar ice cap strongly dependingon these fluctuations. Milankovich
(1930) noted that fluctuation of radiation near the polar circles exert a
particularly strong influence on the movement of ice.
The data of these calculations imply that radiation will sharply
decrease (by two-thirds the decrease of the last Wiirm glaciation)
10,000-15,000years within the "critical latitudes" of the northern hemisphere.This decreasein radiation can lead to the onsetof a new ice age
with continental glaciations somewhat less than the glaciations of the
Wiirm period.
We should later expect a marked increase in radiation, which

THE

CLIMATE

OF THE

FUTURE

199

lead to destruction of the glaciers, after which decreasesin radiation in

the critical latitudes of the northern hemisphere will repeat in about
50,000 and 90,000 years. The amplitude of these radiation decreaseswill
grow, approachingthe magnitude of the radiation decreasethat brought
about the last Wfirm glaciation.
Much greater radiation decreaseswill occur in the critical latitudes
of the northern hemisphere in several hundreds of thousands of years,
which indicatesthat immense glaciation may developin these epochs.
The influence

of astronomical

factors on the climate

of the future

may significantly intensify with a further decrease in the content of

carbon dioxide in the atmosphere.Since such a trend has prevailed for
many millions of years, it is highly likely that the amount of atmospheric carbon dioxide will continue to decrease in the future without
man's economicactivity. The rate of this decreasecan be approximated
from data given in Chapter 4 on its decreaseover the last geological
epoch, the Pleistocene.
Extrapolation can be usedto estimate the period of time over which
the carbon dioxide concentration of the atmosphere will decreasefrom
the value it had prior to the epochof rapid industrialization (0.029%)to a
value at which total glaciation of the planet (0.015%) is possible. The
difference between these two values is close to the change in carbon
dioxide concentration that has come about since the beginning of the
development of Quaternary glaciations through the contemporary epoch. Thus, we may concludethat total glaciation of the earth can occur
in one million years.
This period must decrease if we bear in mind that the general
tendency toward the development of glaciation resulting from a decrease in the concentration of carbon dioxide is accompanied by a
periodic increase of ice coversdue to a variation in the position of the
earth's surfacerelative to the sun. Thus total glaciation of the earth will
probably begin in several hundredsof thousandsof years. This estimate
can be used to develop a picture of the possible evolution of the biosphere.
The continuing decrease in carbon dioxide concentration will be
accompaniedby a gradual drop in the productivity of autotrophic plants
and a lower total mass of living organisms on the earth. The zone of
polar glaciation will simultaneously gradually expand, migrating to
lower latitudes as the ice ages begin.
The productivity of autotrophic plants and the total volume of
biomass on the earth will decreaseby at most one-half when the ice
cover reachesthe critical latitude. The ice cover will then spread to

200

CLIMATIC

CHANGES

equator as it spontaneouslydevelops.As a result, total glaciation of the

planet and a very stable glaciation due to the low negative temperatures
over the entire globe will set in.
Total glaciation couldlead to the cessationof all biologicalprocesses
on the planet. This hypothesis is based on the fact that no living
organism that could permanently exist in the central regions of an
Antarctic glacier has appearedduring the long history of the glaciers.It
would be difficult to expect any living organismsto arise or to adapt to
such unfavorable

conditions as the climatic

conditions of Central

Ant-

arctica spreadover the entire globewithin a comparativelyshort period

of time.

Total glaciation of the planet correspondingto the temperature

conditionsindicated in Fig. 8 by point B will be extraordinarily stable,
since air temperature at all latitudes will be much below fleezing point.
However, it is possiblethat such glaciation can vanish in the distant
future as a result of a gradual increasein the solar constant. A number
of astronomicalstudieshave expressedthe hypothesisthat the temperature of the sun's surface and the amount of energy radiated by it may

increasein the courseof its evolution(Schwarzschild,

1958;./ingstrom,
1965; White, 1967).

It is evident from Fig. 8 that an increase in the solar constant by

roughly 40%, which can occur, accordingto these works, within several
billion years, is required to upset planetary glaciation. In this case,ice
will melt at all latitudes and the air temperature will increase by more
than 100, reaching a mean value of about 80C.Since the temperature
of the earth's

surface will continue

to rise with

a further

increase

in the

solar constant, the melting of ice will hardly lead to conditionsfavorable

for the reappearance of life on earth.
This path of developmentof the earth's thermal regime is extremely
hypothetical, sinceit is not entirely clear how the sun has evolved. It is,
however, of some interest that if this variation holds, the thermal
regime will pass through several stages correspondingto the large
segment of the hysteresis curve in Fig. 8, i.e., from the ice-free conditions of the Mesozoic (somewhat above point E) to contemporary conditions (point 3), to total glaciation (from point B to A) and to the
destruction of glaciation (point A ). This estimate of time within which
total glaciation of the planet can occuris very approximate.The accuracy of the calculation is substantially limited by the method used for
determining the rate at which the carbon dioxide content decreasesin
the atmosphere, which is highly tentative. This is in addition to the
highly schematicnature of the model describingthe thermal regime of
the atmospherethat was used to obtain this

THE

CLIMATE

OF

THE

FUTURE

201

This result is of interest chiefly as an illustration of a previously

expressedconceptionof the specific nature of Quaternary glaciations
(Budyko, 1971). It seemspossibleto us that, unlike the Permo-Carboniferous and other ancient glaciations, Quaternary glaciations were not
temporary episodesin the evolution of the earth, but represented the
beginning of a transition from stable ice-free climatic conditionsto even
more stable total-glaciation conditions of the planet. The duration of
this transitional period in which the existence of the biosphere may
concludeis much smaller than the total duration of life on our planet,
accordingto the estimate presentedabove.
The probability of total glaciation in the course of the earth's
natural evolution increasesif we take into account existing tendencies
toward an increase in the elevation of the continents in a number of

high- and middle-latitude regionsof the northern hemisphere.

Saks (1960) and other authors have presented paleogeographic
maps showingthat by the end of the Jurassicthe oceansoccupiednearly
half the latitude circle at 70N, which.promotedfree circulation of sea
water betweenthe middle and polar latitudes. At the beginning of the
Cretaceous, the areas occupiedby the continents began to gradually
increase at this latitude, and this has continued in our time, as a result
of which the continents now occupyfive-sixths of this latitude circle,

The width of the straits connectingseasof middle and high latitudes in the northern hemispherewas reducedto one-third. There was
thus a marked decreasein meridional heat transport by sea currents to
polar latitudes, and the prerequisites for the development of arctic
glaciation were created.
There are no groundsfor believing that the near future will seean
end to the increasingisolation of the northern polar basin, a trend that
has alreadylastedmany millionsof years. If this trend continues,it will
becomemore likely for the glaciers to reach the critical latitude in the
future.

The influenceof changesin the earth's surfaceon the climate of the

future will probably manifest itself after a decrease in the carbon
dioxidecontentin the atmospherehas set in and may therefore turn out
to be insignificant. However, since the correspondingestimates are
highly approximate,the comparativeimportanceof thesetwo factorsfor
the climatic conditions of the future is not entirely clear.
A hypotheticalschemefor future variationsof the mean globalair
temperature at the earth's surface in the light of these hypothesesis
presentedin Fig. 32, which also depictfluctuationsin the mean global
temperature in the past.

For the sake of clarity, the scaleof the time axis in Fig. 32 for

CLIMATIC

202

CHANGES

Tp(C)

100

-5O

--100

-10

-10

10

I0(yrs)

Fig. 32. Variations of the mean air temperature at the earth's surface.

previous and next million years is 10a times greater than for more
distant times. The range of mean temperatureswhich are most favorable for biologicalprocessesis shownby the hatched area. We should
emphasizethat this scheme is highly tentative in the light of the
hypothesesassumedin constructingit.
Inadvertent Climate Modification. The development of civilization, an exceptionallyrapid event on the time scaleof geologicalhistory,
has radically changedthe outlookfor the further existenceof the biosphere.
We need only note that the concentrationof carbondioxidein the
atmospherehas increasedby 0.003%,i.e., by 10%of its magnitudeover
recent decades,as a result of the burning of different types of fuel. This
increase compensatesfor its decreaseover the previous 200,000years.
Thus, human activity has changedthe directionof the processby which
the concentrationof atmosphericcarbon dioxide has varied and has
increasedits rate by a few thousand times.
Thoughin this caseman's effecton climate has beenunintentional,
it has nevertheless already becomeof great importance for preventing
the further development of glaciation.
Let us imagine the highly unlikely case that man's effect on the
atmospheremay ceasein the future. Under these conditions,the increasein carbon dioxide concentrationin the atmospherethat has been
attained over the last century will defer planetary glaciation by thousandsof years. Obviously,globalglaciationwouldno longerbe possible
if the contemporaryrate of man'seffecton the atmosphereis maintained
and, especially, if it

THE

CLIMATE

OF

THE

FUTURE

203

In addition, the rapidly increasing influence of man's economic

activity on climate will make possiblein the comparatively near future
a significant change in the global climate.
This change, of course,will not be as catastrophicas a total planetary glaciation,thoughit may createa number of substantialdifficulties
for further economicdevelopment.The outlook for anthropogenicclimatic changes requires great attention.
Let us consider the influence

of three fundamental

factors on the

climatic conditions of the next century:

1) a growth in the productionof energy consumedby man;

2) an increase in carbon dioxide content in the atmosphere;
3) fluctuations in the concentration of atmospheric aerosol.
We noted in our first work dealing with this question (Budyko,

1962)that a 4 to 10% increasein the annual productionof energy can

equalize the heat created by man with the radiation balance of the

entire surface of the continents within 100-200 years. It is obvious that

enormous climatic changes could then be expected over the entire
planet.
The data presented in Fig. 33 provide some idea of the possible
influence of a growth in energy productionon the climate of the future
(Budyko, 1972). Here, curve I represents the secular variations of a
deviation of mean air temperature from the norm for the northern
hemisphere,based on observationaldata. Curve 2 gives the results of
calculationsof changesin mean planetary temperature with an increase
in energyproductionof 6% per year. This calculationis basedon the use
of our previousmodelunder the assumptionthat a temperature change
occursas a result of a variation in heat influx, all other factors influencing the climate, including the albedoof the earth-atmospheresystem,
being kept constant.As we have shownabove, this hypothesisdoesnot
take into accountthe possibilitythat polar ice couldmelt, which leadsto
an underestimate of future temperature increases.
It is evident from Fig. 33 that the rise in temperature due to a
growth in energy productionduring the first half of the twenty-first
century will exceedthe temperature changesthat occurredin the first
half of the twentieth century as a result of natural causes.Temperature
will subsequentlygrow rapidly, which will lead to great changes in
global climate. Thus a growth in the productionof energy consumedby
man may exert a substantial influence on the climate of the future.
The secondfactor that may significantly changethe climate is the
growth in the carbondioxideconcentrationin the atmosphere.Observational data indicate that it has been mounting yearly at a rate of about
0.2% (of its total amount) over the most recent decades because of

204

CLIMATIC

CHANGES

A
T(C)
3-

21-

c./

197_0 ! 70

I ///
9so

oO

2
oso

Fig. 88. Secular variations of air temperature anomalies at the earth's

surface. (1) Variations of mean temperature; (2) variations due to growth of
energy production; (8) variations due to growth in carbon dioxide concentration.

burning of a growing amount of fuel. This processmay have accelerated

somewhat over recent years. On the basis of these data, Machta and
other authors

have concluded

that the carbon dioxide

concentration

in

the atmosphere will increase by 15-20% between 1970 and 2000 (Inadvertent Climate Modification, 1971).
We indicated in the preceding chapters that an increase in the
amount of atmospheric carbon dioxide will lead to a change in the
radiation regime of the atmosphere. The transparency of the atmosphere to long-wave radiation will decrease,and this will increasethe air
temperature at the earth's surface. Assuming such an increase in carbon dioxide concentration, Manabe (1970) found that (neglecting fluctuations in the area of the snow-ice cover) the mean planetary air
temperature at the earth's surface in the year 2000 may be approximately 0.5higher, and the mean temperature at high latitudes approximately 1higher than in 1970. Similar results are obtainedby applying
the dependence, given in Chapter 4, of air temperature on carbon
dioxideconcentrationwith constantearth-atmospherealbedo.A change
in global temperature correspondingto this calculation is shownin Fig.
33 by the unbroken segment of curve 3.
It is evident from Fig. 33 that by the year 2000 a change in
temperature due to the increased carbon dioxide concentration

THE

CLIMATE

OF

THE

FUTURE

205

exceedthe anomalies in the natural fluctuations of the global temperature observed during the first half of the twentieth century. If the
carbon dioxide concentration continues to grow in the twenty-first century, more telling climatic changesmay ensue.
The change in the concentrationof atmosphericaerosolis an additional factor resulting from man's economicactivity that may lead to
climatic fluctuations. Data were presented in Chapter 6 implying that
this increase affects the contemporary climate. Obviously a further
growth in aerosolconcentrationcan entail a sharp climatic change.It is,
however, rather difficult to predict the amount of atmosphericaerosolin
the future.

It is worth noting that the latitudinal distribution of absolute

values of temperature variations are similar with increasing carbon
dioxide as well as with aerosolconcentration. Since these changesare of
oppositesign, we may supposethat in particular casesthey compensate
each other to some degree.
Contemporary anthropogenicclimate changes, which have somewhat abated as a result of the counterinfluence of changes in the
concentrationof atmosphericcarbondioxide and aerosolon the thermal
regime, constituteone example of such counterbalancing.
It has been proposedin the author'spreviouswork (Budyko, 1972b)
that, when arrangementsto protectthe atmosphereagainst its pollution
by man's activity are realized, the growth of the anthropogenicaerosol
masshas to stop.In this connectionit was assumedprobablethat future
climatic changeswould be considerablydependenton the influence of
the growth of carbon dioxide concentrationon the atmospherethermal
regime, which would soonlead to developingthe new warming (according to the curve 3 in Fig. 33).
It was noted in the above work that the new warming could be of
great practical importancedue to deterioration of precipitation conditions over much of the continents in moderate latitudes (that will result

in droughts) and due to decreasingthe area of polar sea ice.

In passingto the possibleclimatic conditionsof the future, let us
note that the practical value of this questionis a functionof the scaleof
possibleclimatic changesand the time at which they may begin. If these
changeswill be sufficiently great and will occurin the not too distant
future, the predictionof future climatic changesclearly becomesone of
the most important problemsof contemporarymeteorology.
The importanceof this problemis particularly evident in the strong
dependenceof the national economyof all countries on contemporary
climatic conditons.A sharp climatic change requires enormouscapital
investment to accommodateeconomicactivity to these new

206

CLIMATIC

CHANGES

The time by which it is necessaryto have information on climatic

changesis apparently comparablewith the likely lifetime of currently
planned industrial and agricultural buildings and systemswhosefunctioning is subjectto climatic conditions.This period is at least 100years
for the most durable buildings. If climate will substantially changein
the future, this possibility must obviouslybe taken into accountto some
degree in planning these buildings.
The time required to prepare and carry out measures to control
climatic changesand adjust the national economyto these changesis a
secondcriterion that can be used to estimate the time by which it is
desirableto have information on possibleclimatic changes.This time is
probably at least several decades, since such measures require the
solution of many complexscientificand technologicalproblemsin order
to be implemented.Thus, it is desirableto have informationon changes
in climatic conditionsthat may occurwithin 100 years in the future.
The problem of predicting climatic changes as a result of human
activity substantially differs from weather forecasting. While we may
limit ourselvesin the latter caseto an analysis of the physicalprocesses
in the atmosphereand hydrosphere,the former problem requires that
we also bear in mind the changesthat time may bring to the indicesof
man's economic activity.

Thus the problem of predicting climatic changesinvolves two fundamental elements:predictingthe developmentof a number of aspects
of economicactivity (growth in the use of fuel, increasingconcentration
of carbondioxidein the atmosphere,growthin the productionof energy,
etc.) and the evaluation of thoseclimatic changesthat may changethe
correspondingindicesof man's activity.
This contributesto two important features involved in thesepredictions. On the one hand, they will be unavoidably provisional. Man's

economicactivity is not independent of his influence on climatic conditions. In particular, if this activity leads to extremely unfavorable
climatic changes,it is highly likely that it will have altered its character
beforethese changesbegin. Thereforethe problemof a climatologistis
not to predict the actual climate of the future, but to calculate the
parameters of this climate for a number of possibleforms of economic
activity. We may optimize long-rangeplanning for the developmentof
the national economy,taking measuresagainst unfavorable changes
and taking into accountresultsof sucha calculation.Thus predictionof
possibleclimatic changesmay serve as a basis for taking measuresto
modify climate.
The second feature of predictions of the future climate

THE

CLIMATE

OF

THE

FUTURE

207

their probable error. Since it is difficult for a number of reasons to

quantitatively predict economicactivity for several decadesin advance,
the precisionof such a prediction cannot be high. Thus we are apparently justified in using schematic models of climate theory, which,
however, must yield a correct estimate of the direction and order of
magnitude of possibleclimatic changes.
The practical importance of suchestimates lies in the fact that they
allow us to identify those forms of economicactivity that can lead to
great (i.e., the most consequential to the national economy) climatic
changesexceedingthe error of the correspondingcalculations.
We must first discuss the effect of a further growth in energy
production,which we spokeof above, in order to deal with the possible
climatic changes in the next century.
It is evident from Fig. 33 that even disregarding the feedback
betweenthe polar ice regime and air temperature, we may concludethat
a rapid increasein planetary temperature will begin in the middle of the
twenty-first century if energy productionannually grows by 6%. This
increasewill be accompaniedby enormousclimatic changes,which may
lead to catastrophicconsequences
for the national economiesof many
countries.Thus uncontrolledgrowth in heat productionwill lead to a socalled heat barrier" in the developmentof energetics.
We should note that the assumptionsmade in deriving curve 3 in
Fig. 33 are not entirely unrealistic. A 6% annual increase in energy
productionhas been the case over recent years (Inadvertent Climate
Modification, 1971). The magnitude of the additional heat influx per
unit area reachedby 2070obtainedin this calculation correspondsto the
amount of heat now released in some regions of the industrially most
developedcountries.
A review of a number of works in which the possibleuse of energy in
the future is estimated can be found in an article by Kellogg (1974).
These data imply it is highly likely that the global use of energy will
increasemore than 100-foldin the comparativelynear future. Such an
increasewill significantly changeclimatic conditions,particularly as a
comparativelysmall increasein the air temperature beginsto affect the
polar ice, whoserecessionwill lead to an additional sharp temperature
change at high latitudes.
Thus the possibility that a heat barrier will be reached within a
period of at most 100 years must be consideredone of the fundamental
problemsthat will confrontthe technologyand energeticsof the not too
distant

future.

The methodsto solveit require specialdiscussionoutsidethe

CLIMATIC

208

CHANGES

of this book. Here we may only recall N. N. Semenov'sadvocacyof

widespreaduse in the future of solar energy for economicpurposesin
order to limit overheating of the earth's atmosphere.
The state of the ice coverin high latitudes is of great importancein
studyingclimatic changesthat canbe attributed to an increasein global
temperature.In our previousworks (Budyko,1962a,1971)we offereda
number of methods to estimate the influence of changesin meteorological factorson polar ice. One suchmethod,basedon a calculationof the
heat balance of the ice cover, allowed us to calculate a change in ice
thicknessfor a given air temperatureincrease.This demonstratedthat
polar seaice will completelymelt in severalyearsif the air temperature
in the warm part of the year increasesby 4C.
Since the feedback between

the area of the ice cover and the air

temperaturewasnot taken into accountin this calculation,it is entirely

evident that the magnitude of the temperature anomaly sufficientfor
completemelting of ice is overstated.Nevertheless,we may conclude
that the temperature increase which will be reached in roughly 100
years accordingto the data of Fig. 33 will lead to total thawing of polar
sea ice.

Obviously, sea ice may partially melt before a heat barrier is

reached, i.e., before the mean global temperature increasesby several
degrees.
Let us use equation (3.1) to estimate changesin the area of polar
ice. We will assume that the magnitude of the relative variation in
global radiation in this formula can be expressedin terms of the change
in global temperature correspondingto it, which is determined in accordancewith the semiempirical theory of the thermal regime with
constantalbedo. In this case,the area of polar ice can be calculatedfor
different values of a mean global temperature increasingas a result of
an increase in carbon dioxide concentration and energy production in
accordancewith the data depictedin Fig. 33.
Since the influence of energy productionon the thermal regime of
the atmosphere will be comparatively slight prior to 2000, the air
temperature of the earth's surface will be chiefly determined by the
growth in carbon dioxide concentration if the mass of atmospheric
aerosol remains

constant.

The mean global air temperature in the local Manabe-Wetherald

model (1967) will increase by 0.5 if carbon dioxide concentration increases by 20%, and it will increase by 0.7 according to a threedimensional model of climate theory that takes into account the influence of the snow cover on the thermal regime (Smagorinsky, 1974).
A similar calculation using our semiempirical model of the

THE

CLIMATE

OF

THE

FUTURE

209

pheric thermal regime under the hypothesisthat the polar ice is quasistationary yields a temperature increase of 0.6.
All these models imply that an increase in the mean annual air
temperature at high latitudes is roughly twice as great as at low
latitudes.

Results of a calculation of ice boundary changes are shown in Fig.

34, where the mean latitude of the sea ice boundary in the northern
hemisphereis laid out along the vertical axis (Budyko, 1972b).
It is evident from Fig. 34 that the mean boundary of polar ice will
recede roughly 2 to the north by the year 2000. This change in the
boundaryof polar ice correspondsto a much greater decreasein its area
than occurredduring the warming period in the first half of the twentieth century and also to a stronger increase in mean air temperature
than its level at the end of the last century. A decreasein the mean area
of polar ice will exert a noticeableinfluence on air temperature at high
latitudes.

It is necessary to conjecture what might be the carbon dioxide

concentrationin the atmospherefor the twenty-first century in order to
estimate further changesin the boundaryof polar ice. It was assumedin
the calculations that the mean air temperature will increase linearly
under the influenceof a growth in carbondioxideconcentrationafter the
year 2000. Due to the rapid growth in energy productionin the twentyfirst century, the assumptionthat carbon dioxide concentration will
change in this century will not greatly influence the results of the
calculations.

Errors due to the inaccuracy of equation (3.1) for great changesin

the area of ice may exert a more substantial influence on the computation of the time necessaryfor completethawing of polar ice. As we noted
above, it is not precludedthat polar ice is unstable when comparatively

80-

75-

1950

2000

2050

Fig. 34. Variations of polar sea ice boundary in the northern

210

CLIMATIC

CHANGES

small in area and can thaw without any additional increase in the
temperatue of the lower air layers. In such a case, the data presentedin
Fig. 34 will overstate the period of time necessaryfor completethawing
of ice.

It is evident from Fig. 34 that under these hypothesesthe time

necessaryfor complete thawing of sea ice in the Arctic is roughly 80
years.

That this may be an overestimate follows from a calculation of the

increase in mean global temperature correspondingto the moment at
which sea ice has completelythawed. This value is approximately 3.5C,
i.e., closeto the value presented above, which insures thawing of polar
ice without taking into accountthe feedback of a change in the ice cover
on the thermal regime.
A number of indices describingchangesin climatic conditionsassociated with the thawing of polar ice are shown in Fig. 35. Temperature
changesin middle and high latitudes of the northern hemisphereby the
year 2050, i.e., when thawing of the polar ice will end, are depictedin
Fig. 35 as calculated from the equations of the semiempirical theory of
the thermal regime. It is evident from this figure that completethawing
of polar ice will lead to enormouschangesin the thermal regime in the
high latitudes.
A comparisonof the data of Fig. 35 with paleoclimatic data demonstrates that, assuming ice-free conditions in the Arctic, the thermal
regime in middle and high latitudes of the northern hemispherewill be
similar in many respects to the regime at the end of the Tertiary,
millions of years ago. Thus the transition to a new thermal regime can
be represented as a return to the climatic conditionsof the pre-Quater-

hr(oc)

20"

I0

19

60

70

80

90

Fig. 35. Changes in mean latitudinal temperature with thawing of polar

ice. (1) Warm half-year; (2) cold

THE

CLIMATE

OF

THE

FUTURE

211

nary. The rate of this return will be enormous- roughly 105greater than
the rate of natural cooling over the recent geologicalperiod.
Obviously, a climatic changecausedby the thawing of polar sea ice
will not be limited to an increasein air temperature. A gradual breakup
of the Greenland and Antarctic glaciers, leading to an increase in the
level of the world ocean, is one possibleconsequenceof the thawing of
sea ice. A changein the hydrologicalcycleon the continentsmay alsobe
of substantial importance.
We indicated in Chapter 3 that the hydrological cycle significantly
changedon the middle-latitude continentsof the northern hemisphere
when only a comparatively small part of the Arctic ice cap thawed in the
courseof the warming during the 1920sand 1930s.Changes in the cycle
led to an increasein the frequencyof drought in a number of regionsand
to a change in the run-off and levels of land-locked seas.
Many aspectsof the climaticchangeoccurringin the first half of the
twentieth century were of practical importance. The climatic change
that may be reached in the last decade of the twentieth century may
therefore strongly influence various aspects of economicactivity in a
number of countries. A further intensification of climatic changesdue to
thawing of polar ice may turn into a natural catastrophe if special
preventive measuresin the national economyare not taken.
These arguments on the climatic condition of the future must be
complementedwith an estimate of the influence of anthropogenicaerosol on the thermal regime.
Though, as we notedabove,it is difficult to predictthesechanges,it
is nevertheless unlikely that the amount of anthropogenic aerosol will
significantly grow in the future, since this would lead to conditions
highly unfavorable for the life of the population in regionswhere the
most powerful sourcesof this aerosolexist, that is, in cities and industrial centers.The growth in the massof anthropogenicaerosolat the end
of the twentieth century and, in particular, in the twenty-first century,
will thus hardly compensate the tendency toward an increase in air
temperature due to a growth in energy productionand an increase in the
concentration of carbon dioxide in the atmosphere.
If we assume that the amount of anthropogenic aerosol will decrease over the next decades as a result

of advances

in the control

of

atmosphericpollution, the processby which polar sea ice recedesand

the climatic changesassociatedwith this processwill then accelerate.
Anthropogenic changes of the heat and water balances could be
accompaniedby a change in the amount of oxygen in the atmosphere
due to man's economic activity, this oxygen being consumed in the
burning of different types of fuel and various production processes
(Davitaya,

212

CLIMATIC

CHANGES

Calculations demonstrate that enormous climate changesas a result of the growth in the concentrationof carbondioxide and an increase
in energy production, which constitute an additional sourceof heat for
the atmosphere, will occur before the amount of oxygen noticeably
decreases.Thus the problemof preservingatmosphericoxygencan arise
only at the last stagesof an ecologicalcrisis createdby man's activity.
These possible changes in the next century are hypothetical in
many respects.It is necessaryto repeat that the extrapolation used in
these calculations of the substantial rates of growth of energy production and carbon dioxide concentration in the atmosphere is only a
possible, but not a necessary, condition for economic activity in the
future.

Moreover, while further studies may confirm the conclusionthat

sharp climatic changesare possiblein the next century, measureswill
be taken to limit the influence of industrial developmenton climatic
conditions. This question will be consideredin the next sectionsof this
chapter.
We should therefore emphasizethat the data presentedhere were
obtained as a result of highly approximate calculations.These calculations are basedon highly schematizedmodelsof the thermal regime and
involve many approximating assumptions. For this reason, the data
presented in Figs. 33, 34 and 35 are saddled with significant errors.
Though it is rather difficult to estimate the magnitudesof these errors,
some conjecturesmay be made as to the accuracy of the results.
Reliability of Results. The reliability of available estimates of
possiblechanges in the global climate warrants much attention since
they could be of practical importance.Two different approacheshave
been used in treating this question.
The first approach is based on an analysis of the physical foundation of numerical models used for calculating climate changes in the
past and in the future, and on a comparisonof results of these calculations using different numerical models.The secondapproachconsistsin
checking the numerical models against empirical data on climatic
changesover time.
The semiempirical model utilized here for the thermal conditionsof
the atmosphere takes into account all the componentsof the heat
balance that influence the horizontal temperature distribution. This
modeltakes into accountthe mostimportant feedbackrelations between
air temperature and other meteorologicalfactors, including the dependenceof air temperature on its absolutehumidity (due to the influence of
the absoluteair humidity on the outgoingradiation, which is

THE

CLIMATE

OF

THE

213

FUTURE

for in equation (2.3)) and the dependenceof air temperature on the

position of polar ice. Both these feedback relations intensify the influenceof changesin climate-forming factors on the thermal regime, which
decreasesclimate stability.
Among the assumptionmade in justifying our model, mention must
be made of the loose schematization

of the meridional

redistribution

of

heat in the atmosphere and hydrosphereand of the fact that the influence of possible changes in cloud cover on the thermal regime of the
atmosphere was not taken into account.
As we noted above, the empirical relation used for a parametric
descriptionof meridional heat redistribution holds within a broad range
of variation of the distribution of mean latitudinal air temperatures
attained in the courseof a year. We may thus concludethat the use of
this relation will not lead to great errors in calculating changesin the
thermal regime smaller than its fluctuations in the courseof a year. It is
lessclear whether this dependencecan be usedto study drastic changes
in the thermal regime due, for example, to the developmentof planetary
glaciation.
The influence of changesin the cloudcoveron the thermal regime of
the atmospherehas been discussedin many studies,this influence being
overstated in a number of cases when no provision was made for a
decreasein outgoing long-wave radiation in the presenceof a cloud
cover.

It is evident from the results of calculations carried out in Chapter 2

that for mean global conditions changes in the cloud cover have a
negligible influence on the temperature at the earth's surface, since in
this case a decrease in absorbedradiation accompaniedby a growth in
the cloudcoveris nearly compensatedfor by a decreasein outgoinglongwave radiation. The equations presented in Chapter 2 imply that the
influence of a cloud cover on air temperature at the earth's surfacewill
strongly depend on the magnitude of solar radiation. That is, a growth
in cloudcoverwill decreaseair temperature for high values of radiation
influx to the external boundary of the atmosphere and will increase
temperature if the radiation influx is low.
It was established by Schneider (1972) that the influence of cloud
cover on air temperature likewise dependson the thickness of the cloud
layer, since cloudswith a high upper boundary more strongly decrease
outgoing radiation. Since the albedo of the earth-atmosphere system
increaseswith increasing thickness of the cloud layer, under particular
conditionsa decreasein absorbedradiation can compensatea decrease
in out-goingradiation. Schneider noted (1972, 1973) that the

CLIMATIC

214

CHANGES

between the temperature field and the cloud cover can influence climate
stability, though even the sign of this relation is unknown at the present
time.

Some data indicate that changes in cloud cover rather weakly

influence the distribution of the mean latitudinal air temperature. This
conclusionfollows, in particular, from the data of Fig. 36, which depicts
the distribution of the mean latitudinal temperature in the northern
hemisphere for mean yearly conditions,found by means of the semiempirical model describedin Chapter 2 (curveTo). The influenceof deviations of the mean latitudinal

values of the cloud cover from its mean

planetary value of 0.50 was ignored in this calculation.

Sincethe mean values of cloudinessat different latitudes are highly
variable, the fact that the results of this calculation are quite near the
observationaldata (curve T) indicatesthe limited influence of changes
in the

cloud cover on the distribution

of the

mean

latitudinal

air

temperatures. This conclusiongives us reason to believe that fluctuations in the cloud cover can be disregardedin a number of caseswhen
calculating changesin the thermal regime.
This assumptionwas further corroboratedby studiesof correlation
betweenvariations in cloudinessand air temperature carried out using
satellite data (Budyko, 1975;Cess, 1976).
Let us considerresults of applying different modelsfor studying
climatic changes.
In Chapter 2 we noted that semiempirical models of the thermal
T(C)
3O

2O

I0-

To
\

-10

-20

20

,,

40

60

80 50
o

Fig. 36. Distribution of mean latitudinal air temperature. T, observational

data; To, results of

THE

CLIMATE

OF

THE

FUTURE

215

regime of the atmospherethat have been developedby various authors

yield essentially similar dependenceof changesin mean latitudinal air
temperatures on external climate-forcing factors. These dependences
sometimes reasonably agree both qualitatively and quantitatively,
while in other casesthere exist quantitative differencesbetween them,
which can be explained by the distinctive features in the parameterization of particular atmosphericprocessesin the correspondingmodels.
A comparison of results from the semiempirical models and more
general models of climate theory that directly postulate circulation
processesin the atmosphere is of significant interest.
Earlier chapters comparedcalculations of changes in the thermal
regime of the atmosphere according to the semiempirical model and
Manabe-Wetherald's
three-dimensional
model as a function of fluctuations of the solar constant and of carbon dioxide content in the atmos-

phere. This comparisondemonstratedthat results of these calculations

are in satisfactory agreement.
Let us considerone more comparisonof data from calculations of
changesin the thermal regime using more general modelsand semiempirical modelsof climate theory.
The change in the air temperature field in the troposphereas a
result of the additional heat influx created by man's economicactivity
has been calculated by Washington (1971) using a three-dimensional
model of climate theory. This heat influx, equal to 50 cal/cm. day, was
uniformly distributed over the surface of the continents.
Washington found that the mean air temperature in this case will
grow 1-2 in the tropics and 8-10 in high latitudes of the northern
hemisphere.
A calculationusing the semiempiricalmodelof the thermal regime
of the atmosphereset forth in Chapter 2 yields an increaseof 3 in the
mean globaltemperature for theseconditions,which agreesreasonably
well with Washington's estimates (we note that in both cases the
dependence
of the albedoof the earth-atmospheresystemof temperature
changeswas not taken into account).
In his next work Washington (1972) calculated temperature
changesfor an additional heat influx equal to one-sixthof that usedin
the first calculation, this influx being distributed accordingto population density. A change in the mean global temperature for these conditions was small comparedto its natural fluctuations for periods of
several tens of days (these fluctuations, according to Washington's
calculationsand using the samemodel of climate theory, amountedto
several degrees).
Leaving aside the question of whether such natural fluctuations

216

CLIMATIC

CHANGES

the mean global temperature are realistic, let us note that Washington's
conclusionagrees with results of calculationsusing the semiempirical
model, which yield a changein mean global temperature of 0.5for this
case, i.e., a magnitude significantly less than the natural fluctuations
characterizing his model.

In discussing
r,esultsof calculationsof climaticchangesby different
models, we should emphasizethat these results must be comparedin
terms of how problems in correspondingstudies were formulated.
Sharp differencesin several surveysmay be thought of as a consequence of particular features of the models used, whereas these differencesare basically explained by the fact that the initial values for the
calculationsin these studieswere not equal.
Thus, we may concludethat different modelsof climate theory yield
similar results in most caseswhen estimating climate fluctuations for
identical changesin climate-forming factors.
Some differences in these estimates, occurring, for example, in
calculations

of the influence

of carbon dioxide concentration

on air

temperature, are apparently often explained not by the particular details of given modelsof climate theory, but by the features of different
atmospheric processes,which are taken into account only parametrically in all climate theories.A secondreasonfor possibledivergencesin
estimatesof climate changesconsistsin differencesof the values given
in the calculationsof such physical parameters as, for example, the
albedoof the earth-atmospheresystemwith and without a snowcover.
In a number of casescomparativelysmall variations of thesevalues lead
to significantdifferencesin estimatesof changesin the thermal regime.
In verifying the modelof the thermal regime usedin this bookby
the data on climatic changes in the past, we find that this model is
sufficiently precise (see Chapters 3 and 4).
These results also indicate that this model apparently takes into
accountall the factorsthat are sometimesthought to ensurethe stability of the contemporaryclimate. It is easyto understandthat otherwise
the calculationsof contemporaw climate changeswould give highly
overstatedvaluesof fluctuationsof climatic conditions.However,empirical data demonstrate that the results of calculations are similar to data
obtained from observations.

A similar conclusioncanbe drawn aboutthe majorclimatic changes

that occurredin the Quaternary period.
In discussingthe reliability of theseestimatesof the climatic conditions of the future, we must bear in mind that the laws for climatic

changesas man's economicactivity rapidly intensifies are of a special

THE

CLIMATE

OF

THE

FUTURE

217

In particular, Fig. 34 impliesthat oncethe polar icebeginsto thaw,

the rate of its decreasewill itself rapidly increase.This law allowsus to
obtain a comparativelypreciseestimateof the time necessaryfor complete thawing, even assuming great errors in the calculation of the
curve shown in Fig. 34. This time will change about 10%, which is
insignificant from the practical point of view, if the ordinates of this
curveare either doubledor halved.In otherwords,it is difficult o make
a great error in calculatingthe time at which the polar ice will thaw,
evenusinghighly schematicmodelsof climate, dueto rapid changesin
climatic conditionsas the heat barrier is approached.
Various hypothesesconcerningthe carbondioxideconcentrationin
the atmospherein the twenty-first century also scarcely affect the
results of the calculation (the hypothesisthat the carbondioxideconcentration after the year 2000 will remain constant increases the time
within which ice will thaw by a comparativelysmall value).
Sincethe resultspresentedhere provideonly a general picture of
the possibleclimatic conditionsof the future, it is necessaryto derive
new independentdata that couldbecomparedwith the resultspresented
here.

The needfor suchdata is primarily a functionof the greatpractical

importanceof future climatic changes.Data on the climate of the future

can change the outlook for the developmentof the most important
branchesof the national economyin many countries.

Theseresultsyield onlya generalandhighly schematized

pictureof
possibleclimaticchanges.Obviously,they must be complemented
with
more details, including, in particular, estimatesof changesof the hy-

drological cycle in different areas.

The onlymethodof obtainingsuchdetaileddata requiresextremely
general modelsof climate theory, which have been successfullydevel-

opedin studiesby Smagorinsky,Manabe, and their associates

(Smagorinsky, Manabe, and Holloway, 1965; Holloway and Manabe, 1971;
Manabe and Wetheraid, 1975),in studiesby Minz, 1965;Shvetset al.,
1970 and other authors.

This problemwill probablybe solvedin the comparativelynear

future in spite of great difficultiesinvolvedin the wide applicationof
modelsof general climatic theory. The recently derivedconclusions
that
the newglobalwarminghasstartedare ofimportancefor evaluatingthe
reliability of the results of the calculationsgiven here,
In the early 1970sit was universally acceptedthat over the last
decadesclimatecoolingdeveloped.
Sincethe signof temperaturetrends
changesrather seldom(fromthe endof the nineteenthcenturyup to the
middle of the twentieth century it changedonly once),almost all

218

CLIMATIC

CHANGES

investigators of climatic variations thought that in the immediate future the temperature lowering would continue. Lamb (1973)mentioned
that in the early 1970sthere were available more than twenty forecasts
of climatic changes which all predicted the cooling to be continued
during the next decades. He indicated that those forecasts had no
adequate scientific foundations. Only two years after this work had been
published Lamb obtained the first data pointing to the fact that the
coolingof climate would end (Lamb et al., 1975).Thesedata characterize
the thermal regime in the North Atlantic, where beginning with the
1970-71 winter an earlier prevailing tendency for lowering temperature
had changedto one of warming.
To study present-dayclimatic changes,the data on the air temperature of the Northern Hemisphere were supplementedwith materials for
recent years including 1975(Borzenkovaet al., 1976;Budyko, Vinnikov,
1976). Fig. 37 presents the mean air temperature variation for the
Northern Hemisphere for 1965-1975 (the curve A, temperature values
are subjected to five-year running averaging except for the last two
years). As is seenfrom Fig. 37, the mid-1960sbrought the temperature
increase which

accelerated

in the first half of the 1970s.

The aboveworks present data indicating that warming took place

in all latitudinal belts of the Northern Hemisphere, its amplitude
increasing with increasing latitude. The warming was accompaniedby
decreasing mean meridional temperature gradient and lowering the
mean precipitation amount in a number of regions of the mid-latitude
continents. It was connectedwith great droughts of 1972-1976in the
vast territories of Europe, Asia and North America. Considerable decrease in the area of polar sea ice in the Atlantic and Arctic was the
other consequenceof the warming.
The secondwarming of the twentieth century along with the consequencesindicated here was initially predicted (Budyko, 1972b)and only
later was it found from observationaldata. This showsthat it is possible
to use modern methodsof physical climatology for forecastingclimatic
changes.
It should be noted that Fig. 37 presents data on the most likely
mean air temperature variation of the Northern Hemisphere which
could occur up to the end of the twentieth century. This assumptionis
basedon computingthe CO2concentrationgrowth effect upon temperature (Curve T), taking into accountthe possibleinfluence of the atmospheric aerosolmass variation on temperature which is representedby
the interval between the curvesT, and T2 (Budyko, Vinnikov, 1976).
It is obvious that

this forecast

from observational data on air

can be checked in the near future

THE

CLIMATE

OF

THE

FUTURE

219

O.8

jj J

J j

ij jj!
J

i$ 8 o

19.0o

oo o

Fig. 37. Secular trend of anomalies of mean air temperature.

7.2 The Possible Ecological Crisis

Ecological Crises of the Past. The balanced nature of the biosphere was the basic feature of the Pre-Holocene epoch. This made
possiblethe prolonged existence of a number of speciesof living organisms for many millions of years.
Global disturbancesof the balancedecologyof the biosphereduring
critical epochsin geologicalhistory that were consideredin Chapter 5
were comparatively rare, occurringwithin intervals of time comparable
with the duration of geologicalperiods. Local disturbancesof the balancedecology,resulting from the isolation of ecologicalsystemof different scalesbeing violated, were far more frequent. An example of one
such disturbance

at the end of the Pliocene

after North

America

had

beenjoined to South America by the Isthmus of Panama was mentioned

in Chapter 5; an exchange of many members of the fauna of these
continents then took place.
Though changesin environmental conditionsstrongly

22O

CLIMATIC

CHANGES

the evolutionof living organisms,great disturbancesof the balanced

ecologyof the biospherewere comparativelyrare prior to the appearance of man, which is explainedby the high stability of Pre-Holocene
ecological
systemsbecauseof the mutual adaptationof their organisms.
It is well known that man's contemporaryactivity greatly influencesmanynatural processes
formingthe geographic
mediumthroughout nearly all regionsof the world.
The flora and fauna of our planet have varied quite strongly.Many
animal species
havebeencompletelyexterminatedby man, andan even
greaternumberof species
existwith sharplydecreased
populations
and
are in danger of becomingextinct. Characteristically,these species
includenot only nearly all the large land animals,but alsomany species
ofmarine animals,includingseamammals,whichhavesufferedparticularly from ruthless pursuit by man.
The plant coverof dry land has experiencedenormouschangesover

the greaterpart of the continents.Wild vegetationhasbeendestroyed

and replacedby farm land over large areas, the forestsstill existing
beingmostlysecondary
growth,i.e., they differhighlyfromthe natural
plant coveras a result of man'sinfluence.Great changeshave also
occurredin the plant coverof manysteppeand savannaregionsdueto
intensive grazing by domesticcattle.
Man's influence on the natural plant coverhas been particularly

acuteonthe soildevelopment
process
in specificregionsandhasledto a
changein the physicaland chemicalpropertiesof the soil. The soil of

farm land has been altered to an even greater extent as a result of its
systematiccultivation,the useof fertilizers,and the removalof a major
portion of the biomassof growing plants.
The influenceof man's activity on the hydrologicalregime of dry
land has grownrapidly. The run-offof, not only small, but alsomany

large, rivers has substantiallychangedas a result of human control

throughthe creationof water engineeringworks.A large part of the

water from river run-off has been removedto provide for the needsof

industryandthe urbanpopulationandto irrigatefarmland.The building of man-madelakes,whosearea in somecasesis comparable

to the
area of natural seas, has sharply changedevaporationand drainage
conditions over great areas.

Man's pollutionof the environment,whichhasspreadto the waters

of the continents,oceans,and atmosphere,is assumingan ever-increasing scale.

Until recently, the viewpointprevailedthat man's effect on the

environmentbeganto manifestitself at greater intensitiesbasically
over the most recent decadeswhen industrial development

THE

CLIMATE

OF

THE

221

FUTURE

increased and the size of the population of our planet began to grow
rapidly.
Recent studieshave demonstratedthat even thousandsof years ago
man's activity led to great changesin the natural environment, which
sometimesthreatened the continuing existenceof human society. The
first such case apparently resulted from the development of Upper
Paleolithic culture in Europe, Asia, and America.
It is known that the Upper Paleolithic was the first culture created
by modern man (Homo sapiens sapiens). The economicbasis of this
culture lay in hunting large animals by using toolsthat enabled man to
capture even such animals as the mammoth and woolly rhinocerous.
The Upper Paleolithic culture lasted quite long-more than half the
entire periodof existenceof modernman equal to 35,000-40,000years.
As we noted in Chapter 5, many large animals that had inhabited
Europe, Asia, and America became extinct during the Upper Paleolithic, which coincided in time with the last Wfirm. A number of
researchers have suggestedthat these animals were exterminated by
Upper Paleolithic hunters.
We should, however, note that most authors doubt that human
activity could have exerted a substantial influence at such an early
epochon the disappearanceof such large and widespreadanimals.
It has been noted in a survey of the literature on this question
(Butzer, 1964), that the animals that became extinct were highly specialized and thus not able to adapt to changesin the natural conditions
during the late Wfirm. Butzer also noted that a number of predators
that had not been the objectof systematichunting by primitive man also
became extinct at the end of the Pleistocene, at the same time as did the
large herbivores.
In addition, Butzer asserted that not all the hoofed animals of
Europe disappeared at this time and that until recently many herds of
large herbivoresremained in tropical lands. They continuedto live and
multiply in spite of being intensively hunted by the local population.
These considerations,in Butzer's opinion, contradict the viewpoint that
man could have exterminated a number of animal speciesat the end of
the Upper Paleolithic.
Nevertheless, Butzer recognized that the abundant remains of
large animals in many campsitesof the Upper Paleolithic indicate that
man could have exerted a definite
biomass of now extinct animals.

influence

on the total

balance

of

Colbert (1958, p. 459) has written, in discussingthe causesfor the

extinction of mammoths, Why did the mammoths becomeextinct? This
question, like so many of the questionshaving to do with problems

222

CLIMATIC

CHANGES

extinction, is extremely difficult to answer. In fact it is probablethat we

shall never know the real reason for the disappearanceof the mammoths a few thousand years ago, after their successfulreign through
Pleistocenetime. Very likely the extinction of the mammoths was the
result of complexcauses.Man may have had somethingto do with it,
but we can hardly believe that primitive man was of prime importance
in bringing an end to these numerous, widely spread, and gigantic
mammals."

It should be noted that existing studies have limited their analysis

of reasonsfor the marked change in the animal world at the end of the
Upper Paleolithic to qualitative statements difficult to prove or refute
without a quantitative treatment of the problem. For example, the high
specializationof animals that became extinct at the end of the Upper
Paleolithic

nevertheless

did not interfere

with their existence under the

sharp changesin the environmentthat repeatedlyoccurredthroughout

the Pleistocene.

The extinction of large predatorsat the sametime as that of a large

number of herbivores does not prove the absence of any influence of
man's activity on the change in the animal world, sincethe disappearance of these predators can be explained by a disturbance in their food
chains as specificherbivore became extinct both as a result of natural
causesand as a consequenceof man's activity.
The herds of large herbivoresthat still remain in the tropicsdo not
convincingly demonstrate that similar animals could have been exterminated in Europe by Upper Paleolithic hunters.
Let us discussthe possibleinfluence of man's activity on the extinction of animals at the end of the Pleistoceneon a quantitative basis,
taking into account the basic factors influencing animal populations
(Budyko, 1967).
If such a calculation confirms that this activity substantially influenced the process of extinction, other factors, including climatic
changes, may be consideredsecondary.
Modern man first appeared in Europe about 35,000 to 40,000 years
ago, where he replaced the Neanderthal man and created an effective
hunting system for large herbivores based on a more advanced technique of processing stone and bone tools. The population of Europe
greatly increased during the Upper Paleolithic, the level of material
culture noticeably increased, and the first stepswere made toward the
development of imitative arts (for example, the famous paintings of
animals in the caves of France and other European countries). The
Upper Paleolithic ended 10,000 to 13,000 years ago, which approximately coincideswith the end of the Wfirm

THE

CLIMATE

OF

THE

FUTURE

223

The end of the Upper Paleolithic correspondedto a significant

change in the pattern of life of the European population, indicated by
the spread of the Mesolithic culture of the Middle Stone Age; many
outstanding achievements of Upper Paleolithic culture, including the
art of cliff paintings, were lost to some degree at this time with the
appearance of a new technique of making stone tools.
The end of the Paleolithic may be said to be characterized by a
sharp break in the early history of mankind. The break was apparently
due to the cessation of systematic hunting for large herbivores, as a
result of which Upper Paleolithic man could no longer provide himself
with food and many important materials. Consequently, it was necessary for many to find new sourcesof supply for his existence.
Let us present here a computational scheme, allowing us to estimate how hunting influenced the population of pursued animals during
the Upper Paleolithic.
The schemeis basedon an equation that characterizesthe variation
in the population of an animal speciesover time:
dn
dt

- an - fin,

(7.1)

where n is the population of animals per unit area, dn/dt is the rate at
which population changes over time, e, is the birth rate (relative increase in population as a result of the birth of animals), and fi is the
death rate (relative decreasein population as a result of their death per
unit time).

The parameters e, and fi may depend on the population of animals

per unit area, i.e., on n, and these dependencesmay differ for different
animal species.
Calculations basedon equation (7.1) as well as on field studies have
demonstrated that the population of different animal species usually
fluctuates under natural conditions within rather broad ranges. These
fluctuations have periods commensurate with the lifetime of a single
animal generation, which can be explained by changes in food conditions caused by the weather conditions, outbursts of epidemics and
complex interactions between the populations of herbivores and preda-

tors, which have been shown by Volterra to be of the auto-oscillating

nature in a number of cases (Severtsev, 1941; Laek, 1954).

In addition, the mean population of a given animal specieswill be

more or less constant over a period of time markedly greater than the
mean lifetime of the animal under constant (on the average) external
conditions.Changes in this mean population can occur either as a result
of a change in the natural conditionsor in conjunctionwith the

224

CLIMATIC

CHANGES

tionary development of the given animal speciesor the speciesinteracting with it (Severtsev, 1941; Shmal'gauzen, 1946).
The laws of natural variations of an animal population considerably
change when the animals are being systematically hunted. In this case
equation (7.1) can be represented in the form
dn
dt

- an -/n

- g,

(7.2)

where g is the number of animals killed by hunters within a given

period of time.
Obviously, the coefficient/ that characterizes death from natural
causeswill be less than the coefficient/, other conditions being equal,
for animals that were pursued by hunters during the Paleolithic, since
hunters could more easily capture sick and weakened animals than
healthy and strong ones.
Since the products of hunting were the chief food source of Paleo-

lithic man in Europe, we should expect that g = m, where rn is the

populationof man inhabiting a unit area, and is the relative amountof
biomassof a pursuedanimal consumedby a single personwithin a given
period of time (ratio of weight of consumedbiomassto mean weight of a
single animal).
Thus equation (7.2) becomes
dn

dt

- an -/n

-m.

(7.3)

We should note that the "principle of interactions" proposedby

Volterra for studying the interrelation betweenpopulationsof predators
and herbivores

cannot be used to calculate

the decrease in the number of

animals resulting from systematic hunting.

In accordancewith this principle, the number of slain animals can
be consideredproportional to their population, which is a natural hypothesis for small animals who are able to conceal themselves from a
pursuer. Man, after mastering the technique of systematic hunting, was
probably always able to find his prey within the area he inhabited,
particularly such large animals as the mammoth, who lived in the open
spacesof tundra and steppe. Thus, the term rn in equation (7.3) can be
consideredindependent of n as long as the number of pursued animals
insured the current needs of a particular tribe over its area.
In calculating the population dynamics of large herbivores during
the Upper Paleolithic, it is necessaryto estimate the variation of population over this period. Obviously, this variation is determined by

THE

CLIMATE

OF

THE

225

FUTURE

equation similar to (7.1), i.e.,

dm
dt

- am - bm,

(7.4)

where a and b are the correspondingbirth and death rates.

These coefficientsdependon a number of factors,though in view of
the lack of suitable data, we will limit ourselvesto a rough estimate of
the mean difference of these coefficients, c = a - b, assuming it to be
independent of time and population density.
In this case, we obtain from equation (7.4):
rn = moect,

(7.5)

where rn is populationsize at time t, and mo is populationsize at the

initial

time.

Someidea of the populationdensity in Europeat the beginningand

end of the Paleolithic may be gained from data of recent ethnographic

studies of tribes at various levels of historical development.

Various

authors

who have used this method

have been able to

concludethat the populationdensity in Europe at the end of the Upper

Paleolithic fluctuated from roughly 5 to 50 peopleper 100 km2. These
values must be comparedto the much lower density of population of
those peoplethat did not master the technique of systematichunting,
which has been assumed to be about one person per 100 km on the
average (cf. Braidwood and Reed, 1957; Butzer, 1964).
Since a correspondingincreasein population occurredthroughout
the Upper Paleolithic, which lasted roughly 25,000years, we find from
equation (7.5) that c is between 0.64.10-4 and 1.56.10-4yr -1.
These coefficientsfor the population increase, extremely low from
the modern point of view, are characteristicfor early periodsof human
history. Their values agree reasonablywell with the low mean lifetime
of Upper Paleolithic man establishedfrom studiesof their skeletons.
Let us now calculate changesin the populationof a speciesof large
herbivoresthat inhabited the European tundra during the Upper Paleolithic.

If this speciesdominatedother herbivore speciesin a given locality,

it could use the basic mass of food resources, which, according to
available data, was sufficient for maintaining an animal population in

the tundra correspondingto roughly 800 kg of biomassper I km .

This value determined the upper limit for the possible number of
these animals. We may assume that this limit was reached under
favorable conditions for such animals as the mammoth, whose population was nearly independent of the activity of

226

CLIMATIC

Let us consider

how the number

CHANGES

of mammoths

must

have

varied

once man had begun hunting for them on a broad scale. This calculation
requires that we solve an equation obtained from (7.3) and (7.5)'
dn
dt

- an - ln -- moe
ct.

(7.6)

Multiplying all terms of the equation by the mean weight of a single

animal, we arrive at the form
dN
dt

-- OZJV- 131N- Fmoect,

(7.7)

where N is the biomass of the given animal per unit of occupiedarea,

dN/dt is the rate at which this biomass varies over time, and F is the
biomass of pursued animals consumedby a single person per year.
This equation must be solved under the initial condition N = No at t
- 0, under which the biomass of the animals at the start of the Upper

Paleolithic was equal to the limit determined by feeding conditions.

In order to simplify the calculation, we will assume that the right
side of equation (7.7) vanishes when Fmoet < aNo, since it is difficult to
estimate the coefficient/1, i.e., the mean number of animals remains
constant if lossesfrom hunting are less than the increase in biomass.
Then we shall neglect the term /iN when Fmoet increases to values
greater than or equal to aNo, and we will use equation (7.7) in the form
dN
dt

- aN-

Fmoet.

(7.8)

Obviously, both these simplifications lead to some understatement

in the calculation of the rate at which an animal population decreasesin
comparison with the actual state of affairs.

We can determine this value from the equation

Fmoe=

aNo,

(7.9)

denoting by the time during which hunters scarcely affected the

number of animals they pursued, from which we find that
I
aNo
= -In.
c

r/n o

(7.10)

We must solve equation (7.8) in order to determine the length of

time during which the mammoth population decreased and finally
became completely

THE

CLIMATE

OF THE

FUTURE

227

AssumingN = No when t - , we obtain a dependenceof the

biomassreserve of given animals on time:

N Fmect+eC(t-"[No
Fmo
]
= --

o-c

o-c

(7.11)

From this equation, we find the period of time after which N is

equal to zero, which corresponds
to the extinctionof the given animal
species.

A calculation using these equations of changesin the mammoth

population during the Upper Paleolithic requires, in addition to the
parameterswhosevalues are given above, knowledgeof the values of
the coefficients a and F. The first coefficient characterizes

the ratio of

number of mammothsborn in a given year to the total populationof the

herd. This coefficientis 0.05 accordingto data for modern elephants.
The coefficient F can be approximated from data of ethnographic
observationsfor northern tribes that chiefly feed on large animals. Since
F characterizes

the direct use of biomass for food and for economic

purposesas well as unproductivelosses,it may be estimatedas roughly

500-1000kg/yr per person.This figure for biomassconsumptionhas been
consideredsomewhat overstated by Bibikov (1969), who acceptedthe
basic conclusionsof our calculation. Though we agree that the actual
use of animal meat and fat was much less than this value, our estimate

is neverthelesshighly probabledue to significantbiomassconsumption

for economicpurposes(animal fat, hide, and bones)and due to its great
losses,particularly associatedwith the difficultiesof prolongedstorage.
We calculated the variations of mammoth population for two values
of c (i.e., 0.64-10-4 and 1.56.10-4 yr -) using the above equations and
assumingthat F = 750kg/yr and letting No = 800kg/km2 and mo - 0.01
km - in accordance

with

the above estimates.

If we calculate using equation (7.10), it becomesapparent that the

mammoth population varied comparatively little for 10,000 to 25,000
years (dependingon the given value of c) after they beganto be hunted.
The mammoth populationthen began to decreaserapidly as is evident
from equation (7.11), and the mammoths became completely extinct
within several centuries after the end of the period of stable population.
In evaluating the resultsof this calculation,we must bear in mind
that its precisionis limited by the approximatenature of the parameters
used. In addition, our equations can be seen to imply that significant
variations in the values of the parameters do not change the general
laws of the processand do not weaken the basic conclusionthat

228

CLIMATIC

CHANGES

large animals as the mammoth could have become extinct within a

period of time commensurate with the duration of the Upper Paleolithic.
It can be easily demonstrated that this conclusiondoesnot change if
we replace the dependenceof population on time used above by simpler
linear

relations.

We note that our conclusion

is confirmed

also whenever

a given animal species (for example, the mammoth) is not the only
herbivore subsisting on a given food base. For this purpose we must
assume that the ratio of the biomass of a given species to the total
biomass of the pursued animals approximately correspondsto the similar ratio for the biomass of animals slain by hunters.
Our conclusioncan also be obtainel without any complicated calculations, simply by comparing available data on the biomass balance of
large herbivores in Europe during the Upper Paleolithic. A comparison
of the annual increment of this biomass to the size of the population
occupying this area is of great importance here.
If the increment was at most 4000 }g per 100 }r 2, as in the case of
the mammoth, it becomes obvious that it was sufficient to maintain a
population of at most several people per 100 }r 2 even if it was completely usedby the hunting tribes. Clearly, however, the populationwi]]
have continued to grow after reaching this size, since for a long time
there was a sufficient amount of food not only due to the increment in
biomass, but also as a consequenceof the extermination of part of the
main herd of hunted animals. Finally, prior to complete extermination
of the animals, the increment in biomasswi]] gradually decrease,which
is in accordance

with

the results

of our calculation.

We may supposethat oncethe mammoths ha(] been eliminated by

man over most of man's area, subsequentherds could have been ]eft in
underpopulated regions in Northern Asia. However, in regions with
such a severeclimate, large herbivorescould have a supply of foodonly
during relatively warm periods. A succeedingclimatic change toward
cooling would have led to the extinction of the mammoths in these
regions as we]].
Using these equations, we may explain why primitive man was
able to eliminate the mammoth but not the kindred elephants, which
until recently were numerous in a number of tropical regions.
It is evident from equation (7.10) that the time an animal species
has ]eft to it, if it is being hunted down, dependson its possiblebiomass
reserve F and the relative use of this biomassby hunting tribes. Available data show that the biomass of large herbivores in tropical savannas
is 10 times the correspondingbiomassin tundra. In addition, in tropical
latitudes primitive man's food could have included a much

THE

CLIMATE

OF THE

FUTURE

229

amount of vegetation than did the food on which Paleolithic hunters in

Europe subsisted.Here it was limited to game, due to glaciation. Since,
for hunters in the tropics, F is at least half the value we have assumed

for Upper Paleolithic conditionsin Europe, and taking the value of No

correspondingto savanna conditions,we find from equation (7.10) that
for the tropics, must have been roughly three times that found above
for Europe during the Paleolithic epoch.
Thus the population of elephants, even when constantly hunted,
couldhave been maintained at a high level for several tens of thousands
of years after mammoths had been exterminated.
This theory allows us to explain why only the largest herbivores
becameextinct in Europe during the Upper Paleolithic. Since the population of any speciesof these animals was limited by the foodresources
of the given geographiczone, the increment in biomassof a specieswas
thereby determined in turn by the birth rate a. It is well known that
birth rate is minimal for the largest animals and increases with decreasing animal size. Species of large herbivores whose biomass increment
per unit area was least were therefore the least protectedagainst Upper
Paleolithic hunters. The causesof the biological catastrophe that overtook this speciesis therefore quite simple.
The increase in the size of large herbivores in the course of their
evolution freed them to a certain degree from the threat of predatory
attack, which made a significant decreasein birth rate both possibleand
expedient. When Upper Paleolithic man began to hunt for these animals, he put himself in the role that wolves and other predators hold
toward the smaller herbivores. However, while the smaller herbivores
could compensatefor their lossesby means of a high birth rate, this was
not possible for large animals, and it was a matter of time before their
extinction by pursuit, a pursuit for which their evolution had not
adapted them.
It should be noted that this explanation differs from the one given
by Simpson (1944), who suggested that large animals experienced
greater difficulty in adapting to the changing natural conditionsof that
time due to their longer lifetime and a comparatively slower change of
generations.

The actual periodof time from the start of systematichunting to the

point at which large herbivoresbegan to die out in different regions of
Europe was probably significantly briefer than that which can be obtained from a calculation using our scheme.This discrepancyis due to
the assumptions made in the calculation and also to the fact that
fluctuations in animal population as result of a epizooticdiseasesand,

CLIMATIC

23O

CHANGES

particular, changes in the food base, which could have been extremely
significant under the conditions of the last Wfirm, were not taken into
account

in the calculation.

The existence of such fluctuations could have accelerated the disappearance of various large herbivores within a fixed region over the
course of many thousands of years. On the other hand, it would have
been difficult for a speciesto re-establish its population in this region by
migration from other regions, since Upper Paleolithic hunters would
have continued to pursue it.
From this point of view, the largest animals with lowest birth rate
would have vanished before other animals in the course of systematic
hunting. Their disappearancewould have facilitated a temporary increase in the population of smaller herbivores, their food base increasing. However, a further increasein the rate of hunting couldhave led to
the disappearanceof herbivores with a high birth rate as well, which
was, in fact, the case of reindeer in a number of regions of Western
Europe. We should mention here that lossesdue to man's pursuit for
these animals were complementedby substantial lossesdue to pursuit
by predators, which reducedthe stabilizing effect of a high birth rate for
maintaining the population of a species.
A number of other results of this computational schemeshould also
be noted.

The processin which the population of pursuedanimals decreased

was at first slow and gradually accelerated,and the last, still comparatively numerous herds of animals were exterminated quite rapidly. In
addition, more or less long periodsduring which pursued animals were
absent as a result of natural fluctuations of their population, migration
of different herds, and so on must have occurredlong before complete
extermination of any animals.
Thus, as the population of Europe increasedduring the Paleolithic,
the existenceof this population,which basically dependedon systematic
hunting, was quite often threatened. This accountsfor the unusual
attention Upper Paleolithic man paid to animal paintings, which were
of the nature of magic, as a way of reviving vanished herds. However,
the general processof extermination of large herbivores increasingly
acceleratedwith systematic hunting and its final phasewas particularly
rapid. This put the hunters of a tribe in a difficult position, sincethey
lacked sufficient time to gradually switch over to other methodsof food
gathering. Thus the transition to the Mesolithicwas probablya painful
experiencefor primitive societyand may have been accompaniedby a
temporary decreasein the population.
As we noted above, the end of the Paleolithic in America was

THE

CLIMATE

OF

THE

FUTURE

231

accompaniedby the extinction of large animals, which confirms the

causal relation between these phenomena. It is also entirely likely that
important changes in natural conditionstoward the end of the Paleolithic presentedmore difficulties for the existenceof animals that were
becomingextinct and helped to acceleratean already ongoingprocess.
Thus Paleolithic culture ended in Europe partly as a result of an
insoluble conflict between the technique of systematic hunting of large
animals that Upper Paleolithic man had invented and which both
ensured a temporary abundanceof foodand made possiblean increase in
population, and the limited natural resourcesthat could be hunted in
this way and which were in fact exhausted in the courseof time.
We may therefore ask ourselveswhat could have disturbed the
balanced ecology of the Pre-Holocene biosphere, an ecology in which
animal speciespursued by predators may have existed for hundreds of
thousandsand millions of years.
The chief causeof this disturbancemay have been man's unusually
high rate of evolution in comparisonwith the rate of evolution of his
prey.

Large animals with a low birth rate were less able to maintain their
population at a stable level when pursuedby predatorsthan were small
animals with a higher birth rate.
Large body size therefore gave animals an advantage in the struggle for existenceonly when it made animals less accessibleto attack by
predators. But if animal size did not protect them from attack (as was
the caseeven for mammoths during the Upper Paleolithic), an increase
in birth rate would be the only method for maintaining their population.
Such a change of pursued animals at the rate of evolution characteristic of the Pleistocenerequired far more time than the period it took
for their

extermination.

We noted in Chapter 5 that a similar approachwas used by Martin

(1973) to study the extinction of many speciesof large animals in North
and South America

at the end of the Pleistocene.

It is known

that man

appeared on these continents comparatively late (at the end of the last
Wiirm glaciation), and that the processby which America was settled
has not been studied very thoroughly. Martin proposedthat a small
group (about 1000 people) crossedthe ice of the Bering Strait about
11,000years ago and reached the northwestern regions of North America, where they found enormousherds of large animals that had never
encountered man and which lacked protective instincts. Thus the hunting conditionswere more favorable in America than in Europe or Asia,
where wild animals had been in contact with man at various stagesof
his evolution for hundreds of thousands of years, during which

232

CLIMATIC

CHANGES

these animals had developedvarious protective mechanisms against

pursuit by hunters.
As a result, the first small group of inhabitants of America had an
unlimited amount of foodat their disposaland, after rapidly increasing
their population,settledall of America, exterminatingmostof the large
animals they encountered.Martin's calculationsdemonstratedthat the
wave of first settlers could have reached Tierra del Fuego 1000 years
after completelyexterminating about 30 genera of large animals still
existing in America at the end of the Pleistocene(mammoth, mastodon,

megatherium,horse,camel,andmany others).The huntersofthe tribes

were forcedto switch over to other methodsof foodgathering oncethese

animals had been exterminated, which led to a sharp decreasein the
populationof America.

This problemwassubsequently
studiedby Mosiman(personalcom-

munication), who investigatedthe time at which large animals pursued

by Upper Paleolithichuntersbecameextinct, as a functionof variations
of the parameters and the relations of this model. He thereby established that the basic conclusions obtained

in these calculations

remain

in forcewhether the parametersof the modelare varied within broad

limits or if a number of modifications are introduced into the model

itself in order to refine its initial assumption.

Results of these investigationshave indicated what was the probability that an ecologicalcrisis materialized at the end of the Upper
Paleolithic, spreadingover a large portion of the area populatedby man.

This crisisapparentlyled to a decreasein man'spopulationwherever

the large animals that had been the basis for the Upper Paleolithic
economybecame extinct.
It is quite possiblethat in a number of casesthis crisisresultedin a
gap between the end of Upper Paleolithic culture and the emergenceof
Mesolithic culture, which has been noted by archaeologicalinvestigations in Europe and other regions.
The idea that the first and longestculture in the history of modern
man ended in an ecologicalcrisis is of interest as an illustration of the
theoretical incompatibility betweenman's spontaneousactivity and the
maintenance of a balanced ecology.Other casesare well known when
the balancedecologyof the biospherewas disturbedin the most recent
epochin the historyof the developmentof human society.Thesecases
confirm the statement made by K. Marx that culture, if it has developed
spontaneouslyand is not consciously
directed,leavesa desertafter itself
(K. Marx and F. Engels. Letters. Works, vol. 32, Moscow,1964,p. 45)
Ecological Crisis of the Future. Though man's contemporary
activity, like his activity in the past, has substantially changed

THE

CLIMATE

OF

THE

233

FUTURE

environment over a large part of the planet, until recently these

changeswere only the sum of many local affectson natural processes.
They assumeda planetary nature not as a result of man's alteration of
natural processes
on a globalscale,but becauselocal (or regional) effects
have spreadover large areas. In other words, a changeof the fauna in
Europe and Asia did not affect (or affected only slightly) the fauna of
America, control of the run-off of American rivers did not change the
run-off of African rivers, etc. Man's effect on global natural processes
began only very recently and this change has influenced the natural
conditionsof the entire planet (Gerasimov and Budyko, 1974).
The hypothesishas been recently put forward that a further developmentof man's economicactivity couldmarkedly changethe environment, leading to a general crisis in economicsand a sharp decreasein
man's population.
This point of view has been substantiated in greatest detail by
Forrester (1971) and Meadows et al. (1972).

As was done in a previous work dealing with the economicsof the

Upper Paleolithic (Budyko, 1967),Forrester and Meadowsand his colleagueshave solvedan equationthat determinesthe increaseof population consistentwith a change in the resourcesutilized by man. Like the
calculation for the Upper Paleolithic, the analysis carried out by the
American

scientists

for the modern

era led to the conclusion

that

further growth in population will issue in a crisis that will sharply

decreasepopulation.Accordingto this study, sucha crisiswill manifest
itself in only about 100 years.
Models for the modern era have taken

into account a number

of

factors influencing population size, including food and mineral resourcesand environmental pollution. Since the model for the Upper
Paleolithic was limited to food resources (reserves of mineral raw mate-

rials did not then affect population and environmental pollution did not
exist at all), a calculationusing this model couldbe carried out analytically. The far more complexsystemsof equations usedby the American
scientists could be solved only numerically by using computers.
Let us discussin detail the results of the prediction of changes in
economic conditions based on the use of Forrester's

model.

This calculation was carried out for the period between 1900 and
2100 and used dependencesof birth and death rates on various factors
obtained from contemporary empirical data.
When only scattered data were available (for example, for estimating the influence of environmental pollution on the death rate), the
correspondingdependencesfor the calculation were hypothesized.
Results of the calculation demonstrated that industrial

CLIMATIC

234

CHANGES

and the productionof foodper capita will growthrough the beginningof

the twenty-first century, after which they will rapidly fall far below
their level at the beginning of the twentieth century.
Reservesof mineral resourcesat the beginning of the twenty-first
century will decreasesharply and will be practically exhaustedby the
middle of the century. Environmentalpollutionwill rapidly growin the
first half of the twenty-first century and will again decreasetoward the
middle of the century due to a drop in industrial production.
Populationwill grow through the middle of the twenty-first century. It will then beginto decrease,and will be nearly halvedby the end
of that century.The exhaustionof mineral resources,whichwill lead to
a sharpincreasein the death rate due to a decreasein the foodproduction and a worseningin the medical care of the population,will be the
principal factor in this crisis.
Results of this calculation

in the book Limits

to Growth

were

comparedto a number of other calculationsthat employedvarious

additional assumptions.However, in mostcasestheseassumptionsdid
not changethe basicconclusionobtainedin the initial calculation.
For example, the assumptionthat currently known reservesof
mineral resourceswill doubleleads to a still sharper drop in population
by the middle of the twenty-first century,sincein this casethe level of
environmental pollution will catastrophicallyincrease.
The possibility that environmental pollution will be effectively
controlleddoesnot excludea dropin populationdue to insufficientfood
resources.

Resultsfrom subsequentcalculationsshowthat a complexof measuresis required in order to keep populationfrom decreasingand to

maintain a high level of industrial production.Theseinclude a halt in
the growth of population,a sharp decreasein the use of mineral resources,a cessationin the growth of industrial production,strict control
of environmentalpollution,an increasein foodproduction,and others.
Such calculations imply that measures for stabilizing economic
conditionsin the twenty-first century must be introducedrapidly, and
not put off to the end of the twentieth centurywhen it will be toolate to
prevent the onset of the crisis.
These authors have indicated that the impending economiccrisis
may end in "socialdisintegration,"i.e., in the breakdownof the existing
organizationof society,sinceit doesnot seemvery realistic to assume

that these measures will be carried out in the modern era.

A critical evaluation of these studies has been given by Fedorov

(1972). In his book he noted the fundamental drawbacks of the above

studies,includingespeciallythe fact that the numericalmodelsused

THE

CLIMATE

OF

THE

NUTURE

235

not take into consideration those qualitative changes that are introduced by technical progress into the economicdevelopment of society
and into the manner

in which man's demands

are satisfied.

In addition, Fedorov indicated that the conclusionsof the American

authors regarding the impending crisis in the interaction between society and nature correctly reflect the laws of spontaneousdevelopmentof
bourgeoissociety. The fundamental social and economicfeatures of the
capitalistic system prevent the rational use of natural resources,ham-

per the solution of the food problem, and complicatethe systematic

struggle against environmental pollution. Under such conditions, a

crisis in man's interrelation with nature is unavoidable and many
features of this crisis can be clearly seen in the modern era.
Fedorov

notes that

the difficulties

that

will

arise with

mankind's

technologicaland economicdevelopmentcan be overcomeonly through

a planned socialist economy.
The conclusionthat long-term economicplanning is necessarycan
also be found in the work of American authors, though the measures
proposedhere for preventing the onsetof the crisis are clearly Utopian.

It may be possibleto find more realistic ways of overcomingthese

difficulties

even within

the framework

of the limited

numerical

models

used in their studies. For example, simple calculations demonstrate

that most of the obstaclesto technological and economicgrowth will
have been eliminated in the next century by greater control of environmental pollution, massiveuse of syntheticmaterials, and at least partial advancesin the chemical synthesis of food products.
It is necessaryto study how currently available methods of numerical modeling of economicdevelopmentcan be perfectedin order to solve
particular problems in long-term planning of technical progress.
We can formulate in advancethese mechanical problemsby using
numerical modelsthat allow us to predict probablechangesin population as economic development proceeds and as natural conditions
change.
One of the most important of such problems is that the global
climate of our planet may be modified as a result of economicactivity, a
question that was not consideredin the American studies.
This is particularly important, since such a change will substantially influence man's economicactivity earlier than any other global
ecologicaldisturbances.
This influence will be felt on a large scale. We need only note that
the warm period of the 1920sand 1930swas accompanied,as can be seen
from the data of Chapter 3, by a noticeable decreasein the amount of
precipitation in a number of intracontinental regions. This

236

CLIMATIC

CHANGES

tially worsenedconditionsfor agricultural production, river transport,

and other branches

of the national

economies of a number

of countries.

The data of this chapter imply that the influence of man's economic
activity on climate can lead in the comparatively near future to a
warming comparablewith that of the first half of the twentieth century,
and will then greatly exceedit. Thus, a climatic change can be the first
indicator of the global ecologicalcrisis mankind will encounter in the
courseof fundamental developmentsin technologyand economics.
The redistribution of the amount of precipitation in various regions
of the world, in conjunctionwith a noticeabledecreasein many regions
of unstable humidity, will be the chief cause of this crisis in its first
stage. Sincemost grain cultures are grown in these regions,a changein
precipitation conditionscan substantially complicate the problem of
providingsufficient foodsuppliesfor the rapidly increasingglobal population.

For this reason, the task of overcoming undesirable changes in

globalclimate becomesone of the mostimportant ecologicalproblemsof
the contemporary age.
7.3

Climate

Modification

Methods of Affecting Climate. Different measures have been

taken in order to minimize unfavorable climatic changes arising from
man's economicactivity; atmosphericpollution is now being controlled
on a broad scale.

It should be noted that the level of air pollution has recently

decreased in a number

of cities as a result of numerous

measures taken

in many developedcountries, including air purification at industrial

plants, on vehicles, by heating units, and so on. However, in many
regions air pollution has increased. This indicates how great are the
difficulties involved in preventing a growth in the amount of anthropogenic aerosolin the atmosphere.
It would be even more difficult (and the relevant problemshave yet
to be raised) to prevent an increase in the carbon dioxide content in the
atmosphereand a growth in the heat releasedin transformations of the
energy used by man.
There are no simple technical means for solving the first and,
especially,the secondproblem, other than limits on the use of fuel and
most types of energy, which is probably incompatible with further
technical progressover the next decades.
Thus in the near future climate modification will becomenecessary
in order to maintain current climatic conditions. Obviously, if

THE

CLIMATE

OF

THE

FUTURE

237

becomespossible,it could also prevent natural climatic fluctuations

unfavorable to the national economy, and, subsequently, modify the
global climate in the interests of mankind.
Possibleways of affecting climate have been discussedin a number
of workscontainingdifferent ideasfor modifyingclimatic conditions.
It has been repeatedly noted by Fedorov (1958, 1972)that modificationsof large-scaleatmosphericprocesses
characterizedby low stability,
i.e., that are highly sensitiveto comparativelysmall changesin external factors,are the mostpromising.Suchan approachhas yielded good
resultsin developingmethodsfor dealingwith clouds,precipitation,fog,
and other weather phenomena.

Fedorovgave an interesting example in which the radiation balance of a large area was noticeablychangedby cloud seeding.Similar
experimentshave demonstratedthat weather modificationcan change
climatic conditionsif applied over a long period of time. Thus Fedorov
stated that it was theoretically possibleto modify the climate of the
entire globeor a significant part of it.
Various projectsof weather modificationare exploitedin a number
of studies.

One of the largest suchprojectswas intended to eliminate Arctic ice

in order to increasethe air temperature at high latitudes. A number of
studiesof the relation betweenthe polar ice regime and general climatic
conditions have been carried out to deal with this question (Budyko,
1961, 1962; Fletcher, 1966).
The results of these studies do not exclude ice-free conditions once

the polar ice is eliminated, but apparently indicate that suchconditions

are highly unstable.
It should be noted that very few methods of actually eliminating
polar ice have been developed.A number of calculationshave shown
that Arctic sea ice may thaw within several years if the cloud cover
decreasesin the northern part of the Atlantic ocean and if the rate of
evaporation from the ocean surfacedecreaseshere. Obviously, though
there are certain methodsof controlling cloudsand evaporation from
water reservoirs, it is quite unrealistic to expect that these actionswill
be carried out over vast areas. Some authors believe that polar ice can
be thawed by building a dam in the Bering Strait together with a
powerful system to pump water from the Bering Sea into the Arctic
Ocean. Other authors suggestthat polar ice can be broken up using
atomic energy.
As we indicated above, the disappearance of polar ice will have a
complexinfluence on climate and will not be favorable in all respectsto
human activity. Not all the consequencesof this processcan now

238

CLIMATIC

CHANGES

predictedto a sufficient degree of accuracy.We may therefore suggest

that even if it becomespossibleto thaw the polar ice cover, it will not be
advisable

to do so in the near future.

Methods of altering large-scale atmosphericmovementsis another

possiblyfruitful way of modifying climatic conditions.Until recently, it
was assumedthat movements of air massesinvolve such great energy
consumptionthat man's influence on them couldnot yield any results at
the present time. However, in many cases atmospheric motions are
unstable so that they can be affected using a comparatively small
amount of energy. It is thus of interest to determine how the intensity of
a low-level cyclone could be changedby using a system of vertical air
flows created over the cyclone-formingarea. These calculations imply
that such effects are theoretically possibleat the contemporarylevel of
developmentof energetics,though the technical means for it can be
describedonly in most general terms (Yudin, 1966).
An analysis of a number of possibilitiesfor climate modificationwas
generalizedin the summary report of the Panel on Weather and Climate
Modification of the Committee on Atmospheric Sciencesof the USA
(Weather and Climate Modification, 1966). Some methods of microclimate modification referred to in this report related to agrometeorological problems. These include different methodsof protecting plants from
night frosts, shading plants in order to protect them from overheating
and excessmoisture evaporation, planting of forest strips, and so on.
Obviously, these methodswill not affect the climate of large areas, and,
except for protective forest strips, they do not ensure a permanent
change of local climate.
The idea of modifying the climate of dry regions was consideredfor
the caseof the desertsof southern Peru, which are situated near gently
sloping mountains. Air descendingalong these slopes is heated and
becomesless saturated, which impedescondensationand precipitation.
It was proposedthat the desert surfacebe irrigated with a small amount
of water in the morning, which would lead to a drop in surfacetemperature. This would result in an improvementof conditionsfor the development of daytime convectivecloudinessand precipitation.
Another plan for modifying precipitation conditionswas suggested
for deserts with air characterized by a high dust content. The air is
cooledbecause of the effect of massesof dust on long-wave radiation,
which impedes the development of ascendingcurrents and decreases
precipitation. It was suggestedthat the dust content of the atmosphere
couldbe decreasedin these desertsby restoring the plant cover on their
surface,after which a long-term re-establishmentof the plant cover

THE

CLIMATE

OF

THE

FUTURE

239

be made possibleby an increase in the amount of precipitation. Such a

project was developedfor the desertsof northwestern India, which are
apparently the result of the destruction of the plant cover by excessive
cattle grazing (Bryson, 1972).
Other plans for climate modification have been mentioned in a
number of publications. These include affecting oceancurrents through
the construction of enormous dams. For example, it has been proposed
that the regime of the Gulf Stream couldbe modified by constructinga
dam between Florida and Cuba, or that the direction of the Labrador
Current could be modified by constructing a dam on the Newfoundland
Bank. None of these plans have been sufficiently argumented from a
scientific point of view and it remains quite unclear how they may

influence

climate.

Other plans include proposals for large water-storage basins.

Among them is the old idea of creating enormousseasin Central Africa
and the Sahara by damming up a number of rivers in Africa. These seas
would occupy10% of the total area of the African continent. We should
note, leaving aside the question of whether such a project is possible,
that none of the climatic changesthat could result have been studied to
any great extent.

It is probable that some of these plans of climate modification over

limited areas will becomefeasible for the technologyof the near future if
they prove to be profitable for mankind.
Far greater difficulties are encountered in modifying the global
climate, i.e., the climate of the entire planet or of a large part of it.
A method based on increasing the aerosol content in the lower
stratosphereis apparently the most accessibleone for modern technology (Budyko, 1972b, 1974;Budyko et al., 1974).
Modification of the Aerosol Layer of the Stratosphere. It has
been established in a number of studies by Jung (1963), Rozenberg
(1968), Cadle (1972), and Karol' (1973) that the lower stratosphere
contains an aerosol layer whoseparticles measure tenths of a micrometer and which consistschiefly of sulphur compounds.
Estimates were given in Chapter 3 for the influence of this layer on
the flux of short-wave

radiation

that reaches the earth's surface. These

estimates imply that comparatively small fluctuations of the aerosol

mass in the lower stratospherenoticeably vary the meteorologicalsolar
constant, and, correspondingly, the mean temperature at the earth's
surface.

Let us consider the possibility of changing climatic conditionsby

increasing the concentration of aerosol particles in the lower

240

CLIMATIC

CHANGES

sphere.For this purposewe will calculatethe aerosolmasswhich has to

be ejectedinto the stratosphere

in orderto inducea climaticchangeof

the scaleobservedduring the warming periodof the 1920sand 1930s,but

in the oppositedirection,i.e., ensuringa dropin meantemperatureat
the earth's

surface.

It is evidentfrom the data of Chapter3 that this will requirea 2%

decreasein direct radiation, which correspondsto a 0.3% decreasein

total

radiation.

Since the mean value of the aerosol mass that decreases the total

radiation by 1% is equal, accordingto the data presentedabove,to

0.8.10-6 g/cm2, we find that it is necessaryto increasethe aerosolmass
in the northern hemisphereby 600,000tons.

Different technicalmeansare possiblefor increasingthe aerosol

mass.Let us consideronesuchmethod,basedon the sulphatenature of
stratosphericaerosolparticles.
To increasethe concentrationof sulphateparticles,we needonly
deliver a fixed amount of sulphur dioxide to the level of the aerosol
layer, whichwill then be convertedinto dropletsof sulphuricacid.The
creationof 600,000tons of sulphuricacid requiresabout400,000tons of
sulphur dioxide, which can be formedby burning 200,000tons of sulphur.

We will assumethat the mean lifetime of aerosolparticlesin the

stratosphere
amountsto two years.It will then benecessary
to annually
deliver 100,000tonsof sulphurinto the lower stratospherein orderto
maintain the necessaryconcentrationof aerosolparticlesthat consistof
pure sulphuric acid. However, since concentratedsulphuric acid is
hygroscopic,
dropletsof it will absorbwater vaporin the air, whichwill
increasetheir massand decreasetheir specificweight. This effectwill
leadto a significantlesseningin the amountof reagentnecessary
for a

given decrease in total radiation. Since the exact concentration of sul-

phuricacidin stratosphericaerosoldropletsis unknown,we will take its

valueas 50%.In this case,about40,000tonsof sulphuris required.This
amountsto lessthan half the massof sulphur indicatedabovedue to a
decreasein the specificweight of the aerosolas the sulphurdioxideis

diluted

with water.

Sincemodernhigh-altitude aircraft carrying a loadof about 15tons

can reach the level of the aerosollayer, this massof reagent can be
transportedto the lowerstratosphere
by severalaircraft operatingevery
day equippedwith a devicefor burningsulphurin the atmosphere.
The amountof reagentintroducedinto the atmosphereby meansof
this method is negligibly smaller than the mass of matter that enters
the atmosphereas a result of man'scontemporaweconomic

THE

CLIMATE

OF

THE

FUTURE

241

This mass is 10-4 of that due to man's activity, which, according to

contemporary data, constituteshundreds of millions of tons per year
(Inadvertent Climate Modification, 1971).

The amount of reagent finally settling on the earth's surfacewill be

equal in this caseto about 0.2 mg of sulphur per square meter per year,
which is roughly 10-3 of the amountof sulphur that is naturally washed
out of the atmosphere by precipitation (Budyko, 1974). Obviously, such
amounts are not at all important in environmental pollution.
Though our calculationsare highly approximate, refinements in
succeedingstudies will probably not change the theoretical possibility
for climate modification by means of contemporarytechnical methodsof
aerosol control in the stratosphere. Such modification will be possible
even if it is necessaryto increasethe amount of reagent ten times, which
exceedsthe probable error of the influence of aerosol concentration on
the radiation flux and of the meteorological solar constant on the air
temperature at the earth's surface.
The method of transporting the reagent into the stratosphere by
airplanes is probably not the only one;other technical means, including
rockets and different types of missiles, are also possible.
It is necessaryto investigate the prospectof using other reagents
besides sulphur for increasing the mass of stratospheric aerosol. The
possibility of lateral effectsof the reagent on physical processesin the
stratosphere, in particular, on the ozone layer must be studied. This
influence is comparatively slight in the range of sulphate aerosol observed under natural conditions. However, it may become substantial
when its concentration reacheshigher levels or when new reagents not
found under natural conditionsare introducedinto the stratosphere.
The choiceof the region where the reagent is scatteredis of limited
importancesincedata on the dispersionof productsof volcaniceruptions
demonstrate that reagent from any point outside the tropical zone
rapidly spreadsover the entire hemisphere. Circulation in the lower
stratosphere can be of importance in selecting optimal regions and
periodsof time for ejectingthe reagent to ensureits mosteffectiveuse.
These measuresof climate modification are intended for preventing
or weakening climatic changesthat may ensue in several decadesas a
result of man's economicactivity. Such modification, however, is not
beyond the capacity of modern technology. In the near future it will
apparently be possibleto modify the climate, creating climatic changes
ten times greater than the fluctuations of the 1920s and 1930s, i.e.,
producing a drop in global temperature of several degrees.
Effects on such a scale may be necessary in the twenty-first century, when the significant growth in energy productionmay lead to

242

CLIMATIC

CHANGES

substantial increase in the temperature of the lower layers of the

atmosphere.Under these conditions,a decreaseof the transparencyof
the stratospherewill prevent undesirableclimatic changes.
We shouldnote that it may be feasibleto modify the aerosollayer of
the stratospherein the near future, while the influence of man's economic activity on climate will still be comparatively slight.
Modifications that decrease the mean temperature of the lower
layers of the atmosphereby several degreeswill lead to an enormous
changein natural conditionswhich may be quite unfavorablefor man's
agricultural activity over the greater part of the globe, and thus in the
near future make it impossibleto continue this activity.
It is quite realistic to expect that modifications which decreasethe
mean temperature by several tenths of a degree will be economically
feasible.As we noted above,the total amount of precipitationin many
middle-latitude regions characterizedby insufficient moisture will noticeably decreasewith suchclimatic fluctuations. The ability to prevent
such climatic changescan therefore have practical importance.
These modificationscan apparently be made without any forecasts
of natural changesin the concentrationof atmosphericaerosol,using
instead

current

information

about this

concentration

obtained

from

high-altitude aircraft soundingsor from ground actinometric observations. Since, under natural conditions,the mean aerosol concentration
in the stratosphere varies comparatively slowly, the rate at which
modificationsare utilized to maintain this concentrationat an optimal

level can be controlled.

Let us now discussthe possibilityof a climatic change resulting

from large numbersof aircraft flights in the stratosphere.This has been
dealt with in a collective monographby American scientists (Man's
Influence on the Global Environment, 1970), where it was pointed out
that about 500 aircraft flying at altitudes between 18 and 20 km will be
used in civil aviation in the years 1985to 1990.The end-productsof the
fuel consumptionof these aircraft will annually supply 63,000 tons of
sulphur dioxide to the lower stratosphereof the entire globe, which
correspondsto a massof sulphate particles of about 100,000tons.
In consideringthis questionwe should bear in mind that, as noted
above,a long-termchangein radiation will exert a far greater influence
on the thermal regime of the atmospherethan a similar changeover a
short period of time.
Since most of the stratosphericflights will occur in the northern
hemisphere, the annual increase in the mass of sulphate particles in
this region of the stratospherecan be 70,000-80,000tons, the total mass
of aerosolbeing at least 200,000

THE

CLIMATE

OF

THE

FUTURE

243

Stratospheric flights will increase the concentration of particles of

the aerosollayer in the northern hemisphere by roughly 400,000 tons if
the mean lifetime of these particles is two years; this increase is more
than half the amount of aerosolthat can causethe type of substantial
climatic changeconsideredabove.Obviously, this is hardly an insignificant quantity.
Climate modification by altering the sulphur content of fuel in
stratosphericaircraft is possible.The above calculation was basedon the
assumptionthat fuel with a comparatively low sulphur content of 0.05%
will be used in stratosphericflights (Man's Impact on the Global Environment, 1970). This calculation also assumed that the use of fuel with
sulphur content up to 0.3% and fuel in which the sulphur content was
decreasedto 0.01% was possible,though the latter type of fuel is apparently more expensive than fuel with a higher sulphur content.
These data imply that a change in the sulphur content in fuel from
0.01% to 0.30% correspondsto a change in the mass of aerosol particles
in the stratosphereof the northern hemisphere from 80,000to 2,400,000
tons. While the first of these values is not linked to any noticeable
climatic change, the second exceedsthe change in aerosol content
necessaryfor climate modification. Thus, the prospectof climate modificationby varying the type of fuel usedby stratosphericaviation becomes
promising. In order to maintain stable climatic conditions, it is feasible
to use fuel with high sulphur content when the concentrationof aerosol
particles decreasesas a result of natural causes,and to use fuel with low
sulphur contentwhen particle concentrationis high.
Thus stratosphericflights can be considereda promisingelement in
a system of global climate modification. This problem requires additional study sinceit is necessaryto clarify the influenceof stratospheric
flights on other elements of the weather conditions,including the ozone
content in the stratosphere.
Large numbers of flights in the stratospherecan lead to changesin
climatic

conditions

that must be estimated

beforehand.

It is evident from these data that a comparativelyslight changein

the balance of the mass of stratospheric aerosol can substantially
change the air temperature at the earth's surface. It should be noted
that such an influence is a few hundred times greater than that of
changesexerted by the mass of troposphericaerosol on the thermal
regime of the lower air layers.
This differenceis due to a number of causes,for example:
1) the lifetime of a particle in the stratosphereis 50 to 100 times
greater than that of an aerosolparticle in the troposphere;
2) the stratosphere is essentially lacking in gigantic

244

CLIMATIC

CHANGES

which are responsible for about half the mass of tropospheric aerosol
and, due to their slight influence on the radiation regime, decreasethe
optical activity of aerosol in the troposphereto about half;
3) the influence of troposphericaerosol on the radiation regime is
comparatively slight in the presenceof clouds, which on the average
cover nearly half the earth's surface; this also decreasesthe optical
activity of the aerosol to about half as compared to aerosol in the
stratosphere;
4) absorptionof short-wave radiation by aerosolin the troposphere
partially compensatesfor the temperature drop at the earth's surface
due to a decrease in the radiation flux as the radiation is scattered by
aerosol. This effect is chiefly of a local nature in the stratosphere and
doesnot strongly influence the temperature at the earth's surface,since
long-wave radiation (on which the temperature of the earth's surface
finally depends) departing into outer space remains nearly invariant
above the level of the tropopause.
For these reasons, the stratosphere is a favorable zone for climate
modifications, and modern advancesin stratosphericaviation will help
make these modifications highly promising in the near future.
If we agree that it is theoretically possibleto produce a noticeable
change in the global climate by using a comparatively simple and
economicalmethod, it becomesincumbent on us to develop a plan for
climate modification that will maintain existing climatic conditions,in
spite of the tendency toward a temperature increase due to man's
economic activity.
The possibility of using such a method for preventing natural
climatic fluctuations leading to a decreasein the rate of the hydrological
cycle in regions characterized by insufficient moisture is also of some
interest.

The perfection of theories of climate is of great importance for

solving these problems, since current simplified theories are inadequate
to determine all the possiblechangesin weather conditionsin different
regions of the globe that may result from modifications of the aerosol
layer of the stratosphere. Obviously, any modificationswould be premature before all the consequencescan be exactly precalculated.
It is also obvious that the practical realization of this method of
climate modification requires the solution of many technical problems
associated with the development of optimal reagents and the most
feasible methods of ejecting them into the atmosphere.
Thus much preparatory study must be carried out. Methods of
preventing unfavorable climate fluctuations should be developed

THE

CLIMATE

OF

THE

FUTURE

245

out delay since anthropogenicchangesof global climate may signifi-

cantly intensify over the next decades.

Conclusion

We may conclude from the data set forth in this book that in the

modernera the globalclimatehas alreadybeenchangedto someextent

as a result of man's economicactivity. Thesechangesare chieflydue to
an increasein the massof aerosoland carbondioxidein the atmosphere.
Contemporaryanthropogenicchangesin the global climate have
been comparativelysmall, which is partially explainedby the counterinfluence of the growth in the concentration of aerosol and carbon

dioxide on air temperature. Nevertheless,these changesare of some

practicalimportance,chieflyin connection
with the influenceof precipitation on agricultural production.

Simple calculationsdemonstratethat if the contemporaryrate of

economicdevelopmentis maintained, anthropogenicchangescan rapidly grow and at the end of the twentieth century exceedthe natural
climate fluctuationsthat have occurredover the past century.

Climatic changeswill intensify from now on under these conditions

and in the twenty-first centurymay becomecomparablewith the natural climatic fluctuations that have occurred over tens of thousands of

years of ice agesand interglacialperiods.Obviously,suchsignificant

climatic changescan enormouslyinfluenceour planetary environment
and many aspectsof man's economicactivity.
Thus it becomesnecessaryto predict the anthropogenicclimatic
changesthat will occurunderdifferentformsof economicdevelopment
and alsoto developmethodsof climatemodificationto preventundesirable climaticchanges.In the light of theseproblemsthe importanceof
studies of climatic changesand, in particular, of the causesof these
changes,acquiresa new meaning. Until now such studieshave been
theoreticalto a significantextent. However, it is clear that they must
becomean important tool in planningthe developmentof the national
economy.

Refinementand perfectionof numericalmodelsof climatetheoryin

order to better predictclimatic changesconstituteone of the chief lines
of research.Sinceall suchexistingmodelsare to somedegreeapproximate, it becomes
extremelyimportantto verify the modelsby empirical

data. Data on natural climaticchangesthat have occurredin the past

are particularly important in this respect.
Thus the prediction of the climatic conditions of the future

246

CLIMATIC

CHANGES

intimately associated with the quantitative explanation of climatic

changesthat have occurredin the past.
An attempt was made here to use numerical models of climate
theory to solve this problem, which resulted in conclusionsabout the
laws for climatic changesin the past and in the future that differ in a
number of respectsfrom previously suggestedopinions.
The use of more general modelsof climate theory may allow us to
verify these conclusionsand to obtain many new data on the laws of
climatic changes. Studies along this direction should be greatly expanded in view of the possibilitythat noticeablechangesin the global
climate may occurin the near future. The lack of early detailed data on
such changesmay create enormousdifficulties for human society.
The problemof modifyingthe global climate is of particular importance. The data set forth in this book indicate

that climate

modification

by contemporarytechnical meanscan, in theory, be carried out, which

increasesthe urgency of further studies in this field.
The developmentof methodsfor predicting the climatic conditions
of the future will open horizonsfor creating numerical models of economic development that take into account the influence of man's economic activity on global climate. Such models can optimize economic
planning and eliminate unfavorable changes in environmental conditions with further technologicaldevelopment.
The problem of anthropogenicclimatic changesis of international
significanceand it becomesparticularly important in developinglargescaleprojectsof climate modifications.Modification of the global climate
will lead to a change in climatic conditionsover many countries, the
nature of these changes differing in different regions. It has been
repeatedly noted by Fedorov (1958) that any large project of climate
modificationis possibleonly on the basisof international cooperation.
It appears that there are now grounds for signing an international
conventionfcrbidding the implementation of climate modificationsthat
have not been agreed to. Such modificationsmight be carried out only
after being approved by the proper international institutions. This
agreement must regulate both measuresin the developmentof climate
modificationsand those forms of man's economicactivity that can lead
to inadvertent changes in global climatic conditions.
Summary

The application of physical climatology in studying climatic

changesis the main problem presentedin this book. It deals with

THE

CLIMATE

OF THE

FUTURE

247

principal features of modern climate as well as climates of the past.

Certain factors that influence climate are particularly discussed
A semiempirical theory of the atmosphere's thermal regime is
proposed.This theory enablesus to draw conclusionsas to the intransivity of the modernclimate regime and its high sensitivity to the changes
of climate-forming factors.
The semiempirical climate theory is usedfor a quantitative elucidation of contemporaryclimatic changesand for a study of the development of the Quaternary glaciation.
Presumably, the climatic variations of the geological past were
mainly causedby differencesin the distribution of continentsand in the
compositionof the atmosphere.
The stability of the climate depreciatedunder the influence of the
following factors: a drop in temperature occurredin high latitudes and
polar ice was formed; this stimulated the developmentof continental
glaciationsin high latitudes due to periodicreductionsof solar radiation
causedby astronomical factors.
Comparatively minor variations in the modern climate are mainly
the result of changesin the concentrationof stratosphericaerosolparticles that influence solar radiation flux. These changes are brought
about by volcanic activity.
The bookalso deals with the effectof climatic changeson biological
processesincluding the evolution of living organisms.
In the last decadesglobal climate has been affectedby anthropogenic factors which include an increase of carbon dioxide and aerosol
concentrationsin the atmosphere.
This influence is rapidly augmenting and may in the near future
result in drastic changesof climate of great practical significanceto
man. It is very essential, therefore, to developmethodsfor controlling
climate modifications.

One such method is to maintain

aerosolconcentration in the

a certain level of

THE

CLIMATE

OF

THE

FUTURE

245

out delay since anthropogenicchangesof global climate may signifi-

cantly intensify over the next decades.

Conclusion

We may conclude from the data set forth in this book that in the

modernera the globalclimatehas alreadybeenchangedto someextent

as a result of man's economicactivity. Thesechangesare chieflydue to
an increasein the massof aerosoland carbondioxidein the atmosphere.
Contemporaryanthropogenicchangesin the global climate have
been comparativelysmall, which is partially explainedby the counterinfluence of the growth in the concentration of aerosol and carbon

dioxide on air temperature. Nevertheless,these changesare of some

practicalimportance,chieflyin connection
with the influenceof precipitation on agricultural production.

Simple calculationsdemonstratethat if the contemporaryrate of

economicdevelopmentis maintained, anthropogenicchangescan rapidly grow and at the end of the twentieth century exceedthe natural
climate fluctuationsthat have occurredover the past century.

Climatic changeswill intensify from now on under these conditions

and in the twenty-first centurymay becomecomparablewith the natural climatic fluctuations that have occurred over tens of thousands of

years of ice agesand interglacialperiods.Obviously,suchsignificant

climatic changescan enormouslyinfluenceour planetary environment
and many aspectsof man's economicactivity.
Thus it becomesnecessaryto predict the anthropogenicclimatic
changesthat will occurunderdifferentformsof economicdevelopment
and alsoto developmethodsof climatemodificationto preventundesirable climaticchanges.In the light of theseproblemsthe importanceof
studies of climatic changesand, in particular, of the causesof these
changes,acquiresa new meaning. Until now such studieshave been
theoreticalto a significantextent. However, it is clear that they must
becomean important tool in planningthe developmentof the national
economy.

Refinementand perfectionof numericalmodelsof climatetheoryin

order to better predictclimatic changesconstituteone of the chief lines
of research.Sinceall suchexistingmodelsare to somedegreeapproximate, it becomes
extremelyimportantto verify the modelsby empirical

data. Data on natural climaticchangesthat have occurredin the past

are particularly important in this respect.
Thus the prediction of the climatic conditions of the future

246

CLIMATIC

CHANGES

intimately associated with the quantitative explanation of climatic

changesthat have occurredin the past.
An attempt was made here to use numerical models of climate
theory to solve this problem, which resulted in conclusionsabout the
laws for climatic changesin the past and in the future that differ in a
number of respectsfrom previously suggestedopinions.
The use of more general modelsof climate theory may allow us to
verify these conclusionsand to obtain many new data on the laws of
climatic changes. Studies along this direction should be greatly expanded in view of the possibilitythat noticeablechangesin the global
climate may occurin the near future. The lack of early detailed data on
such changesmay create enormousdifficulties for human society.
The problemof modifyingthe global climate is of particular importance. The data set forth in this book indicate

that climate

modification

by contemporarytechnical meanscan, in theory, be carried out, which

increasesthe urgency of further studies in this field.
The developmentof methodsfor predicting the climatic conditions
of the future will open horizonsfor creating numerical models of economic development that take into account the influence of man's economic activity on global climate. Such models can optimize economic
planning and eliminate unfavorable changes in environmental conditions with further technologicaldevelopment.
The problem of anthropogenicclimatic changesis of international
significanceand it becomesparticularly important in developinglargescaleprojectsof climate modifications.Modification of the global climate
will lead to a change in climatic conditionsover many countries, the
nature of these changes differing in different regions. It has been
repeatedly noted by Fedorov (1958) that any large project of climate
modificationis possibleonly on the basisof international cooperation.
It appears that there are now grounds for signing an international
conventionfcrbidding the implementation of climate modificationsthat
have not been agreed to. Such modificationsmight be carried out only
after being approved by the proper international institutions. This
agreement must regulate both measuresin the developmentof climate
modificationsand those forms of man's economicactivity that can lead
to inadvertent changes in global climatic conditions.
Summary

The application of physical climatology in studying climatic

changesis the main problem presentedin this book. It deals with

246

CLIMATIC

CHANGES

intimately associated with the quantitative explanation of climatic

changesthat have occurredin the past.
An attempt was made here to use numerical models of climate
theory to solve this problem, which resulted in conclusionsabout the
laws for climatic changesin the past and in the future that differ in a
number of respectsfrom previously suggestedopinions.
The use of more general modelsof climate theory may allow us to
verify these conclusionsand to obtain many new data on the laws of
climatic changes. Studies along this direction should be greatly expanded in view of the possibilitythat noticeablechangesin the global
climate may occurin the near future. The lack of early detailed data on
such changesmay create enormousdifficulties for human society.
The problemof modifyingthe global climate is of particular importance. The data set forth in this book indicate

that climate

modification

by contemporarytechnical meanscan, in theory, be carried out, which

increasesthe urgency of further studies in this field.
The developmentof methodsfor predicting the climatic conditions
of the future will open horizonsfor creating numerical models of economic development that take into account the influence of man's economic activity on global climate. Such models can optimize economic
planning and eliminate unfavorable changes in environmental conditions with further technologicaldevelopment.
The problem of anthropogenicclimatic changesis of international
significanceand it becomesparticularly important in developinglargescaleprojectsof climate modifications.Modification of the global climate
will lead to a change in climatic conditionsover many countries, the
nature of these changes differing in different regions. It has been
repeatedly noted by Fedorov (1958) that any large project of climate
modificationis possibleonly on the basisof international cooperation.
It appears that there are now grounds for signing an international
conventionfcrbidding the implementation of climate modificationsthat
have not been agreed to. Such modificationsmight be carried out only
after being approved by the proper international institutions. This
agreement must regulate both measuresin the developmentof climate
modificationsand those forms of man's economicactivity that can lead
to inadvertent changes in global climatic conditions.
Summary

The application of physical climatology in studying climatic

changesis the main problem presentedin this book. It deals with

THE

CLIMATE

OF THE

FUTURE

247

principal features of modern climate as well as climates of the past.

Certain factors that influence climate are particularly discussed
A semiempirical theory of the atmosphere's thermal regime is
proposed.This theory enablesus to draw conclusionsas to the intransivity of the modernclimate regime and its high sensitivity to the changes
of climate-forming factors.
The semiempirical climate theory is usedfor a quantitative elucidation of contemporaryclimatic changesand for a study of the development of the Quaternary glaciation.
Presumably, the climatic variations of the geological past were
mainly causedby differencesin the distribution of continentsand in the
compositionof the atmosphere.
The stability of the climate depreciatedunder the influence of the
following factors: a drop in temperature occurredin high latitudes and
polar ice was formed; this stimulated the developmentof continental
glaciationsin high latitudes due to periodicreductionsof solar radiation
causedby astronomical factors.
Comparatively minor variations in the modern climate are mainly
the result of changesin the concentrationof stratosphericaerosolparticles that influence solar radiation flux. These changes are brought
about by volcanic activity.
The bookalso deals with the effectof climatic changeson biological
processesincluding the evolution of living organisms.
In the last decadesglobal climate has been affectedby anthropogenic factors which include an increase of carbon dioxide and aerosol
concentrationsin the atmosphere.
This influence is rapidly augmenting and may in the near future
result in drastic changesof climate of great practical significanceto
man. It is very essential, therefore, to developmethodsfor controlling
climate modifications.

One such method is to maintain

aerosolconcentration in the

a certain level of

Bibliography

Adhmar, J. F., Les rvolutions de la mar. Dluges priodiques, Paris, 184 pp.,
1842.

Ahlman, H. W., Glacier variations and climatic fluctuations, Amer. Geogr,

Soc., Ser. 3, N.Y., 51 pp., 1953.
Albrecht, F., Doe Aktiongebietedes Wasser- und W/rmehaushaltesder Erdoberfi/che, Z. Meteorol., 1, 4-5, 97-109, 1947.
Allen, C. W., Solar radiation, Quart. J. Roy. Meteorol. Soc., 84, 362, 307-318,
1958.

Alyea, F. N., Numerical simulationof an ice agepaleoclimate,Atmos.Sci.Pap.,

No. 193, Colorado State Univ., 120 pp., 1972.

AngstrSm,A., Atmosphericturbidity, globalillumination and planetary albedo

of the earth, Tellus, 14, 4, 435-450, 1962.

AngstrSm, A., The solar constantand the temperature of the Earth, Progressin
Oceanography,V. 3, Permagon Press, Oxford, pp. 1-5, 1965.
AngstrSm, A., Apparent solar constant variations and their relation to the
variability of atmospherictransmission, Tellus, 21, 2, 205-218, 1969.
Arrhenius, S., On the influence of the carbonic acid in the air upon the
temperature of the ground, Philos. Mag., 41, 237-275, 1896.
Arrhenius, S., Lehrbuch der kosmischenPhysik, Hirzel, Leipzig, 1026S., 1903.
Atlas of Heat Balance (in Russian),ed. M. I. Budyko,Leningrad, Izd. G.G.O., 41
pp., 1955.
Atlas of Heat Balance of the World (in Russian), ed. M.I. Budyko, Moscow,

Mezhduvedomstvennyigeofizicheskiikomitet, 69 pp., 1963.

Atwater, M. A., Planetary albedochangesdue to aerosol,Science,170, 3953,6466, 1970.

Auer, V., The Pleistoceneof Fuego-Patagonia,Part I, The ice age and intergla
cial ages, Suomal. Tiedeakat. Toim. Helsinki, Ser. A, III, Geologica-Geographica, pp. 50, 239, 1956.
Axelrod, D. I., Quaternary extinctionsof large mammals, Univ. of Calif. Publ.
Geol. Sci., 74, 42 pp., 1967.
Axelrod, D. I., and H. P. Baily, Cretaceousdinosaur extinction, Evolution, 22, 3,
595-611, 1968.

Barrett, E., Depletion of short-wave irradiance at the ground by particles

suspendedin the atmosphere, Solar energy, 13, 3, 323-337, 1971.
Bellaire, A., and R. Garrington, The World of Reptiles,London,153pp., 1966.

BIBLIOGRAPHY

249

Berger,A. L., Thorieastronomique

despaleoklimats,
V. 1, 197pp., V. 2, 108

pp., Univ. catholiquede Louvain,Louvain,1973.

Berlyand,M. E., andK. Ya. Kondrat'ev,CitiesandClimateof the Planet(in
Russian),Leningrad,Gidrometeoizdat,
40 pp., 1972.
Berlyand,
M. E., et al., Complex
studyoftheperculiarities
ofthemeteorological
regimein a large city (in Russian),Meteorol.Gidrol., 1, 14-23,1974.
Bibikov,S. N., Certain aspectsof the paleoeconomic
modelof the Paleolite(in
Russian), Sov. Archeol., 4, 5-22, 1969.

Blinova,E. N., On the mean annualtemperaturedistributionin the earth's

atmosphere
with consideration
of continents
and oceans(in Russian),Izv.
AN USSR, Ser. Geogr. Geofiz., 11, 1, 3-13, 1947.

Bogert,C. M., Howreptilesregulatetheirbodytemperature?,

Sci.American,
200, 4, 105-108, 1959.

Bolin, B., and W. Bischof,Variationof the carbondioxidecontentof the

atmosphere
in the northernhemisphere,
Tellus,22, 4, 431-442,
1970.
Bolin,B., andC. D. Keeling,Large-scale
atmospheric
mixingasdeduced
from
the seasonaland meridional variations of carbon dioxide, J. Geophys. Res.,
68, 13, 3899-3920, 1963.

Borzenkova,
I. I., Onthepossible
influence
ofvolcanic
dustontheradiationand

thermal regime (in Russian),Tr. GGO, Vyp. 307, 36-42, 1974.

Borzenkova,
I. I., K. Ya. Vinnikov,L. P. Spirina, and D. !. Stekhnovskiy,
Changeof air temperatureof the NorthernHemispherefor 1881-1975(in
Russian), Meteorol. Gidrol., 7, 27-35, 1976,

Bowen,R., Paleotemperature
Analysis,Elsevier,265pp.,Amsterdam,London,
New York, 1966.

Braidwood,
R. I., andC. A. Reed,Theachievement
andearlyconsequences
foodproduction,
ColdSpringHarborSymp.Quant.Biol.,V. 22,Population
Studies,Animal EcologyandDemography,
pp. 19-31,New'York, 1957.
Brooks,C. E. P., ClimateThroughtheAges,2nded.,ErnstBenn,London,1950.
Bryson,
R. A., All otherfactors
beingconstant
.... Weatherwise,
21,2, 56-62;1968.

Bryson,
R. A., Climatemodification
byair pollution,
Report1,UniV.ofWisconsin, Madison, 16 pp., 1972.

Budyko,M. I., Heat Balanceof the Earth'sSurface(in Russian),

Leningrad;
Gidrometeoizdat, 255 pp., 1956.

Budyko,M. I., On the thermalzonalityof the earth (in Russian),Meteorol.

Gidrol., 11, 7-14, 1961.

Budyko,M. I., Polariceandclimate(in Russian),

Izv. AN USSR,Ser.Geogr.,6;
3-10, 1962.

Budyko,M. I., Certainmethods

ofinfluencing
the climate(in Russian),
Meteorol. Gidrol., 2, 3-8,

1962a.

Budyko,M. I., Climaticchanges

(inRussian),
Meteorol.
Gidrol.,11,18-27,1967,
Budyko,M. I., Reasons
for the extinction
of certainanimalsat the endof the
Pleistocene (in Russian), Izv. AN SSR, Ser. Geogr. 2, 28-36, 1967a.

Budyko,
M.I., Ontheorigin
oficeages
(inRussian),
Meteorol.
Gidrol,
11,3-12,

250

CLIMATIC

CHANGES

Budyko,M. I., ClimaticChanges(in Russian),35 pp., Gidrometeoizdat,

Leningrad, 1969.

Budyko, M. I., The effectof solar radiation variationson the climate of the
earth, Tellus, 21, 5, 611-619, 1969a.

Budyko, M. I., Commentson "A global climatic modelbasedon the energy

balance of the earth-atmospheresystem",J. Appl. Meteorol., 9, 2, 310-311,
1970.

Budyko,M. I., The Climate and Life (in Russian),470 pp., Gidrometeoizdat,
Leningrad, 1971.

Budyko,M. I., Comments

on"Intransitivemodelofthe earth-atmospheric-ocean
system",J. Appl. Meteorol., 11, 7, 1150-1151,1972.
Budyko, M. I., The future climate, EOS, 53, 10, 868-874, 1972a.
Budyko,M. I., Man's Influenceon Climate (in Russian),47 pp., Gidrometeoizdat, Leningrad, 1972b.
Budyko,M. I. AtmosphericCarbonDioxideand the Climate (in Russian),32
pp., Gidrometeoizdat,Leningrad, 1973.
Budyko,M. I., Man andthe biosphere(in Russian),Vopr. Fil., 1, 61-70, 1973a.
Budyko,M. I., Methodof influencingthe climate(in Russian),Meteorol.Gidrol.
2, 91-97, 1974.

Budyko, M. I., Climate of the future (in Russian),Meteorol.Gidrol., 9, 3-15,

1974a.

Budyko,M. I., The dependence

of the mean air temperatureon the changeof
solar radiation (in Russian), Meteorol. Gidrol. 10, 3-10, 1975.

Budyko,M. I., O. A: Drozdov,and M. I. Yudin, The influenceof man'seconomic

activity on the climate, in ContemporaryProblemsof Climatology (in
Russian),pp. 435-448,Gidrometeoizdat,Leningrad, 1966.
Budyko,M. I., and Z. I. Pivovarova,Effectof volcaniceruptionsonthe influx of
solar radiation at the earth's surface (in Russian), Meteorol. Gidrol., 10, 37, 1967.

Budyko,M. I., and M. A. Vasishcheva,Effectof astronomicalfactorson Quaternary glaciations(in Russian),Meteorol Gidrol., 6, 37-47, 1971.
Budyko, M. I., and K. Ya. Vinnikov, ContemporaryClimatic Changes (in
Russian), Meteorol. Gidrol., 9, 3-13, 1973.

Budyko, M. I., and K. Ya. Vinnikov, Global warming (in Russian),Meteorol.

Gidrol., 7, 16-25, 1976.

Budyko,M. I., and M. I. Yudin, On the level fluctuationsin drainlesslakes (in

Russian), Meteorol. Gidrol., 8, 15-19, 1960.

Budyko,M. I., et al., Changesin the Climatein Connection

with the Plan of
Nature Transformation (in Russian), 206 p. Gidrometeoizdat,Leningrad,
1952.

Budyko,M. I., et al., Prospects

of changingthe globalclimate(in Russian),Izv.
AN SSSR, Ser. Geogr., 2, 11-23, 1974.
Burdecki, F., Phenomena after volcanic eruptions, Weather, 19, 4, 113-114,
1964.

Butzer, K. W. Environment and Archeology,524pp., Methuen, London,1964.

Cadle, R. D., Compositionof the stratospheric"Sulfate layer", Proceedings

BIBLIOGRAPHY

251

the survey conferenceFeb. 15-16, 1972, U. Department of Transportation,

1972.

Callender, G. S., The artificial productionof carbon dioxide and its influence on
temperature, Quart. J. Roy. Meteorol. Soc., 64, 27, 223-240, 1938.
Cess,R. D., Climatic change:an appraisal of atmosphericfeedbackmechanisms
employing zonal climatology, J. Atmos. Sci. (in print), 1976.
Chamberlin, T. C., A group of hypotheses bearing on climatic changes, J.
Geology, 5, 653-683, 1897.
Chamberlin, T. C., The influence of great epochsof limestone formation upon
the constitution of the atmosphere,J. Geology, 6, 609-621, 1898.
Chamberlin, T. C., An attempt to frame a working hypothesisof the causeof
glacial periodson an atmosphericbasis, J. Geology,7, 545-584, 1899.
Charlson, R. J., and M. J. Pilat, Climate: The influence of aerosols, J. Appl.
Meteorol., 8, 6, 1001-1002, 1969.

Chizhov, O. P., and A.M. Tareeva, Possibilitiesof estimating the ice regime of
the Arctic basin and its variation, in Materials of glacioiogicalstudies (in
Russian), 15, 57-73, 1969.

Colbert, E. N., Evolution of the Vertebrates, 479 pp., John Wiley & Sons, N.Y.,
1958.

Colbert, E. N., The Age of Reptiles, 228 pp., Weidenfeld & Nicolson, London,
1965.

Darwin, Ch., The Origin of Speciesby Means of Natural Selection, Murray,

London, 1859.
Darwin, Ch., The Descent of Man and Selection in Relation to Sex, Murray,
London, 1871.
Davitaya, F. F., On the possibleinfluenceof atmosphericdust on the melting of
glaciers and warming of the climate (in Russian), Izv. AN SSSR, Ser.
Georgr. 2, 3-23, 1965.
Davitaya, F. F., History of the atmosphereand the dynamics of its gaseous
composition(in Russian), Meteorol. Gidrol., 2, 21-28, 1971.
Donn, W. L., and D. M. Shaw, The heat budgets of an ice-free and ice-covered
Arctic Ocean, J. Geophys. Res., 71, 4, 1087-1093, 1966.
Dorodnitsyn, A. A., B. I. Izvekov, and M. E. Shvets, Mathematical theory of
general circulation (in Russian), Meteorol. Gidrol., 4, 32-42, 1939.
Drozdov, O. A., and A. S. Grigor'eva, Moisture Circulation in the Atmosphere
(in Russian), 314 pp., Gidrometeoizdat,Leningrad, 1963.
Drozdov,O. A., and A. S. Grigor'eva,Long-term Cyclic Fluctuations of Atmospheric Precipitation over the USSR (in Russian), 158pp., Gidrometeoizdat,
Leningrad, 1971.
Drozdov, O. A., and T. V. Pokrovskaya, On the role of random variations of the
water balance in the fluctuations of certain lakes (in Russian), Meteorol.
Gidrol., 8, 43-48, 1961.

Dwayr, H. A., and T. Peterson, Time-dependentenergy modeling, J. Appl.

Meteorol., 12, 36-42, 1973.

Dyer, A. J., and B. B. Hicks, Stratospherictransport of volcanicdust inferred

from solar radiation measurements, Nature, 208, 5006, 131-133,

252

CLIMATIC

CHANGES

Dyer, A. J., and B. B. Hicks, Global spread of volcanic dust from the Bali
eruption of 1963, Quart. J. Roy. Meteorol. Soc., 94, 402, 545-554, 1968.
Elton, Ch., Animal Ecology and Evolution, 96 pp., Clarendon Press, Oxford,
London, 1930.

Emiliani, C., Pleistocenetemperature, J. Geol., 63, 6, 538-578, 1955.

Emiliani, C., Isotopic paleotemperatures, Science, 154, 851-857 1966.
Ensor, D. S., et al, Influence of atmospheric aerosol on albedo J. Appl. Meteorol., 10, 6, 1303-1306, 1972.

Erikson, E., Air-ocean-ice cap interactions in relation to climatic fluctuations

and glaciation cycles,Causesof climatic change INQUA, USA, 1968.
Ewing, D. M., and W. L. Donn, A theory of ice ages, I, Science,123, 1061-1066,
1956.

Ewing, D. M., and W. L. Donn, A theory of ice ages, II, Science,127, 1159-1162,
1958.

Faegre, A., An intransitive model of the Earth-atmosphere-oceansystem, J.

Appl. Meteorol., 11, 1, 4-6, 1972.
Fairbridge, R. W., Ice-agetheory, in The Encyclopaediaof AtmosphericSciences
and Astrogeology,pp. 462-474, Reinhold Publ. Corp., New York-Amsterdam-London, 1967.

Fedorov, E. K., Man's influenceson meteorologicalprocesses(in Russian), Vopr.

Fil., 4, 137-144, 1958.
Fedorov, E. K., Interaction of Mankind and Nature (in Russian), 88 pp.,
Gidrometeoizdat, Leningrad, 1972.
Fletcher, J. O., The Arctic heat budget and atmosphericcirculation, Proc. of the
Symposiumon the Arctic heat budget and atmosphericcirculation, pp. 2332, Rand Corp., Santa Monica, Ca]if., 1966.
Flint, R. F., Glacial Geology and the Pleistocene Epoch, New York-London,
1963.

Flohn, H., Zur meteorologischeInterpretation der Pleistoznen-Kilmaschw/n-.

kungen, Eiszeitalter und Gegenwart, 14, 153-160, 1963.
Flohn, H., Grundfragen der Paloklimatologie in Lichte einer theoretischen
Klimatologie, GeologischeRundschau,54, 5, 504-515, 1964.
Flohn, H., Ein geophysikalischesEiszeit-Modell, Eiszeitalter und Gegenwart,
20, 204-231, 1969.
Flowers, E. C., and H. J. Viebrock, Solar radiation-an anomalous decrease of
direct solar radiation, Science, 148, 3669, 493-494, 1965.
Forrester, J. W., World Dynamics, 142 pp., Wright-Allen Press, Cambridge,
Mass., 1971.
Fuchs, V. E., and T. T. Patterson, The relation of volcanity and orogeny to
climatic change, Geol. Mag., 84, 6, 321-333, 1947.
Gaevskaya, G. N., Absorptionof short-wave radiation by aerosol (in Russian),
Probl. Fis. Atmos., 10, 86-91, Izd. LGU, Leningrad, 1972.
Gandin, L. S., B. M. II'in, and L. V Rukhovets, On the influence of variations of
external parameters on the atmospheric thermal regime (in Russian), Tr.
GGO, 315, 21-38,

BIBLIOGRAPHY

253

George, W., Biologist Philosopher, 320 pp., Abelard-Schuman, London-TorontoNew York, 1964.

Gerasimov, I. P., Nature and evolution of primeval society(in Russian), Izv. AN

USSR, Ser. Geogr. 1, 5-8, 1970.
Gerasimov, I. P., and M. I. Budyko, Actual problems of man's interaction with
nature (in Russian), Kommunist, 10, 79-91, 1974.

Gerasimov, I. P. and K. K. Markov, The ice Age and Territory of the USSR, 462
pp., Izd. AN USSR, Moscow-Leningrad, 1939.
Gol'tsberg, I. A., Expedition for the study of atmospheric turbulence in forest
strips (in Russian), Tr. GGO, 29, 5-10, 1952.
Gordon, H. B., and D. H. Davis. The effect of changes in solar radiation on
climate, Quart. J. Roy. Meteorol. Soc., 100, 423, 123-126, 1974.

Grigor'ev, A. A. The Sub-Arctic (in Russian), 233, pp., Geografgiz, Moscow,

1956.

Grigor'ev, A. A., and M. I. Budyko, On the periodical law of geographical

zonality (in Russian), AN USSR, 110, 1, 129-132, 1956.
Grobecker, A. J., Assessmentof Climatic Changes due to Flights in the Stratosphere, 16 pp., Dept. Transportation, Washington, D.C., 1973.
Hantel, M., Polar boundary conditions in zonally averaged global climatic
models, J. Appl. Meteorol., 13,752-759, 1974.
Held, I. M., and J. M. Suarez, Simple albedo feedback models of the ice-caps,
Tellus, 26, 6, 613-630, 1974.

Hoffman, E., Der Dritte KarbonstratigraphischeKongressin Heerlen, Phyton,

3, 3-4, 193-209, 1951.

Holloway, J. L., and S. Manabe, A global general circulation model with

hydrology and mountains, Mon. Weather Rev., 99, 5, 355-370, 1971.
Humphreys, W. J., Volcanic dust and other factorsin the productionof climatic
changesand their possiblerelation to ice age, J. Franklin Inst., 1976, 131
pp., 1913.

Humphreys, W. J., Physicsof the Air, 2d ed., 654 pp., McGraw-Hill Book Co.,
New York, 1929.

Huntington, E., and S.S. Visher, Climatic Changes. Their Nature and Causes,
329 pp., Yale Univ. Press, New Haven, Conn., 1922.
Inadvertent Climate Modification (SMIC). The MIT Press, Cambridge, Mass.,
and London, 306, 1971.
Junge, C., Atmospheric Chemistry and Radioactivity, Academic Press, New
York, 1963.

Kagan, R. L., and K. Ya. Vinnikov, On the assumption of heat inflow during
numerical experiments with a thermotropic model (in Russian), Tr. GGO,
256, 98-107, 1970.

Kalesnik, S. V., Principles of Physical Geography (in Russian), Uchpedgiz,

Moscow-Leningrad, 1947.
Kalitin, N. N., On the time of the onset of the optical anomaly in 1912 (in
Russian), Izv. GFO, 1, 11-17, 1920.

Kaplan, L. D., The influence of carbon-dioxidevariation on the atmospheric

heat balance, Tellus, 12, 2,209-216,

CLIMATIC

254

CHANGES

Karol', I. L., Radioactive Isotopesand Global Transfer in the Atmosphere (in

Russian), 365 pp., Gidrometeoizdat, Leningrad, 1972.
Karo!', I. L., The size of radioactive aerosoland its transfer in the troposphere
and stratosphere (in Russian), Meteorol. Gidrol., 1, 28-38, 1973.
Kellogg, W. W., Mankind as a Factor in Climate Change. The Energy Question,
Univ. of Toronto Press, 1974.

Kimball, H. H., Volcanic eruptions and solar radiation intensities, Mon.

Weather Rev., 46, 8, 355-356, 1918.

Kochin, N. E., Compiling model for the zonal circulation of the atmosphere (in
Russian), Tr. GGO, 10, 3-27, 1936.
Kondrat'ev, K. Ya., and N.Y. Niilisk, On the question of carbon dioxide heat
radiation in the atmosphere, Geophys. Pura Appl., 46, 216-230, 1960.
Kondrat'ev, K. Ya., et al., The Influence of Aerosol on the Transfer of Radiational; Possible Climatic Consequences(in Russian), 266 pp., Izd. LGU,
Leningrad, 1973.
KSppen, W., and A. Wegener, Die Klimate der geologischenVorzeit, 256 S.,
Berlin, 1924.
Kurten, B., Evolution in GeologicalTime, in Ideas in Modern Biology, ed. by J.
A. Moore, 563 pp., Natural History Press, Garden City, N.Y., 1965.
Kurten, B., Time and hominid brain size, Comment Biol., 36, 1-8, Helsinki,
1971.

Kuznetsova, L. P., and V. Ya. Sharova, Amount of Precipitation (in Russian),

God. Fiz. Geogr. Atlas Mira, 42-43, 1964.
Laek, D., The Natural Regulation of Animal Numbers, Oxford, 1954.
Lamb, H. H., The role of atmosphere and oceansin relation to climatic changes
and the growth of ice sheetson land, Problems in Paleoclimatology,332-348
pp. Interscience Publishers, London-New York-Sydney, 1964.
Lamb, H. H., The Changing Climate, 236 pp., Methuen, London, 1966.
Lamb, H. H., Volcanic dust in the atmosphere . . . Phil. Trans., Roy. Soc. of
London, Appl. Math. Phys. Sci., 266, 1178, 425-533, 1970.
Lamb, H. H., Climate, Present, Past and Future, V. 1, 612 pp., Methuen,
London, 1973.
Lamb, H. H., Climatic change and foresight in agriculture: the possibilities of
long-term weather advice, Outlook on Agriculture, 7, 5, 203-210, 1973.
Lamb, H. H., The Current Trend of World Climate- a Reporton the Early 1970's
and a Perspective, Univ. of East Anglia, Norwich, 28 pp., 1974.
Lamb, H. H., S. A. Malberg, and J. M. Colebrook, Climatic reversal in the
northern North Atlantic, Nature, 256, 5517, 479, 1975.

Landsberg, H. E., The Climate of Towns. Man's Role in Changing the Face of
the Earth, pp. 584-606, Univ. Chicago Press 1956.
Landsberg, H. E., Note on the recent climatic fluctuation in the United States,
J. Geophys.Res., 65, 5, 1519-1525, 1960.
Landsberg, H. E., Man-made climatic changes,Science,170, 1264-1265,1970.
Landsberg, H. E., and T. N. Maisel, Micrometeorologicalobservationsin an
area of urban growth, Boundary-Layer Meteorol. 2, 365-370, 1972.
Landsberg, H. E., C. S. Yu, and L. Huang, Preliminary Reconstructionof

BIBLIOGRAPHY

255

Long Time Series of Climatic Data for the Eastern United States, 20 pp.,
Univ. of Maryland, Inst. of Fluid Dynamics and Appl. Math. Techn., Note
BN-57, 1968.

Lieth, H., The role of vegetation in the carbon content of the atmosphere,J.
Geophys.Res., 68, 13, 3887-3898, 1963.
Lorenz, E. N., Climatic determinism, Meteorol. monogr., 8, 30, 1-3, 1968.
Lukashevich, I.D., On the causesof the ice age (in Russian), Priroda, 959-979,
St. Petersburg, July-August, 1915.
Lyell, C., Principles of Geology, V. 1,511 pp., 1830;V. 2,330 pp., 1832;V. 3,398
pp., 1833, London, 1830-1833.
Manabe, S., The Dependenceof AtmosphericTemperature on the Concentration
of Carbon Dioxide. Global Effects of Environmental Pollution, 218 pp.,
Reidel Publ. Co., Dordrecht, Holland, 1970.
Manabe, S., and K. Bryan, Climate calculation with a combined ocean-atmosphere model, J. Atmos. Sci., 26, 786-789, 1969.
Manabe, S., and R. T. Wetheraid, Thermal equilibrium of the atmospherewith
a given distribution of relative humidity, J. Atmos. Sci., 24, 241-259, 1967.
Manabe, S., and R. T. Wetheraid, The effectsof doubling the CO.concentration

on the climate of a general circulation model, J. Atmos. Sci.,J32, 1, 3-15,

1975.

Man's Impact on the Global Environment (SCEP), MIT Press, Cambridge,

Mass., and London, 318 pp., 1970.
Markov, K. K., Essayson the Geographyof the Quaternary Period (in Russian),
347 pp., Geografgiz, Moscow, 1955.
Markov, K. K., Paleography (in Russian), 2nd ed., 268 pp., Izd. MGU, Moscow,
1960.

Martin, P.S., Pleistocene ecology and biography of North America, Amer.

Assoc.Adv. Sci. Publ., 51,509 pp., 1958.
Martin, P.S., Africa and Pleistocene overkill, Nature, 212, 5060, 339-342, 1966.
Martin, P.S., Prehistoric overkill. Pleistocene extinctions, Proc. III Congr.
Intern. Assoc. Quart. Res., 75-120, Yale Univ. Press, New York-London,
1967.

Martin, P.S., The discovery of America, Science, 179, 969-974, 1973.

Mathew, W. D., Climate and Evolution, 2nd ed., 1915, Spec. Publ., New York
Acad. Sci. 1,223 pp., 1939.

Mayr, E., AnimalSpecies

andEvolution,797pp., Cambridge,
Mass.,1963.

McCormic R. A., and J. H. Ludwig, Climate modification by atmospheric

aerosols, Science, 156, 1358-1359, 1967.

Meadows, D. H., et al., Limits to Growth, 205 pp., Universe Book, New York,
1972.

Milankovich, M., Thorie mathmatique des phnomnes thermiques produits

par la radiation solaire, 339 pp., Paris, 1920.
Milankovich, M., Mathematische Klimalehre und astronomischeTheory der
Klimaschwankungen,KSppen-Geiger,Handb. Klimat., I, A., 176pp., Berlin,

256

CLIMATIC

CHANGES

Milankovich, M., Kanon der Erdbestrahlungund seineAnwendungam Eiszeitproblem, 133, 633 pp., Royal Serbian Academy, 1941.
Mintz, Y., Very long-term global integration of the primitive equation of atmospheric motion, WMO Technical Note, 66, 141-161, 1965.

Mitchell, J. M., Recentsecularchangesof globaltemperature, Ann. N.Y. Acad.

Sci., 95(1), 235-250, 1961.

Mitchell, J. M., On the world-wide pattern of secular temperature changes,

UNESCO, Arid Zone Research, 20, 161-180, Paris, 1963.
Mitchell, J. M., The solar inconstant, Proc. Seminar on PossibleResponsesof
Weather Phenomena to Variable Extraterrestrial Influences, NCAR Techn.
Note, TN-8, 155-174, Boulder, Colorado, 1965.

Mitchell, J. M,, Theoreticalpaleoclimatology.The Quaternaryof United States,

881-901, Univ. Press, Princeton, 1965a.

Mitchell, J. M., Concludingremarks. Causes of climatic changes, 155-159,

iNQUA, USA, 1968.

Mitchell, J. M., The effect of atmosphericaerosolson climate with special

referenceto temperaturenear the Earth's surface,J. Appl. Meteorol., 10, 4,
703-7!4, 1971.

MSller, F., Vierteljahreskarten des Niederschlagsffir die ganze Erde, Pettermanns Geogr. Mitt., 95, 1, 1-7, 1951.
MSller, F., On the influence of changesin the CO, concentrationin air on the
radiation balanceof the Earth's surfaceand on climate, J. Geophys.Res.,
68, 13, 3877-3886, 1963.

Mfiller, D., Kreislauf des Kohlenstoffers,Handbuch,die Pfianzenphysiologie,

12, 2, 934-948, Berlin, 1960.

Newell, R. L., The globalcirculation of atmosphericpollutants, ScientificAmerican, 224, 1, 32-42, 1971.

Newmen, J., and A. Cohen, Climatic effectsof aerosollayers in relation to solar
radiation, J. Appl. Meteorol., 11, 4, 651-657, 1972.
North, G. R., Analytical solution to a simple climate model with diffusive heat
-transport,J. Atmos. Sci., 32, 7, 1301-1307, 1975.
North, G. R., Theory of energy balance climate models,J. Atmos. Sci., 32, 11,
2033-2044, 1975a.

Opik, E. J., On the causesof the paleoclimaticvariations and of the Ice Ages in
particular, J. Glacio!., 2, 13, 213-219, 1953.
Opik, E. J., Solar variability and paleoclimaticchanges,Irish Astron. J., 5, 4, 597, 1958.

Pidoplichko,I. G., Contemporaryproblemsand the study of the history and

environmental

conditions of faunas in natural

conditions and faunas in the

past (in Russian), 1, 152 pp., Izd. AN Ukr. SSR, Kiev, 1963.
Pivovarova,Z. I., Long-termvariations of the solar radiation intensity basedon
observations at actinometric stations (in Russian), Tr. GGO, 233, 17-29,
1968,

Plass, G. N., The carbondioxide theory of climate change, Tellus, 8, 2, 140-154,

BIBLIOGRAPHY

257

Pokrovskaya, T. V., On the radiation factors of climatic variations (in Russian),

Meteorol. Gidrol., 10, 31-38, 1971.

Rakipova, L. R., Thermal Regime of the Atmosphere (in Russian), Gidrometeoizdat, 183 pp., Leningrad, 1957.
Rakipova, L. R., Climatic changesas a result of the modification of the Arctic ice
(in Russian), Meteorol. Gidrol., 9, 28-30, 1962.

Rakipova, L. P., Changesof the zonal distribution of atmospherictemperature

as a result of active climatic modifications in Contemporary Problems of
Climatology (in Russian), 358-384, Gidrometeoizdat, Leningrad, 1966.
Rakipova, L. R., and O. N. Vishnyakova, The dependenceof the atmospheric
thermal regime on the carbondioxide concentration (in Russian), Meteorol.
Gidrol., 5, 23-32, 1973.

Ramsay, W., Orogenesisund Klima. Ofversigt of Finska Vetensk. Soc. FSrh., 52

A, 11, 1910.
Raschke, E., F. MSller, and W. Bandeen, The radiation balance of the Earth-

atmospheresystem over both polar regions obtained from radiation measurementsof the Numbus II meteorologicalsatellite, SverigesMeteorol. och

Hydrol. Inst. Medellanden, Ser. B, 28, 104, 1968.

Raschke,E., et al., The radiation balanceof the Earth-atmospheresystemfrom

NimbusIII radiationmeasurements,
NASA TechnicalNote, D-7242,73 pp.,

Washington, 1973.
"
Rasool,S. I., and S. H. Schneider,Atmosphericcarbondioxideand aerosoleffect
of large increaseson global climate, Science,173, 138-141, 1971.

Robinson,C. D., Surfacemeasurements

of solarand terrestrialradiationduring
the IGY and IGG, Ann. Intern. Geophys.Year, 32, 17-61, 1964.
Romanov, V. V., Hydrophysicsof Swamps (in Russian), 359 pp., Gidrometeoizdat, Leningrad, 1961.
Romer, A. S., Vertebrate Paleontology,687 pp., Univ. of ChicagoPress,Chicago, 1945.

Ronov,A. B., On the post-Cambriangeochemicalhistoryof the atmosphereand

hydrosphere (in Russian), Geokhim., 5, 397-409, 1959.

Ronov, A. B., General tendenciesin the evolution of the compositionof the

Earth's crust, ocean and atmosphere, Geokhim., 8, 715-743, 1964.

Ronov, A. B., Evolution of rock compositionand geochemicalprocessesin the

Earth's lithosphere(in Russian),Geokhim., 2, 137-147,1972.
Rosenberg,G. V., Optical studies of atmosphericaerosol (in Russian), Fiz.
Nauk, 95, 1, 159-208, 1968.

Rubinshtein,E. S., On the problemof climatic changes(in Russian),Tr. NIU,

GUGMS, 1, 22, 3-83, 1946.

Rubinshtein,E. S., The Structureof Fluctuationsof Air Temperaturein the

Northern Hemisphere(in Russian),34 pp., Gidrometeoizdat,Leningrad;
1973.

Rubinshtein,
E. S., andL. G. Polozova,
Contemporary
Climaticchanges
(in

Russian), 268 pp., Gidrometeoizdat,Leningrad, 1966.

Rukhin, L. B., Problemsof the origin of continentalglaciations(in Russian),

Izv. VGO, 1, 25-38,

258

CLIMATIC

CHANGES

Rukhin,L. B., Principlesof GeneralPaleogeography,

2nded., 628pp., Gostoptekhizdat, Leningrad, 1962.

Rutten, M. G., The Origin of Life by Natural Causes,Elsevier Publ. Co.,

Amsterdam-London-New York, 1971.

Saks, V. N., The QuaternaryPeriodin the SovietArctic (in Russian),2nd ed.,

Vodtransizdat, 267 pp., Leningrad-Moscow,1953.

Saks,V. N., Geologicalhistoryof the Arctic Oceanduring the Mesozoicera (in

Russian),Intern. Geol.Congr.XXI session,
ReportsofSovietgeologists,
pp.
108-123, Gosgeoltekhizdat,Moscow, 1960.
Saltzman, B., and A.D. Vernekar, Note of the effect of Earth orbital radiation
variations on climate, J. Geophys.Res., 76, 18, 4195-4197, 1971.

Sarasin, P., and F. Sarasin, La temperaturede la priode glaciare, Verh.

Naturforsch.Ges., 13, 603-615,Basel, 19)1.
Savinov,S. I., Maximum valuesof the solar radiationintensity accordingto
observations
in Pavlovskbeginningin 1892(in Russian)Izv. AN USSR, 6, 7,
12, 707-720, 1913.

Sawyer,J. S., Notesonthe response

ofthe generalcirculationto changesin the

solar constant.Changesof climate, Proc. Rome Symp. UNESCO, 333-338,

1963.

Sawyer,J. S., Possiblevariationsof the generalcirculationof the atmosphere.

World climatefrom 8000to 0 B.C., Proc.Intern. Symp.18-19April, 1966,
Roy. Meteorol. Soc., 218-229, 1966.

Schell,J. J., On mathematicaland "natural" modelsfor the study of climate

changes,J. Appl. Meteorol., 10, 6, 1344-1346,1971.

Schneider, S. H., Cloudinessas a global climatic feedbackmechanism:the
effectson the radiation balanceand surfacetemperatureof variations in
cloudiness,J. Atmos. Sci., 29, 1413-1422, 1972.

Schneider,S. H., and Tzvi Gal-Chen, Numerical experimentsin climate stability, J. Geophys. Res., 78, 27, 6182-6194, 1973.

SchwarzbachM., Das Klima der VSrzeit. Eine Einfiihrung in die Paliioklimatologie, 211 S., Stuttgart, 1950.
Schwarzbach,M., Neue Eiszeithypotesen,
Eiszeitalter and Gegenw., 19, 250261, 1961.

Schwarzschild,
M., Structureand Evolutionsof the Stars, 296 pp., Princeton
Univ. Press, Princeton, 1958.

Sellers,W. D., PhysicalClimatology,272pp., Univ. ChicagoPress,1965.

Sellers, W. D., A global climatic model based on the energy balance of the
Earth-atmospheresystem,J. Appl. Meteorol., 8, 3, 392-400, 1969.
Sellers,W. D., The effectof changesin the Earth'sobliquityon the distribution
of mean annual sea-leveltemperature,J. Appl. Meteorol.,9, 960-961,1970.
Sellers, W. D., A new global climatic model,J. Appl. Meteorol., 12, 2, 241-254,
1973.

Sellers,W. D., The effectof solarconstant

variationonclimatemodeling,Proc.

Workshop:The Solar Constant and the Earth's Atmosphere,Calif. Inst.

Technol. p. 206-209, 1975.
Severtsev,S. A., PopulationDynamicsand the Adaptation Evolution of Animals (in Russian), 316 pp., Izd. AN SSSR,

BIBLIOGRAPHY

259

Sharaf, Sh. G., and N. A. Budnikova, Secular variations of the elements of the
Earth's orbit and the astronomical theory of climatic changes (in Russian),
Tr. Inst. Teor. Astron., 14, 48-85, 1969.

Shifrin, K. S., and I. N. Minin, On the theory of nonhorizontal visibility (in

Russian), Tr. GGO, 68, 5-75, 1957.

Shifrin, K. S., and N. F. Pyatovskaya, Tables of the Visibility Range and

Daylight Sky Brightness (in Russian), 211 pp., Gidrometeoizdat, Leningrad, 1959.
Shmal'gauzen,I. II., Evolutionary Factors (in Russian), 396 pp., Izd. AN USSR,
Moscow, 1946.

Shnitnikov, A. V., Intrasecular Variability of the General Humidity Components (in Russian), 245 pp., Nauka, Leningrad, 1969.
Shuleikin, V. V., Physics of the Sea (in Russian), 1083 pp., Izd. AN USSR,
Moscow-Leningrad, 1941, 1953, 1968.
Shvets, M. E., et al., Numerical modelof general atmosphericcirculation over a
hemisphere (in Russian), Tr. GGO, 256, 3-44, 1970.
Simpson, G. C., Past climates, Quart. J. Roy. Meteorol. Soc., 53, 221, 213-226,
1927.

Simpson, G. C., Ice Ages, Nature, 141,591-598, 1938.

Simpson, G. C., Tempo and Mode of Evolution, New York, 1944.
Sinitsyn, V. M., Introduction to Paleoclimatory (in Russian), 232 pp., Nedra,
Leningrad, 1967.
Smagorinsky,J., General circulationexperimentswith the primitive equations.
1. The basic experiment, Mon. Weather Rev., 91, 99-164, 1963.
Smagorinsky,J., Global atmosphericmodelingand the Numerical simulation of
Climate. Weather Modification, ed. W. N. Hess, John Wiley & Sons, New
York, 1974.

Smagorinsky, J., S. Manabe, and I. L. Holloway, Numerical results from ninelevel general circulation model of the atmosphere,Mon. Weather Rev., 93,
727-768, 1965.

Spirina, L. P., On the influence of volcanicdust on the thermal regime of the

northern hemisphere (in Russian), Meteorol. Gidrol., 10, 38-45, 1971.
Strakhov, N.M., Historical Geology(in Russian),2,428 pp., Izd. MGU, Moscow,
1937.

Strakhov, N.M., Types of climatic zonesduring the post-Proterozoichistory of

the earth and their significancefor geology (in Russian), Izv. AN USSR,
Ser. Geol, 3, 3-25, 1960.

Strokina, L. A., On the comparisonof calculated and observedvalues of the

ocean'sradiation balance (in Russian), Tr. GGO, 193, 44-47, 1967.
Suarez, J. M., and I. M. Held, The effect of seasonallyvarying insolation on the
simple albedo-feedback
model, Proc. WMO/IAMAR, Symp. on Long-Term
Climatic Fluctuations, WMO, 421,407-414, 1975.
Takahashi, Carbondioxidecyclein the seaand atmosphere.The Encyclopaedia
of AtmosphericSciencesand Astrogeology,Ed. R. W. Fairbridge, Reinhold
Publ. Corp., New York-Amsterdam-London,131-135, 1967.
Teis, R. V., and D. P. Naidin, Paleothermometryand isotopecontentsof oxygen
in organogeniccarbonates(in Russian), 255 pp., Nauka, Moscow,

260

CLIMATIC

CHANGES

The Solar Constant and the Earth's Atmosphere, Proc. Workshop, Big Bear
Solar Obs., Big Bear City, 1975.
Tyndall, I., On the absorptionand radiation of heat by gasesand vapoursand on
the physicalconnectionof radiation absorptionand conduction,Phil. Mag.,
22, 144, 167-194, 273-285, 1861.

Velichko, A. A., Natural Processin the Pleistocene (in Russian), 256 pp.,
Nauka, Moscow, 1973.
Vernadsky, V. I., The Biosphere, 146 pp., ONTI, Leningrad, 1926.
Vernekar, A.D., Long-period global variations of incoming solar radiation,
Meteorol. Mon., 12, 34, 21 pp., 1972.
Vinogradov, A. P., Introduction to the Geochemistryof the Ocean (in Russian),
213 pp., Nauka, Moscow, 1967.
Vinogradov, A. P., Atmosphericchangesunder the influence of human activity
(in Russian), Geokhim., 1, 3-10, 1972.
Voeikov, A. I., Climates of the World, particularly of Russia (in Russian), St.
Petersburg, 1884.
Voeikov, A. I., Climatic fluctuations (in Russian), Meteorol. Vestn., 1, 1-12,
1902.

Wallace, A. R., Origin of human races and the antiquity of man deducedfrom
natural selection, Anthrop. Rev., 2, CLVIII, 1864.
Wallace, A. R., Geologicalclimates and origin of species,Quart. Rev., 126 pp.,
1869.

Washington, W. M., On the possibleuse of global atmospheric models for the

study of air and thermal pollution. Man's Impact on the Climate, pp. 265276, MIT Press, Cambridge, Mass., 1971.
Washington,W. M., Numerical climatic-changeexperiments;the effectof man's
productionof thermal energy, J. Appl. Meteorol., 11, 5, 768-772, 1972.
Weather and Climate Modification, Washington, 1966.
Weatheraid, R. T., and S. Manabe, The effectsof changing the solar constanton
the climate of a general circulation model, J. Atmos. Sci., 32, 11, 2044-2059,
1975.

Wexler, H., Radiation balance of the Earth as a factor of climate changes.

Climatic change:evidence, causesand effects, Harvard Univ. Press, 1953.
Wexler, H., Variations in insolation, general circulation and climate, Tellus, 8,
480-494, 1956.

White, O. R., Sun. The encyclopaediaof atmosphericsciencesand astrogeology,

ed. R. W. Fairbridge, Reinhold Publ. Corp., New York-Amsterdam-London,
1967.

Williams, J., R. G. Barry, and W. M. Washington, Simulation of the climate at

the last glacial maximum using the NGAR global circulation model. Inst.
Arctic and Alpine Res., Occas.Paper, 5, 23 pp, 1973.
Wilson, A. T., Origin of Ice Ages: an ice shelf theory for Pleistoceneglaciations,
Nature, 201, 477-478, 1964.

Wrist, G., Oberfichensalzgehalt,Verdunstungund Niederschlagauf dem Weltmeere. Lnderkundliche Forschung, Festschrift Norbert Krebs,

BIBLIOGRAPHY

261

Yamamoto, G., and M. Tanaka, Increaseof global albedodue to air pollution, J.

Atmos. Sci., 29, 8, 1405-1412, 1972.

Yudin, M. I., The influence of foreststrips on the turbulent exchange,and their

optimal width, in Problems of HydrometeorologicalEfficiency of Afforestation for Field Protection (in Russian) 84 pp., Gidrometeoizdat, Leningrad,
1950.

Yudin, M. I., On the possibilitiesof controlling large-scaleatmosphericmotions

in Contemporary Problemsof Climatology, (in Russian), pp. 393-412, Gidrometeoizdat, Leningrad, 1966.
Zeuner, F., The Pleistocene Period, Hutchinson, London, 1959.
Zubenok, L. I., Water Balance of the Continents and Oceans (in Russian), Dokl.
AN USSR, 108, 5, 829-832, 1956.
Zuev, V. E., L. S. Ivlev, and K. Ya. Kondrat'ev, New results in studies of
atmosphericaerosol (in Russian), Izv. AN USSR, Fis. Atmos. Okeana, 9, 4,
371-386,