Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 25
Water Supply and
the Future Climate
1
STEPHEN H. SCHNEIDER and RICHARD L. TEMKIN
National Center for Atmospheric Research
WHAT IS CLIMATIC CHANGE?
This chapter addresses a variety of questions related to
issues of understanding the climate and how it might
change and, additionally, how climate is related to the
water supply. Many conflicting aspects of climate have
recently been discussed in the news. One hears about
approaching ice ages on the one hand and the melting of
the ice caps on the other with both natural and human-
induced postulated causes. For example, Reid Bryson (in
Alexander, 1974) has said that "there is very important
climate change going on right now, and if the trend
continues, will affect the whole human occupation of the
earth—like a half billion people starving." On the other
hand, former U.S. Secretary of Agriculture Earl Butz has
been quoted to the effect that such statements are at best
without scientific bases and are at worst apocalyptic
nonsense. Obviously, there is much confusion on the
issue. Although the following discussion may add little to
reduce the uncertainties, it is an attempt to show the
range of arguments used, with a further attempt to tie the
issues to questions of water supply in order that one may
25
gain a feeling for the types of uncertainties that decision
makers may have to face in the future.
Figure 1.1 shows some long-term temperature records
extracted from ocean-sediment cores over the last 700,000
years (Emiliani, 1972~.
There remains considerable dispute about the mag-
nitude of the temperature fluctuations, which are on the
order of 5°C, but the figure does, nonetheless, show an
important point: over this several-hundred-thousand-year
period there were fairly large excursions in temperature,
with the cold and warm periods known as glacials and
interglacials, respectively, these being separated in time
by some 10,000 to 100,000 years. From a statistician's
point of view, this record might appear to resemble a
stationary time series. However, from the perspective of a
human lifetime, or for that matter all of human history, the
record contains dramatic climatic changes. Hence one
may raise the question: "What is climatic change?" In
reply, it should be stated that climate is a time average of
the instantaneous state of atmosphere (i.e., the weather
events) and that the weather itself is unpredictable in
detail past a few weeks (see CARP, 1975, for a review).
OCR for page 26
26
- 304
25-
c~
_
20-
0 100 200 300 400
TODAY ~ TIM E ( X 1000 YEARS AGO)
FIGURE 1.1 Long-term temperature trends extracted from
ocean sediment cores (after Emiliani, 1972).
Some people, in fact, believe it to be unpredictable in
practice after only a few days, but it is theoretically
possible that some skill of weather prediction exists up to
the period of a few weeks. The atmosphere scrambles
itself to a point where there is practically no recognition
of its initial condition after some two to four weeks.
Therefore, any climatic average that one takes that is
longer than that predictability period is in essence av-
eraging a fluctuating time series of unpredictable weather
events. Does this mean that there can be no climatic
predictability a month or a season or a decade ahead? The
answer is clearly "yes" for weather but maybe "no" for
climate, since there is no established theoretical reason
why a time average of weather must also be unpredicta-
ble. From this average one would not be able to predict,
for example, on what day in the next month and where a
storm is going to take place; but one might be able to
predict a few months into the future the number of storms
that may pass through a given region in a given time. This
latter possibility still remains theoretically conceivable
with above-zero probability.
In 1974, at the Global Atmospheric Research Program
Conference on Climate Modeling (held in Stockholm), a
large number of experts attempted to set up a plan for
future climate research (GARP, 1975~. On the first day of
the conference, about a dozen modelers, who were
mainly interested in discussing questions about parame-
terizing their models and deciding what physical theories
and observations were needed for their verification, de-
cided that in the first half hour they would dispose of the
issue of a definition of climate. Hours later they were still
working on this definition. Essentially, the definition
began from the obvious fact that climate is just a time
average of weather. However, this being rather imprecise
implied the need for some subdefinitions, e.g., the
"climatic time series," which was defined as the time
series of some fluctuating climatic parameter, say, rain-
fall. A "climatic sample" was also defined. It is the length
of time over which one computes statistics for that time
series. The "climate" is simply the statistical properties of
the climatic time series taken over a specified sample
period. When one takes two climatic samples and finds
that they are different, "climatic change" is further de-
fined. One still needs to investigate the statistical signifi-
cance of those differences. That is, one must ask: Does the
STEPHEN H. SCHNEIDER and RICHARD L. TEMKIN
difference in the climate of the two samples occur be-
cause of unpredictable fluctuating components, e.g., the
daily weather, or because of changes in the long-time
statistics (or an ensemble mean) traceable to some varying
physical forcing mechanism, e.g., usually a boundary forc-
ing factor such as a volcanic dust veil?
From the above it is clear that in studying climatic
change the length of the averaging period and the size of
the averaging region are quite important. In fact, the
physical processes that are most influential on the short
time scales could well be different from those that operate
on the long ones. Thus, a result of the Stockholm meeting
(as seen from the participants' lengthy attempt to define
climate) is that anyone referring to "climate" should be
certain that He sample length is more than three weeks
(which is more than the period of weather predictability)
and that extreme care should be used in specifying the
averaging period and statistical procedure. It is important
that precision be used in the specification of the defini-
tion of climate. In the remainder of this paper, samples
and averages will be considered for much shorter terms
than ice ages and for periods longer than three weeks.
Figure 1.2 shows that the climate fluctuates on all time
scales (U.S. Committee for the Global Atmospheric Re-
GENERALIZED TRENDS IN GLOBAL CLIMATE
Air Temperature
COED WARM
1960 ' ' <'
1920 _
880 - ~~N
O .2 ~ .6
to C
900
COED WARM
1900 _ ~
~500 _~i
1300 _ ~
1100 - (EASTERN-
- ~EUROPE-
(a)THE LASTIOO YEARS I ~1.5°C I
(b)THE LAST 1,000 YEARS
Mid-Latitude Air Temperature
COLD WARM
- ~
o
In
~ 10
o 15
In
C] 20 .
In
0 25
I
30 .
LEGE N D
1. Thermal Maximum
of 1940s
2. Little Ice Age
3. Cold Interval
4. Present Interglacial
(Holocene)
5. Last Previous Inter-
glacial (Eemian)
s
COLD WARM
o
CO
In
LU
o
.: IOC
In
0 12e
2e
5C
7e
150 \
1 1 1 1
~ 10°C ~ 10°C
(c) THE LAST 10,000 YEARS (d ) THE LAST 100,000 YEARS
FIGURE 1.2 Generalized trends in global climate represented
by approximate surface-temperature patterns that prevailed over
a variety of time scales for the past lSO,OOO years.
OCR for page 27
Water Supply and the Future Climate
search Program, 19751. In the past 1000 years, tempera-
ture has fluctuated on the order of 1~/2°C. This may seem
to be of minimal magnitude, but, from Figure 1.1, it is
recalled that the magnitude of the glacial to interglacial
oscillation was only about 5°C globally, although actually
many times this at the high latitudes (Matthews et al.,
1971~. The cool period in Figure 1.2(b) between about
1400 and 1800 is known as the Little Ice Age in Europe
and shows a decrease in temperature of only about 1°C. A
change of this magnitude for food crops can be put in
perspective when one notes that with a couple of degrees
movement northward in latitude through the central part
of the United States one would find that a 10-day drop in
growing season is matched roughly by a 1°C drop in
average temperature. The implication here is that a
large-scale areal change in temperature of some 1°C is not
necessarily trivial for people. Finally, the recent record
EFigure liars shows a hemispheric warming of about half a
degree to 1940, followed by a cooling. Again a problem
exists with this record in that it is really an estimate of a
hemispheric average. Even though it is based on instru-
mental observations taken at many points, there still re-
main large areas left uncounted or uncovered by mea-
surements. When referring to it as a hemispheric average,
sampling error bars on the order of tenths of a degree
should probably be placed on this record. It is clear then
that difficulties exist even in pinning down past climatic
changes to a very high degree of accuracy. It has also been
mentioned that human societies are vulnerable to small
changes in the climatic system and that a hemispheric
temperature change of only a few degrees, or even
perhaps tenths of a degree, can be important.
PHYSICAL FACTORS CAUSING CLIMATIC
CHANGE
Figure 1.3 (from Schneider and Mesirow, 1976) contains a
very simple overview of the mechanisms that drive our
climate. The straight arrows represent the energy coming
from the sun, of which annually about 30 percent is re-
flected (the reflectivity is called the earth's albedo). The
wavy arrows represent the outgoing planetary infrared
radiation. It is well known that the tropics are warm
because they receive more heat from the sun than they
emit as infrared radiation, and the poles are cold because
of a greater average solar zenith angle and the presence of
more highly reflective ice- and snow-covered surfaces.
Thus in a simple picture the warm air rises in the tropics
and moves poleward. As the air rises, it takes with it the
fast rotational speed of the equator, which provides the
momentum that creates the westerly winds in the mid-
latitudes. Because the circulation is driven by the
equator-to-pole temperature difference, the system is
much more vigorous in the winter hemisphere than in the
summer hemisphere.
This equator-to-pole temperature difference driven cir-
culation is quite important, since a variation in global
27
Terrestrial Infrared Radiation
Reflected Sunlight
~ ~ <
Ode \ Incoming Sunlight
Winds . ~
~>4
,,, ~ \
Ode
Winds
-Equator Tropics-
\ ~~: /
my\
Terrestrial Infrared Radiation
Reflected Sunlight
FIGURE 1.3 Schematic illustration of how the major weather
systems of the earn are driven by the unequal heating between
the equator and the poles. The tropics intercept a much larger
fraction of Me incoming solar energy than do the polar zones,
thus giving rise to the motions that regulate We climate.
mean temperature is usually amplified at the poles. For
example, from Figure 3.6 of Matthews et al. (1971), it is
seen that the major recent temperature change (in the
northern hemisphere at least) occurred in the higher
latitudes. Some theoretical studies (e.g., Manabe and
Wetherald, 1975) also suggest that a change in global
temperature from changed surface heating will be
amplified at the poles and thus will result in a change in
the equator-to-pole temperature gradient. Since the circu-
lation systems depend on this gradient, a seemingly small
change in global temperature may thus latitudinally shift
the average location of the circulation systems. Based on
both some theoretical considerations and observations,
Bryson (1974) argues that if the north polar region cools
more than the equatorial region, the Asian and African
monsoon belts get compressed, that is, they would move
slightly equatorward, and the midlatitude baroclinic
zones, the westerly belts, would also move toward
the equator. Comparable evidence (e.g., Kellogg, 1977)
exists suggesting from reconstructions of climatic warm
periods that the opposite occurs when the poles warm
more than the equator. This hypothesis is beginning to
appear to have some theoretical support, since some re-
cent numerical modeling experiments also show a tem-
perature gradient monsoon effect. For example, Gates
(1976) and Williams et al. (1974) have shown with numeri-
cal simulations that during the last (Wisconsin) ice age,
the monsoon rains were weakened in their simulations.
Nevertheless, the connection between circulation re-
gimes and equator-to-pole surface-temperature gradient
is still based on fragmentary evidence and thus remains
somewhat controversial and qualitative.
OCR for page 28
28
The point to be made here is that marginal, seemingly
small changes in zonal or hemispheric means could
in turn, cause large changes in local regions that
are near boundaries of different circulation regimes. But
even if one of these shifts did occur, it would not neces-
sarily mean that the "end of the world" has come. If the
globe warmed up a degree or two and if the U.S. corn belt
then moved from the Iowa area northward by a few
hundred kilometers, the result might be but a small per-
turbation from a global evolutionary point of view to
world food supplies and the earth's carrying capacity.
However, in the present world situation in which people
are locked into national boundaries, and there is little
global food reserve, such marginal shifts could be serious
(see the discussion in Schneider and Mesirow, 19761.
Since crops are usually planted based on the pre-existing
climatic conditions and their expected continuance, slow
and gradual changes can be anticipated before crops are
planted, thus perhaps avoiding a crisis. One of the main
points to make here is that if changes should come
quickly, a catastrophe might well result.
What causes the climate to change? Obviously, the
output of the sun is very much implicated. Figure 1.4
(Thompson, 1973) is a plot of the double sunspot cycle.
Since there are not negative sunspots, the part below the
X axis is just the second half of the double sunspot cycle
when the magnetic polarity of the spots has reversed. The
figure further shows that there have been droughts in the
U.S. high plains that correspond to alternate minima in
the double sunspot cycle. A question arises as to whether
a new drought will be beginning in the late 1970's, since
the sun has recently passed through another minimum.
Problems, however, arise as to why there may be a good
correlation here. One could ask: Why do something like
droughts in the U.S. plains correlate with sunspots and
not some other atmospheric variable represented by a
long-term continuous record?
These droughts occur, apparently, in different parts of
the plains, and they move around. Nor is their occurrence
On
lL
id
o
A
an
J
43:
Z 1740 1760 1780 1800 1820 1840 1860 1880 1900 1920 1940 1960 1980
YEAR
l
Solar Cycle And Drought In Nebraska
Drought Periods H 1 H H I I H H H ?
1 1 1 1 1 1 1 1 1 1 1
FIGURE 1.4 A plot of the double sunspot cycle versus drought
in Nebraska. The graph suggests that droughts in the U.S. Great
Plains tend to occur in a 22-year cycle centered near the minima
of that cycle. If this relationship holds, then the next such
drought is "due" in the last half of the 1970's (after Thompson,
1973).
STEPHEN H. SCHNEIDER and RICHARD L. TEMKIN
precisely timed with alternate sunspot minima. But, why
do they occur in the plains? Others have looked for
periodicities to match with the sunspot cycles, of which
there are thousands of possibilities. The problem is this:
without a physical connection one can get oneself in
statistical trouble. Figure 1.5 (Stetson, 1937) shows an
example of the sort of work that has been done in this
area. Here, the sunspot number is correlated with the
Dow tones stock market average, and the quality of wine
vintage, or number of automobiles. In Figure 1.6 (Stetson,
1937) there is an even more fascinating "correlation"—
sunspot number and rabbit population. This type of corre-
lation attempt may be disturbing to us as scientists; but
perhaps vegetation is affected by sunspots. If one has a
physical theory, then one has some confidence in such
statistical correlations that is, one could take a time
series of the variability of the sunspot cycle and look
around for geophysical phenomena like wine harvests or
ant
1924 1926 1928 1930 1932 1934
<,, 500
~t 450 _
(I 400 _
O 350 _
C) 300 _
A: 250 _
of 200 _
I 150 _
100
10O
Pn~
40
20
O _
h
1925 1927 1929 1931 1933
.... .....
1925 1 927
1931 1933 1935 1937
FIGURE 1.5 Plots of the Dow Jones stock market averages,
wine vintages, and automobiles compared with the number of
sunspots (after Stetson, 1937).
OCR for page 29
Water Supply and the Future Climate
70
60
50
40
30
20
In
A:
J
o
11
o
z c/)
—z
O
J J 1
I< J
e - ... ...
1923 1925 1927 1929
100—
An
an
on
O-
1935 1 937
1925 1927 1929 1931 1933 1935 1937
FIGURE 1.6 Plots of business, rabbit population, building con-
tracts, and tree growth compared with the number of sunspots
(after Stetson, 1937).
droughts, and then one might find some fit. But if one
looked at a hundred of these possibilities and found one
with 99 percent confidence, then one should question
this kind of correlation on both statistical and physical
grounds. On the other hand, there may be something
physical there, nonetheless. But one must be extremely
cautious.
The fact is, of course, that sunspots are a visible man-
ifestation of change on the element that provides us with
our primary energy source. Could there also be a change
in the "solar constant" with sunspots? Nobody has mea-
sured the solar constant to better than 1-2 percent. Yet
present models indicate that a change of this magnitude
29
would be sufficient to cause the kind of climatic changes
seen in Figure 1.2. Eddy (1976), in his recent analysis of
historical records, has shown that the sun may have lost
its spots from about 1640 to 1700, which corresponds with
the fairly consistent and strong temperature drop on most
of the paleoclimatogical records for this time. Whether or
not this is indicative of change in the solar constant is still
an open question (see Schneider and Mass, 1975~. Again,
why are the droughts seemingly localized near Nebraska,
and why is the cycle not reflected in many other global
records?
Some have invoked physical mechanisms other than
solar constant variation with sunspots (e.g., Dickinson,
1975), namely, changes in the magnetic field of the earth,
which modulate galactic cosmic rays, which create parti-
cles that form in the stratosphere, which, in turn, might
make cirrus clouds, which can perturb radiation fluxes in
the climatic system with enough energy to explain some
terrestrial climate fluctuations. The real problem is that in
the absence of either adequate measurement or adequate
theory, one is basically left with the statistical correla-
tions; and in the absence of theory, a fight is quite likely.
At least at present there are some people seriously look-
ing at the theoretical aspects of a solar-climate relation-
ship.
Another possible cause of climatic change with sup-
porting evidence is volcanic eruptions. Figure 1.7 (Ellis
and Pueschel, 1971) shows the apparent transmittance of
solar direct beam radiation at Manna Loa in Hawaii at the
3000-m height observatory. In 1963, there was about a 2
percent drop in the direct solar radiation reaching the
observatory, although 75 percent of this attenuated direct
beam would still reach the surface because of scattering
in the forward direction. This decrease occurred im-
mediately following the Agung volcanic eruption. Note
that roughly a half percent decrease in total solar energy
reaching that station occurred and lasted for several years
following that eruption. If model calculations are right, a
decrease in solar radiation of that magnitude might drop
the earth's global temperature on the order of several
tenths of a degree. Volcanism is thus another potential
mechanism of climate change.
On short time scales, and maybe even longer ones,
there is always the possibility of internally caused cli-
matic changes. The atmosphere could be viewed as a very
fast oscillating device connected to the oceans by some
spring, with the oceans being an even larger mass con-
nected by a bigger spring to another mass, which are the
glaciers. This whole system is an oscillatory one. By
analogy, there is difficulty in determining whether some
of the observed climatic changes are due just to the
internal vibrations of this system, i.e., the redistribution
of energy among the main reservoirs the atmosphere,
the oceans, and the glaciers or are, in fact, due to exter-
nal pushes by variations on the sun or by volcanoes. A
problem lies in separating out the external factors the
volcanoes, the sun, or even carbon dioxide from human
activities—from the internal redistributions. Based on
OCR for page 30
30
IL
An
He
93
en ~
Z Z
~ cry 92
llJ
~ Q
By _
111
1 1 1 1 1 1 1 1 1 1 1 1
94 ~ ~ to - LIT
1-1I[14=~ II I S~.
, . . ~ ~ ~ ~ ~
It I I f I t I I ~ I I
1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970
T I ME (YEARS)
91
90-
essentially intuition, one might speculate that the
shorter-time-scale fluctuations are internal, but the long-
term changes are externally caused. The problem is that
no quantitative theory of climate exists to end the intui-
tive speculations.
CLIMATIC MODELS AS ESTIMATORS
OF CLIMATIC CHANGE
There is one other important "externality" in the system,
and that is people. Technically, this is not the traditional
economic usage of the term "externality" (i.e., external
economies or diseconomies). Since people are part of the
biota, which are in our definition part of the climatic
system, people are not truly an externality to the climatic
system. However, in an important sense the social cost of
many peoples' activities (e.g., those that release carbon
dioxide to the atmosphere) does represent a situation
where the producers (or users), who benefit from the
activity that generates, say, this carbon dioxide, do not pay
for the costs at the time of energy usage. Instead, the costs
will be "externalized" to future generations. Thus, the
analogy to the economic meaning of externality is valid.
Figure 1.8(a) is a graph from a projection made in 1971
(Machta, 1971) of the carbon dioxide concentration in the
atmosphere. Burning of fossil fuels puts carbon dioxide
into the atmosphere, of which roughly a half to three
fourths stays in the air, with the remaining fraction taken
up by the oceans and biosphere. It also appears rea-
sonably certain that an increase in carbon dioxide would
affect the radiation balance of the earth in such a way as to
warm the earth's surface. But quantitatively, how much of
a heating might occur is another unanswered question.
One has to go to a model to make such estimates because
Here are no available historical data to perform an actuar-
ial study. There is no analogy in geological history (re-
cent at least) from which one can obtain a scaling factor
for the climatic response to anthropogenic physical labo-
ratory experiments (which in fact do not really exist in this
case).
It is probably true that these theoretical models of the
STEPHEN H. SCHNEIDER and RICHARD L. TEMKIN
FIGURE 1.7 Plot of apparent atmospheric
transmittance at Mauna Loa Observatory.
Note the years of major volcanic eruptions
(after Ellis and Pueschel, 1971).
CO2 radiation effect have some observational justifica-
tion. After all, one can predict the surface temperatures
and the vertical temperature profiles of the CO2 atmo-
spheres of Mars and Venus, albeit not perfectly, but Hey
still agree somewhat with the observations of those
planets. This gives some indication that the models are
not two orders of magnitude off. However, the issue as to
He quantitative accuracy of He models remains. Models
show that the increase in temperature from about 1900 to
1975 should have been on the order of a few tenths of a
degree globally because of the increased carbon dioxide
(see Figure 1.81. Therefore, one could argue that the
carbon dioxide theory has been proved wrong by the fact
Hat the northern hemisphere, at least after 1945, has
cooled even though CO2 increased exponentially. How-
390 1 T ~ ~ i I
Mode! Calculation Of Atmospheric CO2
Frarn rnmh'~clinn Of Fr`.R;! FIIaI.
(a)
, I , r I ~ 1
Mods' Coicuiation Of Atmospheric CO2
_ From Combustion Of FOSRjI Fuels
380
~ 370
o
, 360
350
of
o
~ 340
in
i 330
cat
o
c,
° 3~0
he
° 300
290 _
280 ~ 1 1 1 1 1 1 1 1 1 1 1
1860 1880 1900 1920 1940 1960 1980 2000 1860 1880 1900 1920 1940 1960 1980 2000
/
. — _
/~ ~ MCL O LOO Observed j
YEAR YEAR
(b)
FIGURE 1.8 Projections of atmospheric carbon dioxide con-
centration from fossil fuels calculated by models of the carbon
dioxide cycle. Projection (a) shows an early estimate of 375 parts
per million (ppm) CO2 concentration by the year 2000, whereas
an updated model (b) predicts a CO2 concentration of near 390
ppm. Projection (a) is from Machta (1971), and projection (b) is
from Machta and Telegadas (1974).
OCR for page 31
Water Supply and the Future Climate
ever, since the fluctuations in global temperature that
have occurred naturally before humans could have had
any real impact have been on the order of l/2 °C, or perhaps
even larger, the cooling does not eliminate the CO2 warm-
ing argument. Therefore, when a model predicts a change
on the order of a few tenths of a degree, it is impossible to
distinguish this carbon dioxide temperature "signal" from
the natural climatic "noise." Thus it cannot be said
categorically that the theory is incorrect. However, car-
bon dioxide in the atmosphere is increasing exponen-
tially. According to most projections, the next 10 percent
increase would occur in about two decades with the sub-
sequent 10 percent increase in about one decade. As
Broecker (1975) indicated, one may then certainly expect
the CO2 effect to exceed the climatic noise level; that
is, if the modeling predictions are correct, the signal will
become dramatically detectable very quickly (he projects
sometime after 1980~.
A test for the verification of these models is for them to
show a high degree of faithful reproduction of atmo-
spheric variability. Yet, on the other hand, a test of these
models for sensitivity (to CO2 increase, e.g.) is very dif-
ficult to devise. One is left with the terrible dilemma that
in order to verify this kind of climate change (i.e., a
detectable, significant climatic change from CO2 possibly
as early as a generation away) one uses a tool that itself is
not entirely verifiable. Perhaps the only way, in a sense,
to eliminate this problem is to have the atmosphere itself
"perform the experiment" and verify the models. Aside
from the environmental and social risks of such a happen-
ing, this implies that a model is substantially complete or
at least contains the predominant physics, chemistry, and
other mechanisms. This example of CO2 uncertainties is
typical of a large class of climatic problems at present.
FOOD-CLIMATE CONSIDERATIONS
Consider at this point some of the food-climate issues.
Figure 1.9 shows a record of corn yield in Missouri
(Decker, 1974~. The solid line is an average trend over the
last 70 years; the dots are the individual yields, that is, the
number of bushels per acre for the state's area, and the
squares represent the yields in the drought years. One
notices that a tremendous increase in yield occurred after
about 1940, and this is particularly noticeable in the
1950's, 1960's, and early 1970's. There is no question that
this increase was caused by technology, that is, crop
strains had been developed that could have higher yields
and that were especially responsive to fertilizer applica-
tions or the use of pesticides. It is also seen on this figure
that the relative percentage of yield variability (especially
during the dust bowl period and the last drought in the
1950's) was very high relative to the variability in crop
yield that has occurred recently (195~1973~.
Technology has been given credit by some for both of
the above improvements. But McQuigg et al. ~ 1973)
showed that the period from 1956 through about 1973 was
31
120
100
O 80 _
9
-
c~
lo' 60 _
> 40
20 _-
· Years without major drought
· Years with major drought
_ ~
·~ ._ ~ _
· ;.
~ ~~e
·~. ~
.
I , I ·~1 , 1 , 1
1900 1920 1940 1960 1980
YEAR
FIGURE 1.9 Solid line is the trend in corn yield per acre in
Missouri (after Decker, 1974). The circles represent individual
yearly yields for years without major drought, and the squares are
for years with drought. Note that since 1956, not only have yields
increased significantly but variability in yields from one year to
the next has been reduced.
also an extremely unusual one weatherwise. As Figure
1.10 (Gilman, 1974) shows, that period differs from many
other periods in climatic history. For example, the sum-
mer rainfall and summer temperature in the five major
wheat states in the United States during the 1930's "dust
bowl era" were below normal and above normal, respec-
tively, and this combination is bad for crops. One should
note that there still is much noise in the system including
another drought in the 1950's (coincidentally 20-some years
away from the previous one). But there is an unobtrusive
looking 16-year period from 1957 to 1973, which may be
called the "high-yield era" and shows (with one excep-
tion) all years with normal or above-normal rainfall and
normal or below-normal temperature, which is in fact
abnormally good for crops. However, there is no physical
reason to expect that this abnormal trend will continue.
When one plans for food at least, one cannot look only at a
period of 15 years; one needs to look at periods longer
than that in consideration of the future (see the discussion
in Chapter 4 of Schneider and Mesirow, 1976).
OCR for page 32
32
ABOVE:
NORMAL]
BELOW]
ABOVE1
NORMAL:
BELOW]
FIVE "WHEAT BELT" STATES
(Oklahoma, Kansas, Nebraska, S. Dakota and N. Dakota)
DUST BOWL ERA HIGH YIELD ERA
1' ~ k .1
I , I , I r I ~ l T I 1~ r r T I
SUMMER RAINFALL
. B~ nit n ~~} new
, ~d ~ mu ~— ~ Lit L{'
SUMMER TEMPERATURE
1 , I , I , I , I , I , I , I
00 1910 1920 1930 1940 1950 1960 1970
YEAR
FIGURE 1.10 Seventy-five-year record of summer average tem-
perature and rainfall in the five major wheat-producing states of
the United States, compiled by Donald Gilman of the National
Weather Service and showing the 10-year drought period (i.e.,
high temperatures and low rainfall) of the dust bowl era and
the 15-year recent high-yield era (i.e., above average rainfall and
below-average temperature).
RELATION TO WATER-SUPPLY ISSUES
Two conclusions are now applicable to the water-supply
issue. The first suggests that one should accumulate data
over a fairly long period in order to obtain some actuarial
frequencies of the kinds of fluctuations that one might
expect. Perhaps this would be centuries for reliable statis-
tics, but it is at least 20 years. The second conclusion is
that even though a time series may look stationary if
viewed from "far enough back," a closer view might
reveal that over a much shorter and more recent climatic
sample period changes could be occurring because of
human effects. Furthermore, these changes could be
rather substantial in the next 20 to 50 years and may well
change not only climatic means but also the frequencies
of the shorter-term fluctuations. In essence, the real mes-
sage here is that if we must contend with climatic uncer-
tainty due to the natural fluctuations, that uncertainty will
probably be greater in the future because there is a good
possibility that the climatic system's boundary conditions
are also changing—perhaps from human activities.
There should now be added one other point. In the
only perfect forecast ever made by Joseph in the Book of
Genesis he warned of seven years of feast and seven
years of famine and, of course, proposed the solution of
prudence in the face of environmental variability
namely, a food reserve. Now, whether what Schneider
and NIesirow (1976) have called "The Genesis Strategy"
for food reserves also applies by analogy for water-supply
planning as part of the solution to an expectation of
climatic uncertainty or whether planning should rather be
to make society less vulnerable to the kinds of fluctua-
tions that have occurred in the past few centuries are
STEPHEN H. SCHNEIDER and RICHARD L. TEMKIN
issues that need to be debated further. The one thing that
is very clear is that climatic variability has been the rule
in the past, but now there are additional unknowns pro-
duced by people, so it certainly would be prudent to
expect considerable variability in future climate.
FINDINGS AND RECOMMENDATIONS
1. The climate varies on all time scales, and the mean-
ing of climatic change depends on the defining period for
the climatic average.
2. The climate of the past century or two is not neces-
sarily typical of the climate over the past few thousand
years, particularly on a regional scale.
3. Theory is yet unable to predict the future climate, so
an actuarial analysis of recent past records, while not
guaranteed valid, is probably the best quantitative way to
estimate the range of future climatic variability. However,
there is considerable numerical modeling evidence to the
effect that human activities, i.e., production of CO2, could
detectibly change the "equilibrium" climate by as early
as A.D. 2000 and that such a disruption could also change
the patterns of climatic variability as well as climatic
means.
4. World food supplies are very dependent on climatic
stability, and world food supplies and needs are currently
in a precarious balance that depends on large food trans-
fers and stable supplies of fertilizer, water, and seeds.
Population growth in the face of unstable or fluctuating
food supplies, i.e., the classical Malthusian problem, may
change agricultural demand for water in the future. At
least, the implications of this tight margin for reserves of
both food and water need to be examined.
5. The implications of these long-term growth (de-
mand) and needs (supply) projections should be
examined in the context of hedging, and the choice must
be clarified as a value judgment fundamentally contrast-
ing short-range economic benefits versus long-range
catastrophic risks.
REFERENCES
Alexander, T. (1974). Ominous changes in the world's weather,
Fortune 89, 90.
Broecker, W. S. (1975). Climatic change: Are we on the brink of a
pronounced global warming? Science 189, 460.
Bryson, R. A. (1974). A perspective on climatic change, Science
184, 753.
Decker, W. (1974). The climatic impact of variability in world
food production. Prepared for the 1973 Annual Meeting of the
American Association for the Advancement of Science, San
Francisco. Reprinted in Am. Biol. Teacher 36, 534.
Dickinson, R. E. (1975). Solar variability and the lower atmo-
sphere, Bull. Am. Meteorol. Soc. 56, 1240.
Eddy, J. A. (1976). The Maunder minimum, Science 192, 1189.
Ellis, H. T., and R. F. Pueschel (1971~. Solar radiation: Absence
of air pollution trends at Mauna Loa, Science 172, 845.
OCR for page 33
Water Supply and the Future Climate
Emiliani, C. (1972). Quaternary paleotemperatures and the dura-
tion of the high temperature intervals, Science 178, 398.
GARP (1975~. Publication Series No. 16, The Physical Basis of
Climate and Climate Modelling, WMO, Geneva, Switzerland.
Gates, W. L. (1976). Modeling the ice-age climate, Science 191,
1138.
Gilman, D. (1974~. Paper presented at the 140th meeting of the
American Association for- the Advancement of Science, San
Francisco, Feb. 27.
Kellogg, W. W. (1977~. Global precipitation patterns during the
altithermal period compared with the present. In preparation.
Machta, L. (1971~. The role of the oceans and the biosphere in
the carbon dioxide cycle, Nobel Symp. 20, Gothenburg, Swe-
den, Aug. 1~20.
Machta, L., and K. Telegadas (1974). Inadvertent large-scale
weather modification, in Weather and Climate Modification,
W. N. Hess, ea., John Wiley and Sons, Inc., New York,
pp. 687-725.
Manabe, S., and R. T. Wetherald (1975). The effects of doubling
the CO2 concentration on the climate of a general circulation
model,J. Atmos. Sci. 3Q, 3.
Matthews, W. H., W. W. Kellogg, and G. D. Robinson, eds.
(1971~. Inadvertent Climate Modification, Study of Man's
Impact on Climate (SMIC), MIT Press, Cambridge, Mass.
33
McQuigg, J. D., L. Thompson, S. LeDuc, M. Lockard, and G.
McKay (1973). The influence of weather and climate on
United States grain yields: bumper crops or droughts, The
Associate Administrator for Environmental Monitoring and
Prediction, National Oceanic and Atmospheric Administra-
tion, U.S. Dept. of Commerce, Dec. 14.
Schneider, S. H., and C. Mass (1975~. Volcanic dust, sunspots,
and temperature trends, Science 190, 741.
Schneider, S. H., and L. E. Mesirow (1976~. The Genesis Strat-
egy, Plenum Press, New York.
Stetson, H. T. (1937~. Sunspots and Their Effects, McGraw-Hill
Book Co., New York.
Thompson, L. (1973). Cyclical weather patterns in the middle
latitudes,J. Soil Water Conservation 28, 87.
U.S. Committee for the Global Atmospheric Research Program
(1975~. Panel on Climatic Variation. Understanding Climatic
Change: A Program forAction, National Academy of Sciences,
Washington, D.C.
Williams, J., R. G. Barry, and W. M. Washington (1974~. Simula-
tion of the atmospheric circulation using the NCAR global
circulation model with ice-age boundary conditions,J. Appl.
Meteorol. 13, 305.
Representative terms from entire chapter:
future climate
