Climate has always influenced human affairs. There are now increasing signs that man may in turn be altering global climate. This could change economic, political and even military relations among nations.

Climate drives agricultural and forest production, and it largely governs the way people live and work. Unexplained natural climatic variation has characterized all of human history and prehistory. Many climatologists now believe that in the present advanced stage of industrialization the addition of carbon dioxide and particulate matter to the atmosphere through burning of fossil fuels and clearing of land has become a significant agent of climatic change that could measurably raise the temperature of the earth by the end of this century. This would bring about appreciable shifts in the global pattern of activities dependent upon climate: agriculture, forestry, residential heating and cooling, water-dependent industry, recreation, and many more. It could affect the level of the seas.1

Movement of climatic zones is likely to be at least as important as the global temperature change itself. The ultimate result of higher temperature, and of the more active atmospheric water circulation that will probably accompany it, could well be a net increase in global biological productivity. But the impacts will not be felt equally. Some regions and nations will gain; others will lose.

If the geophysical assumptions are correct, the process of climatic change due to industrialization is probably almost irreversible; the only practical strategy is adjustment. Technology will help to ease that adjustment, but the institutional rigidities of our advanced societies may correspondingly make the response to changed climate stickier than in earlier, simpler ages.

No one can yet say surely that the cumulative burning of fossil fuels at high rates will alter climate significantly. The answer lies at the end of a long cascade of uncertainties. There is uncertainty about the rates at which carbon dioxide, particulate matter, and other materials will be added to the atmosphere. How much will remain, and how much will be removed by physical and biological processes? There is still some disagreement whether additional atmospheric carbon dioxide will actually increase global temperature. If so, it is uncertain exactly how much temperature change will result from a given level of carbon dioxide. There is further uncertainty about how temperature change will be translated into shifts in global climatic zones. Ecologists are unable to say accurately how the postulated climatic shifts will affect cultivated and noncultivated plant and animal communities. And the social and political implications are still more problematical.

Yet none of the uncertainties is absolute. Estimates, within some range of error, can be made for each. The sum of these estimates suggests a fairly high probability that measurable man-induced climatic change is in the offing. Thus, it seems worth taking a look at the possible long-term implications for human institutions.


Before one can do so, it is necessary to have some understanding of the steps in the causal chain leading from burning of coal, oil, and natural gas to alteration of global climate.

Carbon dioxide, comprising little more than .03 of one percent of the earth's atmosphere, plays a disproportionate role in global processes. All life depends upon the carbon dioxide that plants use for photosynthesis. Atmospheric carbon dioxide is also an important regulator of global temperature, through its impact on the earth's radiation balance.

The earth as a whole receives a vast amount of heat from the sun each day. If it is not to grow continuously hotter, it must on the average radiate away to outer space an amount of heat equal to that received. Energy comes from the sun predominantly in the form of shortwave visible and near-infrared radiation. Because the earth is much cooler than the sun, and because the wavelength of emitted radiation is dependent on the temperature of the radiating body, the energy given off by the earth is at a much longer wavelength than that which it receives.

Carbon dioxide is almost completely transparent to the wavelengths of light given off by the sun. However, it strongly absorbs long wave radiation at certain wavelengths where other atmospheric gases are transparent. The earth's emitted radiation happens to be centered at just the wavelengths that carbon dioxide best absorbs. So the gas takes up some of the earth's radiant energy and is warmed. It then radiates its absorbed heat partly on out to space and partly back to earth again. The net result is that the earth is about 10° centigrade (C) warmer than it would be without the carbon dioxide envelope. Because a greenhouse is similarly transparent to incoming solar radiation but impedes outgoing heat radiation, the process to which carbon dioxide contributes in the atmosphere is commonly called the "greenhouse effect."

Water vapor also absorbs at some of the wavelengths at which the earth radiates, and is indeed apparently more important than carbon dioxide in producing the greenhouse effect. This fact is at the root of the controversy over the extent to which additional carbon dioxide will affect global climate.

The amount of carbon dioxide in the atmosphere has unquestionably increased since the late nineteenth century. Accurate and regular measurements have been made since 1957 at the Mauna Loa Observatory on the island of Hawaii. At an elevation of 3,400 meters on a remote oceanic island, Mauna Loa is far from most industrial sources of carbon dioxide and so can be fairly said to represent a global average. The measurements are corrected for any local sources that may interfere from time to time. More sporadic measurements at other remote localities around the world confirm the reliability of the Mauna Loa data.2

This 20-year series of observations shows a regular cycle that reflects the annual march of photosynthesis and plant decay in the northern hemisphere. Since plants absorb carbon dioxide, its atmospheric level is greatest in mid-April when plant growth is getting started, and lowest in fall after the year's uptake is completed. Superimposed on the annual fluctuations, however, is an unmistakable long-term upward trend, now running at about 0.7 percent per year. The current value of atmospheric carbon dioxide is about 330 parts per million (ppm) compared with an estimated level of 290-300 ppm in the preindustrial era before about 1890.

There are only two plausible sources for the added carbon dioxide. One is burning of fossil fuels. The other is a net reduction in the world's stock of organic carbon stored in plants, in soil humus, and in plant and animal material suspended and dissolved in the oceans. There is some argument over the relative size of the two sources, but no disagreement that the first is highly important. In fact, calculations based on the estimated annual release of stored carbon through burning of fossil fuels suggest that this alone is about twice the measured annual increase in atmospheric carbon dioxide. An obvious question is what happens to the rest, a point to which we will return.

Projections based on anticipated annual burning of fossil fuels, and on the fraction of the resulting carbon dioxide expected to remain in the atmosphere, lead to forecasts that atmospheric concentrations of carbon dioxide will range between 375 and 400 ppm by the end of this century. These models attempt to estimate cumulative levels of unabsorbed carbon dioxide under several plausible scenarios of future use of fossil fuels (coal, of course, as well as oil and gas). All assume a steady growth, at alternative rates, in overall consumption. The best analyses indicate that a doubling of the preindustrial level can be expected sometime between 2020 and 2040, depending upon which of the alternative use rates finally prevails.3

These models assume that the earth's plant cover - the biosphere - is constantly taking up some of the added carbon dioxide through increased growth, or at least is not adding to the total in the atmosphere. However, recent analyses present several lines of evidence that the biosphere is actually a source of carbon dioxide about as large as the fossil fuel source. The reason is that net destruction of plant life and soil organic matter releases carbon dioxide. The new analyses suggest that the rate of forest clearing, particularly in the tropics, is much greater than had previously been suspected.4 This conclusion is still controversial. If it is correct, a doubling of atmospheric carbon dioxide would be much closer in time. This is an example of the kind of uncertainties that can be reduced only through a concerted international effort.

Whatever the time scale, how will the projected increase in carbon dioxide translate into changes in global temperature? Because of the complexity of competing processes in the atmosphere, it is impossible to use direct observations, relating historic temperature changes to the known historic increases in atmospheric carbon dioxide. Resort must be to complex calculations based on radiation properties of atmospheric gases and the observed structure of the atmosphere. Even the most realistic of these models involve many unknown factors, omit important feedback mechanisms, and require much computer time for their solution.

The most generally accepted models all predict that increased carbon dioxide will produce an increase of temperature near the ground. The temperature response is nearly logarithmic; that is, each proportionate increase in the concentration of carbon dioxide produces about the same absolute increase in global temperature. Stephen Schneider of the National Center for Atmospheric Research has critically reviewed the models in existence up to 1975.5 He concluded that those which best reflect current geophysical knowledge converge on an increase of between 1.5°C and 3°C for each doubling of carbon dioxide - the increase predicted in 2020 to 2040, roughly 40 to 60 years away - with lesser increases appearing in the meantime. Small as such changes may appear, they are enough to make a substantial difference in many economically important processes at the earth's surface, as will be seen later.

The general neglect of feedback mechanisms for clouds and ocean circulation in these models opens the possibility that the estimates may be substantially in error in either direction. However, Schneider concluded that there is no strong evidence that present models are more likely to overestimate the rise than to underestimate it.

There is one major problem with these model estimates. They do not precisely accord with observations over the last century. Careful analysis of worldwide temperature records indicates that global mean temperature apparently rose by a total of about 0.6°C between 1850 and 1940. Since then, instead of continuing to increase as would have been predicted by the carbon dioxide theory, temperatures have apparently dropped about 0.3°C to the present. Some climatologists believe that since about 1971 global mean temperature has been increasing again, but this is too short a time to firmly establish a trend.

A number of respected atmospheric scientists cite the lack of warming during the industrial era since 1940 as strong evidence that the postulated effect is not important, and that global cooling still predominates. However, the occurrence of an apparent trend contrary to that predicted by theory, even over a period of 30 years, is in itself neither surprising nor a refutation of the theory.

Temperature fluctuations, up or down, of 1°C or more per century throughout the past have been inferred from a variety of historical and paleoecological records. It is easy to argue that the period 1940-70 was one of normal cooling, on which was superimposed some warming due to the carbon dioxide effect. The cooling was therefore less than it would have been in the absence of industrial carbon dioxide. The verdict must remain "not proven." But if the warming effect of carbon dioxide over the next 40 to 60 years should be toward the higher (3°C) range in the models reviewed by Schneider, its magnitude would far outweigh the apparent 1°C range of normal fluctuation, even assuming that the latter continues to be in the direction of cooling.


What is clear is that any warming that does occur will not be uniformly distributed. That is one point on which all the experts seemingly agree. Both the climatic simulation models and past observations confirm this estimate.

Temperature increases will be greater at high latitudes than near the equator. This is borne out by experience during warming and cooling trends in the past century. A 1°C increase in global mean temperature might result in a change of 0-0.5° between 10°N and 10°S latitude, 2-4°C at 60°N and 60°S latitude, and even more than that closer to the poles.6 The impact of a 2°C mean increase would be roughly twice as great, and so on.

This is important in two ways. In the first place, the tropics, already near the limits of tolerance for many organisms including man, will not become unbearably hotter even in the face of a considerably higher mean temperature. Neither is there likely to be much change in precipitation in the wet tropics. The greatest impact in the equatorial zone is likely to be a second-order one, resulting from changes in demand and prices for agricultural exports and imports, due to climatically induced economic changes elsewhere.

At high latitudes, on the other hand, both agricultural productivity and general livability are likely to be improved, particularly where the length of the growing season is now marginal or sub-marginal for crop production. Of the great subarctic land masses, those in the Soviet Union would appear to stand a better chance of reaping significant benefits than those in Canada or Alaska or the Scandinavian countries. The reasons for this conclusion will be considered in more detail below.

This is a good point at which to consider the impact of higher poleward temperatures on the polar icecaps and on sea level. A complete melting of glacial ice, almost all of which is in Antarctica, would raise the level of the oceans by more than 50 meters, but this is most unlikely on a time scale of less than many centuries, if then. Even with a polar warming of several degrees, air temperatures over Antarctica would still be below freezing most of the time. Furthermore, dry air alone is a relatively inefficient melting agent compared with solar radiation, which would be little altered.

More complex scenarios can be visualized. Some have suggested that a moderate warming of the Antarctic ice, without melting, would accelerate glacier flow and discharge of ice into the ocean. Or warm polar water could induce a flow of ice from the continental shelf into the sea, possibly raising sea level by five meters in 300 years.7 A rise of much smaller magnitude could be catastrophic for a nation such as the Netherlands, but this does not seem a major near-term threat.

Melting of the Arctic ice is perhaps slightly more likely. The North Polar ice floats on the Arctic Ocean, fluctuating in thickness from about 15 meters in winter to sometimes no more than a meter or two in summer. Because it is already floating, its disappearance would not alter sea level, but there could be drastic and unpredictable climatic feedbacks. An open ocean would absorb more solar radiation than ice. The Arctic Ocean has apparently not been open for the last million years or so, yet it appears that the Arctic ice could in fact be melted under some circumstances.8

As important as shifts in temperature bands may be changes in the pattern of precipitation, particularly the subtropical monsoon. From the equator to the poles, the earth is roughly characterized by an equatorial zone of heavy rainfall, a rather abrupt transition to a belt of deserts and semideserts, a relatively moist zone of prevailing westerly winds in the temperate latitudes where most of the world's advanced nations are concentrated, and increasing dryness toward the poles. This general pattern is greatly modified by continental land masses.

The annual monsoon is particularly important to some of the world's less-developed nations. During summer, the zone of moist tropical air moves north, bringing heavy rainfall to the Indian subcontinent and lesser but still essential amounts to much of Africa south of the Sahara. When the monsoon rains are abundant, crops and water supplies are plentiful. When they fail, hardship follows, as evidenced by the widely publicized drought in the Sahel region of sub-Saharan Africa in the early 1970s.

Reid Bryson of the University of Wisconsin has shown that the position of the boundary between moist monsoon air and dry desert air is extraordinarily sensitive to small changes in the north-south temperature gradient.9 What apparently happens is that when polar temperatures are relatively warm, there is a strong pattern of global circulation, and the north-south oscillation of the westerly winds that sweep around the temperate regions is confined within a relatively narrow range of latitudes. Farther south, however, the tropical wind patterns are pulled well to the north, and allow the monsoon rains to reach high latitudes. Historically, this kind of strong circulation and northward penetration of the monsoon seems to be associated with unusual warmth throughout the northern hemisphere. This is at least suggestive that the same may happen as a result of general global warming. The monsoon areas of the world could thus be substantial gainers from the climatic effects of worldwide industrialization.

In the mid-latitudes, on the other hand, the effect is more mixed. There the north-south oscillation of the westerly winds largely governs the distribution of precipitation. When global circulation is relatively weak, it is easier for large masses of cold air to spill southward, where they meet warm subtropical air and increase precipitation. Warm air is then forced northward in spinning storm systems that move eastward through the temperate latitudes. With the stronger circulation patterns associated with polar warming, these oscillations tend not to push so far south. Rainfall over the Mediterranean region and the Middle East is suppressed, and there tends to be higher than normal precipitation in the British Isles and at similar latitudes.

Because evaporation is related to temperature, global warming would tend to create a greater flux of water vapor from the land and oceans into the atmosphere. Computer models suggest that the degree of warming estimated from a doubling of carbon dioxide could increase total global precipitation by an average of about seven percent. Again, such an increase would not be uniformly distributed. Much of the added moisture might fall back into the ocean.

Finally there is the factor of variability. What we call climate is just the average, over months and years, of individual weather events. Even if world climate should become warmer and wetter, there will still be cold spells and dry spells, as sharp as in the past. These will be overshadowed, though, by more frequent and perhaps more marked warm episodes.

Neither will all parts of the world warm simultaneously. Historical experience clearly indicates that warm periods in one part of the world will be accompanied by cold in another, just as the bitter winter of 1977 in the eastern United States was matched by the warmest weather in a century in Vienna.

Year-to-year weather variability is important to human comfort, to industry, and to agricultural productivity. Agriculture and industry can more easily adjust to changed conditions if the new weather regime is relatively uniform from year to year. Increased variability makes agriculture in particular a more uncertain venture. Opinions - evidence is too strong a word - differ as to whether global warming will mean more or less climatic variability. The preponderance of opinion seems to be that the increased intensity of global circulation will bring with it the possibility of more frequent and larger deviations from the mean. Fluctuations in agricultural production, and consequent wide swings in grain prices and markets, could follow. The effect on global and national economies and on inflation rates of similar swings in the early 1970s are a vivid memory.10


The emphasis so far has been on global warming through increased atmospheric carbon dioxide. At least two other potential human impacts on global climate could be influential: discharge of particulate matter and release of heat into the environment from burning of fossil and nuclear fuels (called heat emission).

The transparency of the global atmosphere, as measured at many places, is steadily decreasing due to continued addition of fine particles. These arise chiefly from three man-made sources and one natural: burning of fossil fuels, wind-borne agricultural dust, and smoke from slash-and-burn agriculture in the tropics, plus intermittent injections into the stratosphere of debris from major volcanic eruptions. Particles reflect and scatter solar radiation back to space, reducing the amount available for heating of the surface. They also absorb solar radiation, and thereby heat the atmosphere. Recent evidence suggests that the second effect predominates in the lower atmosphere, at least up to the point where particle loading would be a serious health hazard. Although the impact of particulate matter on global climate is not negligible, the Energy and Climate Panel of the National Academy of Sciences concluded that man-induced particulate loading is unlikely to increase to the point that it is a serious threat in this respect.11 (The blunt fact is that, before then, impacts on human health and on agricultural productivity would require stringent control measures.)

The panel reached similar conclusions with regard to the impact of heat emission on climate. Even a future world population of ten billion people, with a per capita energy use several times greater than at present, would release an amount of heat equivalent to only a thousandth of the daily radiation received from the sun. If evenly dispersed, this would have little measurable effect on climate.

Concentrated heat emissions could, however, trigger changes in climate if, as some atmospheric scientists believe, global circulation is sensitive to small changes in inputs at crucial times and places. That this may be so is suggested by computer simulation experiments carried out by the International Institute for Applied Systems Analysis (in Vienna) and the United Kingdom Meteorological Office. Economically significant changes in simulated temperature and rainfall were indicated almost everywhere in the northern hemisphere as a result of concentrated heat discharge from assumed large "energy parks" (specifically, centers for the production and distribution of electricity). The present models do not simulate climate in a fully realistic way, so that these results must be interpreted cautiously. Other climate simulation models, when more fully developed, should permit better prediction of the effects of carbon dioxide increase and also help to give better answers to the questions raised by heat emission.

A further reason for giving most attention to carbon dioxide is that its climatic effect, if real, will be less reversible than that of the other two factors. Particles in the troposphere - the layer of the atmosphere where active weather processes occur - have a mean residence time of perhaps 10 days. They stay a bit longer in the stratosphere, three years or so. In either case, stringent measures to reduce particle emission would be reflected in the atmosphere in weeks or at most years. The same is true of concentrated heat sources, which could be turned off instantly if one was willing to disregard other social consequences.

Carbon dioxide is different. Even if burning of fossil fuels were suddenly to be banned all over the world - a far-fetched possibility - it would take decades, maybe centuries, for the concentration of carbon dioxide in the atmosphere to decrease significantly. To see why this is so, some atmospheric processes must be understood.

The preindustrial atmospheric concentration of carbon dioxide was the result of a complex interplay over geologic time of carbonate rocks, growing and decaying plants, and solution in the ocean. On a time scale of thousands of years in the past or in the foreseeable future, the oceans are clearly dominant. Atmospheric carbon dioxide is in long-term equilibrium with that dissolved in the oceans. The carbon dioxide in the atmosphere is a minute fraction of the oceanic amount. The excess in the atmosphere will inevitably dissolve in the ocean. The oceanic reservoir can easily absorb all that will ever be added by burning of the earth's entire stock of fossil fuel. The question is one of time.

The ocean is separated into two layers that exchange water only slowly: a warm active surface layer perhaps 200 meters thick, and the colder depths that hold an almost infinitely greater amount of water. The active layer, in regular contact with the air, is almost saturated with carbon dioxide. It, and perhaps other sinks, are now apparently absorbing about half of the annual addition of carbon dioxide. Uptake of the rest is possible only to the extent that it is transferred to the depths. Best estimates by oceanographers are that the exchange time is on the order of a thousand years or so, although there is great uncertainty about this.

It appears, therefore, that a substantial fraction of all carbon dioxide added by burning of fossil fuel will remain in the atmosphere for many human lifetimes. Eventually, in the fullness of geologic time, the level will return to what it was before the industrial age. The present rate of consumption of fossil fuel, and the attendant increase of atmospheric carbon dioxide, has been referred to as the Hubbert Pimple, from the American geologist and student of mineral resources, M. King Hubbert. Hubbert points out that on a geologic time scale, the period from the start to the finish of man's consumption of the entire stock of stored fossil fuel resources is a mere blip. But it is a blip we and our children are living in.

It seems unlikely that even the most advanced future technology will make feasible the direct removal of carbon dioxide from the atmosphere. Only about a one-percent annual increase in the world's stock of plant material would be enough to take care of the carbon dioxide produced by the current rate of fossil fuel consumption. This level of annual increase would not be impossible to arrange, but would do no good unless the new material could be protected from decaying or being eaten, and thus discharging its carbon back to the atmosphere. Carbon could be stored for long periods in wood, but the potential increase in the world's forest volume is surely limited - by how much is not known. Or the annual increment could be placed out of contact with the atmosphere. Even if a place could be found to put it, the materials handling problem alone would be staggering. Current world consumption of fossil fuel is about five billion metric tons per year. Because organic matter contains more oxygen and water, the plant equivalent would be even heavier, and far more bulky. Similar materials-handling problems, to say nothing of the energy required, seem to rule out direct scrubbing of the air.

An imaginative engineer at the International Institute for Applied Systems Analysis near Vienna has proposed that carbon dioxide be collected at the source, in the stacks of power plants, and disposed of by injection into sinking oceanic currents. He has identified the Mediterranean outflow entering the Atlantic at Gibraltar as one such current which would have sufficient capacity to deal with all carbon dioxide produced in Europe even in the year 2100. A rough evaluation indicates a cost to the consumer of 20 percent of the fuel value for disposal of 90 percent of the carbon dioxide. Whether such a scheme could ever be engineered is not known. Nor has anyone shown how to attain the degree of international cooperation and discipline that would be required for it to work - a problem that plagues almost all aspects of man-induced climatic change.


With these preliminaries out of the way, what can be said about the impact of global climatic change on nations and regions? Any conclusions must necessarily be highly tentative and speculative; some may be wholly wrong. The principal aim is to point out directions for analysis and evaluation, assuming a 1°-2°C global warming. The emphasis in this discussion is on agriculture, forestry, and livestock production. Let us start with the Soviet Union.

The Soviet grain region lies mostly between 50° and 60° N latitude. This close to the pole, a 2°C global temperature rise would probably translate into an increase in mean temperature of 3-4°C. From a crop production standpoint, the principal impact would be on length of the growing season, since this is the factor that chiefly limits agriculture at high latitudes. Even if midsummer temperatures are high enough for optimum growth, as they often are in the interior of continents, successful harvests are impossible if there is too little time between spring and fall frosts to complete the process of germination, growth, and maturation.

A rule of thumb often used by agronomists is that a temperature change of 1°C causes an approximate 10-day change in length of the frost-free period. This relation, however, is dependent upon the initial length of the frost-free period. A 1°C increase in mean temperature in northern Canada apparently lengthens an 80-day growing season by about 20 days, and a 120 to 130-day season by about six days. Thus the impact in presently colder areas is accentuated.

It is apparent from these considerations that global warming is likely to make available for crop production substantial areas of the northern U.S.S.R. where successful harvests are now marginal or wholly impossible. The question is whether soils are suitable. The soils of the Soviet northern forest and forest-steppe zones are mostly heavily leached and impoverished, compared with the fertile black-earth soils of the central Soviet grain belt. Nevertheless, there are abundant opportunities for extension of agriculture toward the north, particularly in the alluvial soils of the Yakutian plain along the Lena River.

At the southern boundary of crop production almost everywhere, the limiting factor tends to be precipitation, with some of the growing season going unused. Most of the precipitation in the Soviet grain region falls in summer, and almost all originates in the North Atlantic. If global warming accentuates east-west circulation and suppresses north-south oscillations, as discussed above, it appears that rain might penetrate more readily to the eastern-most parts of the grain region. However, suppression of southward storm excursions could reduce rainfall in the southern portions of the grain zone, especially in the already trouble-prone region of Kazakhstan and in parts of the Caucasus. This, coupled with higher temperatures and greater evaporation, could appreciably reduce yields in parts of the country.

These conclusions are partially confirmed by an unclassified study by the U.S. Central Intelligence Agency on the impact of recent climatic change on grain production in the U.S.S.R.12 On the basis of a careful statistical analysis of grain yields and weather over the last 15 years, the CIA concluded that there was a steady improvement in the climate of the Soviet grain belt between 1962 and 1975. During that time, average temperatures in the grain region increased by about 0.75°C, although worldwide temperature was continuing the downward trend that apparently culminated about 1971. The southward shift of global circulation during the period of global cooling moved the moister northern climate southward about 2° of latitude, increasing agricultural production in the new lands of Kazakhstan and central Asia. If this pattern of global cooling is now reversing itself, then the favorable effect experienced in the new lands would likewise be reversed - and production from these always-precarious areas could drop sharply.

There is a distinct difference between the impact of long-continued and generally predictable climatic change and the effects of chance vagaries in year-to-year weather. In the first case there will eventually almost certainly be a shift in production patterns; in the second case the best strategy is simply to wait out the change. Just how continued global warming would be reflected in Soviet agriculture is far from clear. A reasonable speculation would be acknowledged or unacknowledged failure of the southern new-lands schemes, coupled with highly publicized similar endeavors in the north.

The major grain regions of Canada lie in the same general latitudinal zone as those of the U.S.S.R. The scope for northward extension of grain production is more limited in Canada, however. Only in the Peace River Valley of northern Alberta and in a few other scattered localities are there extensive areas of good soils where temperatures are too low and the growing season too short for successful grain production. Most of the rest of Canada's northern margin is underlain by the vast Laurentian Shield, with thin glacial soils unproductive in any climate. More favorable climate could, however, reduce the variability of Canadian grain yields.

Similarly, local increases in food production in Scandinavia may aid those countries to feed themselves, but the area of suitable soil is too small to be important on a world scale.

The impact on the United States, the world's largest grain exporter, could be significant. The southern limit of the North American grain region is at a latitude of about 35°N, the same as the southern Mediterranean. Here the controlling factor is availability of soil moisture, not length of the growing season. The American grain belt has attained its preeminent position chiefly because of a fortunate combination of excellent soils and suitable weather. A northward movement of temperature and precipitation belts would leave Kansas, Oklahoma, and other states at similar latitudes dangerously exposed to threats of diminished yields because of drought. A permanent change would mean that the highly productive soils of this region would be underutilized, a loss hard for the United States to make up by shifts of production to other localities.

In terms of current major American food crops, recent research indicates that it is very hard to generalize about the effect of climatic change either on any crop or on any natural ecosystem. Upward changes in any one U.S. area are often offset by production decreases elsewhere.13

Both physiological studies of processes of corn growth, and observed correlations of corn crops and weather, suggest that the U.S. corn production will decrease about 11 percent for each 1°C increase in average temperature during the growing season. Warmer and drier weather, such as might occur in the American Corn Belt from a global temperature increase, is in general predicted to depress corn yields. Soybeans would also be adversely affected by temperature increase almost everywhere in the United States - which is, of course, significant because of the U.S. position as the world's chief supplier of soybeans.

Cotton, another major export crop, would not be materially affected by either a 2°C global increase or decrease in temperature, either in the United States or elsewhere in the world. Cotton tolerates a rather wide range in temperatures above a certain minimum, and its production is largely governed by other factors than weather variations.

Rice, one of the major worldwide grain crops, is of particular interest because of its importance in many developing countries. World rice production has been analyzed by J. Stansel of Texas A & M and Robert Huke of Dartmouth College. Increases in mean temperature would generally increase the areas where rice could be grown, thereby increasing world production. More significantly, however, higher temperatures might allow multiple cropping in the vast rice areas of the intermediate latitudes where only single crops can now be grown. This might be offset somewhat by decreases in rice yields in the tropics because of excessive temperatures, but as noted above, temperatures near the Equator will be affected relatively little. The impression that emerges from this analysis is that China might be the principal beneficiary of increased rice yields due to higher temperatures. There are too few data to permit assessment of possible compensating effects on other crops in China. India and Bangladesh would also benefit somewhat, especially if the annual monsoon became more reliable.

The productivity of noncultivated forests and rangelands is directly dependent on temperature and precipitation, but the quantitative relationships are poorly known on a global scale. As part of the Department of Transportation study, I endeavored, together with colleagues at six American institutions, to combine several computer simulation models in an investigation of possible impacts of climatic change on natural ecosystems. This experiment was designed to see whether common patterns could be discerned in the simulated ecosystems, rather than to make quantitative predictions of ecological change.

The principal conclusion was that primary plant production is expected to increase linearly with temperature increase everywhere except in the driest climates. The increase will be greatest in ecosystems well supplied with water throughout the growing season. Associated with productivity changes there are likely to be nonlinear effects on insects and other herbivores that are more severe and less predictable than the effects on the host plants. Finally, it appears that interactions among climatic variables may result in responses that are quite different from the responses expected from the sum of several independent climatic changes.

A real deficiency in these and similar models is that they address only short-term effects, on the order of four to ten years. They do not accommodate changes in species composition or other long-term responses to climatic change. Nevertheless, there is at least a suggestion that wood production might be increased somewhat in the hardwood forests of the eastern United States and perhaps in the Pacific Northwest. The same is likely to be true of the great boreal forests of Canada, Scandinavia, and the U.S.S.R. Positive effects are less likely in the pine regions of the southeastern United States and in California. Climatic impacts on the tropical rain forest are likely to be negligible.

Carbon dioxide itself may have some effect on agricultural and forest productivity apart from its impact on climate, but this is likely to be small. Carbon dioxide is occasionally a limiting factor in well watered and fertilized crops, and the addition of carbon dioxide to intensively managed greenhouses is a common practice among flower growers. Under field conditions, however, carbon dioxide is rarely seriously deficient. Most estimates are that any increase in photosynthesis from this source will not be nearly proportional to the amount of carbon dioxide to the atmosphere.

Water resources are highly sensitive to climatic change. Some river systems are likely to have larger and more regular flows than in the past; others will be depleted, even if, as suggested, total global precipitation increases by some small amount. Particularly hard hit could be the Colorado River system of the western United States. Any diminution of precipitation due to a northward shift of climatic zones would seriously reduce the flow of the Colorado, already developed beyond its capacity to supply the demand for high quality water in California, Mexico and the southern Rocky Mountain states. The nasty conflict between the United States and Mexico over American failure to deliver water of a quality required by treaty could become even worse.

Oceanic fisheries are sensitive to climatically induced shifts in temperatures and currents. Fisheries are not considered here, however, mostly because I am convinced that failure of national and international fishing controls will result in the collapse of most fisheries long before climatic influences are evident.

Any number of industrial processes will likewise be affected by climatic change. In an article intended chiefly to lay out the nature of the problem, these possibilities cannot be explored in detail. This is an area for future investigation, particularly the international ramifications.


Carbon dioxide enters the atmosphere from so many sources that any effective social control of its emission on a global scale is most unlikely. As a practical matter, it is stoppable only at the source, by worldwide prohibition of coal mining, peat cutting, and extraction of oil and natural gas. Short-term economic and social consequences are almost sure to rule out the required unanimous international consent. Fossil fuels are so convenient for so many purposes, and so easily extracted, that they are almost certain to be used to the limit of their availability even if there should be a global commitment to emphasize nuclear, solar, and geothermal energy resources in their place. As we have already noted, because of the long residence time of excess carbon dioxide in the atmosphere, even an appreciable decrease in the rate of fossil fuel burning will only delay the time of maximum climatic effect. It will not prevent it.

If the world community is unwilling or unable to take the stringent measures necessary to stop carbon dioxide emission, society must simply adjust to changing climate. Technology may aid in that adjustment, but it also may make the transition more difficult. Witness how a moderate snowfall, little more than a minor inconvenience a couple of generations ago, ties up an automobile-bound modern city.

The official position of the U.S. Department of Agriculture has been that new and emerging agricultural technology will more than keep up with any shifts in climate. Among the suggested climate-insulating technologies are extension of irrigation, increased use of fertilizer, mechanization, improved cultural practices, and development of new seed varieties adapted to altered weather conditions. It should be noted, though, that irrigation requires water, and expansion of water supply in a time of drought is something of a contradiction in terms. Irrigation, fertilization, and mechanization are all energy-intensive technologies whose future on a world scale is far from certain.

Intensified crop breeding carries with it the danger that narrowly adapted genetic strains will be highly susceptible to unanticipated environmental stresses. This is borne out by the nearly disastrous experience a few years ago in the southern United States, with corn varieties that all carried a single wholly incidental gene that made the plants almost unresistant to a newly virulent strain of corn blight. Geneticists are well aware of such possibilities - partly because of lessons learned in the corn blight episode - but plant breeding, in the light of past experience, can hardly be considered a trouble-free panacea. In particular, there is danger that new varieties designed to accommodate temperature and rainfall outside the normal experience of the crop may lose enough genetic diversity to make them particularly vulnerable to temporary weather fluctuations.

Grain yields in the United States have increased dramatically since about 1945, mostly because of increased fertilization and better tillage, and the development of new crop varieties able to benefit from the improved practices. The year-to-year percentage deviation from the accelerating trend has also declined, at least since the mid-1950s. Thus, not only have yields increased but their relative variability has decreased. Agronomists in the U.S. Department of Agriculture and elsewhere have contended that improved technology accounts for most of the lessened variability, and conclude that agriculture can now readily adapt to climatic change. However, Dean Louis Thompson of the Iowa State University College of Agriculture and others point out that the time of low variability in yield coincided with 15 nearly consecutive years of unusually favorable crop weather, a pattern unlikely to be perpetuated.14 There remains considerable controversy about the degree of protection that improved agricultural technology provides against the potential impact of climatic change.

Technology and social patterns will influence the way that human institutions react to climatic change. The extinction of the Norse colony in Greenland during the medieval cooling period is unlikely to be duplicated in an era of rapid communication and efficient transportation. One can surmise, though, that the death of the Norsemen was due in part to their unwillingness or inability to adapt to the subsistence techniques of the native population, which did survive. How much more adaptable are modern societies?

The widespread suffering during the Sahel drought of the early 1970s is at least indicative of the ways in which new technology and social change can make whole populations more, not less, vulnerable to climatic shifts. The nomadic livestock culture of that part of Africa was well adapted to periodic drought. Livestock were pastured near the few permanent watering places during dry periods, and dispersed throughout the region during wet weather. This natural rotation allowed recovery of the forage around the permanent water holes when the nomads left them, and conversely permitted regrowth of the more distant vegetation during the time that animals were concentrated around the water.

In an effort to provide for more livestock, a need brought about in part by increased population due to newly introduced public health measures, deep wells were drilled throughout the region. This permitted wider dispersion of livestock at all seasons, and increased the number that could be supported, at least for a time. The resultant overgrazing led to depletion of the forage resource, which could not recover as it had in the past because the natural rotation cycle had been broken. So when the monsoon failed for several years in a row, first livestock and then people starved.15 It is not at all certain that the actual percentage of the human population affected was greater than in past droughts, but there is no question that a larger absolute number of people faced malnutrition and starvation than in previous episodes.

This somewhat oversimplified account of an extreme case in a region where development was just beginning may not be representative of the situation in the industrialized world, but it is illustrative. Will social institutions in more developed regions endeavor to resist and counteract the effects of climatic change, or will they flow with the tide, making incremental adjustments as they go? The former strategy might work for a while, but eventual disruption and breakdown would seem inevitable. The difference may be important for the peace of the world.


Not only is there great uncertainty about the nature and magnitude of the consequences of global warming, but its principal effects will be felt only in decades to come. Yet actions taken, or not taken, today are likely to affect significantly our children and grandchildren. What is our obligation to future generations, and how shall that obligation be carried out? That is of course an age-old question; the difference now seems to lie in the scope and rapidity of human impact on the globe. Man has apparently become a major geological and geophysical agent in his own right, able to influence the physical and biological conditions of the future, deliberately or inadvertently, in a way not open to our ancestors.

There is an urgent need for intensified research to limit some of the uncertainties which now make informed political choice almost impossible. We must learn more about the carbon cycle itself, particularly the quantitative fluxes of carbon dioxide between the atmosphere and the land and between the atmosphere and the oceans. We must be able to predict more accurately the climatic effect of increased levels of atmospheric carbon dioxide. This is now the major uncertainty in assessing the environmental impact of fossil fuel consumption. We must also learn to anticipate the ecological, economic, social, and political consequences of climatic change. This is a formidable interdisciplinary and international research task whose dimensions are only beginning to be seen. There are heartening indications of a growing international consensus on the need for cooperation to provide solutions.

But research can be only a prelude to action. What actions are called for, or even possible?

In a talk before the American Physical Society, Thomas Moss, principal assistant to Representative George Brown, Chairman of the House Subcommittee on Environment and Atmosphere, pointed out that climatic change is a virtual prototype of a problem poorly matched to existing human institutions. Its time span is longer than a political leader's career. The potential effects are enormous, conceivably dwarfing those of normal man-made technical and social change. This kind of problem presents an almost insurmountable challenge to institutions designed for times when societies were less complex, man's abilities for doing "good" and "bad" much more limited, and thinking more restricted in time and space.

In both its spatial and temporal aspects, global climatic change stands almost alone among the world's environmental problems. Many pollution and natural resource issues do not respect national boundaries, but the carbon dioxide question is unique in that regardless of how much sources may be localized, the atmospheric concentrations will be the same everywhere. It is likewise unique in that its impact will persist long after the sources are eliminated.

And unlike most environmental impacts, this one could in the long run appreciably benefit some nations and regions while harming others. This will make international consensus even more difficult than with other forms of environmental change. Essential to a global consensus is a better understanding of the causes and consequences of climatic change, an understanding that can come only through a truly international multidisciplinary effort.


1 For a semi-popular discussion of climatic change, see Stephen H. Schneider, The Genesis Strategy, New York: Plenum Press, 1976. A more technical discussion of the carbon dioxide question is given by C. F. Baes, et. al., "Carbon Dioxide and Climate: The Uncontrolled Experiment," American Scientist, May-June 1977, p. 310. The Geophysics Study Committee of the National Academy of Sciences has produced a detailed problem analysis with much background information in Energy and Climate, Washington: National Academy of Sciences, 1977.

2 Charles D. Keeling and Robert B. Bastacow, "Impact of Industrial Gases on Climate," Energy and Climate, op. cit., p. 110.

3 Ibid.

5 Stephen H. Schneider, "On the Carbon Dioxide-Climate Confusion," Journal of the Atmospheric Sciences, November 1975, p. 2060.

6 Joseph Smagorinsky, "Modeling and Predictability," Energy and Climate, supra, footnote 1, p. 229.

7 T. J. Hughes, "West Antarctic Ice Streams," Review of Geophysics and Space Physics, January 1977, p. 1.

8 Schneider, The Genesis Strategy, supra, footnote 1, p. 19.

9 R. A. Bryson, "A Perspective on Climatic Change," Science, May 17, 1974, p. 753.

11 Energy and Climate, supra footnote 1, p. 2.

12 Central Intelligence Agency, "USSR: The Impact of Recent Climatic Change on Grain Production," Report ER 76-10577 U, October 1976.

13 The most complete estimate of the long-term effects of climatic change on both cultivated and noncultivated land is probably that carried out as part of the Climatic Impact Assessment Program sponsored by the U.S. Department of Transportation. This three-year study was intended specifically to evaluate the environmental impact of SST aircraft, but the information compilation was designed to be broadly applicable to assessment of climatic change from any cause. Although emphasis was on the United States, attention was given to worldwide impacts. The effects of climatic change on production were assessed by means of literature reviews, by statistical regression models relating monthly temperature and precipitation to yield, by computer simulation, and by studies of how historical climatic changes have affected crop production and the boundaries of natural ecosystems. Because particulate matter from SSTs in the stratosphere was expected to lower global temperature, most emphasis was on a reduction of temperature, but temperature increase was also considered. For further detail, see Impacts of Climatic Change on the Biosphere, Department of Transportation, Climatic Impact Assessment Program (CIAP), CIAP Monograph 5, DOT-TST-75-55, September 1975.

14 L.M. Thompson, "Weather Variability, Climatic Change, and Grain Production," Science, May 9, 1975, p. 535.

15 M. Glantz, "The Sahelian Drought: No Victory for Anyone," Africa Today, April 1975, p. 57. See also Jean Mayer, "Coping with Famine," Foreign Affairs, October 1974, p. 112.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now
  • Charles F. Cooper, a plant ecologist, is Professor of Biology and Director of the Center for Regional Environmental Studies at San Diego State University. During 1977 he was a visiting scientist at the International Institute for Applied Systems Analysis near Vienna and a Fellow of the Woodrow Wilson International Center for Scholars in Washington, where this article was written.
  • More By Charles F. Cooper