We now have reached the point at which we can begin to focus on the central issue of this course. What is the impact of the rising concentrations in greenhouse gases on the climate of the earth? All of the information in the preceding chapters has served as the foundation of a structure upon which we can now construct our answer, or at least present a number of possibilities.
Before we begin to address this question, let's take stock of what we can state with certainty and what aspects of this issue should be considered with some caution.
* The global surface temperature is .5[o]C warmer now than it was a hundred years ago. The uncertainty in this figure is small (about 10%).
* As a result of the present concentrations of the greenhouse gases (water vapor, CO2, etc.) and of clouds the earth is 34[o]C warmer than it would otherwise be. We are very confident that the greenhouse effect exists.
* Simple models of the greenhouse effect indicate that as the concentration of greenhouse gases increases, the effective radiating level (ERL) of the earth-atmosphere climate system increases in altitude. This results from the increased opacity of the atmosphere to infrared radiation. Because the temperature of the troposphere decreases with height, a higher ERL implies that the amount of energy emitted to space is reduced. Since the climate system must remain in equilibrium with the incoming solar radiation, the temperature of the atmosphere (from the surface to the ERL) must increase. This model reproduces most of the key physical principles involved in rising concentrations of greenhouse gases; however, complex interactions within the climate system necessitate the use of climate models to calculate this temperature change. We will investigate this further in this chapter.
* Clouds have a major impact on the climate system. At the present time, clouds act overall to cool the earth's surface. The potential feedback effects of clouds in the climate system represent one of the largest sources of uncertainty for determining the response to rising CO2 concentrations. Changes in cloud height and amount as a response to rising levels of greenhouse gases could alter our climate dramatically.
* Models of the atmosphere require accurate knowledge of the initial state of the atmosphere in order to make useful forecasts of weather. However, we can never measure the atmosphere well enough to eliminate all of the errors at the start of a model's forecast. The small errors at the outset grow as a result of nonlinear interactions and in a matter of days destroy the usefulness of the model's forecast. Climate simulations of the atmosphere suffer from the same fate. However, if we yield our interest for specific weather information in the future and settle for information about the average conditions or trends relative to the present, then climate simulations for periods of a hundred years or more can be quite useful. The uncertainty related to a number of aspects of these climate simulations is quite large, as we shall see in this chapter.
* The concentration of carbon dioxide and other greenhouse gases has increased at an alarming rate as a result of human activities. The amount of CO2 in the atmosphere has increased from 280 ppm in the pre-industrial era to present concentrations of about 355 ppm. There is no question that this rise in CO2 has resulted from anthropogenic emissions.
* The residence time of a typical carbon dioxide molecule in the atmosphere is short. The combined atmosphere and mixed layer of the ocean adjust within a decade or so to anthropogenic emissions. The residence time of carbon in the deep ocean is long (over 1000 years). Thus, the deep ocean is not in equilibrium with the rising emissions of CO2 that have taken place in the past 150 years. At the present time, the deep ocean is absorbing all of the carbon that becomes available to it. Eventually the deep ocean will become saturated (in equilibrium) with respect to carbon and it will no longer continue to absorb ever increasing amounts. This in turn may lead to faster rates of accumulation of CO2 in the atmosphere. Similar inbalances with respect to carbon probably exist now in the terrestrial portion of the carbon cycle. There is considerable uncertainty about where all of the carbon dioxide emitted in the past 150 years has gone in the carbon cycle.
* Reforestation of temperate and tropical forests at a rate sufficient to slow significantly the rate of addition of CO2 to the atmosphere is unlikely to occur. Difficult decisions that have major economic consequences are required to reduce the rate of CO2 emissions from their present levels.
Radiative transfer models that we made reference to in Section 7.2 have also been helpful at identifying a number of potential feedback processes that may enhance or diminish the effect of rising greenhouse gas concentrations on the radiative equilibrium of the climate system. Three-dimensional climate simulations are nearly as difficult to evaluate as the operation of the real climate system because the models have so many processes occurring at once. One-dimensional radiative transfer models can be used to evaluate specific physical processes selectively. Then their results can be examined more thoroughly with more elaborate climate models.
One of the most significant feedback processes results from the effect of the amount of water vapor in the atmosphere on temperature. We have said repeatedly that human activity has little direct effect on the concentration of the largest greenhouse gas, water vapor. What about indirect effects? One possible indirect effect is shown in Fig. 8.1. Rising concentrations of carbon dioxide lead to reduced emissions of infrared to space, which in turn requires that the temperature of the surface (and troposphere) increase in order to maintain radiative equilibrium. However, the increase in temperature of the surface and in the troposphere allows for more water vapor to evaporate from the oceans and more water vapor to be held in the troposphere (see Fig. 4.11). A larger amount of water vapor in the troposphere renders it more opaque to infrared radiation, which results in a higher altitude of the effective radiating level and less infrared emitted to space. Thus, the surface would have to warm more to continue to maintain radiative equilibrium. In this manner, a positive feedback loop develops that can enhance the effect of rising concentrations of carbon dioxide.
By how much does this water vapor feedback affect the radiative forcing?
Current estimates indicate a further reduction of between 2 and 3 W m[-2]. Using the Stefan-Boltzman Law, we can determine that this
may result in an additional warming of the surface of almost another 1[o]C.
Figure 8.1. Rising carbon dioxide concentrations lead to increases in the concentration of water vapor in the atmosphere.
We now introduce a parameter that is used to estimate the amplifying effect of feedback processes, the climate sensitivity parameter that will be specified by the Greek letter [[lambda]]. It is the measure of how much the surface temperature changes for a given net change in radiative forcing at the top of the atmosphere. Imagine for the moment that no feedback processes are acting in the climate system. It has been determined from one dimensional radiative transfer models that [[lambda]] equals .33 [o]C per W m[-2]. In other words, if doubling CO2 causes the net radiative forcing at the top of the atmosphere to change by 4 W m[-2], then the value of [[lambda]] indicates that the surface temperature should increase by 1.2[o]C. This is slightly larger than the estimate of a 1[o]C increase that we obtained earlier in this section. Radiative transfer models indicate that when the water vapor feedback effect is included, then [[lambda]] is equal to .454. The higher value indicates that the climate system is now more sensitive to (that is, responsive to) the radiative effects that result from a doubling of CO2. Since the radiative forcing at the top of the atmosphere due to CO2 doubling remains equal to 4 W m[-2], we can estimate that the surface temperature must increase by 1.8[o]C (4 times .454) as a result of the radiative effects of doubling CO2 and the water vapor feedback.
Another positive feedback process has already been discussed in the context of the onset and demise of the ice ages (see Section 5.1). If rising CO2 emissions lead to warmer temperatures, then the Antarctic polar ice cap may begin to melt at an accelerated pace[*]. The ice-albedo feedback process (Fig. 5.3) may then work in reverse and lead to further warming and further melting, and so on. While several estimates of the ice-albedo feedback in the context of the global warming problem have been made with radiative transfer models, this process is so complex that it requires climate models that can simulate the atmosphere and oceans together. Cloud feedbacks also require the use of 3-dimensional climate models to understand their behavior.
The best projections about what may happen to the climate system in the future
are based on the results of three-dimensional simulations of the atmosphere.
As we described in Section 3.3, such climate simulations suffer from a number
of problems. Some of these problems result from the inherent difficulties in
simulating the fluid motion of the atmosphere (predicting the development and
decay of clouds for instance) and some occur as a result of the need to
complete the simulation in a reasonable amount of time using the computer
resources available at the present time (poor resolution of the horizontal
extent of land and sea, for example).
Figure 8.2. Observed surface temperature (solid line) as a function of latitude compared to temperatures obtained from several climate model simulations.
In order to develop confidence that these climate models are capable of predicting what may occur in the future, we must first examine how well they simulate the present climate. Figure 8.2 compares the observed distribution of surface temperature as a function of latitude to that simulated by several different models developed by different research groups. It doesn't matter here how the various models differ; some use more complicated treatments of clouds and radiation than others. This particular comparison is for the winter season of the Northern Hemisphere, so the observed temperature is colder at the north pole than at the south pole. All of the models are able to simulate within reasonable accuracy the variation of temperature with latitude, except for the conditions over Antarctica (right edge of figure). Because of the presence of the deep ice sheets in this region, the models have a tougher time simulating the surface temperature in this region.
The ability of the same models to simulate the observed latitudinal
distribution of precipitation is shown in Fig. 8.3. The observed variation of
precipitation with latitude is more complex than that of temperature. There is
a peak in rainfall near the equator which derives from the lifting of warm air
in the tropics (see Section 3.1), with secondary peaks in the mid-latitudes of
the Northern and Southern Hemispheres. The simulated precipitation exhibits
larger deviations from the observed precipitation than those found for
temperature. Some models predict much more rainfall near the equator compared
to the observed while others predict less. As we have stated before in Section
3.3, models used for weather forecasting or climate simulations are less
capable of predicting the formation and decay of clouds (and the precipitation
that falls from them) than predicting other model parameters.
Figure 8.3. Observed precipitation (solid line) as a function of latitude compared to the precipitation obtained from several model simulations.
As a further indication of the models' sensitivity to clouds and cloud feedbacks in the climate system, Fig. 8.4 compares the sensitivity parameter [[lambda]] determined from 14 different climate model simulations. We showed in the previous sub-section that the larger the value of the sensitivity parameter, the more sensitive the globally averaged surface temperature is to changes in radiation at the top of the atmosphere. We have arranged the values determined from each model from lowest sensitivity (on the left) to highest sensitivity (on the right). Thus, model 1 (with a value of .39 [o]C per W m[-2]) is nearly three times less sensitive to changes in radiation at the top of the atmosphere than model 14 (with a value of 1.11 [o]C per W m[-2]). Assuming again that the effect of doubling CO2 is 4[ ]W m[-2], the range of values indicate that the surface temperature could change by as little as 1.6[o]C or as much as 4.4[o]C[*].
The large range of climate sensitivity parameter values from model-to-model indicates that there is considerable uncertainty as to the actual value of this parameter. One way to determine an appropriate value for it is to average all of the available ones as shown on the far right of Fig. 8.4. This average value is .68[o]C per W m[-2]. This value is higher than determined in the previous sub-section. Thus, the atmosphere, when simulated in all three dimensions, is capable of developing feedbacks that make the climate system more susceptible to increasing concentrations in CO2 than 1-dimensional radiative transfer models would indicate that it is.
What causes the sensitivity parameter to span such a large range of values? As
a means of answering this question, the sensitivity parameter for each model
has been computed for clear-sky conditions only. This procedure removes the
effects of cloud feedbacks found in each model. These values are shown as
circles in Fig. 8.4. Notice that the range of clear-sky values [[lambda]]c
(from .42 to .57 [o]C per W m[-2]) is
much smaller than the range of values of [[lambda]] . All of these models
predict similar sensitivities to changes in radiative forcing at the top of the
atmosphere when the effects of clouds are neglected. We can then infer that
the wide range in [[lambda]] is a consequence of the different methods adopted
in different models to simulate clouds.
Figure 8.4. Climate sensitivity parameter [[lambda]] (diamonds) and clear-sky sensitivity [[lambda]]c (circles) for 14 different climate models. The respective averages (MN) of all of the values are shown at the far right. The values are in [o]C per W m[-2.]
There is one other disconcerting aspect of Fig. 8.4. The first three models have values of [[lambda]] that are less than their values of [[lambda]]c. This means that those models are more sensitive to radiative forcing under clear-sky than under cloudy conditions. Put another way, these models suggest that clouds act as a negative feedback process that would dampen the effects of rising greenhouse gas concentrations. On the other hand, the rest of the models have values of [[lambda]] that exceed those of [[lambda]]c. All of these models suggest that clouds act to increase the sensitivity of the climate system to changes in radiative forcing brought about by increasing concentrations of greenhouse gases. It is troubling, and a source of uncertainty in the results of climate simulations, that we cannot state for sure whether clouds act to enhance or diminish the effects of changes in greenhouse concentrations.
Syukuro Manabe and co-workers ran in 1968 the first climate model simulation of the effect of increasing the CO2 content of the atmosphere. The design of their research has served as a plan for many climate model simulations that have appeared since. First, a model simulation referred to as the control simulation is made for present CO2 content. The model is normally run for a long enough period that it can be assumed that the atmosphere has attained equilibrium. In the early days, most models were run over a period of only 45 to 120 days and were expected to reach equilibrium after a period of 15 to 30 days. The average of the weather simulated by the model over the last 30 to 90 days of the simulation were assumed to represent the equilibrium conditions of the control simulation. Thus, the climate state of the model simulation was typically a month to a season in length. It is also important to note that in these early models, only the fluid motion of the atmosphere was allowed to evolve; the oceans and ice-covered regions remained fixed with time.
More recently, models of the climate system have incorporated fluid motion in
both the atmosphere and ocean as well as allowing the extent of sea ice to
expand and contract. Recall that in our earlier discussion of the ocean, we
divided it into two parts: the mixed layer and the deep ocean. As we saw in
Section 4.2 water from the mixed layer sinks into the deep ocean in high
latitudes. It is now appropriate to remark that this process is capable of
removing heat from the atmosphere and storing it (for 1000 years) in the deep
ocean. Up until very recently the most advanced climate models included only
the mixed layer of the ocean and not the underlying deep ocean. This situation
was dictated jointly by lack of computing power and uncertainties regarding the
details of the deep ocean circulation that an ocean model should simulate. We
will consider in the next section what one of the very recent simulations with
a model that includes a deep ocean says about global warming. In this section
we consider the results of a simulation where the only part of the ocean that
is included is the mixed layer.
Figure 8.5. The change in surface temperature (anomaly simulation minus control simulation) predicted by a climate model as a result of doubling the concentration of carbon dioxide. Areas in which the change in temperature exceeds 4[o]C are shaded.
First the combined global atmospheric model with global mixed-layer ocean is run into the future with atmospheric CO2 maintained at its present level. This is the control run. Then the combined model is run again with atmospheric CO2 content at twice its present level. This simulation is referred to as the anomaly simulation, since the change in CO2 content constitutes a departure from the present climate state as defined by the control simulation. The difference in surface temperature (or some other parameter) between the anomaly and control simulations averaged over some period represents the model's estimate of the response of the climate system to a doubling of CO2.
Over 20 CO2-doubling experiments have been performed with various different models, each consisting of a global, three-dimension climate model coupled to a mixed-layer ocean. These experiments indicate that doubling the concentration of CO2 would lead to changes in the globally averaged temperature between 1.9[o]C and 5.2[o]C. On the basis of these experiments, the general consensus among scientists working in this field is that a doubling of the CO2 concentration would lead to an increase in the globally averaged surface temperature of 2.5[o]C.
Up until the beginning of this section we have been generally satisfied to view the climate system as a whole in terms of the globally averaged surface temperature. However, the advantage of three-dimensional models is that they provide us with much more detailed information than that. In what regions of the earth will the temperatures rise the most? Who will get more rain? Who will get less? Which ocean currents in the mixed layer will be affected? All of these questions (and many more) can, in principle, be answered by examining in detail the output from the model's simulations. However, just because it is possible to answer the question in the context of such model simulations does not mean that the answers are right. One must keep in mind the considerable uncertainty that exists as a result of the models' inabilities to simulate clouds and other physical processes accurately. With this caveat in mind, let's now address some more specific questions about the potential changes in our climate's future.
What areas of the
globe may experience
the largest increases in
surface temperature? We have repeated Fig. 1.2 here as
Fig. 8.5 to show one estimate of the regional changes in temperature as a
result of doubling the concentration of CO2. This figure shows the conditions
during the Northern Hemisphere winter season. This particular model predicts
an increase in global-average surface temperature of about 4[o]C; thus, it predicts a global warming that is higher than
the present consensus of 2.5[o]C.
Figure 8.6. Change in precipitation due to doubling CO2 concentrations for the Northern Hemisphere summer season. Contours are shown at +/-0, 1, 2, and 5 mm per day. Areas in which precipitation decreases are shaded.
The largest changes in surface temperature evident in Fig. 8.5 occur in the polar latitudes, where in some places the temperatures are 8[o]C higher than in the control run. As viewed on this map projection, the areas in which the temperatures have increased by more than 4[o]C cover well over half of its area. Remember that on the spherical earth, the polar latitudes constitute a small fraction of the earth's surface. However, increase in surface temperature of 8[o]C in the polar latitudes would certainly seem to have dire consequences for melting the ice sheets of Greenland and Antarctica and the sea ice of the polar seas.
All of the climate simulations in which CO2 concentrations have been doubled predict similar changes in surface temperature, although the specific details of the pattern of temperature change over the globe may differ. The similar predictions of all of the models adds confidence to their estimates. However, there is always the possibility that all of the models are making similar but incorrect predictions.
Which regions will experience more/less rainfall? Figure 8.6 shows the change in precipitation predicted by the same climate model. The polar latitudes tend to have slightly higher precipitation rates in the anomaly experiment compared to the control. Most of the central and eastern United States is predicted to have less rainfall than at present, while regions of the tropics (portions of Central America, India, etc.) are predicted to have greater rainfall. Unfortunately, each model predicts different regional precipitation patterns resulting from doubling the CO2 concentration. Thus, the simulations with 15 different models produce no basis for consensus on this particular aspect of climate change.
There is also no agreement among the model simulations as regards other fields that are important for assessing the impact of rising CO2 concentrations. For example, the amount of soil moisture available during the growing season is predicted in many models to decrease over portions of the central United States. However, none of these models simulate accurately the processes that control soil moisture, and we have doubts about the significance of these results.
Figure 8.7. The difference in globally averaged surface temperature (in [o]C) between the transient response simulation, in which the CO2 concentration rises gradually, and the control simulation. The effect of an instantaneous doubling of CO2 in a simulation without a deep ocean is shown for comparison in terms of an x at 70 years, the time at which CO2 concentrations at a rate of increase of 1% per year will have doubled.
The simulations we have just described were undertaken with atmospheric models coupled to an ocean consisting only of a mixed layer. The addition of a deep ocean below the mixed layer is an advancement that permits heat storage that has the potential to lessen somewhat the global warming of the atmosphere due to CO2 doubling and also alter the pattern of temperature change over the globe. In order for this fully coupled ocean-atmosphere model to evolve realistically, the CO2 increase and accompanying warming must occur gradually rather than all at once, as in the simulations just discussed. For this reason, the response of a fully coupled model to rising C02 concentration is referred to as the transient response to a C02 increase. Several such simulations have been made. We will discuss here the one which uses the same atmospheric model from which Fig. 8.5 was obtained. First, a model simulation was made in which the CO2 was held fixed at its present value (the control simulation). Then, another simulation was made in which the concentration of CO2 in the atmosphere was increased by 1% each year (the anomaly simulation).
Figure 8.7 shows the global average transient response in surface temperature as a function of time. This model simulation predicts the globally averaged surface temperature would rise at a rate of about .35[o]C per decade to attain an increase of 3.5[o]C after 100 years. Notice that after 70 years, the surface temperature is expected here to rise by only 2.3[o]C compared to the over 4[o]C rise obtained in the simulation where the concentration of CO2 is doubled instantaneously and there is no deep ocean. This suggests that the simulations described in the previous sub-section may overestimate the global warming; that is, that heat storage in the deep ocean could significantly moderate global warming.
Figure 8.8 shows the regional changes in temperature as a result of doubling CO2 in the transient response simulation. Compare the magnitude of the temperature changes here to those shown earlier in Fig. 8.5. The change in surface temperature is now greater than 4[o]C only in the polar latitudes of the Northern Hemisphere. An even larger difference in the respective simulations is seen in high latitudes of the Southern Hemisphere, where the warming has now diminished considerably. This relief is a consequence of storage of heat in the deep ocean, which is a particularly efficient process in the region of downwelling water that characterizes the model ocean in the region encircling the Antarctic continent. The model ocean also captures the phenomenon of downwelling in the North Atlantic that was highlighted in Fig. 4.6, and this lowers the simulated increase of surface temperature there and in high northern latitudes in general.
Figure 8.8. The difference in surface temperature between a model simulation in which the CO2 increases gradually and the control simulation. The contouring and shading are the same as in Fig. 7.7.
Based on the information presented in the previous section, we must conclude that there is a definite risk that surface temperatures will be much higher during the middle of the next century than they are at present. Figure 8.9 shows estimates of how rapid the rise in globally averaged surface temperature could be as a function of the four policy scenarios introduced in Section 6.2. If no significant changes are made to present policies (Scenario A), we may see increases in the globally averaged surface temperature of more than 2[o]C within 50 years from now. If more stringent controls are placed on emissions, then the increase in temperature may be limited, on a global basis, to less than that.
We turn now to weigh the risks associated with rising CO2 concentrations against the uncertainties that exist in our knowledge of how the climate system will respond to human activity.
Figure 8.9. Estimates of the increase in global mean temperature relative to that in 1765 due to increases in greenhouse gas concentrations in the atmosphere. Scenarios A-D reflect different levels of controls on gas emissions.
In the past decade, it has been stated many times that humans have embarked on a climate experiment for which the outcome is uncertain. Assume for the moment that by the year 2010 the globally averaged temperature has increased by .5[o]C compared to 1990 values. There is now little question that global warming is underway. Industrial nations have begun to control the rate of growth of emissions of greenhouse gases, but the concentration of CO2 and other gases continues to increase. What might we expect our climate to be?
Beginning with our original definition of climate, we have focussed throughout this text on average conditions, such as the globally averaged surface temperature. In this context, we have viewed weather as underlying the climate state, but we have never considered whether a change in the climate state affects the nature of our weather.
To understand how the slight increase in globally averaged temperature over the next few decades may affect the weather experienced at Salt Lake City or elsewhere, it is necessary to view rapidly varying weather, such as the daily average surface temperature, for example, as being superimposed on the slower changes in the climate system. Figure 8.10 shows schematically how a shift in the climate state may affect the weather on day-to-day time scales. The heavy curve reflects the present conditions. Most of the time the temperature is likely to be within a few degrees of the climate state; in other words, the frequency of occurrence of temperatures near the climate state is high. Every so often, however, the temperature may fall below -5[o]C (about 23[o]F) or greater than 35[o]C (about 95[o]F). The likelihood that these temperatures will occur is small (as shown by the fact in Fig. 8.10 that their frequency of occurrence is low).
Now imagine that the increase in CO2 results in a rise of 1[o]C for the climate state in our locale. Most of the time, the weather would not be any different than before; the temperature will lie within relatively common values between 10[o ]and 20[o]C. However, the frequency of occurrence of extremely hot temperatures is now slightly higher, as shown in Fig. 8.10 by the slight difference between the heavy and thin lines. In addition, extremely cold temperatures are slightly less likely to occur. Thus, most of the time the weather that we may experience as a result of a climate change would not be significantly different from what we observe now, except for the possibility that higher temperatures are slightly more likely.
Such a subtle change in our local climate would, from most points of view, be inconsequential. Air conditioners would have to work a little harder perhaps. But, is that all that may happen if global warming occurs? We have little hard information to go on, but it is certainly likely that many other things could happen. Slight changes in the equator-to-pole temperature contrast might lead to variations in the location of the jet stream. Since storms are carried along by the jet stream, this could cause regions to experience more frequent droughts or floods, depending on where the jet stream lies. The climate model simulations made to date exhibit too much variation from model-to-model to be able to determine changes in the location of the jet stream with any degree of certainty.
What about hurricanes? A slight increase in the surface temperature of the
oceans would provide conditions more favorable for the occurrence of
hurricanes. Climate simulations are incapable of predicting even the existence
of hurricanes at this time, as a result of the poor horizontal resolution in
the models.
Figure 8.10. The frequency of occurrence of daily average temperatures at a location. The climate state is here equal to 15[o]C. Present conditions are shown schematically by the heavy line. If the climate state increases by 1[o]C, then the temperatures are distributed according to the thin line.
Nearly every aspect of our weather in Utah and the United States could be affected by global warming. Most of these changes would be so subtle that they would not be apparent to most people. However, it is possible that global warming could affect the frequency of occurrence of the most violent weather (hurricanes, tornadoes, severe storms, etc.). In addition, subtle changes to the positions of the jet streams and other factors that control the weather could lead to heartbreaking floods and droughts. Whether we want to be or not, we are all part of this climate experiment; our environment, econonmy, and lives are potentially at risk.
At the beginning of this text, we stated that some members of the scientific community feel that the uncertainties associated with global warming are so large that the potential threat from it is not as important as the cost to curtail it. There is even a smaller number of scientists who suggest that global warming may be a good thing. As odd as this may seem, let's take a moment and investigate their viewpoint.
As we discussed in Section 6.1, plants consume CO2 in the photosynthesis process. Growth of plants can be accelerated over a period of weeks and months when CO2 levels are higher than that found in the present atmosphere. Further, when CO2 levels are higher, the tiny apertures in plant leaves that admit CO2 do not have to open so far. This conserves plant water and therefore leads to a more effective use of another diminishing resource: fresh water. Given rising population levels worldwide, there is tremendous pressure to improve the productivity of the remaining agricultural areas in our increasingly industrialized world. Our inadvertent climate experiment could actually be doing some good.
While the above information sounds superficially to be a positive aspect of global warming, there is considerable uncertainty about what may happen to plant communities as a result of gradual increases in CO2 levels in the atmosphere. Coupled with the potential changes in regional climates, it is possible that plant species could become increasingly stressed as a result of flood and drought cycles and other weather changes.
Now let's turn to several of the issues upon which the global warming debate has been centered and which focus attention on the key areas of uncertainty.
* CO2 has been increasing since the onset of the industrial age, yet global warming is not detectable unequivocally in the present climate record. We have shown in Chapter 4 that there are many natural forms of climate variability from year-to-year that may help to mask long-term trends in the climate record. ENSO episodes and volcanic eruptions, among other forms of natural variability, make it impossible at this point to state with certainty that global warming has taken place. Nevertheless, the globally averaged temperature since the late 1970's has averaged about .5[o]C higher than that found one hundred years ago. If we ascribe all of the changes in the past hundred years to natural variations, then we would need to see a further rise of about .5[o]C to have confidence that we are detecting a global warming signal. Based on the four scenarios of greenhouse gas emissions introduced in Section 6.2, this could happen shortly after the year 2000 or thiry or more years later. Thus, the debate about whether global warming is underway may continue for many more years.
* Climate simulations tell us more about the models than about the atmosphere. There is no question that climate models are presently incapable of simulating accurately the future state of the atmosphere and other aspects of the climate system. Just because several models predict the same thing does not mean that the prediction is correct. Present climate models are particularly poor at simulating clouds; thus, the significant effect of clouds on the radiation balance of the climate system could be substantially in error in the models.
However, out of the dozens of simulations that have been made with a wide array of models, there has never been a credible case where increasing the concentration of greenhouse gases did not lead to higher globally averaged temperatures, with the most pronounced warming taking place in the polar regions. The greenhouse effect is real; the only problem is that cloud feedback processes in the atmosphere and heat storage in the deep ocean may mitigate its effect.
* The problem could take care of itself; we don't know for sure where all of the CO2 put into the atmosphere since the beginning of the industrial age has gone anyway. Uncertainty exists as to whether the deep oceans can continue to absorb a significant fraction of the CO2 emitted into the atmosphere. There is uncertainty as to whether some of this carbon will continue to be sequestered in the soils. Thus, instead of the problem taking care of itself, it could get worse.
* Scientists are always crying wolf; they just want more money. Scientific research has become embroiled in controversy in recent years. Funding for major projects in the United States often depends more on public relations efforts and influence within Congress than scientific merit.
The global warming problem has been studied by atmospheric scientists for decades; the potential problems are real. The debate among scientists as to the validity of certain aspects of the global warming problem is healthy because it focusses attention on aspects of our climate system that need to be studied further. Unfortunately, special interest groups often find plenty of material in this scientific debate that helps to justify their particular perspectives.
It is now time for you to assess the information presented on the global warming problem. For students who have been out of high school for only a couple of years (age 20 or so), you can expect to live for another 50 years or more. Why should you be concerned about global warming? You must live your entire life with the effects of the ongoing industrial age and the dependence of the world economy on fossil fuel burning as the major source of energy.
Will the climate system in the next century be so totally altered that life on earth is threatened? Such an outcome is unlikely, at least for the highly adaptable human species. The climate system has shown remarkable resilience over the past million years. We can take some comfort in the fact that the climate system has a bewildering array of check-and-balances that control the size of climate fluctuations. Most of these feedback processes are just beginning to be understood.
Will the climate system in the next century be the same as the one today? Our feeling is that it will be altered. However, the extent of the alteration is difficult to evaluate at this time. Subtle shifts in the jet stream over the United States could plunge Utah into a drought for a decade or more. As the population along the Wasatch Front continues to increase, such an outcome would have significant personal and economic consequences.
What can you do about it? The most obvious thing is to remain informed about the problem and its consequences as it evolves. Unfortunately, news coverage of global warming tends to focus on narrow issues (and often ones of marginal consequence). You shoul be able to place the new information in context. What does it mean overall? Has some new insight come to light that has significant impact?
Is recycling your trash and planting an extra tree or two enough of a contribution to solve the problem? In our opinion, no. It certainly helps, but global warming will not go away as a result of such personal initiatives. The United States is by far the largest producer of greenhouse gases per person. Steps could be taken to reduce this unenviable distinction. None of these steps can take place without effects on our economy and personal lives. There are no easy solutions to global warming. That does not free us from the necessity to search for them.
1. Which of the four scenarios for future emissions do you think will be closest to that which is eventually observed? Why?
2. What is the predicted warming (in [o]C) due to a doubling of atmospheric carbon dioxide alone? By how much (in [o]C) is this warming increased dueto water vapor feedback? What is the consensus figure (in [o]C) as to what will be the warming due to a doubling of atmospheric carbon dioxide when all of the feedbacks are taken into consideration? What are these feedbacks?
3. What role will the ocean play in the global warming arising from a doubling of atmospheric carbon dioxide?
4. According to the "business as usual" scenario, how long will it take for atmospheric carbon dioxide to double from its present value?
5. What problems do climate models have in predicting the present climate?
6. Why do we need to make a control experiment as well as one (the anomaly experiment) in which CO2 is allowed to increase?
7. Identify what factors you believe have the greatest risk associated with the global warming problem.
8. Identify what factors you believe have the greatest uncertainty associated with global warming.
[*] These sensitivity parameters are not determined from simulations of future climates or ones in which CO2 has been doubled. They were determined from simulations of the current climate.