scientific community began to come to grips with the issue of global climate
change more than three decades ago. While it is safe to say that researchers
have made considerable progress in advancing our understanding of the climate
system and its sensitivity to human-induced increases in the concentrations
of greenhouse gases, it is equally safe to say that this progress has not always
lead to a reduction in the uncertainty that clouds our vision of how the future
In northwest Iowa, terraces on the uplands and watershed dams help control erosion and flooding one way to respond to climate changes over time. Photo by Lynn Betts, USDA Natural Resources Conservation Service.
No matter what, different communities are going to have to confront change in their environments, and they will do so in different ways. But profound and persistent uncertainty will undermine the ability of the conventional cost-benefit approach to inform their decisions adequately. This conclusion is unsettling for people who are comfortable comparing costs and benefits in their own lives and in their communities, but an alternative approach exists: Apply the same risk-management approach that has inspired people to buy insurance against the intolerable consequences of events such as house fires and serious disease and has inspired their governments to create social safety nets.
Estimates of the increase in global mean temperature associated in equilibrium
with a doubling of atmospheric concentrations of greenhouse gases have been
published by the Intergovernmental Panel on Climate Change (IPCC) since its
First Assessment Report in 1990; these authors put the range between 1.5 and
4.5 degrees Celsius. Enormous effort has been devoted to trying to reduce this
uncertainty over the intervening 15 years, but to no avail. Indeed, the most
recent estimates have expanded to an upper tail of the distribution beyond 9.0
degrees Celsius. This one specific, but absolutely fundamental, example shows
how increasing our knowledge base can sometimes expand the range of futures
with which we may have to cope.
Difficulty in understanding the science of the climate system has not been the only source of uncertainty. The research community has worked through this complexity to reach a consensus view that human activities have been driving increases in atmospheric concentrations of greenhouse gases and therefore spawning changes in local and global climate regimes, but answers to the corollary question So what do we do about it? have been elusive.
Notwithstanding the complexity of the climate system, it is people that make the So what? question so confounding. People respond to change and stress, of course, and it turns out that their adaptation can matter. Estimates of the economic cost of change and stress must therefore take these responses into account an imperative that is perhaps best illustrated by estimates of the economic cost of greenhouse-induced sea-level rise along developed coastlines.
Early work that focused on the United States demonstrated that decisions about whether or not to retreat from rising seas, if done efficiently, could significantly reduce the cost that could legitimately be attributed to climate change. In some instances, the benefits of protecting property with levies or beach nourishment would be larger than the cost of these projects. Investing in protection would therefore be worthwhile because it would reduce exposure to coastal flooding events at least over the medium-term, when sea-level rise would be most predictable. In other cases though, policies imposing geographic limits on which damaged properties could be repaired, for example, would be more appropriate.
These policies do not necessarily pass a cost-benefit test, but they do recognize longer term risks over which we will never have control, and are designed explicitly to reduce coastal sensitivity to those risks. This is not to say that adaptation would eliminate the considerable damage caused by coastal storms; it is rather to say that adaptation to rising seas could significantly reduce the added portion of that damage that could be attributed to climate change.
In its Third Assessment Report, the IPCC recognized this distinction when it defined any systems vulnerability to climate change and climate variability in terms of not only its exposure to the impacts of climate and its baseline sensitivity to those impacts but also its adaptive capacity, the ability to reduce either exposure or sensitivity. This approach quickly revealed the fundamental role of adaptive capacity in defining societal thresholds of tolerance to climate-related stress. Perhaps more importantly, their approach focused immediate attention on the critical role that local circumstances play in determining a communitys ability to adapt.
Some communities will respond efficiently to climate-induced stress by reducing exposure. A town might, for example, build a dam to reduce variability in river flow and thereby lower the likelihood of flooding; but it could undertake this adaptation only if it had access to the necessary resources and could work within a receptive governance structure. The community might, however, take advantage of this reduced exposure in ways that actually increased its sensitivity to climate stress. Some citizens might build houses and schools in the flood plain because they felt more secure. But in so doing, they would increase the likelihood that the community would suffer greater damage in the event of a severe flood that the dam could not handle. Other communities might choose simply to reduce their sensitivity to floods by building levies or abandoning property. And still others might do nothing at all.
In response to this type of complexity, the IPCC came to the conclusion in its Third Assessment Report that it is probably infeasible to systematically evaluate lists of adaptation measures for various communities, but that does not mean that all is lost. The take-home message is simply that future research has a long way to go if it is to come to grips with the diversity of the social-political-economic systems within which people all over the world will try to adapt to climate change.
Climate researchers have tried to draw a distinction between mitigation and adaptation: treating the disease by slowing the rate at which greenhouse gases accumulate in the atmosphere versus treating the symptoms of the disease by minimizing the consequences of those accumulations. As they think about the roles that each approach might play in an integrated response to the climate problem, researchers are coming quickly to the realization that global decision-makers may need to adopt an entirely new approach in their attempts to answer the So what? question.
Techniques that compare costs with benefits have framed much of the climate policy debate ever since William Nordhaus, an economist at Yale University who has been engaged in climate research for nearly 30 years, authored his seminal policy paper entitled To Slow or Not to Slow? in 1991. He computed a series of taxes on the carbon content of fossil fuel through the year 2100 that would make the economic benefits of mitigation, expressed in terms of damages avoided, as large as possible when they were compared to the economic cost of increasing the cost of energy.
Nordhaus accommodated the reality that society would be paying to reduce greenhouse gas emissions well before it would see the positive effects of slowed climate change. His model, however, assumed a world with a global decision-maker who would be fully informed about how the future would unfold, how climate impacts would be distributed, and how communities and individuals would adapt. But what if we do not know how the future will unfold? What if we cannot tell how climate impacts will be distributed? What if we cannot foresee how people would adapt? And what if there is no global decision-maker?
There may come a time when analysts have this information for most of the worlds communities, but that time is not now. It is, therefore, becoming increasingly clear that a risk-management approach to climate policy should be adopted by researchers and decision-makers alike.
Unlike the cost-benefit framework, risk-management techniques were designed explicitly to inform decisions that must be made under conditions of uncertainty. They are, nonetheless, supported by the same efficiency criteria that many find so appealing about an approach that weighs costs against benefits. And the techniques are familiar, if perhaps not quite as obviously so, to most people who buy insurance to protect themselves against unpredictable events, such as fire, floods or car accidents.
To inform decisions based on such an approach, researchers need to identify possible thresholds of intolerable outcomes of climate change. They will need to quantify how frequently specific communities will cross those thresholds, and they will need to describe how adaptation could alter those thresholds by reducing exposure or ameliorate the associated harm by reducing sensitivity. They also will need to determine how mitigating climate change itself might reduce the chance of crossing these critical thresholds.
Already, some researchers are having success in providing this information in specific locations. Roger Jones, an environmental specialist with the CSIRO Atmospheric Research Group in Australia, has reported results from some studies in which a risk-based approach was employed to portray vulnerability in terms of the likelihood of crossing specific thresholds.
Coral bleaching, shown here at Great Keppel Island during the 2002 mass bleaching event in the Great Barrier Reef Marine Park in Australia, can lead to coral mortality, if ocean temperatures exceed critical limits for even short periods of time. Modeling global warming in a given area may be able to show thresholds of tolerance for coral. Copyright of Ove Hoegh-Guldberg, Centre for Marine Studies, University of Queensland.
In one case, Jones described how climate change could influence the frequency and severity of coral bleaching at a specific location along the Great Barrier Reef. Coral mortality can occur from ocean temperatures exceeding critical limits for even short periods of time, and Jones has demonstrated how global warming over the next several decades could overwhelm those thresholds.
In another case, Jones reviewed studies of the sensitivity of water supplies in an Australian catchment to climate-induced changes in precipitation and rates of evaporation. Representing the links between changes in climate and changes in precipitation adds another profound dimension of uncertainty. Nonetheless, Jones was able to compute trajectories displaying the relationship between climate change and the likelihood of enduring periods of flooding or drought.
All of these examples show how climate uncertainty, described in terms of specific impacts felt in specific locations, can be characterized to frame questions about adaptation. But the risk-management approach can also bring adaptation and mitigation policies to bear on the same locally defined climate problem.
Work that I have recently completed with Kenneth Strzepek from the University of Colorado has examined the effect of reducing global carbon emissions on the ability of flood-control adaptations along the Brahmaputra and Ganges rivers in India across a wide range of possible future scenarios. In scenarios where climate change would produce smooth, monotonic and manageable effects, reducing greenhouse gas emissions increased the ability to reduce the likelihood of crossing critical flooding thresholds through adaptation. For scenarios where climate change would produce variable impacts over time, though, it turned out to be possible that mitigation would make adaptation less productive for decades at a time. Finally, for scenarios where climate change would produce enormous impacts, adaptation failed regardless of how much mitigation was applied.
What about situations in which we do not even know enough about impacts to think plausibly about what sort of adaptations might be contemplated? The potentially sudden collapse of the Atlantic thermohaline circulation, for example, is on the top of the list of such nonlinear climate impacts that the IPCC highlighted in its Third Assessment Report as significant sources of concern.
Most of Europe is 6 to 8 degrees Celsius warmer, on average, than other regions that lie at the same latitude because it enjoys the warming effects of a warm ocean current that is the part of the Atlantic thermohaline circulation that flows north along the eastern coast of the United States before turning right and traveling to the western shores of Europe. Although we do not know for sure the physical, natural and economic effects of the collapse of the thermohaline system, we do know that having life without the thermohaline circulation is not an experiment that should be tried on our planet.
It is difficult to evaluate how to respond to such a risk, given the profound uncertainty in our understanding of climate sensitivity and the circulation process. The first step, though, is to assess the likelihood of a collapse. Working with Michael Schlesinger and Natasha Andronova of the University of Illinois at Urbana-Champaign, I recently attached the current variant of the economic model employed by Nordhaus in his first mitigation paper to a state-of-the-art model of the thermohaline circulation. The results are sobering.
Current scientific knowledge supports estimates that the likelihood of a collapse of the circulation over the next 200 years is larger than 2 chances in 3, if we do nothing. The likelihood declines with mitigation, but even the most rigorous climate policy would leave a 1 in 4 chance of a collapse; and waiting 30 years to act increases the odds to more than 1 chance in 3. Put another way, it is more likely than not that the circulation will collapse even if increases in global mean temperature are held below 2.5 degrees above the 1900 level.
These estimates should attract considerable attention among those people who calculate risks associated with such events, even if the consequences, while known to be significant, are not particularly well-defined. We do not need to wait for a more thorough understanding of how adaptation might work to see that a risk-based approach can highlight sources of concern and a need to buy insurance against potentially extraordinary costs.
We can be certain that adaptation can work to reduce some of the damages associated with climate change even if we do not know exactly when and where. By understanding what defines a communitys ability to adapt, we can also try to come to grips with facilitating adaptation in locations where it is now difficult.
Most importantly, we need to recognize that profound uncertainty of this magnitude is often inappropriately used by analysts tied to a cost-benefit approach as a reason not to do anything. Indeed, uncertainty becomes the reason to do something in the near-term, if adopting a risk-management perspective in which mitigation and adaptation can be complementary tools in a portfolio of policy options tools that treat the disease and the symptoms at the same time.