Other reports in this collection
This section discusses the strengths and limitations of the analytic approaches used to address the reasons for concern, mainly with regard to whether they can, with the confidence levels given, indicate the severity of impact or risk as a function of increase in global mean temperature. This discussion identifies key uncertainties inherent in each method and offers directions for future research that could improve our confidence in the results produced with each approach.
The organization of this section parallels that of the previous sections of this chapter. The strengths, limitations, uncertainties, and directions for each approach are discussed in the same order in which they were discussed in the preceding sections. However, integrated assessment frameworks are considered separately from aggregate approaches. Last is a discussion of integration across methods and reasons for concern.
Advantages: Because observations are based on observed effects rather than models, they can be used to indicate whether climate change is causing impacts and whether impacts lead to positive, negative, or indeterminate outcomes. They also can be used to validate hypotheses and models that formalize hypotheses on cause and effect.
Disadvantages: The problem with relying on observations to determine the severity of impacts or risk from climate change is that there has been only 0.7°C of mean global warming over the past century (although some regions have experienced much more warming). Because many impact thresholds may not be crossed until greater magnitudes or rates of warming are reached, it is not clear how to interpret an observed effect of warming or a group of such observations. Such observed impacts to date often will be of only minor consequence, even though they may tend to confirm our understanding of impact processes. Moreover, lack of observed impacts may be simply because climate change has not yet reached critical thresholds for such effects. Finally, attribution of causality is very difficult with observed effects or groups of effects. One must be able to demonstrate that a regional change in climate is a significant cause of an observed effect and that the regional change in climate is linked to global climate change.
Uncertainties: Uncertainties include the magnitude of climate change that has occurred, the extent to which impacts can be attributed to climate change that has occurred, and whether the relationship between climate change and possible impacts is linear or nonlinear and continuous or discontinuous.
Research Needs: For climate change impact detection to advance, there is a need for continued, improved, and augmented data collection and further development of analytical techniques. Geographical diversity is needed to balance the current bias of study locations in North America and Europe; more observation studies are needed in developing countries, with emphasis on those where physical, biological, and socioeconomic systems have higher vulnerability to climate change (see Chapter 18).
Because climate and impact systems are linked over a range of temporal scales, longer time series of data allow better understanding of the relative magnitudes of short- and long-term responses (Duarte et al., 1992; McGowan et al., 1999). Large-amplitude temporal changes usually involve large spatial dimensions, so broad-scale spatial/temporal studies are necessary as well. Satellite measurements of the Earth's surface provide a very useful monitoring capability for ocean, ecosystem, and land-cover changes. For example, satellite measurements of the Earth's surface offer the potential for aggregation of observed impacts with regard to broad-scale ecological responses such as vegetative responses to increasing lengths of growing seasons (e.g., Myneni et al., 1997), complemented by meteorological and vegetation data (e.g., Schwartz, 1998).
For ecosystem impacts, continuing observations are needed at sites where studies already have been conducted, at long-term ecological research sites (e.g., Chapin et al., 1995), and in protected areas. Programs that provide continued long-term monitoring of marine and terrestrial environments also are important (Duarte et al., 1992; Southward et al., 1995). Large-scale spatial/temporal ecosystem studies are necessary because effects from local changes cannot be extrapolated to large areas without evidence (McGowan et al., 1999; Parmesan et al., 1999).
Definition of indicator species or systems is a useful element of detection studies (e.g., Beebee, 1995; Nehring, 1998; Cannell et al., 1999). Coupled with monitoring programs, such data may then provide a consistent set of evidence with which to study past, present, and future impacts of climate changes.
A further critical research need is to strengthen analytical tools for understanding and evaluating observed climate change impacts. Robust meta-analyses of studies that present good quality, multivariate data from a diversity of settings around the world will help to define further the global coherence among impacts now observed. Care also must be taken to ensure that the sample of studies is representative across time and space, is not biased in its reporting, and uses appropriate statistical tests. Also needed is development of methods to analyze differential effects of climate across a range or sector. Individual and grouped studies need to address possible correlations with competing explanations in a methodologically rigorous manner.
Also needed are refinements in the fingerprint approach (e.g., Epstein et al., 1998), including more precise definition of expected changes and quantitative measurement techniques, similar to that used in detection of climate changes (see TAR WGI Chapter 12). For climate, fingerprint elements include warming in the mid-troposphere in the southern hemisphere, a disproportionate rise in nighttime and winter temperatures, and statistical increases in extreme weather events in many locations. These aspects of climate change and climate variability have implications for ecological, hydrological, and human systems that may be used to define a clear and robust multidimensional "expected impact signal" to be tested in a range of observations. A more refined and robust fingerprint approach may aid in the study of difficult-to-detect, partially causal climate effects on socioeconomic systems such as agriculture and health.
Assessments of unique and threatened systems tend to be based on studies of particular exposure units such as coral reefs, small islands, and individual species.
Advantages: These studies contain richness of detail and involve many researchers, often from developing and transition countries. In contrast to aggregate studies, studies of exposure units can be used, at least in principle, to analyze distributional effects by focusing on impacts on particular systems, species, regions, or demographic groups.
Disadvantages: One of the main disadvantages is that exposure-unit studies often are not carried out in a consistent manner. Exposure-unit studies often examine related sectors in isolation and do not examine linkages or integration among sectors and regions; for example, studies of the effects of climate change on ecosystems or individual species often are conducted without examining the potential effects of societal development on such systems. Local processes and forces (e.g., urbanization, local air pollution) often can be more important than global ones at the local scale, complicating the task of measuring the influence of global climate change at the local scale.
Another key disadvantage is incompleteness of coverage. For example, in spite of many and extensive country studies, there still are many gaps in coverage in terms of countries, regions within countries, and unique and important potential impacts that have not been assessed. The choice of exposure units may not necessarily cover the most vulnerable systems. Topics such as impacts on biodiversity or unique ecosystems often are not covered. There also has been little attention to impacts on poor and disadvantaged members of society. Even where particular critical exposure units have been covered, there may be just a single study. Drawing conclusions with high confidence on the basis of one study may be inappropriate.
Uncertainties: Uncertainties include the likely magnitude of climate change at the spatial resolution required by the study of the particular unique and threatened system, masking of global change effects by nonclimate factors, the degree of linearity/ nonlinearity in the relationship between stimulus and response, and the degree to which results from individual studies can be extrapolated or aggregated.
Research Needs: It would be desirable to have more studies of individual systems, according to some set of priorities concerning the likely immediacy of the impacts. Additional work on standardizing methods and reporting of results also would be extremely useful. It also would be useful to devote more effort to integration of results from existing studies. Again, it would be especially useful to increase monitoring of changes in organisms, species, and systems that have limited range now or are near their limits and to try to separate out or consider other causal mechanisms such as local air pollution, loss of habitat, and competition from invading pests and weeds.
Advantages: Distributional impact studies draw attention to likely heterogeneity in impacts among different regions and social and economic groups. They also help to identify and assess the situation of the "most vulnerable" people and systems. Thus, such studies bring equity considerations to center stage.
Disadvantages: Distributional impact studies require regional climate change projections and impact projections at the regional to local scale, where GCMs may not be very accurate. They also require projections of demographics and socioeconomic structure over a long time horizon.
Uncertainties: Research into the distribution of impacts of climate change is recent (see Section 19.4). There are some findings on which there is virtual unanimity. Some findings are broad conclusions—such as that more resource-constrained regions are likely to suffer more negative impacts, as are people whose geographic location exposes them to the greatest hazards from climate change. (Such people often live in regions with marginal climate for food growing or in highly exposed coastal zones.) Others are more specific but to date have been more conclusive with regard to the direction of different impacts among regions, rather than the magnitudes. For example, we know that impacts in developing low-latitude countries are more negative—in part because those countries tend to be operating at or above optimum temperatures already—and, in some cases, in regions where rainfall will decrease, leading to water stress. There also is limited capacity for adaptation in these areas. In some mid-latitude developed countries, agriculture would benefit initially from warmer conditions and longer growing seasons. Beyond such sweeping statements, uncertainties are vast. Resource constraints and (climatic) marginality are multidimensional and complex phenomena. Currently, it is not known which components of resource constraints or climatic marginality are more important or which components may compensate for others or may have synergistic effects. There are suggestions in other literature, but these have not been systematically applied to the impacts of climate change, conceptually or empirically.
In sum, there is virtual consensus about the broad patterns. There is much less knowledge about the details, although that situation is slowly improving.
Research Needs: Development of appropriate indicators of differences in regional impacts and ways of comparing them across regions and socioeconomic groups would be extremely useful. Improved methods for characterizing baseline demographics and socioeconomic conditions in the absence of climate change or climate change-motivated policies also would be useful. There is a need to quantify regional differences and to develop estimates of the cost of inequity in monetary or other terms (e.g., effect on poverty rates and trade, social and political instability, and conflict). More accurate projections of regional climate change would increase confidence in predictions of regional climate change impacts.
|
Other reports in this collection |
|