Treatment of impacts in these models also varies greatly. Generally, however, impacts are one of the weakest parts of IAMs. To a large extent this is a reflection of the state of the art of the underlying research, but it also reflects the high complexity of the task at hand (see Tol and Fankhauser, 1998, for a survey). Despite the growing number of country-level case studies, our knowledge about climate change and climate change impacts at the regional level remains limited. A coherent global picture, based on a uniform set of assumptions, has yet to emerge. The basis of most global impact assessments remain studies undertaken in developed countries (often the United States), which are then extrapolated to other regions. Such extrapolation is difficult and will be successful only if regional circumstances are carefully taken into account, including differences in geography, level of development, value systems, and adaptive capacity. Not all analyses are equally careful in undertaking this task, and not all models rely on the latest available information in calibrating their damage functions.

The actual functional relationships applied in many integrated models remain simple and often ad hoc. This reflects our still poor understanding of how impacts change over time and as a function of climate parameters. Impacts usually are a linear or exponential function of absolute temperature, calibrated around static "snapshot" estimates (such as 2xCO₂) without distinguishing the different dynamics that may govern impacts in different sectors. Developing a better understanding of these relationships is one of the most important challenges for integrated model development.

Baseline trends—such as economic development, population growth, technological progress, changes in values, natural climate fluctuations, and increased stress on natural ecosystems—have strong repercussions for climate change vulnerability (e.g., Mendelsohn and Neumann, 1999). They must be better understood and their effect incorporated in the models. Unfortunately, these trends are inherently difficult, if not impossible, to predict over the longer term. This generic problem will not go away, but it can be overcome, at least partly, through broad scenario and sensitivity analysis.

Another key challenge is taking adaptation into account. Adaptation can significantly reduce people's vulnerability to climate change, as shown in Chapter 18. However, adaptation can take many different forms and is correspondingly difficult to model (see Section 19.4). To date there are no IAMs available that can adequately represent or guide the full range of adaptation decisions.

IPCC Homepage