In the foregoing sections, impact categories have been treated separately. Obviously, however, the impacts of climate change interact with one another-through the use of resources such as water and land and through economy, society, and politics. An integrated assessment tries to take these interactions into account. In addition, an integrated assessment tries to compare the diversity of impacts with respect to their seriousness (from a human viewpoint) and evaluate overall vulnerability to climate change.
Three approaches to integrated assessment have been applied to Europe: monetization, integrated modeling, and coupled models and expert panels. Each method has its advantages and disadvantages, which are discussed at length in Weyant et al. (IPCC 1996, WG III, Chapter 10). The focus here is on the results. Note, however, that integrated assessment is a young field, without a great number of well-established methodologies, let alone results.
One way of integrating the range of potential impacts of climate change is to derive a comprehensive monetary estimate, which adds all impacts expressed in their dollar value. This approach allows for comparison of the seriousness of climate change with other problems, comparison of vulnerabilities to climate change among regions and sectors, and comparison of the impact of climate change with the impact of greenhouse gas emission reduction. Expressing effects on marketed goods and services (e.g., land loss resulting from sea-level rise, energy savings in winter) in monetary terms is relatively straightforward because the price is known. Expressing damage to nonmarketed goods and services (e.g., wetland loss, mortality changes) in monetary terms can be accomplished by examining market transactions where such goods or services are implicitly traded (e.g., landscape beauty) or by interviewing people about their preferences. That is, human preferences are expressed by people's willingness to pay to secure a benefit or their willingness to accept compensation for a loss. In western Europe, valuation techniques are well established and widely applied. Numerous theoretical and empirical problems remain, however. Based on the SAR (IPCC 1996, WG II, Chapter 6), Fankhauser and Tol (1997) report best estimates for the annual impact resulting from a doubling of atmospheric concentrations of carbon dioxide of about -1.6% to -1.4% of the GDP in western Europe, using a mix of earlier GCM scenarios standardized to a 2.5°C increase in the global mean temperature. Best estimates for eastern Europe and the former Soviet Union vary between -0.4% and +0.4% (a benefit) of GDP. These figures compare with an estimated world impact of -1.8% to -1.2% of GDP. National differences may be hidden by the regional average, but only one country-specific study has been carried out to date. Using similar methods and scenarios as Fankhauser and Tol (1997), Kuoppomaeki (1996a, b) concludes that Finland may gain about 1% of GDP. Many assumptions underlie these best guesses; the uncertainties are large, yet unknown.
An integrated assessment model combines climate change scenarios with models of various impacts of climate change in a single computer code (i.e., hard-linking of models). One advantage of this approach is that consistency is ensured (e.g., land or water used for agriculture is not used in the domestic sector). A disadvantage is that modeling and computational requirements entail simplifications of the state-of-the-art. Only two integrated assessment models are even somewhat useful for climate change impact assessment in Europe; the other twenty-odd models reviewed in IPCC (1996, WG III, Chapter 10) have other aims or are based on monetization. The first integrated assessment model is the ESCAPE model (Rotmans et al., 1994), which was developed specifically for the EU. It includes a variety of impacts on natural and human systems; since its completion in 1991, however, the model has not been updated. The second such model is the IMAGE2 model (Alcamo, 1994). This model is up-to-date. It models the whole world, with a focus on land use. The resolution is 0.5°x0.5°, with east and west Europe grouped into two economic regions. The socioeconomic modeling of agriculture and human infrastructure-the dominant land-use tupes in Europe-requires further work.
Other reports in this collection