2.2.5 Advances in integrated assessment

Integrated assessment represents complex interactions across spatial and temporal scales, processes and activities. Integrated assessments can involve one or more mathematical models, but may also represent an integrated process of assessment, linking different disciplines and groups of people. Managing uncertainty in integrated assessments can utilise models ranging from simple models linking large-scale processes, through models of intermediate complexity, to the complex, physically explicit representation of Earth systems. This structure is characterised by trade-offs between realism and flexibility, where simple models are more flexible but less detailed, and complex models offer more detail and a greater range of output. No single theory describes and explains dynamic behaviour across scales in socio-economic and ecological systems (Rotmans and Rothman, 2003), nor can a single model represent all the interactions within a single entity, or provide responses to questions in a rapid turn-around time (Schellnhuber et al., 2004). Therefore, integration at different scales and across scales is required in order to comprehensively assess CCIAV. Some specific advances are outlined here; integration to assess climate policy benefits is considered in Section 2.2.6.

Cross-sectoral integration is required for purposes such as national assessments, analysis of economic and trade effects, and joint population and climate studies. National assessments can utilise nationally integrated models (e.g., Izaurralde et al., 2003; Rosenberg et al., 2003; Hurd et al., 2004), or can synthesise a number of disparate studies for policy-makers (e.g., West and Gawith, 2005). Markets and trade can have significant effects on outcomes. For example, a study assessing the global impacts of climate change on forests and forest products showed that trade can affect efforts to stabilise atmospheric carbon dioxide (CO₂) and also affected regional welfare, with adverse effects on those regions with high production costs (Perez-Garcia et al., 2002). New economic assessments of aggregated climate change damages have also been produced for multiple sectors (Tol, 2002a, b; Mendelsohn and Williams, 2004; Nordhaus, 2006). These have highlighted potentially large regional disparities in vulnerability to impacts. Using an integrated assessment general equilibrium model, Kemfert (2002) found that interactions between sectors acted to amplify the global costs of climate change, compared with single-sector analysis.

Integration yields results that cannot be produced in isolation. For example, the Millennium Ecosystem Assessment assessed the impact of a broad range of stresses on ecosystem services, of which climate change was only one (Millennium Ecosystem Assessment, 2005). Linked impact and vulnerability assessments can also benefit from a multiple stressors approach. For instance, the AIR-CLIM Project integrated climate and air pollution impacts in Europe between 1995 and 2100, concluding that that while the physical impacts were weakly coupled, the costs of air pollution and climate change were strongly coupled. The indirect effects of climate policies stimulated cost reductions in air pollution control of more than 50% (Alcamo et al., 2002). Some of the joint effects of extreme weather and air pollution events on human health are described in Chapter 8, Section 8.2.6.

Earth system models of intermediate complexity that link the atmosphere, oceans, cryosphere, land system, and biosphere are being developed to assess impacts (particularly global-scale, singular events that may be considered dangerous) within a risk and vulnerability framework (Rial et al., 2004; see also Section 2.4.7). Global climate models are also moving towards a more complete representation of the Earth system. Recent simulations integrating the atmosphere with the biosphere via a complete carbon cycle show the potential of the Amazon rainforest to suffer dieback (Cox et al., 2004), leading to a positive feedback that decreases the carbon sink and increases atmospheric CO₂ concentrations (Friedlingstein et al., 2006; Denman et al., 2007).