2.2.2 Advances in impact assessment
Application of the standard IPCC impact approach has expanded significantly since the TAR. The importance of providing a socio-economic and technological context for characterising future climate conditions has been emphasised, and scenarios assuming no climate policy to restrict greenhouse gas (GHG) emissions have been contrasted with those assuming GHG stabilisation (e.g., Parry et al., 2001; see also Sections 220.127.116.11 and 18.104.22.168). The use of probabilities in impact assessments, presented as proof-of-concept examples in the TAR (Mearns et al., 2001), is now more firmly established (see examples in Section 2.4.8). Some other notable advances in impact assessment include: a reassessment of bioclimatic niche-based modelling, meta-analyses summarising a range of assessments, and new dynamic methods of analysing economic damages. Nevertheless, the climate-sensitive resources of many regions and sectors, especially in developing countries, have not yet been subject to detailed impact assessments.
Recent observational evidence of climatic warming, along with the availability of digital species distribution maps and greatly extended computer power has emboldened a new generation of bioclimatic niche-based modellers to predict changes in species distribution and prevalence under a warming climate using correlative methods (e.g., Bakkenes et al., 2002; Thomas et al., 2004; see also Chapter 4, Section 4.4.11). However, the application of alternative statistical techniques to the same data sets has also exposed significant variations in model performance that have recently been the subject of intensive debate (Pearson and Dawson, 2003; Thuiller et al., 2004; Luoto et al., 2005; Araújo and Rahbek, 2006) and should promote a more cautious application of these models for projecting future biodiversity.
A global-scale, meta-analysis of a range of studies for different sectors was conducted by Hitz and Smith (2004) to evaluate the aggregate impacts at different levels of global mean temperature. For some sectors and regions, such as agriculture and the coastal zone, sufficient information was available to summarise aggregated sectoral impacts as a function of global warming. For other sectors, such as marine biodiversity and energy, limited information allowed only broad conclusions of low confidence.
Dynamic methods are superseding statistical methods in some economic assessments. Recent studies account, for example, for the role of world markets in influencing climate change impacts on global agriculture (Fischer et al., 2002), the effect on damage from sea-level rise when assuming optimal adaptation measures (Neumann et al., 2000; Nicholls and Tol, 2006), the added costs for adapting to high temperatures due to uncertainties in projected climate (Hallegatte et al., 2007), and increasing long-term costs of natural disasters when explicitly accounting for altered extreme event distributions (Hallegatte et al., 2006). The role of economic dynamics has also been emphasised (Fankhauser and Tol, 2005; Hallegatte, 2005; Hallegatte et al., 2006). Some new studies suggest damage overestimations by previous assessments, while others suggest underestimations, leading to the conclusion that uncertainty is likely to be larger than suggested by the range of previous estimates.