One approach to mitigation policy is to evaluate targets for reducing GHG emissions on the basis of reductions in vulnerability, rather than GHG concentrations or similar indirect measures of dangerous climate change. By applying existing methods for impact analysis, it is possible to invert the assessment procedure and start with defined sets or windows of impacts that are judged to be tolerable for humankind. This procedure results in emission corridors that embrace all future GHG emissions that are compatible with changes defined to be tolerablethe "safe landing" approach (WBGU, 1995; Alcamo and Kreileman, 1996; Petschel-Held et al., 1999). This approach can be extended to include economic, social, or equity aspectsthat is, to define tolerable windows for climate-related facets of these sectors and obtain emission corridors that simultaneously satisfy all possible windows (Toth et al., 1997).
Such approaches require that climate impacts should be differentiated between smooth changes and thresholds that mark abrupt shifts in the system's functioning. In the latter case, the definition of tolerable windows appears to be quite obvious: Damaging, abrupt shifts should be avoided. In the case of smooth changes, specification of tolerable windows is more difficult, confounded in part by uncertainty about adaptation.
Normative decisions on tolerable windows must be a consultative process involving scientists in close cooperation with stakeholders, decisionmakers, nongovernmental organizations (NGOs), and others. There are various designs for this participatory process, such as policy exercises (Toth, 1986, 1988a,b) or what is known as the Delft Process (van Daalen et al., 1998). Nevertheless, specification of windows remains somewhat arbitrary and preferably is used as an assumption in an "if-then" analysis rather than as an ultimate specification.
Methodological challenges include development and validation of reduced-form models, devising robust damage functions, identifying thresholds in adaptive systems, and concerns for equity in relating the distribution of impacts to systemic vulnerability.
Other reports in this collection