A critical issue is the indicator of climate change against which we measure impacts. A common measure allows consistent discussion about the relationship between climate change and impacts. Several indicators could be used:
Several considerations must be taken into account in selecting an indicator. Using GHG emission levels (1) or even concentration levels (2) implies examining impacts beyond the 21st century. Published estimates of time frames for stabilizing GHG atmospheric concentration levels tend to assume such levels will not be stabilized until after the end of the 21st century (Enting et al., 1994; Wigley et al., 1996; Schimel et al., 1997).
The problem with using such levels as an indicator is that most of the impact literature examines potential impacts only as far as 2100. In addition, most studies are based on scenarios of specific changes in global mean or, more typically, regional climate variables such as temperature or precipitation.1 It is difficult to relate a specific level of GHG concentration to a specific change in global average climate or regional climate. For each GHG concentration level, there is a range of potential changes in global mean temperature (see Box 19-1). And for each change in global mean temperature, there is a range of potential changes in average regional temperature, precipitation, and extreme events.
The problem with indicators 3, 4, and 5 is the inverse of the foregoing problem. For each change in global or regional climate or extreme events, there is a range of levels of GHG concentrations that could cause such a change in climate. Thus, using these indicators makes it more difficult to work back to defining atmospheric concentrations of GHGs, as required by Article 2 of the UNFCCC. In addition, as one gets to finer levels of spatial and temporal resolution, such as changes in regional climates and extreme events, it becomes more difficult to attribute such changes to changes in GHG concentrations.
Thus, whatever the indicator selected, there will be problems in using it to relate impacts to the level of GHG concentrations. The choice of indicator depends on two factors:
We selected change in global mean temperature as our indicator for two reasons. The first is that the impact literature can be directly related to a change in global mean temperature. Many studies are based on specific results from general circulation models (GCMs), which estimate a change in global mean temperature. Other studies can be related to a change in global mean temperature by inversely using the scaling method from Chapter 4. The second reason is that, as discussed in Box 19-1, it is most feasible to relate changes in global mean temperature to GHG concentrations. It is harder to relate the other indicators directly to GHG concentrations. Thus, global mean temperature increase is the indicator that can be used most readily to relate GHG emissions (and emissions control) to changes in climate and impacts.
For any change in global mean temperature, there are many possible changes in regional climate and climate variability, which could have quite different results. Thus, a 2°C increase in global mean temperature may result in a particular region being much wetter or drier or having more or fewer extreme climate events. Whether the region gets wetter or drier or has more severe climate is likely to have much greater bearing on impacts than a change in mean temperature. Hence, although the use of global mean temperature as an indicator is preferable to the other options because it has fewer problems in implementation, it has its own limitations.
This chapter does not address the effect of different rates of change in climate on vulnerability. There is no doubt that a 3°C increase in global mean temperature realized in 50 years could be far worse than the same amount of warming realized in 100 or 200 years. In addition, changes in extreme events such as more intense El Niño-Southern Oscillation (ENSO) events (see, e.g., Timmermann et al., 1999) could lead to more adverse impacts than a monotonic and gradual change in climate. Thus, rate of change is an important factor affecting what climate change is considered to be dangerous. Unfortunately, most of the impact literature has addressed only static or equilibrium changes in climate. These studies have not examined what rates of change various sensitive systems can adapt to. Future research should address this matter.
Other reports in this collection