Key Uncertainties in Setting Goals for Limiting Climate Change
The report ACC: Advancing the Science of Climate Change (NRC, 2010a) provides a detailed discussion of the scientific uncertainties involved in setting targets for limiting climate change. Here we provide a brief overview of some key uncertainties that affect the results presented later in this chapter.
Climate sensitivity. The quantitative relationship between long-term temperature changes and atmospheric GHG concentrations is very difficult to specify, due primarily to the large uncertainty of climate sensitivity. Climate sensitivity is typically defined as the global mean equilibrium temperature response to a doubling of CO2 concentrations. IPCC (2007b) indicated that climate sensitivity is likely to be in the range of 2°C to 4.5°C with a best estimate of about 3°C. It is very unlikely to be less than 1.5°C,1 and values substantially higher than 4.5°C cannot be excluded. Some recent studies have in fact suggested that much higher climate sensitivity values are possible (Hansen et al., 2008; Sokolov et al., 2009). This uncertainty indicates that relying only on the best estimate of 3°C may not be a prudent risk-management strategy.
Because of these uncertainties, the temperature-concentration relationship is often given in probabilistic terms. As noted earlier, recent research (Meehl and Stocker, 2007; Wigley et al., 2009) indicates that limiting global GHG concentrations to around 450 ppm CO2-eq over the long term would result in a 2°C temperature change using a climate sensitivity of 3°C. The best-estimate increase in temperature for a long-term concentration of 550 ppm CO2-eq is 3°C (or 2°C, if the climate sensitivity is at the lowest end of the IPCC range). However, there is significant uncertainty around these point estimates. For example, a study using three models from the EMF22 exercise finds that the probability of staying below 2°C for several “overshoot” scenarios (see below) leading to 450 ppmv CO2-eq ranges between 24 and 72 percent, depending on the degree of overshoot and on which probability distribution from the literature is used (Krey and Riahi, 2009).
A related consideration is that, due to time lags in the climate system, one might allow actual concentrations to temporarily and modestly exceed (or “overshoot”) 450 ppm CO2-eq while allowing temperature to remain below the 2°C temperature goal. However, an overshoot scenario entails ad
ing the century, and with delays in full global participation. For instance, only 2 of the 14 participating models8 were able to produce scenarios that attained the 450 CO2-eq goal without immediate, full global participation, and only then if an overshoot trajectory to the goal was allowed. Without the option to overshoot the goal, and with