An Empirical Model of Global Climate: Implications for Future Temperature

We use an empirical model of global climate based on a multiple linear regression analysis of the past global surface temperature to explore why climate models provide divergent estimates of future temperature. Our focus is on the interplay of three factors: net anthropogenic aerosol radiative forcing, climate feedback (water vapor, clouds, surface albedo) in response to greenhouse gas radiative forcing, and ocean heat export. We show that modeled future temperatures, for regressions constrained to provide acceptable fits to the 110 record of global surface temperature, mimic the behavior of climate models because of a cantilevering of aerosol radiative forcing and climate feedback. If anthropogenic aerosols presently exert small net global cooling, climate feedback must be weak and the future rise in global average surface temperature in 2053, the time CO2 is projected to double according to RCP 8.5, could be moderate. On the other hand, if aerosols presently exert large net global cooling, climate feedback must be large and future temperature when CO2 doubles could be substantial. In our model framework, warming at the time CO2 doubles is nearly independent of ocean heat export, because climate feedback must be adjusted for observed temperature to be properly simulated. If ocean heat export is small, modeled temperature at the time CO2 doubles is close to simulated equilibrium climate sensitivity. Conversely, if ocean heat export is large, temperature when CO2 doubles is much less than equilibrium climate sensitivity. Similar behavior likely occurs within climate models. We therefore suggest the dependence of equilibrium climate sensitivity on ocean heat export be factored into analyses that use this metric.