F. Hugo Lambert

A tropospheric energy budget argument is used to analyze twentieth-century precipitation changes. It is found that global and ocean-mean general circulation model (GCM) precipitation changes can be understood as being due to the competing direct and surface-temperature-dependent effects of external climate forcings. In agreement with previous work, precipitation is found to respond more strongly to anthropogenic and volcanic sulfate aerosol and solar forcing than to greenhouse gas and black carbon aerosol forcing per unit temperature. This is due to the significant direct effects of greenhouse gas and black carbon forcing. Given that the relative importance of different forcings may change in the twenty-first century, the ratio of global precipitation change to global temperature change may be quite different. Differences in GCM twentieth- and twenty-first-century values are tractable via the energy budget framework in some, but not all, models. Changes in land-mean precipitation, on the other hand, cannot be understood at all with the method used here, even if land–ocean heat transfer is considered. In conclusion, the tropospheric energy budget is a useful concept for understanding the precipitation response to different forcings but it does not fully explain precipitation changes even in the global mean.

We investigate the causes of temperature dependent changes in global precipitation in contemporary General Circulation Models (GCMs) subjected to a doubling of atmospheric CO₂ concentration. By analyzing the energy budget of the troposphere, we find that changes are dominated by processes robustly simulated by GCMs. Importantly, shortwave cloud feedbacks, whose uncertainty is largely responsible for the wide range of GCM temperature climate sensitivities, are shown to have little effect. This is because these mainly arise from the scattering of shortwave radiation that has little impact on the tropospheric heating that controls precipitation. Hence, we expect that the range of simulated precipitation sensitivities to temperature will not change greatly in future GCMs, despite the recent suggestion that satellite observations indicate that GCM precipitation changes are significantly in error.