Search

This study seeks to explain the effects of cloud on changes in atmospheric radiative absorption that largely balance changes in global mean precipitation under climate change. The partial radiative perturbations (PRPs) due to changes in cloud and due to the effects of the pre-existing climatological cloud distribution on non-cloud changes, known as 'cloud masking', are calculated when atmospheric CO concentration is doubled for the HadSM3 and MIROC models and for a large ensemble of parameter perturbed models based on HadSM3. Because the effect of cloud on changes in atmospheric shortwave absorption is almost negligible, longwave fluxes are analysed alone. We find that the net effects of cloud masking and cloud PRP on atmospheric absorption are both substantial. For the tropics, our results are reviewed in light of hypotheses put forward to explain cloud and radiative flux changes. We find that the major effects of clouds on radiation change are linked to known physical processes that are quite consistently simulated by models. Cloud top height changes are quite well described by the fixed anvil temperature hypothesis of Hartmann and Larson; cloud base heights change little, remaining near the same pressure. Changes in cloud geographical location and cloud amount are significant, but play a smaller role in driving radiative flux changes. Finally, because clouds are a large source of modelling uncertainty, we consider whether resolving errors in cloud simulation could reconcile modelled global mean precipitation trends of about 1-3 % $$\hbox {K}^{-1}$$ with some estimates of observed trends of 7 % $$\hbox {K}^{-1}$$ or more. This would require the radiative effect of clouds to change from one that increases atmospheric radiative absorption by about $$0.5\,\hbox {Wm}^{-2}\,\hbox {K}^{-1}$$ to one that decreases it by $$-3.5\,\hbox {Wm}^{-2}\,\hbox {K}^{-1}$$ . Based on our results, this seems difficult to achieve within our current rationale for the tropics at least. [ABSTRACT FROM AUTHOR]