On the general circulation of the atmosphere without clouds

An experiment has been performed with an 18‐level, hemispheric, stereographic general circulation model to clarify the radiative role of clouds in the general circulation of the atmosphere. All cloud properties were removed from the radiation scheme in the model, and it was then integrated to generate a new set of quasi‐equilibrium atmospheric statistics which were compared with a control run. Apart from an expected warming of the surface and troposphere the overall changes to the general circulation characteristics following the removal of clouds were surprisingly small. For example, the zonal mean latitude‐height distributions of the zonal wind and eddy kinetic energy were almost identical to those of the control run, while synoptic distributions of most model variables revealed little change in character attributable to the removal of cloud. The largest modification occurred in the mean meridional stream function, with corresponding effects in surface pressure. This result was explicable in terms of the enhanced role of latent heat in the model which occurred at the expense of sensible heat. Despite the individual variations in latent and sensible heat, the poleward flux of total heat by large‐scale eddies was essentially unchanged in the two models. This surprising result indicated that the forcing function for the general circulations of the two models was also unchanged, and moreover that this forcing function was independent of cloud cover. Where possible, explanations are advanced to account for the limited response of the model to the removal of cloud. It appears that the major function of clouds in radiative calculations, at least in extended general circulation simulations, is to provide an obscuration or blocking effect, rather than a more subtle radiative‐dynamical coupling mechanism.