GCM experiments to test a proposed dynamical stabilizing mechanism in the climate system

A dynamical mechanism of sufficient strength to maintain the stability of the observed annual mean global climate against the destabilizing influence of the positive lower tropospheric water vapour/infrared radiative (WVIR) feedback on SST perturbations has recently been proposed (Bates, 1999). The mechanism consists of an evaporative negative feedback associated with the atmospheric angular momentum (AM) cycle and the surface winds which it induces. The mechanism was found from an analysis of a simple two‐zone model of the climate system. Numerical experiments with an atmospheric GCM aimed at testing the conclusions of the simple model are here described. Aquaplanet boundary conditions with a prescribed latitudinally‐varying SST distribution are adopted. An equinoctial distribution of solar radiation is assumed. An equilibrium climate for the GCM is established and perturbation experiments with a uniform 2 K increase in the SST are carried out. In the equilibrium and basic perturbation experiments, the GCM is run with all fields evolving freely. Further perturbation experiments are then carried out with the cloud fields held fixed, and with both the cloud and water vapour fields held fixed, in the GCM's radiation package. These experiments allow the relative influence of cloud and water vapour in the infrared and solar radiative feedbacks at the surface to be estimated. The results of the experiments are compared with those predicted by the simple model for a similar SST perturbation. The GCM results are found to be consistent with the basic assumptions and conclusions of the simple model. In particular, they indicate that a positive WVIR feedback exists in both the tropical and extratropical zones and confirm the existence of a negative evaporative feedback associated with the AM cycle that is of sufficient strength to overcome it. The experiments support the conclusion based on the simple model that it is necessary to consider both the wind factor and the humidity factor in evaluating the evaporative feedback. The cloud/infrared radiative feedback in the GCM is found to be small compared with the WVIR feedback. The experiments suggest that the feedback resulting from the absorbtion of solar radiation by water vapour (which was not included in the simple model) provides a significant additional stabilizing influence on SST perturbations.