On the linear theory of the atmospheric response to sea surface temperature anomalies.

Abstract

A linear, hemispheric, two-level model which is based on the primitive equations is used to study the steady-state response of the atmosphere to sea surface temperature anomalies ( SSTA ). First it is shown by comparison with analytical theories that the model's vertical resolution is not too coarse for such a study. Then the model's response to an idealised heat source in middle latitudes is examined, and it is demonstrated that the main features of the response can be explained in terms of a simple quasi-geostrophic theory. The model is applied to SSTA cases off Newfoundland where statistically significant observations are available. The verification of the model's results against these observations is generally favourable downstream of the SSTA, i.e., the observed low (high) at the surface downstream of a warm (cold) pool is simulated. However, the linear theory predicts a weak surface high upstream of a positive SSTA which has no statistically significant counterpart in the observations. The model is able to reproduce approximately the blocking situation over Europe during the period 1958-60 when the observed SSTA in the Atlantic is prescribed. Realistic pressure and flow patterns in the tropics are obtained for an SSTA in the eastern equatorial Pacific and for an SSTA in the tropical Atlantic. In the first case the influence of SSTA's in the tropics on the mid-latitude circulation turns out to be negligible. This result is not supported by the observations. The SSTA in the Atlantic, however, causes a rather strong response at midlatitudes which is similar to that found in experiments with general circulations models where the same SSTA has been prescribed. The influence of the tropical mean basic wind field and the grid spacing on the midlatitude response is discussed. Finally, it is shown that the linear model must fail when the SSTA is the result of atmospheric circulation anomalies. (A)