Seasonal Differences in the Stationary Response of a Linearized Primitive Equation Model: Prospects for Long-Range Weather Forecasting?

Abstract

A linear steady-state primitive equation model has been developed for the computation of stationary atmospheric waves that are forced by anomalies in surface conditions. The model has two levels in the vertical. In the zonal direction the variables are represented by Fourier series, while in the meridional direction a grid-point representation is used. The equations governing atmospheric motion are linearized around a zonally symmetric state which depends on latitude and height according to Oort (1980). We have studied the amplitude and phase relations of the model response as a function of latitude for a very simple beating, which is sinusoidal in the zonal direction, with zonal wavenumber m (m=1, 10) and constant in the meridional direction, using February mean conditions. The response of the model indicates that a heating in the tropics can have a substantial influence on the middle and high latitudes, provided that part of the heating is in the westerlies. We have compared the model response for such a heating with the results of similar experiments with GCM and a linear barotropic model and also with mean anomaly patterns at middle and high latitudes derived from observations for Northern Hemispheric winters with a warm equatorial Pacific. In all cases we find strong similarities of hemispheric wave patterns. We plan to test the model for the prediction of that part of the anomalies in the monthly or seasonal mean circulation that comes from persistent abnormal surface conditions In order to predict more than a persistent atmospheric response, such an anomaly in the surface conditions must have different effects in different months or seasons. We have tested the hypothesis that due to a changing zonally symmetric state, the response to a prescribed beating will be different in the four seasons. This effect is computed for a heating in the tropics and in the middle latitudes. Both in amplitude and phase the response to exactly the same heating can change significantly from one season to the next.