Seasonality in the Local and Remote Atmospheric Response to Sea Surface Temperature Anomalies

Two apparently contradictory situations are provided by observations of the atmospheric response to sea surface temperature anomalies. These are: (i) The extratropical regions of the winter hemisphere appear to possess strong teleconnections with equatorial forcing but weak or non-existent connections with local (extratropical) heating anomalies. (ii) The extratropical regions of the, summer hemisphere are quite sensitive to local thermal forcing but apparently unaffected by the remote forcing from equatorial regions. An attempt is made to provide a consistent physical picture which simultaneously embraces these two situations. With the aid of a simple linear model it is shown that the summer hemisphere is more sensitive to local forcing than the winter hemisphere because it is closer to the diabatic limit of Webster (1981), thus allowing an efficient energy generation. The winter hemisphere is much closer to the adjective limit. The sensitivity of the midlatitudes to remote (equatorial) forcing is shown to be a function of the relative location of the SSTA (sea surface temperature anomaly) to the zeros of the basic flow and the magnitude of the midlatitude westerlies. A hemisphere will become excited by remote forcing if at least part of the low-latitude sea surface temperature anomaly is located in the weak subtropical westerlies. Given that the latter criterion is met it is shown that the amplitude of the response and the latitude to which a particular mode is transmitted depends upon the distribution of westerly winds. The specific situation of El Niño sea surface temperature forcing is considered relative to realistic seasonal mean zonal wind fields. The model response is compared with the gross features of the observed anomalous atmosphere during El Niño years and a correspondence found. Finally, it is argued that the explanation of seasonality in atmospheric response offered in this paper will allow seasonal climate forecasting to be approached with an a priori physical expectation.