Observed Relationships between Large-Scale Tropical Convection and the Tropical Circulation on Subseasonal Time Scales during Northern Hemisphere Winter

Abstract

The work presented is a correlative study of the interaction between large-scale tropical convection and midlatitude wind anomalies, and the tropical wind field on subseasonal time scales. Outgoing longwave radiation (OLR) is used as a proxy for convection. Correlations are calculated from six years of 5-day averaged data for the December?February (DJF) season. The seasonal cycle and interannual variability are removed before computing the correlations. The results show that it is appropriate to classify the tropics into two regimes based on the direction of the time-mean 200-mb zonal wind at the equator. In the easterly regime there is abundant convective activity, a large standard deviation of OLR, and a small standard deviation of zonal wind. At longitudes of upper-level westerlies there is little convection, a small standard deviation of OLR, and a large standard deviation of wind. These results imply that convection is an important determinant of time-mean flow asymmetries along the equator. The small standard deviation of wind at longitudes of easterlies is interpreted as resulting from a shielding by the easterly winds from disturbances originating in midlatitudes. It is then shown that anomalies of OLR are better correlated with the anomalous wind field at longitudes of easterly winds than at longitudes of westerlies. This is interpreted to result from the larger forcing due to convection at longitudes of easterlies than westerlies, and from a lack of contamination of the signal by disturbances originating in midlatitudes. Finally, one-point correlation maps are presented for OLR anomalies at the equator at 130°E correlated with the near global field of nondivergent and divergent wind at 200 and 850 mb. These results are compared with the output from a linear model which was forced with an idealized heat source. At 200 mb the nondivergent anomalies do not resemble the linear model output except perhaps near the region of heating. It is suggested that the poor correspondence results from either nonlinearities in the real atmosphere or from the slow adjustment of the atmosphere to forcing. An interesting feature is the anomalously strong Asian jet associated with heating to its south. At 850 mb the correspondence with the model is better than at 200 mb. The divergent anomalies show an apparent westward tilt of the large-scale circulation pattern with height. Also interesting is the apparent convergence into the Asian jet entrance region during periods of convection, and the fact that divergent anomalies which are of opposite sign and precede those over Indonesia appear east of the dateline. There is some evidence that wavetrains originating in midlatitudes propagate to the equator at longitudes of westerlies over the Atlantic Ocean, but, surprisingly, not over the Pacific.