COMPARISON OF BAROTROPIC AND BAROCLINIC NUMERICAL FORECASTS AND CONTRIBUTIONS OF VARIOUS EFFECTS

Abstract

500-mb. barotropic and 500-mb. baroclinic numerical forecasts for two cases, a developing baroclinic cyclone and a quasi-barotropic cyclone, are presented and compared. The barotropic forecasts did not indicate accurately the changes in circulation or the magnitude of the height falls ahead of the circulation maxima. 700-mb. forecasts from the same initial times as the 500-mb. barotropic and baroclinic forecasts, for each of the four terms of the frictionless vorticity equation, are presented. These 700-mb. forecasts are compared with each other and in added combinations with the 700-mb. verifications and with 700-mb. barotropic forecasts. These comparisons are then used diagnostically in an analysis of the errors in the 500-mb. barotropic forecasts. Each of the four terms of the vorticity equation is discussed. An explanation for the success of the barotropic forecasting model is suggested. Contributions of the horizontal velocity divergence, vertical advection of vorticity, and twisting terms to errors in the barotropic forecasting model are discussed in some detail. It is concluded that the major problem in developing a successful baroclinic forecasting model to substitute for the existing barotropic forecasting model is that of determining in space and time an accurate approximation of the vertical profile of vertical motion.