Inferring the Diabatic Heat and Moisture Forcing of the Atmosphere from Assimilated Data

Abstract

Different estimates of the global-scale diabatic sources and sinks of atmospheric heat and moisture, obtained from a data assimilation system of the ECMWF, are examined. Estimates arrived at using the budget method are compared with those given by the center's forecast model during very short range integrations. Integrations and initialized analyses for February 1979, produced in the mid-1980s in a reassimilation of the final FGGE lIb dataset, are used as a database. The results are specific to the particular assimilation system used in the reassimilation. All estimates yield qualitatively equivalent and meteorologically reasonable vertical integrals. For the moisture forcing, the vertical distributions are also similar. This is not true for the diabatic heating. In the Tropics, the model's physics gives strong cooling in the lower troposphere, where the budget method implies heating. This unrealistic cooling is present even in regions of intense net condensation. Quantitatively, both the heat and moisture forcing given by the parameterized physics of the model differ significantly from the budget residuals. Model-generated heating and moistening from 2-hour integrations and 6-hour integrations also exhibit differences. The principal source of these differences is probably the presence of systematic tendencies of the model variables during the short-range integrations.