Global Heating Distributions for January 1979 Calculated from GLA Assimilated and Simulated Model-based Datasets

Abstract

Global distributions of atmospheric heating for January 1979 are estimated from two Global Weather Experiment (GWE) Level III datasets generated at the Goddard Laboratory for Atmospheres (GLA). One set utilized data from the full GWE observing system (to be denoted SAT), while the other excluded information either measured or transmitted by satellite (to be denoted NOSAT). These two distributions of heating are compared with the ones predicted by the forecast model (MODEL) during the above SAT and NOSAT GWE assimilations and another one predicted during a wintertime climate simulation (CLIMATE) of the GLA GCM. Through intercomparison of the five distributions, this study with an emphasis on satellite-derived information investigates the global distribution of atmospheric heating and the impact of observations on the diagnostic estimates of heating derived from assimilated datasets. The spatial patterns of heat sources and sinks north of 40°S estimated from the SAT and NOSAT datasets are similar and are consistent with physical and climatological considerations and in general agreement with estimates derived from other GWE data. South of 40°S the relative accuracy of the distributions is uncertain. Substantial differences between the two estimates occur in tropical oceanic regions of deep convection and over the central North Pacific. Over the North Pacific the SAT results depict a mid-latitude oceanic storm track extending from the coast of Asia to the east of the dateline, while the NOSAT heating is confined to the western Pacific. In tropical-subtropical oceanic regions of deep convection, differences occur in the intensity of heating and in the disposition of heating maxima. Overall, the results indicate a substantial impact of satellite information on diagnostic estimates of heating in regions where there is a paucity of conventional observations. Although there are uncertainties, the addition of satellite data provides information on the atmosphere's wind and temperature structure that is important for estimation of the global distribution of heating and energy exchange. The comparison of the SAT and NOSAT diagnostic and corresponding MODEL distributions indicates that in Northern Hemisphere middle latitudes the assimilation of observed data substantially impacted the diagnostic estimates of heating. In tropical latitudes, these comparisons imply a larger influence of the assimilation model's predicted heating on the resulting diagnostic estimates in the NOSAT than in the SAT assimilation. The substantial departure of the NOSAT and SAT MODEL distributions from the climatological heating of the GLA model (CLIMATE) is an indication of the effect of observations on the ensemble of forecasts during the GWE assimilations and provides additional documentation of the impact of observed data on the diagnosed heating distributions.