A split-step IR-advection/model-convection approach to fill temporal gaps in the microwave remote sensing of precipitation

摘要

One of the most important problems in non-convection-resolving atmospheric circulation models is cumulus parametrization, i.e. the problem of determining, from the model's coarse-scale state variables, the triggering, vertical distribution and time scale of convective kinetic energy dissipation. One way to derive an empirically-verified parametrization scheme is to use intensive four-dimensional observations of tropical precipitation to compile the statistics of the joint behavior of the convective available potential energy (CAPE) on one hand and the resulting four-dimensional latent heating on the other hand. While tracking CAPE requires frequent soundings, tracking the latent heating is much harder because it requires frequent estimates of the vertical structure of precipitation: the latter can only be estimated from microwave remote sensing, which is only available in the form of infrequent (at best six-hourly) snapshots from low-Earth-orbiting satellites. The approach proposed in this paper seeks to remedy this problem by combining geo-stationary IR observations with a simplified vertical evolution model to fill-in the vast gaps in the microwave observations. The results will help in developing empirically verified convective parametrization schemes for use in large-scale atmospheric models, and in producing fine temporal-scale precipitation estimates which can be directly assimilated.