A Simple Cloud Parameterization Derived from Cloud Resolving Model Data: Diagnostic and Prognostic Applications

Publisher Website
Google Scholar
Abstract
A simple statistical parameterization of cloud water–related variables that has been originally developed for nonprecipitating boundary layer clouds is extended for all cloud types including deep precipitating convection. Based on three-dimensional cloud resolving model (CRM) simulations of observed tropical maritime and continental midlatitude convective periods, expressions for the partial cloudiness and the cloud water content are derived, which are a function of the normalized saturation deficit Q1. It turns out that these relations are equivalent to boundary layer cloud relations described earlier, therefore allowing for a general description of subgrid-scale clouds. The usefulness of the cloud relations is assessed by applying them diagnostically and prognostically in a mesoscale model for a midlatitude cyclone case and a subtropical case, and comparing the simulated cloud fields to satellite observations and to reference simulations with an explicit microphysical scheme. The comparison uses a model-to-satellite approach where synthetic radiances are computed from the meteorological fields and are compared to Meteosat satellite observations both in the visible and the thermal infrared spectral channels. The impact of the statistical cloud scheme is most pronounced for shallow and deep convective cloud fields (where Q1 < 0), provided that the host models convection parameterization is able to correctly represent the ensemble average water vapor profile in the troposphere. The scheme significantly reduces the biases in the infrared and especially shortwave spectral range with respect to the explicit microphysical scheme. Furthermore, it produces more realistic (smooth) horizontal and vertical condensate distributions in both diagnostic or prognostic applications showing the potential use of this simple parameterization in larger-scale models.

This publication has 26 references indexed in Scilit: