Diagnosis of cirrus cloud occurrence using large-scale analysis data and a cloud-scale model

Abstract

The development of cirrus clouds is governed by large-scale synoptic movements such as updraft regions in convergence zones, but also by smaller scale features, for instance microphysical phenomena, entrainment, small-scale turbulence and radiative field, fall-out of the ice phase or wind shear. For this reason, the proper handling of cirrus life cycles is not an easy task using a large-scale model alone. We present some results from a small-scale cirrus cloud model initialized by ECMWF first-guess data, which prove more convenient for this task than the analyzed ones. This model is Starr’s 2-D cirrus cloud model, where the rate of ice production/destruction is parametrized from environmental data. Comparison with satellite and local observations during the ICE89 experiment (North Sea) shows that such an efficient model using large-scale data as input provides a reasonable diagnosis of cirrus occurrence in a given meteorological field. The main driving features are the updraft provided by the large-scale model, which enhances or inhibits the cloud development according to its sign, and the water vapour availability. The cloud fields retrieved are compared to satellite imagery. Finally, the use of a small-scale model in large-scale numerical studies is examined.