Cloud tracking using satellite data for predicting the probability of heavy rainfall events in the Mediterranean area

Description

The use of cloud tracking techniques and storm identification procedures is proposed in this paper with the aim of predicting the evolution of cloud entities associated with the highest rainfall probability within a given meteorological scenario. Suitable algorithms for this kind of analysis are based on the processing of digital images in the thermal infrared (IR) band from geostationary satellites: a selection of such algorithms is described in some detail together with a few real case applications. Three heavy rainfall events have been selected for this purpose with reference to the extreme meteorological situation observed during Fall 1992 and 1993 over the Mediterranean area. A window from 30 to 60 degrees N and from 20 degrees W to 30 degrees E has been identified for the analysis of data from the radiometer on board the ESA Meteosat platform. In conclusion, the suitability of cloud tracking techniques for predicting the probability of heavy rainfall events is discussed provided that the former are associated with proper modeling of small scale rainfall distribution.

Additional details