Precipitation estimation by a multi-threshold method using cloud optical and microphysical properties from MSG / SEVIRI data

Abstract

The aim of this paper is the estimation of precipitation in northern Algeria using satellite data. To do this, a multi-threshold method based on the exploitation of the clouds optical and microphysical properties is developed. Depending on the availability of the MSG / SEVIRI (M eteosat Second Generation/Spinning Enhanced Visible and Infrared Imaging) satellite channels, the database is divided into daytime data and nighttime data. In the learning phase, daytime and nighttime two-dimensional thresholds are determined from comparisons between the MSG / SEVIRI satellite data of rainy season 2010/2011 and the corresponding data from the Sétif meteorological radar. Using linear regression, an empirical relationship between the SEVIRI spectral data and the precipitation intensities from the radar is determined.To estimate rainfall, the determined empirical relationship is ap plied to the validation dataset collected during rainy season 2011/2012. To evaluate the method, a comparison of the estimation results with reference radar data is performed. The results show that these estimations are in good correlation with those measured by radar.