This paper presents a method to detect precipitation and estimate rain rates using cloud physical properties retrievals from the Spinning Enhanced Visible and Infrared Imager (SEVIRI). The method calculates rain rates from cloud liquid water path (LWP), particle effective radius, cloud thermodynamic phase and cloud top height retrievals that are based on fundamental radiative transfer simulations. The accuracy of the precipitation detection and rain rate retrievals from SEVIRI is evaluated with Weather Radar observations.
For a two month period rain rates from SEVIRI are compared against Weather Radar observations for an area over Northern Europe. The Weather Radar observations are used to validate the instantaneous rain rate retrievals and accumulated rainfall sums (precipitation depths) across the entire study area and period. In addition, we evaluate the ability of SEVIRI to discriminate precipitating from non-precipitating clouds. The results show very high agreement (corr. ~0.90) between amounts of precipitating clouds detected from Weather Radar and SEVIRI observations. Although weaker correlations (corr. ~0.63) are found between the rain rate retrievals from Weather Radar and SEVIRI, the SEVIRI-retrievals still have an acceptable accuracy of about 0.2 mm hr-1 and a precision of about 0.7 mm hr-1. Part of the differences between Weather Radar and SEVIRI are explained by irregularities in the Weather Radar data due to residual sea clutter, and parallax shifts in the SEVIRI data.
In conclusion, the results of this study have shown the potential of SEVIRI retrieved cloud physical properties for the detection of precipitation and the retrieval of realistic rain rates. In future studies we intend to exploit the observations of the European Weather Radar network (OPERA) and repeat this study to entire Europe
R Roebeling, I Holleman. Validation of Rain Rate Retrievals from SEVIRI using Weather Radar Observations
Conference: EUMETSAT Meteorological Satellite Conference, Organisation: EUMETSAT, Place: Darmstadt, Year: 2008, First page: 0, Last page: 0