Applicability of satellite-based rainfall algorithms for estimating flood-related rainfall events in the mid-latitudes. Part II: temporal integration



Christina Görner, Technische Universität Dresden, Institut für Hydrologie und Meteorologie, Professur für Meteorologie, Pienner Str. 23, 01737 Tharandt, Germany

Tel: +49 351 463 31340

Fax: +49 35203 38 31302



In a continuation of Part I on spatial integration, this paper presents research on the applicability of four satellite-based rainfall algorithms to derive hourly rain intensities in the mid-latitudes using highly resolved data from the first Meteosat Second Generation (MSG-1) satellite. It focuses on the suitability of these algorithms to detect and monitor (flash) flood-related rain intensities for different temporal resolutions. The temporal integration is examined using six different integration steps ranging from 1 h to 24 h and three different data-filtering methods. The effects of this temporal integration on the quality of satellite-based rain intensities (amounts and locations) are examined with regard to heavy and possible flood-triggering rain intensities. Also, loss of temporal information and possible applications for flood risk management are considered. For this purpose, Saxony (Germany) is used as an example, and 9 high-rainfall days between May and August 2006 were analysed by comparing satellite rain intensities to rain gauge-adjusted radar data. The results show that temporal integration leads to improvements in evaluation scores, especially for heavy rain intensities, by up to +62%. However, this integration leads to a decreased applicability for the detection of high-intensity rainfall events because they tend to be short in duration.