ADGEO Advances in Geosciences ADGEO Adv. Geosci. 1680-7359 Copernicus Publications Göttingen, Germany 10.5194/adgeo-7-97-2006 Artificial neural-network technique for precipitation nowcasting from satellite imagery Rivolta G. 1 Marzano F. S. 2 1 Coppola E. 1 Verdecchia M. 1 Centro di Eccellenza CETEMPS, Università dell’Aquila, L’Aquila, Italy Dipartimento di Ingegneria Elettronica, Universit`a di Roma “La Sapienza", Rome, Italy 02 02 2006 7 97 103 Copyright: © 2006 G. Rivolta et al. 2006 This work is licensed under the Creative Commons Attribution-NonCommercial-ShareAlike 2.5 Generic License. To view a copy of this licence, visit https://creativecommons.org/licenses/by-nc-sa/2.5/ This article is available from https://adgeo.copernicus.org/articles/7/97/2006/adgeo-7-97-2006.html The full text article is available as a PDF file from https://adgeo.copernicus.org/articles/7/97/2006/adgeo-7-97-2006.pdf

The term nowcasting reflects the need of timely and accurate predictions of risky situations related to the development of severe meteorological events. In this work the objective is the very short term prediction of the rainfall field from geostationary satellite imagery entirely based on neural network approach. The very short-time prediction (or nowcasting) process consists of two steps: first, the infrared radiance field measured from geostationary satellite (Meteosat 7) is projected ahead in time (30 min or 1 h); secondly, the projected radiances are used to estimate the rainfall field by means of a calibrated microwave-based combined algorithm. The methodology is discussed and its accuracy is quantified by means of error indicators. An application to a satellite observation of a rainfall event over Central Italy is finally shown and evaluated.