PRECIPITATION IN THE MEDITERRANEAN REGION OBSERVED WITH TRMM MICROWAVE DATA

Different satellite rain estimations based on microwave (MW), infrared (IR) or combined MW-IR techniques and contrasts them with the mesoscale Bologna local area model (BOLAM) rain analysis or the network based gauge data from the Global Precipitation Climate Centre (GPCC) for a period from 08 to 13 November 2001 over the western Mediterranean Sea during a severe weather event, which resulted in a disastrous flood in Algeria. The PR Adjusted TMI Estimation of Rainfall (PATER) and frequency difference algorithm (FDA) are applied to MW TRMM data, the neural rain estimator (NRE) uses geostationary IR Meteosat data and the combined NRL Turk algorithm uses both, MW data from low orbiting satellites and IR data from a geostationary orbit. The unique gridded data provide an effective basis to compare instantaneous space measurements with different algorithms. Validation results indicate that there is generally a better performance for heavy rain than for weak rain. Both MW algorithms, PATER and FDA, perform rather similarly whereas PATER is applicable exclusively over the ocean and shows some rain detection problems due to thick aerosol loads originating from the desert. The BOLAM model performs rather well in this case study, only a small location error of a heavy rain area was analyzed. The IR based techniques have the advantage of a high temporal revisit but both algorithms, NRL and NRE, have problems with identifying the correct rainy areas compared to MW results. Overall, the results suggest combining both advantages, the well-known rain physics of the MW channels with the high temporal resolution of IR algorithms, to retrieve precipitation operationally from satellite data