Use of satellite data for the improvement of high resolution precipitation forecasts in the Mediterranean
Regional real-time numerical weather prediction is currently performed at a large number of sites worldwide using limited area models. Precipitation forecasts are considered to be among the most important model outputs as precipitation has direct impacts on human activities. In this work the potential of various methods of assimilation and nudging of observational data (mainly satellite products) on the improvement of precipitation forecast skill are investigated. Sixteen cases of heavy precipitation in the Eastern Mediterranean have been selected. The first method applied is a simple humidity adjustment technique that modifies the initial fields used by a limited area model using the near real-time precipitation estimates distributed by the National Aeronautic and Space Administration Goddard Space Flight Center. The second method is the application of a 3DVAR assimilation scheme that introduces both conventional measurements (i.e. surface observations and soundings) and satellite data from sensors on board SSM/I and QuikSCAT. The statistical results obtained through the application of these methods, mainly concerning the improvement o f the precipitation forecast skill, are presented. Finally an application of the first method for the simulation of a flood-producing storm (5 December 2002 in Antalya) is presented.
HyMeX – Hydrological cycle in the Mediterranean Experiment 2010-2020