In autumn deep convection in the Mediterranean region is a commonphenomenon. The local events characterized by deep convection arestill a difficult task even for high resolution numerical weatherprediction. Three flood cases, produced by convection eitherembedded in a large scale system or locally developed, occurringin Italy, are presented. All these case were not correctlyforecasted: Sardinia (Cagliari, 13 November 1999); Calabria (Soverato,7 September 2000) and Sicily (Catania, 16 September 2003). The first caseoccurred during the Mesoscale Alpine Programme (MAP) campaign,therefore a lot of data are available; for the second one onlydata from SSM/I and local rain-gauge are available; the third oneoccurred during the operational experimentation of the TOUGHproject. The last one was not well predicted even using theoperational assimilation of ground based GPS. To improve theforecast of these cases the assimilation of several data istested. The variational assimilation performed using 3DVAR of GPS,SSM/I and surface and upper air data is applied to improve theInitial Conditions of the Sicily case. The Sardinia case isimproved using either GPS and surface data, whereas for theSoverato case only ZTD is assimilated. The experiments areperformed using the MM5 model from Pennsylvania StateUniversity/National Center for Atmospheric Research (PSU/NCAR);the model is initialized using the new Initial Conditions producedby the variational assimilation of conventional and nonconventional data. The results show that the assimilation of theretrieved quantities does produces large improvement in theprecipitation forecast. Large sensitivity to the assimilation ofsurface data and brightness temperature from SSM/I is found.
展开▼
