Results and future perspectives of the application of coupled atmosphere-ocean-waves numerical models for the analysis and prediction of extreme met-ocean events over Mediterranean basin

In a complex Global Warming context, the Mediterranean Sea shows significant anomalies in Sea Surface Temperature (SST), which can potentially influence the intensity and frequency of extreme weather events. The most modern forecasting techniques based on the use of numerical models focus mainly on the implementation of the most realistic possible initial conditions, often neglecting the quality of the SST data, both because of the physical complexity of the air-sea interface processes and for the scarcity of synoptic data with high spatial and temporal resolution. In this work we have investigated some typical atmospheric events of the Mediterranean basin, using numerical atmospheric models with different coupling configurations with the ocean and waves, in order to examine and identify both physical modeling problems that air-sea interaction which they can strongly influence the final event forecast. Briefly recalling the selected case studies concern the Cold air outbreak event of 2012 with a focus on the North Adriatic, the Tropical-Like-Cyclone (TLC) "Rolf" formed on the Balearic Islands in 2011, and the TLC "Ilona" formed over Gibraltar in January 19th 2014, the work presents an example with more detail of flash-floods formed on the Venice lagoon on September 26th 2007. The general results show that the importance of the use of coupled models is fundamental for forecasting purposes, in particular in the case of events caused by strong air-sea interactions or with dynamics driven by intense heat and vapor fluxes. The need for an adequate description of the SST and wave fields, coupled with the atmosphere, is evident especially in the case of coastal atmospheric phenomena, in which the SST acquired by satellite, normally used in operational models, are particularly prone to errors. In fact, the usability kind of data is strongly affected by cloudiness, the presence of sediments dissolved in the sea, and the low resolution of the data itself, particularly in a context characterized by strong small-scale gradients.