Wave Simulations Through Sar Cosmo-Skymed Wind Retrieval and Verification with Buoy Data

In this paper the use of SAR-derived wind fields is proposed to force surface wave models in a coastal environment, together with their validation with a valuable data set provided by both scatterometer and wave buoys data. The test area is the coastal zone of the Southern Tyrrhenian Sea, including the gulfs of Gaeta, Napoli, Salerno and Policastro, which is of great applicative relevance for both oceanographic and coastalmaritime surveillance purposes. The study can be summarized according to the three following stages. The first one consists of preliminary simulations of the wave fields, accomplished by means of both Wave Watch III (WWIII) and Weather Research and Forecasting (WRS) models, respectively, with forcing provided by ECMWF wind field model data. Once the relevant wave parameters have been evaluated on the calculation grid (i.e. significant wave height, Hs, wave direction, Dw, and peak period, Tp), the wave model validation has been performed with respect to some significant wave storms relevant to the summer and the winter seasons of 2010. The second stage of the study concerns the verification of the wave model numerical simulations with forcing provided by SAR-derived wind field estimations. Within such a framework, X-band Level 1B Detected Ground Multi-look (DGM) ScanSAR Huge Region mode VV-polarized COSMO-SkyMed© SAR measurements have been processed. A validation step has been accomplished to compare relevant wave parameters gathered by means of SAR-derived wind field estimations with the ones provided by both ASCAT scatterometer wind data and in situ wave buoy observations. The results were quite satisfactory in terms of the storm trend, less satisfactory in terms of the maximum Hs values. So the third stage of the study is about the development of some techniques to improve the model accuracy, in particular the use of blended data, i.e. ECMWF model data and COSMO-SkyMed© SAR-derived wind field retrievals. The use of blended data improves the model accuracy with respect to the COSMOSkyMed© data alone, although the best results have been obtained with ECMWF data,both in terms of Hs values and directions.