Mitigating the scatterometer-model wind biases in the Adriatic sea

Numerical Weather Prediction (NWP) models are operationally used for meteorological and oceanographic applications. Storm surge prediction models (SSMs), in particular, use the NWP surface wind and mean sea level pressure fields as forcing. It is well known that in basins surrounded by steep orography the surface wind fields supplied by NWP models are often not well represented [1]. In such cases the inaccuracies of the meteorological fields may propagate into the water level forecasts [2], making thus desirable a more faithful representation of the surface wind by NWP models. Accurate observations of the sea surface wind, performed routinely by satellite-borne scatterometers, are nowadays easily available with very short delays and can be used to improve the representation of the surface wind. This work aims to illustrate a methodology that brings the time and space dependent scatterometer-model bias into a tuning scheme of the wind forecast. The results of the study have been successfully used for SSM forecast applications, but are valid in general and can be applied equally well in other regional basins and coastal areas, as well as for other applications (ocean currents and waves).