Search

The Adriatic Sea, a semi-enclosed sea extending about 900 km by 200 km and entirely surrounded by mountain chains, is subjected to high water levels produced essentially by the wind blowing along its major axis (sirocco). The operational storm surge models use the forecast fields produced by ECMWF atmospheric model as meteorological forcing. Recent works showed that in small basins surrounded by steep orography the ECMWF model underestimates the wind. This may introduce inaccuracies in forecasting the water level. Experiments conducted in the framework of the “Storm Surge for Venice” (eSurge-Venice) project, funded by the European Space Agency Data User Element program, were aimed to improve the operational forecasting of storm surge in the Northern Adriatic Sea by using satellite observations. The paper presents a methodology aimed to mitigate the ECMWF model underestimate of the sea surface wind by a tuning procedure based on the scatterometer-model wind speed and direction biases. The mitigation takes into account the spatial characteristics of the biases and is time dependent. This method globally reduces the model underestimate of the surface wind, handling the different areas of the basin differently, according to the spatial structure of the bias. The methodology is applied and verified over thirty-three test cases. It may be considered a tool to improve the present high water level forecast, waiting for the advent of convincing methodologies for scatterometer wind assimilation into the regional atmospheric models. [ABSTRACT FROM AUTHOR]