A statistical approach for SeaWinds data assimilation

Scatterometer measurement potentially contain relevant information on the mesoscales. However, as the resolution increases both the noise and the ambiguity in retrieved winds increases, leading to arbitrary local minima wind solutions. Exploiting the full wind vector probability density function of the wind retrieval, and adopting spatial meteorological balance constraints in a 2D-Var alleviates the problem of arbitrary minima and noise and provides a spatially consistent scatterometer wind field. The method can de adopted in the context of 3D- or 4D-Var data assimilation systems.