Optimal ambiguity removal for scatterometer winds

For many users of wind scatterometer data it is of interest to have an ambiguity removal system, or dealiasing, resolving the directional ambiguity in the observations. As a part of the ground segment, ESA performs such an ambiguity removal for the UWI wind product. The ambiguity removal method at meteorological centres such as DNMI, UKMO and KNMI is tuned and evaluated relying on a high quality first guess windfield (6 hrs forecast). The ESA operational scatterometer processing is however constrained to use a lower quality first guess wind field, that is an ECMWF 18 to 36 hrs forecast. Two algorithms for possible use in the ESA operational environment are evaluated against the currently operational ESA algorithm (CREO) "PRESCAT", which is the scheme in operation at several meteorological centers A new variational approach based an a Modified Successive Corrections (MSC) numerical method The operational experience with the PRESCAT scheme is that it generally works well, and only in a few cases it obviously picks the wrong solution. The more recent variational scheme is derived from Bayesian probability theory as finding the one of the wind possibilities giving rise to the windfield with highest probability. This enables us to set up a cost function to be minimized which takes into account the same features as in the PRESCAT scheme. In addition the new scheme is able to take advantage of the fact that the rotational part of the wind dominates over the divergent. It is concluded that a significant improvement in the operational ESA ambiguity removal can be obtained both by implementing the MSC or the PRESCAT processing, and the MSC scheme performs slightly better.