Scatterometer backscatter uncertainty due to wind variability

Wind retrieval from scatterometer backscatter measurements is not trivial. A good assessment of the different measurement uncertainties inherent in scatterometer systems is very important for successful wind retrieval and quality control. One source of these uncertainties, i.e., the geophysical noise, is dominated by the sub-cell wind variability. Although the latter is known to dominate the total measurement noise at low winds, no attempt to fully model such effect has yet been performed. In this paper, a simple method to derive a model of geophysical noise for the European remote sensing satellite (ERS) scatterometer is proposed. It is assumed that this noise is mainly due to the spatial distribution of the backscatter footprints and the wind variability within the wind vector cell. In a simulation experiment these parameters were varied and the values for which the simulation compares best to real data in the 3D measurement space were selected. The resulting geophysical noise model is dependent on wind speed and across sub-satellite track location. The empirical method presented here is straightforward and could be applied to other scatterometer systems.