Scatterometer Observations at High Wind Speeds

Abstract

Satellite scatterometer winds are commonly validated by comparing them to buoy observations and/or numerical model surface wind analyses. However, the empirical scatterometer algorithm (geophysical model function) has been calibrated against a subset of the buoy and, most likely, numerical model winds. Thus, this approach is only useful for evaluating the robustness of the empirical fit of the scatterometer algorithm to the buoy and/or numerical model winds. It cannot detect the errors in the scatterometer winds due to inherent inaccuracy of the buoy and numerical model winds. Using an alternate approach based on surface pressure measurements, the authors demonstrate the existence of a low bias at high wind speeds (greater than 20 m s?1) in the buoy and numerical model winds that is consistent with previous theoretical results. To assess to what extent this bias is carried over to the European Remote Sensing (ERS) satellite scatterometer winds, a validation method based on the University of Washington planetary boundary layer model is introduced. This method quantifies the accuracy of scatterometer winds using sea level pressure observations. Buoy pressure observations are seen to be more robust than the winds in high wind conditions since they are less affected by high sea states, wave sheltering, buoy motion, and other effects. An empirical correction for the bias in ERS scatterometer winds is provided.