A study of surface winds in tropical storms using modified near real-time processing of QuikSCAT measurements

Performance of the NOAA/NESDIS near realtime QuikSCAT wind retrieval algorithm was assessed in the case of tropical cyclone IRIS, that occurred in Northern Atlantic, in October 2001. For this storm, the estimated wind field failed to produce a circular structure characteristic of tropical cyclones. We established that the Aviation Forecasting Field that was used to initialize ambiguity removal process was the primary cause of error, since it completely missed this feature. We proposed an initialization process that uses the information available from backscatter data to estimate possible regions of tropical cyclones. The proposed procedure allows estimation of wind fields that are less influenced by an external field. However, in the case of tropical storm IRIS the wind retrieval algorithm failed to generate solutions that would allow ambiguity removal to close circulation even when clustered errors in the initial field were broken by the new initialization procedure.