Assessment of the ASCAT wind error characteristics by global dropwindsonde observations

This research focuses on assessing the accuracy of the advanced scatterometer (ASCAT) wind vector around tropical cyclones by using data from dropwindsondes deployed by the surveillance and reconnaissance flights during 2007-2010. There are 987 matching samples for the comparisons, with wind speed up to 50ms(-1). The bias and root-mean-square differences of wind speed between ASCAT and dropwindsonde data are -1.7 and 5.3ms(-1), respectively. Further analyses also indicate that large wind direction differences occur in the low wind speed regime, while large wind speed differences occur in the high wind speed regime. The accuracy of wind vector in weak wind speed and high wind speed regimes are significantly reduced in saturated regions, implying that the rain contamination issue still affects the accuracy of ASCAT wind retrieval. The wind vector in the medium wind speed regime has much better quality, in good agreement with the satellite's designed specification. The ASCAT wind data obviously contain negative wind speed bias, which grows larger as wind speed increases. A regression fit between ASCAT and dropwindsonde wind speed is adopted to correct the bias of the ASCAT wind speed. The error characteristics are largely determined by the magnitude of wind speed and moisture saturation which are highly variable around the storm. Nevertheless, results from this study suggest that ASCAT wind data of velocities measuring between 12 and 18ms(-1) are more reliable and can be applied to determine the radius of 34knot winds, a critical parameter in operational tropical cyclone analysis.