A multidimensional histogram rain-flagging technique for SeaWinds on QuikSCAT

摘要

The SeaWinds scatterometer was developed by NASA JPL to measure the speed and direction of global ocean surface winds. SeaWinds was launched aboard the QuikSCAT spacecraft on June 19, 1999 and has continued to operate successfully since being turned on. Although the initial SeaWinds wind vector products were of excellent quality, they were occasionally degraded by the presence of rain. It soon became obvious that a way to flag wind vector cells for rain contamination was needed. The authors have determined a set of parameters that are sensitive to rain and are computed from the scatterometer measurements. These parameters are: (1) the retrieved wind speed, (2) the retrieved wind direction relative to the satellite ground track, (3) the normalized beam difference, which indicates a statistically significant imbalance in the beams relative to the geophysical model function, (4) the maximum likelihood estimate calculated by the wind retrieval algorithm, and (5) the radiometric brightness temperature. Using these parameters and external rain information (SSM/I) an estimate of the conditional probability of rain given the parameters is developed using a multidimensional histogram technique. This probability estimate is then used to flag rain contaminated wind vector cells using only scatterometer data. This technique is currently employed to generate a rain flag for SeaWinds on QuikSCAT data. In this paper, the effects of rain on SeaWinds data are explored, the rain flagging technique is explained, and the performance of the rain flag is illustrated using a number of metrics.