Toward an Improved Wind Inversion Algorithm for RapidScat

The sea-surface winds from the RapidScat scatterometer (RSCAT) onboard the International Space Station (ISS) have been produced using the Pencil-beam scatterometer Wind Processor (PenWP) since December 2014. An inversion residual or Maximum Likelihood Estimator (MLE-) based quality control (QC) algorithm is included in PenWP to distinguish between goodand poor-quality winds. Generally, the QC-accepted RSCAT winds are in good agreement with both the Advanced Scatterometer (ASCAT) winds and the European Centre for Medium-range Weather Forecasting (ECMWF) model output. In contrast, the QC-rejected winds present an overall positive bias with respect to ASCAT and ECMWF winds, mainly due to the impact of rain. However, it has been recently found that a considerable portion (>5%) of RSCAT QC-rejected contains anomalously low retrieved speeds (ω <; 4 m/s) under medium or high wind conditions (ω > 10 m/s) according to ASCAT/ECMWF. This paper attempts to sort out the cause for these spuriously low winds. A revised MLE inversion with prefiltering the anomalous backscatters is proposed to correct the mentioned inversion issue. The impact of such improved inversion on the retrieved RSCAT winds is evaluated using both the collocated ASCAT and ECMWF winds. The results show that the proposed algorithm improves the wind retrieval of the spuriously low wind cases remarkably, while preserving about 4.7% of the nominal QC-rejected data (0.25% in total).