Inferring Ocean Surface Wind Speed and Direction from Multi-Directional Polarized Reflectance Observation from PARASOL-POLDER Sensor

Both wind speed and direction are important for remote sensing, atmospheric data assimilation and weather forecast. The A-train satellite community provides a number of remote sensing products, however, lacks of surface wind vector retrieval product. We use the multi-directional polarized reflectance observation from POLDER sensor onboard PARASOL to infer the surface wind vector. Breon and Henriot (2006) successfully inferred ocean surface wind vector from the intensity reflectance observation from POLDER 1. We found that including the polarized components of the reflectance further improves the retrieval. We apply our retrieval algorithm to the POLDER 1 sensor and compare the resulting wind vector with scaterometer product from NSCAT, and to the POLDER 3 sensor in the A-train community and compare with the ECMWF reanalysis product. With these comparisons, we show that our retrieval algorithm improves the wind vector retrieval and can be applied to provide surface wind vector product for the A-train community.