Wind speed and sea ice motion retrieval in the marginal ice zone from RADARSAT-2 HH-HV images

We present a new model for wind speed retrieval from HH-HV RADARSAT-2 dual-polarization imagery which have been developed, based on a large body of statistics on buoy observations collocated and coincided with SAR images. The model's predictors are co-polarization (HH), cross-polarization (HV) radar backscatter, instrument noise floor, and incidence angle. We found that the proposed models without wind direction input demonstrated a significantly better accuracy compared to the CMOD5.N model, which requires wind direction input. Furthermore, we showed that the developed wind speed retrieval models, in conjunction with our previously developed ice motion tracking algorithm, can be a useful tool for studying sea ice dynamics processes in the Arctic Ocean marginal ice zone (MIZ).