Improvements in Atmospheric Water Vapor Content Retrievals Over Open Oceans From Satellite Passive Microwave Radiometers

The study is dedicated to the development of an improved algorithm for integrated water vapor content (WVC) retrieval from the advanced microwave sounding radiometer 2 (AMSR2) measurement data over open ocean areas. The algorithm is based on physical modeling of the brightness temperature (BT) of the upwelling radiation of the atmosphere - ocean system. The BT inversion is carried out with neural networks (NNs), trained on an ensemble of BTs, calculated using the dataset of the atmospheric and oceanic parameters. The improvement in the WVC retrieval algorithm as compared to previously developed one is associated, first, with the usage of the new ocean emissivity empirical model and, second, with the exclusion of the vertically polarized BT from NNs inputs. The new model was incorporated into the BT geophysical model and a new set of NNs coefficients was obtained for WVC algorithm. Though the results of the numerical experiment demonstrate higher retrieval accuracy for the old version of the NN model, real-valued validation with global positioning system (GPS) WVC shows better performance of the new NN model with the excluded vertically polarized AMSR2 BT from the NNs algorithm inputs. The advantages of the new algorithm are the most remarkable under conditions of high winds. This is confirmed by the validation based on the database of GPS WVC measurements under extreme winds. The improved WVC retrieval algorithm is mostly beneficial for the studies of extreme weather events such as polar lows, extratropical and tropical cyclones, associated with high winds.