Rainfall-Induced Sea Surface Emissivity Change Inferred From Satellite Passive Microwave Observations and In Situ Ship Measurements

Benefiting from the ability to penetrate clouds, microwave (MW) can theoretically achieve all-sky monitoring, thus making significant contributions to weather forecasts and climate models. Various MW sea surface emissivity (SSE) models have been applied to the products of spaceborne MW radiometers and assimilation models. However, these models do not consider the impact of rainfall on the MW SSE. In this study, rainfall-induced SSE change was inferred based on the brightness temperature (BT) observed by Advanced Microwave Scanning Radiometer 2 (AMSR2) and in situ ship measurement OceanRAIN. The neural network-based fast atmospheric parameter simulators (NN-FAPSs) were constructed for quickly calculating the atmospheric upwelling/downwelling BT and transmissivity and the errors of NN-FAPSs fitting on the SSE were below $0.001~\pm ~0.001$ and $0.01~\pm ~0.03$ at 6.925 and 36.5 GHz, respectively. The SSE estimates show that rainfall causes an increase of SSE (e.g., up to 0.2 at 10.65 GHz), and this phenomenon is more pronounced at low wind speeds (WSs) and high sea surface temperature (SST), corresponding to the effects of rainfall-induced local wind and SST on the dielectric constant. The error analysis shows that at 6.925, 7.3, and 10.65 GHz, the accuracy of the SSE is less affected by the accuracy of the input parameters as well as the fitting errors of the NN-FAPSs ( $\Delta \varepsilon $ was below $0.01~\pm ~0.01$ ). Additionally, polynomial models for 6.925, 7.3, and 10.65 GHz are proposed to provide the first guess of SSE under rainfall conditions for improving the accuracy of cloud and rainfall products. It is believed that this work helps improve the understanding of rainfall-induced MW SSE changes.