A Study Of Sea Surface Rainfall Recognition Based On HY-2A Scatterometer

HY-2A satellite is the first ocean dynamic environment observation satellite in China.Its significance is remarkable,and it fills the blank of ocean dynamic environment remote sensing satellite in China.Although microwave scatterometer can penetrate cloud and rain at all-weather and all-time,rainfall will reduce its measurement accuracy,especially for Ku band scatterometer,which can not be ignored.However,because remote sensing research on marine dynamic environment in China was relatively late,some auxiliary information of its product is not perfect,such as rainfall flag and quality control,so it is temporarily unable to identify contaminated rainfall data by itself.In addition,due to the inconsistency among wind field products of different scatterometer,some existing rainfall recognition algorithms are only applicable to specific satellite sensors,and cannot be directly applied to HY2-SCAT.Therefore,this study will work out a set of suitable rainfall identification scheme based on the practical measuring situations of HY2-SCAT.The research idea refers to the MUDH of JPL used in Sea Winds scatterometer,which build the rainfall recognition model by selecting some parameters sensitive to rainfall.The data sources used in this study include Level-2A and Level-2B products of HY-2A,ECMWF and Tao buoy data.After all the data are spatiotemporal matched,the data with wind speed ranging from 3m/s to 15m/s are selected to study and analyze the response of HY-2A satellite to rainfall.The study finds that the response of HY-2A to rainfall is slightly different from Sea Winds,and the sensitivity of some factors to rainfall decreased.Therefore,the study further analyzed the scattering characteristics of raindrops and the measurement characteristics of HY-2A,added the node number of the wind vector cell,the mean deviation of backscattering coefficient and the beam difference between fore and aft as the sensitivity factors based on MUDH,and the rainfall identification model of HY2-SCAT was build by KNN.After the validation of the verified samples,the identification accuracy of Hy2 RRM is as high as 88%,and about 70% of the data with rainfall can be correctly identified,which shows that the recognition effect of rainfall is significant.After removing the data marked as rainfall,the accuracy of retrieving wind speed and direction is increased by 11% and 4%respectively,and the quality of inversion wind field is significantly improved.The research results are helpful to better understand the characteristics of microwave backscattering,and provide a possible way to further improve the accuracy of wind field inversion of HY-2A scatterometer and other Ku band scatterometers.