Research On The Fusion Of Multi-source Satellite Sea Surface Wind Fiel

With the advancement of technology,the research and application of oceanography and meteorology have put forward increasingly higher requirements for the temporal and spatial resolution of sea surface wind data.Existing remote sensing observations or numerical weather forecast models are no longer able to fully meet the needs of cutting-edge research in the fields of oceanography and meteorology.Therefore,it is urgently needed to develop a sea surface wind dataset with higher temporal and spatial resolution based on existing ocean remote sensing technology,which can meet the needs of studying small and medium-scale ocean dynamic processes and assimilating high-resolution numerical models for forecasting.This paper is based on the remote sensing sea surface wind data obtained from microwave scatterometer and radiometer on board satellites.Using the Two-dimensional Variational Method(2D-Var),we conducted a multi-source satellite sea surface wind field blended study,aiming to develop a high-precision,grid-based global sea surface wind dataset every 6 hours.This dataset provides important technical and data support for cutting-edge research in oceanography and meteorology.The main research contents and conclusions are as follows:(1)The study validates that microwave scatterometer and radiometer wind speeds retrieved in mid-to-high latitudes are more suitable for interpreting as equivalent neutral winds compared to commonly used neutral reference winds.The accuracy of the scatterometer and radiometer used in the study meets operational requirements.In addition,this paper proposes an indicator factor for lateral comparison of wind speed errors,in which ASCAT-A and ASCATB in the C-band have the best accuracy for scatterometers,followed by Ku-band scatterometers,while AMSR-2 and Wind SAT have the best accuracy for radiometers,followed by the SSMIS series.Overall,for wind speeds ranging from 0-25 m/s and time intervals within 6 hours,the accuracy of scatterometers is better than that of radiometers.For wind speeds ranging from 0-25 m/s and time intervals within 30 minutes,the error of scatterometer wind is smaller than that of ERA5 wind.(2)Bias correction of input data can effectively improve the accuracy of the blended wind field.Assimilating satellite wind observation data,the blended wind field of domestic scatterometers and radiometers after assimilation has a certain degree of increase in wind speed accuracy compared to the unfused blended wind field.Under the same input data,the accuracy of the blended wind field with different background wind field error correlation functions is significantly different.The blended wind field under the Empirical Background Error Correlation(EBEC)has higher accuracy than that under the Gaussian Structure Function(GSF),and contains more small-scale information.Overall,2D-Var blended data can accurately describe the cyclonic circulation structure.Specifically,the feasibility of blended wind fields using HY-2 series satellite microwave scatterometers was explored,and the possibility of domestic satellite data blended was discussed.