Retrieval Algorithm Of Sea Surface Wind And Wave Based On SAR Image

Since 1970s,Synthetic Aperture Radar(SAR)has been widely used Radar(SAR)has been applied to the inversion of ocean elements.Its microwave can penetrate through the clouds and fog,and is not affected by day and night.It can image the physical dynamic process of the sea surface all day and all-weather,and the imaging resolution is very high.It can achieve multi polarization,multi angle,multi band object observation,and provide large-scale real-time dynamic information about the ocean surface.Therefore,it is convenient to use these advantages of SAR to retrieve the sea surface wind field and wave parameters.In the chapter of sea surface wind retrieval,we propose are tuning algorithm based on Geophysical Model Function(GMF)C-SARMOD2.We have collected more than 10000GF-3 images of VV polarization and HH polarization in QPS mode and WV mode SAR image,and take these data as input items,and bring them into the C-SARMOD2 algorithm to fit new coefficients.The adjusted algorithm is called C-SARMOD-GF,and then combine with the other three algorithms C-SARMOD,C-SARMOD2 and CMOD7 to carry out wind field inversion,and then compare with the wind speed data of advanced scatterometer(ASCAT)and National Data Buoy Center(NDBC).Under VV polarization,the Root Mean Square Error(RMSE)of C-SARMOD-GF algorithm is about 1.7m/s,and that of the other three algorithms is 1.86-2.35m/s;under HH polarization,the Root Mean Square Error(RMSE)of C-SARMOD-GF algorithm is about 1.8m/s,and that of the other three algorithms is 1.93-2.28m/s.The results show that the performance of C-SARMOD-GF algorithm is better than the other three advanced algorithms,but it is only limited to low and medium wind speed,and its effectiveness in strong wind conditions remains to be verified.In the chapter of ocean wave parameter inversion,we use two methods to invert ocean wave parameters.The first one is the ocean wave inversion algorithm based on theory,which inverts the intensity spectrum of SAR to the spectrum based on SAR wave mapping mechanism.In the previous study,we found that the wave inversion algorithm PFSM is suitable for C-band and X-band SAR with medium and low sea conditions.In this work,we study the performance of PFSM algorithm applied to dual polarization C-band sentinel-1 SAR acquired in EW and IW modes under cyclone conditions.Six VH polarimetric sentinel-1 SAR images are retrieved by using C-3PO model,and the wave parameters,namely significant wave height(SWH)and mean wave period(MWP),obtained from VV polarimetric sentinel-1 SAR images are compared with the simulation results of WAVEWATCH-III(WW3)model.The validation results show that the root mean square error(RMSE)of SWH is 0.69 m and the deviation is - 0.01m;the root mean square error(RMSE)of MWP is 0.62 s and the deviation is - 0.17 s.Although PFSM algorithm relies on high-quality SAR spectrum,this study confirms the applicability of PFSM algorithm in wave inversion of sentinel-1 SAR image.In addition,the accuracy of the inversion results is low in the right area of the cyclone eye where the wind wave directly affects the strong wind sea area,while the PFSM algorithm is better in the left and rear areas of the cyclone eye where the wind wave interaction is poor.The second is artificial intelligence algorithm.In this study,we use GF-3 SAR image data collected in China Sea area to discuss the problem of wave inversion under typhoon condition.Six GF-3 SAR images related to four typhoons are processed by two AI methods.The retrieval results are verified by the significant wave height simulated by ww3 numerical model and the independent significant wave height data obtained by jason-2 altimeter.The results show that the RMSE is 0.1m and 0.61 m,which is within the allowable error range.The results show that the algorithm can better retrieve the wave parameters under typhoon conditions.