Research On Wave Height And Period Extraction From ASAR Wave Mode Data Based On SVM

Synthetic aperture radar(SAR)is a space-borne microwave remote sensor that is used to observe the earth without being influenced by the sunshine,clouds and other environmental factors.The ENVISAT satellite is equipped with advanced synthetic aperture radar(ASAR)in the C-band,which has five modes of operation.ASAR could to access data all day when it operated in wave mode.Based on the data acquired from wave mode,a support vector machine(SVM)regression model is proposed to retrieve ocean wave parameters such as significant wave height(SWH)and mean wave period(MWP)from the ASAR wave mode data.It has been introduced that the background and significance of the inversion of ocean wave parameters based on SAR data and summarized the classical methods.Analyzed the SAR imaging mechanism of sea surface and pointed out the shortcomings of classical inversion methods such as max-planck institute(MPI),partition rescale and shift algorithm(PARSA)and semi-parametric retrieval algorithm(SPRA).Therefore,a new method of using SVM to retrieve the ocean wave parameters was proposed.The main contents of the paper are as follows:1.SVM modeling.Firstly,the preprocess step of ASAR wave mode images and the spectral decomposition process was described in detail.Then the relationship between the characteristic parameters and the SWH and the MWP is analyzed from the spatial and frequency domains.The theoretical derivation of the SVM regression model is given.A global data set of 30771 pairs of ASAR wave mode images and collocated global atmospheric reanalysis ERA-Interim data was used to train the SVM model.2.Validation and analysis of SVM regression model.The extracted SWH and MWP compared with in situ buoy measurements and ERA-Interim data.The RMSE of SWH is 0.49m and 0.4m and the correlation is 0.80 and 0.93,respectively.The retrieved MWP compared with the buoy data and the ERA-Interim data,the RMSE is 1.08s and 0.62s,and the correlation is 0.76 and 0.87 respectively.The results show that the SVM regression model is an effective method to retrieve SWH and MWP from SAR data.3.Further evaluation of SVM model.The paper analyzed a typhoon case and seasonal variation of global sea conditions to evaluate the performance of SVM model.The results show that the proposed model is of great significance for the sea traffic warning,marine storm monitoring and the operational operation of the wave parameters.