Research On Wind And Ocean Wave Information Retrieved From Sythetic Aperture Radar

The traditional marine measurement methods such as coastal station,vessel and buoy can only measure the single point of the ocean, there are several shortcomings: Firstly, there are too few measurements point to obtain large area's information of ocean; Secondly, the measurement expense is faily expensive, so it is impossible to obtain measurements all over the ocean. Thirdly, the traditional measurement methods greatly suffer from the weather condition, thus continuos measurements by traditional methods are impossible in a long term. However, synthetic aperture radar (SAR) could provide all-weather, all day and all night, and high resolution ocean image, which could operate in muti-band, multi-incedence angle and multi-polarization. These advantages are impossible for optical and infrared sensors. So the research on how to retrieve information from SAR image is a very significant science.Since the launch of ENVISAT satellite on March 1st, 2002, advanced synthetic aperture radar (ASAR) imagery has been acquired over the ocean on a continuous basis. Compared with ERS-1/2 SARs, ASAR's important new capabilities include beam steering for acquiring images with different incidence angles, dual polarization and wide swath coverage. The dual polarization capability could provide co-polarization scene and cross-polarization scene simultaneously, thereby considerably increasing the target classification capability and providing more information of the target. This provides the opportunity that eliminates the inherent 180°ambiguity when retrieving wind field using ASAR dual polarization imagery. A new algorithm for retrieving wind vector using dual polarization imagery of advanced synthetic aperture radar (ASAR) is developed and tested. Based on the combination of co-polarization algorithm and cross-polarization algorithm, this new algorithm effectively eliminates 180o ambiguity which occurs when using single imagery of ASAR to retrieve wind vector. This algorithm also solves the problem that the retrieval results will break down on very small spatial scales. The results retrieved from dual polarization imagery of ASAR show that the wind speed and wind direction are in agreement with data from Quikscat and buoy measurements. The root mean square errors (rms) of wind direction and speed between the retrieved results and data from Quikscat are 2.21o and 0.53 m/s, respectively.Ocean wave height retrived from SAR image is also a hotspot. The traditional method is that wave height is retrieved from wave spectrum. The popular models that wave spectrum retrieved from SAR image are MPI (Hasselmann, 1991) and Semi-Parameterization method (Mastenbroek and de Valk, 2000). The main shortcoming of these two model is that additional information such as WAM model result or Scatterometer wind field is needed to eliminate 180o ambiguity. The algorithm applied in this study-CWAVE need not additional information and could retrieve significant wave height (Hs) directly from SAR wave mode imagette. The input parameters are the radar cross section, the image variance, and 20 parameters computed from the SAR image variance spectrum. These parameters select by stepwise regression method, and the coefficients of CWAVE are fitted between 6000 globally distributed ERS-2 wave mode image spectra and collocated WAM ocean wave spectra. Two years of Hs retrieved from CWAVE using about 1 million global imagettes from Sep. 1998 to Nov. 2000. The NOAA buoy data are applied to validate CWAVE. The correlation coefficient is 0.83, and rms error is 0.61 m and the bias is 0.02m. The comparison between CWAVE and ERA-40 data and between CWAVE and altimeter are performed. All of these comparisons prove that CWAVE is an effective algorithm.The SAR results are applied to study the nonlinear Hs distribution. First, The nonlinear ocean elevation,wave height and Hs distribution functions are deduced using dynamical method joint random statistical methods. The validation using NOAA bouy data, wave height data taken from East China Normal University and SAR Hs results are performed respectively. The SAR results are also used to fit a new wind wave growth relationship function: = 0.00854^1.564. The comparison between this new function and other former results indicates that this new function is reasonable.