基于同步风场的机载SAR海浪参数反演方法

摘要

针对机载合成孔径雷达(SAR)对海探测特点，采用多入射角法从 SAR 数据本身得到与海浪参数反演区域时空匹配的同步海面风速和风向，并结合线性变换关系，计算得到海浪初猜谱对应的仿真SAR 图像谱，将仿真 SAR 图像谱和观测 SAR 图像谱输入代价函数中进行迭代运算，通过非线性方程的解算得到最适海浪谱；采用交叉谱法去除海浪传播180°方向模糊，最终得到海浪参数。论文提出的基于同步风场的机载SAR海浪参数反演方法，充分利用了机载SAR海洋环境探测的优势，解决了传统SAR 海浪参数反演中初猜谱构造依赖外部风场的问题，机载同步飞行试验的海浪参数反演结果与浮标观测值的有效波高、波向的均方根误差分别为0.23 m和13.23°，验证了该方法的有效性，可为机载SAR海浪参数反演业务化提供支持。%Airborne synthetic aperture radar (SAR), which has flexible characteristics, is an effective means of ob-taining regional ocean-wave parameters. These regional ocean-wave parameters are the most fundamental and im-portant marine environmental parameters in dynamic marine processes. Based on the sounding characteristics of airborne SAR, the first-guess spectrum is constructed by combining ocean-wave spectrum model, direction function with the synchronous wind speed and wind direction which match with the inversion region of ocean wave parame-ters temporally and spatially. The synchronous wind speed and wind direction can be obtained from the SAR data itself using the multi-incidence angle method. Next, a simulated SAR image corresponding with the first-guess spectrum can be produced based on the nonlinear image relations between the ocean-wave direction spectrum and the SAR image spectrum. The simulated observed SAR image spectrums are combined as an iterative operation, and the most suitable wave spectrum can be obtained by solving the nonlinear equation. Finally, the wave parameters are obtained using the cross-spectra method, which is used for eliminating the 180° direction ambiguity of ocean-wave propagation. This study proposes a method of retrieving ocean-wave parameters from airborne SAR imagery based on the synchronization information of the wind field. This method entirely utilizes the advantage of airborne SAR in the marine environment by solving the problem of the first-guess spectrum structure that is de-pendent on external wind field in traditional SAR wave parameter inversion. Airborne ocean-wave parameter inver-sion results from synchronized flight tests and survey ship observations yield root-mean-square errors for signifi-cant wave height and wave direction of 0.23 m and 13.23°, respectively. These errors verify the validity of the pro-posed method and provide support for the using ocean-wave parameters retrieved from airborne SAR imagery.