An empirical approach for the retrieval of integral ocean wave parameters from synthetic aperture radar data

[1] In this study a new empirical approach to retrieve integral ocean wave parameters from synthetic aperture radar (SAR) data is presented. The idea behind this computationally efficient technique is to estimate integral ocean wave parameters without the intermediate step of retrieving the two-dimensional ocean wave spectrum. The method has the radiometrically calibrated SAR image as the only source of information and is based on a quadratic model function with 22 input parameters. These parameters include the radar cross section, the image variance, and 20 parameters computed from the SAR image variance spectrum using a set of orthonormal functions. The coefficients of the quadratic function were fitted for the estimation of H-s, the mean periods T-m01, T-m02, T-10, the wave power, and the wave heights associated with different spectral bands. The fit procedure is based on a stepwise regression method. A data set of 12,000 globally distributed ERS-2 wave mode image spectra and colocated WAM ocean wave spectra was available for the study. Two separate subsets of 6000 collocation pairs each were used to fit the model and to carry out comparisons of the retrieved wave parameters with numerical model results. Additional comparisons were performed using NDBC buoy measurements. Scatterplots and global maps with the derived parameters are presented. It is shown that the rms of the SAR derived H-s with respect to the WAM H-s is about 0.5 m. For the mean period Tm-10 an rms of 0.72 s with a high-frequency cutoff period of about 6 s is achieved.