The Effect of Radar Ocean Surface Sampling on Wave Spectrum Estimation Using X-Band Marine Radar

In this paper, the effect of the ocean surface sampling process on the ocean wave spectral estimation using the Cartesian Fourier transform (CFT) method on <inline-formula> <tex-math notation="LaTeX">$X$ </tex-math></inline-formula>-band marine radar data is investigated. Our analysis shows that the ocean surface sampling process involves a spatial averaging process that might be described as a 2-D low pass filter. Furthermore, a filter referred to as the inverse sampling averaging filter (ISAF) is proposed to be integrated with the CFT method in order to mitigate the effect of the sampling process. For validation, the CFT-with-ISAF method as well as the CFT-without-ISAF method were used to estimate ocean wave spectra and sea state parameters from <inline-formula> <tex-math notation="LaTeX">$X$ </tex-math></inline-formula>-band marine radar field data. The estimates from both methods were compared to ground truth estimates generated using TRIAXYS wave buoy data. The results show that the ISAF improves the CFT method in estimating ocean wave spectra. The recorded accuracy improvements in estimating the non-directional wave spectrum, the peak wave period, the mean wave period, the zero-crossing wave period, and the peak wave direction were 11%, 12%, 21%, 17%, and 34%, respectively. The performances of significant wave height estimation using the ISAF method and the standard CFT method were validated against ground truth estimates and found to be comparable.