IMPROVED DETERMINATION OF DIRECTIONAL WAVE SPECTRUM FOR SWELL WAVES USING X-BAND RADAR IMAGING

Abstract

Typically, estimation of wave and current information from radar image sequences involves the following steps: 3-D discrete fast Fourier transform (FFT), surface current determination, filtering the 3-D FFT result, integration of the 3-D FFT result over the positive frequency domain, transformation of the 2-D image spectrum to a 2-D wave spectrum using a modulation transfer function, and conversion of the 2-D wave spectrum from the wavenumber domain to the frequency direction space. The spurious clutter speckles appearing below a certain wind speed (e.g. 3 m/s) tend to scatter in all directions in the image frame, while intermitted wave patches move with specific directions. In order to eliminate the drifting spurious speckles all around and retain wave features in directional motions in an image sequence, a backward image frame differencing scheme, a 2-D band-pass image filter, and a 2-D elliptic Butterworth filter are additionally applied to each individual images prior to the 3-D FFT processing step. Compared to the in situ measurements, the 2-D spectrum result of the processed image data lead to better agreements than that of the original image data