Study On The Retrieval Algorithm Of Sea Surface Parameters From X-band Marine Radar Image Sequences

X band marine radar can image sea surface in high spatial and temporal resolution.By using corresponding inversion algorithm,the sea surface physical information can be obtained in high temporal-spatial resolution, such as the significant wave height,wave spread direction, the wave period, sea surface wind speed and so on, and these parameters can be used on the protection of ocean engineering, marine fishery, ocean transportation and the like. This dissertation studies the algorithms to retrieve ocean wave parameters, sea surface wind speed from X-band marine radar image sequences,and also presents a new synergistic retrieval of sea surface wind speed and waves.Firstly, X-band marine radar imaging is influenced by different modulation mechanisms, the conventional linear modulation transfer function (MTF) was derived in deep water conditions, which may not appropriate in shallow waters. Additionally,in consequence of the radar backscatter intensity strongly depend on the radar azimuth and distance, it is not suitable of using the same MTF exponent in all radar area. By comparison of the differences between the radar image spectrum and in-situ buoy measured wavenumber spectrum, we propose new MTF with both range and azimuth dependency based on VV-polarized radar measurements acquired from heterogeneous coastal wave fields. The proposed MTF is used to retrieve the peak wave period and mean wave period from X-band marine radar image sequences, and the root mean square error (RMSE) between them and those measured by the buoy are 0.95 s and 0.48 s while the Bias for them are 0.52 s and 0.26 s, respectively.Secondly, sea surface wind speed has a non-linear relationship with radar backscatter intensity, we developed a new empirical physical model. In comparison to the old model, the accuracy of retrieval wind speeds is improved significantly. The retrieved wind speeds are compared with those measured by anemometer. The root square error is 1.9 m/s and the mean relative error is -0.06.Finally, we established simulated wave spectra by utilizing the retrieval wind speeds, and presented a synergetic retrieval method of wind speed and wave spectrum.We retrieve ocean surface wave spectrum by minimizing the constraint function including the simulated wave spectrum and observed radar image spectrum. The retrieved wave parameters are compared with those measured by anemometer and buoy. The root mean square errors of significant wave height, dominant wave period and wave direction are 0.4m, 1.2s and 9.6°, and the mean relative errors of those are 0.1?-0.1 and 0.05, respectively.