The Algorithm And Applied Basic Study On Remote Sensing Of Ocean Wave Spectrum For Ocean Wave Spectrometer

The instrument SWIM (Surface Waves Investigation and Monitoring) is the first ever space radar concept that mainly dedicated to the measurement of ocean waves directional spectra through multi-azimuth and multi-incidence observations. SWIM is real aperture radar in Ku-band pointing sequentially at six different incidences (from 0 to 10 ). Nadir beam works like satellite altimeter. SWIM instrument will provide global long time observation of three kinds of measurements: the from nadir to 10 , the wave directional spectra, the significant wave height and wind speed. This thesis focuses on SWIM ocean wave remote sensing technology and its application basic research on other information retrieval.At low incidence, the sea surface scattering mechanism and ocean wave remote sensing technique is different from previous active microwave radars. We developed the forward transfer function from directional wave spectrum to radar NRCS based on Kirchhoff approximation and two-scale method. Precipitation radar (PR) on the Tropical Rainfall Mapping Mission (TRMM) provides radar backscattering cross section near nadir. The model is tested using measurement from PR.Based on the scattering mechanism, the thesis presents the algorithm of inversion ocean wave directional spectra from SWIM signal. Radar preferences and output resolution and accuracy are analyzed. Then end-to-end simulations are performed to assess the performances of the system in terms of different radar preferences, sea states and incidence angles. The end-to-end SWIM simulator is developed, which includes radar parameters setup panel, sea states setup panel, radar signal display panel, inversion results display panel and err display panel. NRCS is proportional to the probability density function (PDF) of surface wave slopes at low incidence angles. Two methods are proposed to retrieve filtered slope variance under Gaussian slope PDF assumptions. The relationships between filtered slope variance and surface conditions (wind and waves) are investigated. Due to a lack of SWIM data, PR measurements are used to present the first results from a study of upwind/downwind asymmetry (UDA) and upwind/crosswind anisotropy (UCA) of the low incidence microwave backscattering from satellite observations. Incidence angle and wind speed dependence were a particular focus. The UDA is explained by the use of non-Gaussian statistics of the sea surface slope. Sensitivities to sea state are also analyzed. The application to SWIM wind vector retrieval is also discussed.The potential utility of this new radar for direct wind and wave measurement is of great interest. By using nine years of collocated PR and buoy data, the relationships between Ku-band NRCS and integrated wind and wave parameters (e.g., wave period, significant wave height, wave steepness, wave age) at low incidence angles are analyzed for different sea states. Empirical tabular functions relating to sea surface wind speed above 10 m (U10), significant wave height (Hs), wave steepness ( a) and integral wave age ( a), respectively, are developed for incidence angles from 0 (nadir) to 18 . Besides, the potential for inverting these parameters directly from a single is investigated. Those investigations can serve to design directly models relating SWIM NRCS to ocean wind and wave measurements.The above analyses clearly demonstrate that radar backscattering correlates with both the near-surface wind speed and the sea surface wave slope at low incidence angles. Multi-incidence angles PR data are used to retrieve sea surface wave slope parameter and normalized nadir backscattering. After that, an empirical wind speed model was developed based on those two parameters that attenuates the surface tilting effect. The inversion is defined using a multilayer perceptron neural network with radar-derived backscatter and surface wave slope parameter as inputs. Results show the root mean square err (RMSE) between wind speeds retrieved and in situ buoy observations is 1.36 m/s, bias is nearly zero, revealing good agreements in wind speed estimations. The method can be directly applied to SWIM.