【作者】 储小青；

【导师】 何宜军；

【作者基本信息】 中国科学院研究生院（海洋研究所） ， 物理海洋学， 2011， 博士

【摘要】 海浪波谱仪(surface waves Investigation and Monitoring，swIM)是新型的海浪信息探测雷达，它以真实孔径雷达为基础，配置六个波束圆锥扫描，波束中心入射角分别为O。，2。，4。，6。，8。和10。。其星下点波束的工作机制相当于一个雷达高度计，而侧视旋转波束通过增加相应的工作模式和功能模块，实现海洋波浪方向谱的测量。海浪波谱仪可获得全球大面积、长时间序列的标准化雷达后向散射截面(NRcs，0.)、海浪方向谱、风速和波高信息。本文对海浪波谱仪海浪遥感探测方法及应用基础研究的若干问题进行了研究与探讨。首先，利用电磁波散射理论，建立了从海浪方向谱映射到海浪波谱仪后向散射截面的模型。并利用目前唯一能提供Ku波段小入射角NRcs的传感器一一降雨雷达(Precipitation Radar,PR)的散射测量数据对散射模型进行了检验。基于建立的海浪波谱仪散射模型，论述了由海浪波谱仪散射信号反演海浪方向谱的方法。对不同系统参数、不同海况、不同入射角下的海浪波谱仪海浪方向谱反演进行了仿真。编制了面向对象人机交互海浪波谱仪仿真软件。仿真结果显示，误差部在允许范围内。利用线性回归法和多入射角组合法，反演了高斯斜率分布中的斜率方差和天底点0。。结果显示，这两种不同的反演方法反演得到的观测方向上的斜率方差的差别在O.0025以内。研究了不同海况下，斜率方差和天底点0。与风速和方位向的关系。利用降雨雷达0。数据研究了小入射角下0。随方位向的变化，重点分析了0。的顺风逆风不对称性(uDA)和由于风向不同导致0。变化(ucA)的幅度与海况、入射角的关系，并利用海浪斜率概率密度函数中的高阶项参量一偏度来解释了这种与散射计相反的不对称性。研究了涌浪对uDA和ucA的影响，并讨论了其在海浪波谱仪上的应用。利用降雨雷达与浮标匹配数据，研究了Ku波段HH极化0。与海面风、浪参数(风速、风向、波高、波长、波周期、波陡、波龄)的相关性。建立了不同入射角下0。分别与海面风速(U0)、有效波高(H)、波陡(&。)和积分波龄(Ba)的经验关系。对从单个0。直接反演各个风、浪参数的可行性进行了全面的分析，包括灵敏度分析、最佳的雷达入射角和海况，最佳的辅助信息等。建立了海浪波谱仪多入射角数据获取海面风速的新方法。该方法考虑了海浪斜率对NRcs与风速之间关系的影响，并且不需要波高、波龄等辅助信息。新的风速反演方法均方根误差为1.36 m／s，提高了以往降雨雷达的风速反演精度。更多还原

【Abstract】 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.更多还原