This paper presents a method for joint retrieval of the ocean surface wind and current vectors using the backscatter and the Doppler frequency shift measured by spaceborne single-beam single-polarization synthetic aperture radar (SAR). The retrieval method is based on the Bayesian approach with the a priori information provided by atmospheric and oceanic models for surface wind and currents, respectively. The backscatter and Doppler frequency shift are estimated from the along-track interferometric SAR system TanDEM-X data. The retrieval results are compared against in-situ measurements along the Swedish west coast. It is found that the wind retrieval reduces the atmospheric model bias compared to in-situ measurements by about 1 m/s for wind speed, while the bias reduction in the wind direction is minor as the wind direction provided by the model was accurate in the studied cases. The ocean model bias compared to in-situ measurements is reduced by about 0.04 m/s and 12^∘ for current speed and direction, respectively. It is shown that blending SAR data with model data is particularly useful in complex situations such as atmospheric and oceanic fronts. This is demonstrated through two case studies in the Skagerrak Sea along the Swedish west coast. It is shown that the retrieval successfully introduces small scale circulation features detected by SAR that are unresolved by the models and preserves the large scale circulation imposed by the models.

本文提出了一种利用星载单波束单极化合成孔径雷达（SAR）测量的反向散射和多普勒频移联合检索海面风和流矢量的方法。该检索方法基于贝叶斯方法，并分别具有由大气和海洋模型分别提供的表面风和洋流的先验信息。后向散射和多普勒频移是从沿轨道干涉SAR系统TanDEM-X数据估计的。将取回结果与瑞典西海岸的现场测量结果进行比较。已发现，与原位测量相比，对于风速，取风使大气模型偏差降低了约1 m / s，而风向的偏差减小很小，因为模型提供的风向在研究案例中是准确的。与原位测量相比，海洋模型偏差减少了约0.04 m / s和12^∘分别代表当前速度和方向。结果表明，将SAR数据与模型数据混合在一起在复杂的情况下（如大气和海洋前沿）特别有用。这是通过瑞典西海岸的斯卡格拉克海中的两个案例研究证明的。结果表明，该检索成功地引入了由SAR检测到的小规模环流特征，而这些特征是模型无法解决的，并保留了模型所强加的大规模环流。