A Basic Study on the Estimation of Ocean Wind Vector Using Reflected Radiance and Polarization Data

本文利用多重散射模型对海洋上空的机载POLDER数据进行了分析。假设一个具有Cox-Munk型反射海面的大气-海洋系统，在几种不同的大气模式下，采用加倍法计算了包括多重散射和反射贡献在内的反射率和线偏振度。研究结果表明：1)海洋气溶胶模式不适合用于Medimar数据，而3 < v < 5的Junge型气溶胶模式能够满足观测到的极化特征；2)在多次散射计算的基础上，首次对图像进行了反射率和线偏振度的重建。通过假设一个由瑞利粒子和气溶胶粒子组成的两层大气模式，我们发现观测到的反射率和偏振图像中的许多特征可以在理论上重建，并具有Cox-Munk海表层。3)各向同性Cox-Munk模型不能解释观测到的反射率曲线在垂直平面上相对于观测天顶角rhea =0的不对称性，应被拒绝。各向异性Cox-Munk模型可以解释这种不对称性，在反演地面风场时应采用该模型。这里需要指出的是，为了准确地估计地面风场，对这种模型进行多次散射分析是必要的。5)对于在主平面对称位置上的两个偏振极小点，这是偏振问题中尚未解决的一个主要问题，有待于将来的明确解释。

In this study we have made an analysis of the airborne POLDER data over the ocean by a multiple scattering model. Assuming an atmosphere-ocean system with a Cox-Munk type reflecting sea surface, the reflectance and degree of linear polarization, including contributions from multiple scatterings and reflections have been computed by using the adding and doubling method for several different atmospheric models. Our results based on this study are summarized as follows :1) We found the oceanic aerosol model is not an appropriate aerosol model for Medimar data, but Junge type aerosol model with 3 < v < 5 is able to satisfy the observed polarization characteristics.2) Based on the multiple scattering computations, the image reconstruction for the reflectance and degree of linear polarization was made, for the first time. We showed that many features in the observed reflectance and polarization images can be theoretically reconstructed by assuming a two layrs atmospheric model, composing of Rayleigh and aerosol particles with a Cox-Munk sea surface layr.3) An isotropic Cox-Munk model should be rejected on the ground that it fails to explain the asymmetry in the observed reflectance curve with respect to the viewing zenith angle rheta=0 in the perpendicular plane. An anisotropic Cox-Munk model can explain such the asymmetry and it should be used in retrieving the surface wind field instead. We should point out here that the multiple scattering analysis for such a model is necessary for estimating the surface wind field accurately.5) As a major unsolved problem conceming with polarization, a clear explanation of two polarization minimum points in a symmetric location to the principal plane is left in future.