基于最大似然估计的海面风场反演算法研究

Research on Numerical Wind Vector Retrieval Algorithm Based on Maximum Likelihood Estimation

最大似然估计法被认为是海面风场反演的最佳方法,目前被用来处理SeaWinds散射计数据。风矢量求解算法是风场反演算法的核心内容。最大似然法的目标函数形式决定了在风场反演过程中必须采用数值方法求解风矢量,而传统数值求解方法运算量大。该文对最大似然估计的风场反演方法的基本原理和具体过程进行探讨,根据其目标函数的一般分布特征,提出一种较为高效的数值风矢量搜索算法。用SeaWinds散射计的L2A实测数据和相应的L2B数据验证了该算法的可行性。

The approach based on Maximum Likelihood Estimation is considered to be the best one to ocean surface wind retrieval and is applied to process the SeaWinds scatterometer data at present.The inversion algorithm of wind vector is the kernel content of the wind retrieval algorithm.The objective function of MLE approach determines that numerical method must be adopted in wind vector retrieval.But traditional numerical method for wind vector inversion needs a large amount of computation.After detailed discussion on the rationale and procedure of the MLE approach,an efficient numerical wind vector search algorithm is presented in this paper,according to the general distribution of objective function.Using some SeaWinds L2A and corresponding L2B data the numerical algorithm is validated.The result indicates that the algorithm is feasible for wind retrieval.