基于集合同化方法的南海北部最优观测实验

摘要

In this study, we attempt to estimate and select the optimal observation profiles and observation tracks over the northern South China Sea, using an ensemble Kalman filter based optimized observational scheme and a ROMS regional ocean model for northern part of South China-Sea.Kalman-Filter-based optimal observation scheme is a Bayesian method, which improves observation locations by minimizing the expectation of the root mean square deviation (RMSD) of the analysis field.In another word, this method will minimize the expectation of the uncertainty of the analysis field after assimilating the observations in the selected locations.In this method, only observation locations are required to construct the observation operator, without using the real-time value of each observation.A group of historical snapshots of the model result are served as background ensembles.This method uses the covariance matrix uploading equation in the Kalman-filter theory, to minimize the RMSD of analysis filed.The optimal observation method combines the observation locations with the simulation system, in order to optimize the observation locations and observation strategy before it was conducted.This method can significantly increase the observation efficiency.In this study, the optimal method was used in South China Sea.First, we used the optimal method to select a group of observation locations (profiles) in the northern part of South China Sea.These observation profiles mainly distributed in the areas with high variability and very strong ocean currents.In addition, we used this method to select the best observation tracks.We found that the tracks mainly located around Shantou, Shanwei and the Taiwan shallow shoal.%根据集合同化理论及最优观测方法,应用基于ROMS模式的南海北部区域模拟结果作为集合样本,对该区域的最优观测廓线以及最优观测航次进行了评估和选择.最优观测方法是一种不需要直接同化观测,而通过最小化分析场误差方差理论值来选择最优观测点的方法.最优观测方法根据卡曼滤波理论中协方差矩阵的更新方程,使用一组模式或观测得到的历史样本,寻找使得分析场的不确定性的理论值最小的一个或一组观测位置作为最优观测点,从而达到最优化观测位置的目的.首先应用最优观测方法以及由ROMS模式结果计算出的混合层深度样本选择了一组最优观测廓线,得出的最优观测点主要分布在样本方差较大,流相关较强的区域.在此基础上,应用同一组样本针对南海北部区域实际航次进行了最优化组合和筛选,所选择的最优航次主要集中在汕头、汕尾、台湾浅滩附近.