A new algorithm for reconstructing the three-dimensional flow field of the oceanic mesoscale eddy

A new algorithm for reconstructing the three-dimensional flow field of the oceanic mesoscale eddies is proposed in this paper, based on variational method. Firstly, with the numerical differentiation Tikhonov regularizer, we reconstruct the continuous horizontal flow field on discrete grid points at each layer in the oceanic region, in terms of the horizontal flow field observations. Secondly, benefitting from the variational optimization analysis and its improvement, we reconstruct a three-dimensional flow field under the constraint of the horizontal flow and the vertical flow. The results of simulation experiments validate that the relative error of the new algorithm is lower than that of the finite difference method in the case of high grid resolution, which still holds in the case of unknown observational errors or in the absence of vertical velocity boundary conditions. Finally, using the reanalysis horizontal data sourcing from SODA and the proposed algorithm, we reconstruct three-dimensional flow field structure for the real oceanic mesoscale eddy.     

大气边界层模式中随机参数的反演与不确定性分析

在大气边界层气象中湍流黏性系数是一个很重要的参数,通过直接观测往往无法得到其准确值,仅能通过间接观测获得大致范围.本文选用随机广义Ekman动力近似模式中的湍流黏性系数进行反演研究与不确定性分析.首先利用风速观测数据,并采用基于混沌多项式的集合Kalman滤波方法对系数进行反演,降低其不确定性,缩小可能取值的范围,该方法的核心思想是将集合Kalman滤波方法中求解模式不确定性传播的方法由蒙特卡罗法改为混沌多项式展开,从而避免大规模采样带来的计算资源耗费.然后进行数值实验,结果表明该方法能够有效且快速地求解出湍流黏性的后验概率分布,从而达到降低系数不确定性的目的.根据系数的先验分布计算出风速的先验分布,从而找到风速不确定性大的区域,且揭示了在不确定性大的区域内的观测数据进行系数反演可得到十分明显的效果,这对于观测点位置的选择提供了重要的指导.