非线性优化方法在大气和海洋科学数值研究中的若干应用

Applications of Nonlinear Optimization Method to the Numerical Studies of Atmospheric and Oceanic Sciences

控制大气和海洋运动的模式是复杂的非线性模式,在考虑到线性奇异向量和线性奇异值只能描述切线性模式有效时段内小扰动发展的情况下,介绍了作者们近年来用非线性优化方法数值研究大气和海洋科学的有关工作,其中包括非线性奇异向量和非线性奇异值、条件非线性最优扰动、以及它们在数值天气和气候可预报性研究中的应用· 结果表明,上述非线性优化方法在很大程度上揭示了大气和海洋运动的非线性特征;此外,对可预报性问题的新分类也做了详细介绍,即最大可预报时间、最大预报误差和最大允许初始误差· 这种分类的应用背景是针对数值天气预报和气候预测产品的评价;最后,讨论了数值模式敏感性分析的非线性优化方法,该方法在一定条件下可以定量识别模式误差和初始误差。

Linear singular vector and linear singular value can only describe the evolution of sufficiently small perturbations during the period in which the tangent linear model is valid. With this in mind,the applications of nonlinear optimization methods to the atmospheric and oceanic sciences are introduced, which include nonlinear singular vector (NSV) and nonlinear singular value (NSVA), conditional nonlinear optimal perturbation (CNOP), and their applications to the studies of predictability in numerical weather and climate prediction. The results suggest that the nonlinear characteristics of the motions of atmosphere and oceans can be explored by NSV and CNOP. Also attentions are paid to the introduction of the classification of predictability problems, which are related to the maximum predictable time, the maximum prediction error, and the maximum allowing error of initial value and the parameters. All the information has the background of application to the evaluation of products of numerical weather and climate prediction. Furthermore the nonlinear optimization methods of the sensitivity analysis with numerical model are also introduced, which can give a quantitative assessment whether a numerical model is able to simulate the observations and find the initial field that yield the optimal simulation. Finally, the difficulties in the lack of ripe algorithms are also discussed, which leave future work to both computational mathematics and scientists in geophysics.