首页 | 本学科首页   官方微博 | 高级检索  
     检索      

特征线计算格式下共轭方程两种导出途径的比较
引用本文:杨永增,于卫东.特征线计算格式下共轭方程两种导出途径的比较[J].海洋湖沼通报,2002(2):10-16.
作者姓名:杨永增  于卫东
作者单位:1. 国家海洋局第一海洋研究所,海洋环境科学与数值模拟国家海洋局重点实验室,青岛266061;中国科学院海洋研究所,青岛266071
2. Laboratoire de Meteorologie Dynamique, Paris, France
基金项目:山东省自然科学基金 Q99E05 (2000-2001) 资助项目,中法合作研究PRA T00-03 资助项目。
摘    要:共轭方程的导出是建立资料同化模型的关键,其导出方式有两种途径:AFD形式与FDA形式。在特征线计算格式基础上针对一类较广泛海洋动力控制方程分析了其两种共轭方程(AFD形式与FDA形式)之间的关系,并将理论结果应用于波谱共轭方程的讨论。

关 键 词:特征线计算格式  共轭方程  变分同化  海洋  海浪谱
文章编号:1003-6482(2002)02-0010-07
修稿时间:2001年12月29

DIFFERENCE OF ADJOINT EQUATIONS BETWEEN TWO APPROACHES BASED ON THE CHARACTERISTIC INLAID METHOD
Olivier Talagrand.DIFFERENCE OF ADJOINT EQUATIONS BETWEEN TWO APPROACHES BASED ON THE CHARACTERISTIC INLAID METHOD[J].Transaction of Oceanology and Limnology,2002(2):10-16.
Authors:Olivier Talagrand
Abstract:The characteristic inlaid method is widely used to integrate a general first-order partial differential equation, such as the wave energy propagation equations in large scale ocean phenomena. For data assimilation, there are two approaches which are commonly used to formulate the adjoint to the given control model as stated above. The first involves writing the adjoint equations of the linearing continuous model equations and then designing an appropriate finite-difference formulation. The second derives the adjoint equations directly from the finite-difference equations of the forward model. Here these two adjoint equations are obtained respectively for the general first-order partial differential equation based on the characteristic inlaid computational method. And the difference between these two eguations is also analyzed. Then the theoretical results are applied to discussing the adjoint ocean wave energy spectrum equation, which indicates that the two approaches are almost the same for practical situations if the numerical model is calculated by using the characteristic inlaid method.
Keywords:Characteristic inlaid method  Adjoint model  Variational data assimilation
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号