摘要
变分数据同化方法中,背景误差协方差矩阵(简称B矩阵)作为反映模型不同状态变量间关系的重要数学量,对保证解的唯一性和分析值的平滑性具有重要意义.文中简述了B矩阵的结构特征和相关性质.针对难以获得真实B矩阵的情况,采用National meteorological center(简称NMC)方法对B矩阵进行了近似构造,以Lorenz63模型和三维变分同化算法进行了同化试验,验证了NMC方法的有效性,讨论了B矩阵的数学特性和分布特征.
In variational data assimilation method,as an important mathematical quantity that can reflect the relationship between different state variables of model,background error covariance matrix has great significance to guarantee the uniqueness of solution and smoothness of analysis value.In this paper we have briefly described the structure features and properties of B matrix.For the cases where real B matrices are difficult to obtain,we use the national meteorological center(NMC) method to compute an approximation B matrix based on Lorenz-63 model and to verify the effectiveness of NMC method.Then,we discussed the mathematical characteristics and distribution characteristic of B matrix according to the results.
作者
摆玉龙
孟若玉
马真东
柴乾隆
BAI Yu-long;MENG Ruo-yu;MA Zhen-dong;CHAI Qian-long(College of Physics and Electrical Engineering,Northwest Normal University,Lanzhou 730070,Gansu,China)
出处
《西北师范大学学报（自然科学版）》
CAS
北大核心
2019年第1期50-54,共5页
Journal of Northwest Normal University(Natural Science)
基金
国家自然科学基金资助项目(41861047
41461078)
西北师范大学科研能力提升团队资助项目(NWNU-LKQN-1706)
