SVD方法在场分析和预测中的应用

留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名

邮箱

手机号码

标题

留言内容

验证码

SINGULAR VALUE DECOMPOSITION AND ITS APPLICATION IN ANALYSIS AND FORECAST OF FIELD

摘要: 由预报场与因子场的奇异值分解(SVD),可找到影响预报场的主要物理因子,能提取两个场相互作用的主要耦合信号。借助最优化技术,可实现由因子场对预报场的客观预报。以华中汛期降水场为左场,4月北半球500hPa高度场、海平面气压场、北太平洋海温场组成右场,进行SVD分析和预报试验,其结果令人满意。
- 奇异值分解 /
- 展开系数 /
- 同场相关图 /
- 异场相关图 /
- 最优化技术
Abstract: Conducting singular value decomposition (SVD) of the cross-covariance matrix gives main physical factors relating to forecast fields and isolates a coupled signal from two data fields. A field can be predicted objectively from the other field by optimization technique. Considering the rain field in central China in flood season as a left field, and the Northern Hemisphere 500 hPa height, the Northern Hemisphere sea level pressure and northern Pacific SST as a right field, the relationship between the two fields is studied by SVD. The result of forecast experiments is proved effective.
- SVD /
- expansion coefficient /
- homogeneous correlation map /
- heterogeneous correlation map /
- optimization technique

[1]	CHRISTOPHER S BRETHERTON, CATHERINE SMITH, JOHN M WALLACE. An intercomparison of methods forfinding coupled patterns in climate data [J]. Journal of Climate,1992,5: 541-560.
[2]	丁裕国,江志红. SVD方法在气象场诊断分析中的普适性[J]. 气象学报,1996,54: 365-372.
[3]	JOHN M WALLACE, CATHERINE SMITH, CHRISTOPHER S BRETHERTON. Singular value decomposition of wintertime sea surface temperature and 500 mb height anomalies [J]. Journal of Climate, 1992,5: 561-576.
[4]	江志红,丁裕国. 我国夏半年降水距平与北太平洋海温异常的奇异值分解法分析[J]. 热带气象学报,1995,11: 133-141.
[5]	谢炯光,秦冰冰,王静渊. 奇异值分解方法在季降水预测中的应用[J]. 气象学报,1997,55: 117-123.
[6]	张礼平,扬志勇,陈正洪. 典型相关系数及其在短期气候预测中的应用[J]. 大气科学, 2000, 24: 427-432.