摘要
数值天气预报中分析同化的基本方法先后经历了多项式函数拟合方法、逐步订正方法、最优插值方法、变分方法和集合卡尔曼滤波方法。本文首先根据相关的经典文献力求本色地介绍这些方法的基本思想和实施的具体要点;然后,着重于它们的上下承接关系,试图阐述同化的历史发展脉络,评述这些方法的显著特征和创新性,以期清晰地理解资料同化的循序渐进的内在发展逻辑。此外,从起源上阐明"主观分析"与"客观分析"、"初猜场"与"背景场"、"分析"与"同化"、以及"更新"、"新息"等基本概念,以期准确地理解和把握"大气资料同化"的由来和内涵。
Main basic methods of analysis and assimilation for numerical weather prediction（NWP） has gone through the polynomial function fitting method,the successive correction method（SCM）,the optimum interpolation（01）,the variational method（Var） and the ensemble Kalman filter（EnKF）.This article tries to present a concise and true description with their basic ideas and implementation approaches based on original classical references to these methods.Then remarks on their distinct characteristics and innovative development are made,and it is highlighted,from these remarks,that we can see a step-by-step expansion of useful information and cutdown on limitations along the historical progress of atmospheric data assimilation,which is clearly endowed with a fore-and-aft connection and a cognitively simple internal logic of development.Also,an effort is taken to precisely understand the concept of ＂atmospheric data assimilation＂,along its historical progress and on basis of some essential terms ＂subjective analysis＂ and ＂objective analysis＂, ＂first guess field＂ and ＂background field＂, ＂analysis＂ and ＂assimilation＂, ＂update＂,and ＂innovation＂.
出处
《气象》
CSCD
北大核心
2015年第8期986-996,共11页
Meteorological Monthly
基金
公益性行业(气象)科研专项(GYHY201206007)
国家自然科学基金面上项目(41175034)共同资助
关键词
数值天气预报
客观分析和资料同化
逐步订正方法
最优插值方法
变分方法
集合卡尔曼滤波方法
numerical weather prediction（NWP）
objective analysis and data assimilation
successive correction method（SCM）
optimal interpolation（OI）
variational methods（Var）
ensemble Kalman filter（EnKF）
