Open AccessTime‐expanded sampling for ensemble‐based filters: Assimilation experiments with a shallow‐water equation model
Author(s) -
Xu Qin,
Wei Li,
Lu Huijuan,
Qiu Chongjian,
Zhao Qingyun
Publication year - 2008
Publication title -
journal of geophysical research: atmospheres
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.67
H-Index - 298
eISSN - 2156-2202
pISSN - 0148-0227
DOI - 10.1029/2007jd008624
Subject(s) - data assimilation , covariance , sampling (signal processing) , computer science , algorithm , filter (signal processing) , ensemble kalman filter , mathematical optimization , mathematics , statistics , kalman filter , meteorology , artificial intelligence , physics , computer vision , extended kalman filter
A time‐expanded sampling is proposed for ensemble‐based filters in data assimilation. This approach samples a series of (preferably three) perturbed state vectors from each prediction run in an ensemble of forecasts at properly selected time levels in the vicinity of the analysis time. As all the sampled state vectors are used to construct the ensemble and compute the covariance (with localization), the number of required prediction runs can be greatly reduced and so is the computational cost. As the sampling time interval is properly selected, the proposed approach can improve the ensemble spread and enrich the spread structures so that the filter can perform well even though the number of prediction runs is greatly reduced. Assimilation experiments are performed with a shallow‐water equation model to demonstrate the potential merits and limitations of the time‐expanded sampling in improving the filter performance either without conventional covariance inflation (method‐A) or with optimally tuned covariance inflation and localization (method‐B).