Parallel implementation of data assimilation

被引:3
作者
Bibov, Alexander [1 ]
Haario, Heikki [1 ,2 ]
机构
[1] Lappeenranta Univ Technol, Skinnarilankatu 34, FI-53850 Lappeenranta, Finland
[2] Finnish Meteorol Inst, Erik Palmnin Aukio 1,PO 503, FI-00560 Helsinki, Finland
关键词
dynamical systems; extended Kalman filter; parallelization; stabilized method; probabilistic method; error estimation; QUASI-GEOSTROPHIC MODEL; ENSEMBLE KALMAN FILTER;
D O I
10.1002/fld.4278
中图分类号
TP39 [计算机的应用];
学科分类号
081203 ; 0835 ;
摘要
Kalman filter is a sequential estimation scheme that combines predicted and observed data to reduce the uncertainty of the next prediction. Because of its sequential nature, the algorithm cannot be efficiently implemented on modern parallel compute hardware nor can it be practically implemented on large-scale dynamical systems because of memory issues. In this paper, we attempt to address pitfalls of the earlier low-memory approach described in and extend it for parallel implementation. First, we describe a low-memory method that enables one to pack covariance matrix data employed by the Kalman filter into a low-memory form by means of certain quasi-Newton approximation. Second, we derive parallel formulation of the filtering task, which allows to compute several filter iterations independently. Furthermore, this leads to an improvement of estimation quality as the method takes into account the cross-correlations between consequent system states. We experimentally demonstrate this improvement by comparing the suggested algorithm with the other data assimilation methods that can benefit from parallel implementation. Copyright (c) 2016 John Wiley & Sons, Ltd.
引用
收藏
页码:606 / 622
页数:17
相关论文
共 31 条
[11]  
Dennis JE, 1980, NEW DERIVATION SYMME
[13]  
FANDRY CB, 1984, J ATMOS SCI, V41, P807, DOI 10.1175/1520-0469(1984)041<0807:ATLQGM>2.0.CO
[14]  
2
[15]  
Fisher M, 2014, PRECONDITIONING SADD
[16]  
Fisher M, 2009, INVESTIGATION MODEL
[17]  
FISHER M, 1998, 260 ECMWF
[18]  
Gaspari G, 1999, Q J ROY METEOR SOC, V125, P723, DOI 10.1002/qj.49712555417
[19]   A FAMILY OF VARIABLE-METRIC METHODS DERIVED BY VARIATIONAL MEANS [J].
GOLDFARB, D .
MATHEMATICS OF COMPUTATION, 1970, 24 (109) :23-&
[20]   Parallel Implementation of an Ensemble Kalman Filter [J].
Houtekamer, P. L. ;
He, Bin ;
Mitchell, Herschel L. .
MONTHLY WEATHER REVIEW, 2014, 142 (03) :1163-1182