A data-Driven method for improving the correlation estimation in serial ensemble kalman filters

Michèle De La Chevrotière, John Harlim

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

A data-driven method for improving the correlation estimation in serial ensemble Kalman filters is introduced. The method finds a linear map that transforms, at each assimilation cycle, the poorly estimated sample correlation into an improved correlation. This map is obtained from an offline training procedure without any tuning as the solution of a linear regression problem that uses appropriate sample correlation statistics obtained from historical data assimilation outputs. In an idealized OSSE with the Lorenz-96 model and for a range of linear and nonlinear observation models, the proposed scheme improves the filter estimates, especially when the ensemble size is small relative to the dimension of the state space.

Original languageEnglish (US)
Pages (from-to)985-1001
Number of pages17
JournalMonthly Weather Review
Volume145
Issue number3
DOIs
StatePublished - Jan 1 2017

Fingerprint

Kalman filter
data assimilation
transform
filter
method

All Science Journal Classification (ASJC) codes

  • Atmospheric Science

Cite this

@article{8ae505d69b2b4a439e9de10403335735,
title = "A data-Driven method for improving the correlation estimation in serial ensemble kalman filters",
abstract = "A data-driven method for improving the correlation estimation in serial ensemble Kalman filters is introduced. The method finds a linear map that transforms, at each assimilation cycle, the poorly estimated sample correlation into an improved correlation. This map is obtained from an offline training procedure without any tuning as the solution of a linear regression problem that uses appropriate sample correlation statistics obtained from historical data assimilation outputs. In an idealized OSSE with the Lorenz-96 model and for a range of linear and nonlinear observation models, the proposed scheme improves the filter estimates, especially when the ensemble size is small relative to the dimension of the state space.",
author = "Chevroti{\`e}re, {Mich{\`e}le De La} and John Harlim",
year = "2017",
month = "1",
day = "1",
doi = "10.1175/MWR-D-16-0109.1",
language = "English (US)",
volume = "145",
pages = "985--1001",
journal = "Monthly Weather Review",
issn = "0027-0644",
publisher = "American Meteorological Society",
number = "3",

}

A data-Driven method for improving the correlation estimation in serial ensemble kalman filters. / Chevrotière, Michèle De La; Harlim, John.

In: Monthly Weather Review, Vol. 145, No. 3, 01.01.2017, p. 985-1001.

Research output: Contribution to journalArticle

TY - JOUR

T1 - A data-Driven method for improving the correlation estimation in serial ensemble kalman filters

AU - Chevrotière, Michèle De La

AU - Harlim, John

PY - 2017/1/1

Y1 - 2017/1/1

N2 - A data-driven method for improving the correlation estimation in serial ensemble Kalman filters is introduced. The method finds a linear map that transforms, at each assimilation cycle, the poorly estimated sample correlation into an improved correlation. This map is obtained from an offline training procedure without any tuning as the solution of a linear regression problem that uses appropriate sample correlation statistics obtained from historical data assimilation outputs. In an idealized OSSE with the Lorenz-96 model and for a range of linear and nonlinear observation models, the proposed scheme improves the filter estimates, especially when the ensemble size is small relative to the dimension of the state space.

AB - A data-driven method for improving the correlation estimation in serial ensemble Kalman filters is introduced. The method finds a linear map that transforms, at each assimilation cycle, the poorly estimated sample correlation into an improved correlation. This map is obtained from an offline training procedure without any tuning as the solution of a linear regression problem that uses appropriate sample correlation statistics obtained from historical data assimilation outputs. In an idealized OSSE with the Lorenz-96 model and for a range of linear and nonlinear observation models, the proposed scheme improves the filter estimates, especially when the ensemble size is small relative to the dimension of the state space.

UR - http://www.scopus.com/inward/record.url?scp=85013983689&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85013983689&partnerID=8YFLogxK

U2 - 10.1175/MWR-D-16-0109.1

DO - 10.1175/MWR-D-16-0109.1

M3 - Article

AN - SCOPUS:85013983689

VL - 145

SP - 985

EP - 1001

JO - Monthly Weather Review

JF - Monthly Weather Review

SN - 0027-0644

IS - 3

ER -