An adjoint-based adaptive ensemble kalman filter

Hajoon Song, Ibrahim Hoteit, Bruce D. Cornuelle, Xiaodong Luo, Aneesh C. Subramanian

Research output: Contribution to journalArticle

7 Citations (Scopus)

Abstract

A new hybrid ensemble Kalman filter/four-dimensional variational data assimilation (EnKF/4D-VAR) approach is introduced to mitigate background covariance limitations in the EnKF. The work is based on the adaptive EnKF (AEnKF) method, which bears a strong resemblance to the hybrid EnKF/three-dimensional variational data assimilation (3D-VAR) method. In the AEnKF, the representativeness of the EnKF ensemble is regularly enhanced with new members generated after back projection of the EnKF analysis residuals to state space using a 3D-VAR [or optimal interpolation (OI)] scheme with a preselected background covariance matrix. The idea here is to reformulate the transformation of the residuals as a 4D-VAR problem, constraining the new member with model dynamics and the previous observations. This should provide more information for the estimation of the new member and reduce dependence of the AEnKF on the assumed stationary background covariance matrix. This is done by integrating the analysis residuals backward in time with the adjoint model. Numerical experiments are performed with the Lorenz-96 model under different scenarios to test the new approach and to evaluate its performance with respect to the EnKF and the hybrid EnKF/3D-VAR. The new method leads to the least root-mean-square estimation errors as long as the linear assumption guaranteeing the stability of the adjoint model holds. It is also found to be less sensitive to choices of the assimilation system inputs and parameters.

Original languageEnglish
Pages (from-to)3343-3359
Number of pages17
JournalMonthly Weather Review
Volume141
Issue number10
DOIs
Publication statusPublished - 2013 Oct 8

Fingerprint

Kalman filter
data assimilation
matrix
interpolation
method
experiment
analysis

All Science Journal Classification (ASJC) codes

  • Atmospheric Science

Cite this

Song, H., Hoteit, I., Cornuelle, B. D., Luo, X., & Subramanian, A. C. (2013). An adjoint-based adaptive ensemble kalman filter. Monthly Weather Review, 141(10), 3343-3359. https://doi.org/10.1175/MWR-D-12-00244.1
Song, Hajoon ; Hoteit, Ibrahim ; Cornuelle, Bruce D. ; Luo, Xiaodong ; Subramanian, Aneesh C. / An adjoint-based adaptive ensemble kalman filter. In: Monthly Weather Review. 2013 ; Vol. 141, No. 10. pp. 3343-3359.
@article{03e7baeafbb44397a1415162ab81c15d,
title = "An adjoint-based adaptive ensemble kalman filter",
abstract = "A new hybrid ensemble Kalman filter/four-dimensional variational data assimilation (EnKF/4D-VAR) approach is introduced to mitigate background covariance limitations in the EnKF. The work is based on the adaptive EnKF (AEnKF) method, which bears a strong resemblance to the hybrid EnKF/three-dimensional variational data assimilation (3D-VAR) method. In the AEnKF, the representativeness of the EnKF ensemble is regularly enhanced with new members generated after back projection of the EnKF analysis residuals to state space using a 3D-VAR [or optimal interpolation (OI)] scheme with a preselected background covariance matrix. The idea here is to reformulate the transformation of the residuals as a 4D-VAR problem, constraining the new member with model dynamics and the previous observations. This should provide more information for the estimation of the new member and reduce dependence of the AEnKF on the assumed stationary background covariance matrix. This is done by integrating the analysis residuals backward in time with the adjoint model. Numerical experiments are performed with the Lorenz-96 model under different scenarios to test the new approach and to evaluate its performance with respect to the EnKF and the hybrid EnKF/3D-VAR. The new method leads to the least root-mean-square estimation errors as long as the linear assumption guaranteeing the stability of the adjoint model holds. It is also found to be less sensitive to choices of the assimilation system inputs and parameters.",
author = "Hajoon Song and Ibrahim Hoteit and Cornuelle, {Bruce D.} and Xiaodong Luo and Subramanian, {Aneesh C.}",
year = "2013",
month = "10",
day = "8",
doi = "10.1175/MWR-D-12-00244.1",
language = "English",
volume = "141",
pages = "3343--3359",
journal = "Monthly Weather Review",
issn = "0027-0644",
publisher = "American Meteorological Society",
number = "10",

}

Song, H, Hoteit, I, Cornuelle, BD, Luo, X & Subramanian, AC 2013, 'An adjoint-based adaptive ensemble kalman filter', Monthly Weather Review, vol. 141, no. 10, pp. 3343-3359. https://doi.org/10.1175/MWR-D-12-00244.1

An adjoint-based adaptive ensemble kalman filter. / Song, Hajoon; Hoteit, Ibrahim; Cornuelle, Bruce D.; Luo, Xiaodong; Subramanian, Aneesh C.

In: Monthly Weather Review, Vol. 141, No. 10, 08.10.2013, p. 3343-3359.

Research output: Contribution to journalArticle

TY - JOUR

T1 - An adjoint-based adaptive ensemble kalman filter

AU - Song, Hajoon

AU - Hoteit, Ibrahim

AU - Cornuelle, Bruce D.

AU - Luo, Xiaodong

AU - Subramanian, Aneesh C.

PY - 2013/10/8

Y1 - 2013/10/8

N2 - A new hybrid ensemble Kalman filter/four-dimensional variational data assimilation (EnKF/4D-VAR) approach is introduced to mitigate background covariance limitations in the EnKF. The work is based on the adaptive EnKF (AEnKF) method, which bears a strong resemblance to the hybrid EnKF/three-dimensional variational data assimilation (3D-VAR) method. In the AEnKF, the representativeness of the EnKF ensemble is regularly enhanced with new members generated after back projection of the EnKF analysis residuals to state space using a 3D-VAR [or optimal interpolation (OI)] scheme with a preselected background covariance matrix. The idea here is to reformulate the transformation of the residuals as a 4D-VAR problem, constraining the new member with model dynamics and the previous observations. This should provide more information for the estimation of the new member and reduce dependence of the AEnKF on the assumed stationary background covariance matrix. This is done by integrating the analysis residuals backward in time with the adjoint model. Numerical experiments are performed with the Lorenz-96 model under different scenarios to test the new approach and to evaluate its performance with respect to the EnKF and the hybrid EnKF/3D-VAR. The new method leads to the least root-mean-square estimation errors as long as the linear assumption guaranteeing the stability of the adjoint model holds. It is also found to be less sensitive to choices of the assimilation system inputs and parameters.

AB - A new hybrid ensemble Kalman filter/four-dimensional variational data assimilation (EnKF/4D-VAR) approach is introduced to mitigate background covariance limitations in the EnKF. The work is based on the adaptive EnKF (AEnKF) method, which bears a strong resemblance to the hybrid EnKF/three-dimensional variational data assimilation (3D-VAR) method. In the AEnKF, the representativeness of the EnKF ensemble is regularly enhanced with new members generated after back projection of the EnKF analysis residuals to state space using a 3D-VAR [or optimal interpolation (OI)] scheme with a preselected background covariance matrix. The idea here is to reformulate the transformation of the residuals as a 4D-VAR problem, constraining the new member with model dynamics and the previous observations. This should provide more information for the estimation of the new member and reduce dependence of the AEnKF on the assumed stationary background covariance matrix. This is done by integrating the analysis residuals backward in time with the adjoint model. Numerical experiments are performed with the Lorenz-96 model under different scenarios to test the new approach and to evaluate its performance with respect to the EnKF and the hybrid EnKF/3D-VAR. The new method leads to the least root-mean-square estimation errors as long as the linear assumption guaranteeing the stability of the adjoint model holds. It is also found to be less sensitive to choices of the assimilation system inputs and parameters.

UR - http://www.scopus.com/inward/record.url?scp=84884972541&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84884972541&partnerID=8YFLogxK

U2 - 10.1175/MWR-D-12-00244.1

DO - 10.1175/MWR-D-12-00244.1

M3 - Article

VL - 141

SP - 3343

EP - 3359

JO - Monthly Weather Review

JF - Monthly Weather Review

SN - 0027-0644

IS - 10

ER -

Song H, Hoteit I, Cornuelle BD, Luo X, Subramanian AC. An adjoint-based adaptive ensemble kalman filter. Monthly Weather Review. 2013 Oct 8;141(10):3343-3359. https://doi.org/10.1175/MWR-D-12-00244.1