Details
| Original language | English |
|---|---|
| Pages (from-to) | 354-365 |
| Number of pages | 12 |
| Journal | Advances in water resources |
| Volume | 86 |
| Publication status | Published - 14 Sept 2015 |
Abstract
The ensemble Kalman filter is, due to its computational efficiency, becoming more and more popular as a method for estimating both model states and parameters in hydrologic modeling, also for nonlinear state propagation models. In the ensemble Kalman filter the calculation of the error correlations, and hence the filter update, is done based on the ensemble of model evaluations and can therefore be strongly influenced by a few ensemble members with extreme values. With nonlinear state propagation models, extreme values can be a common phenomenon that can be, especially if there are nonlinearities between the observed variable and the modeled states, problematic during the filter update. An illustrative example of this problem is shown using an unsaturated flow model where the modeled states are pressure heads and observations are water content. It is demonstrated that the ensemble Kalman filter can in this case yield a deterioration of state predictions. We discuss the normal score transform and the transform with the retention function applied to the model states in order to mitigate this problem. It is shown that both transforms improve the estimation of the model states and parameters.
Keywords
- Data assimilation, EnKF, Nonlinear models, Normal score transform, State transformation, Unsaturated flow
ASJC Scopus subject areas
- Environmental Science(all)
- Water Science and Technology
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
In: Advances in water resources, Vol. 86, 14.09.2015, p. 354-365.
Research output: Contribution to journal › Article › Research › peer review
}
TY - JOUR
T1 - The importance of state transformations when using the ensemble Kalman filter for unsaturated flow modeling
T2 - Dealing with strong nonlinearities
AU - Erdal, D.
AU - Rahman, M. A.
AU - Neuweiler, I.
N1 - Funding information: Financial support from the Deutsche Forschungsgemeinschaft (DFG) under NE 824/12-1 and CI 26/13-1 in the framework of research unit FOR 2131 is gratefully acknowledged.
PY - 2015/9/14
Y1 - 2015/9/14
N2 - The ensemble Kalman filter is, due to its computational efficiency, becoming more and more popular as a method for estimating both model states and parameters in hydrologic modeling, also for nonlinear state propagation models. In the ensemble Kalman filter the calculation of the error correlations, and hence the filter update, is done based on the ensemble of model evaluations and can therefore be strongly influenced by a few ensemble members with extreme values. With nonlinear state propagation models, extreme values can be a common phenomenon that can be, especially if there are nonlinearities between the observed variable and the modeled states, problematic during the filter update. An illustrative example of this problem is shown using an unsaturated flow model where the modeled states are pressure heads and observations are water content. It is demonstrated that the ensemble Kalman filter can in this case yield a deterioration of state predictions. We discuss the normal score transform and the transform with the retention function applied to the model states in order to mitigate this problem. It is shown that both transforms improve the estimation of the model states and parameters.
AB - The ensemble Kalman filter is, due to its computational efficiency, becoming more and more popular as a method for estimating both model states and parameters in hydrologic modeling, also for nonlinear state propagation models. In the ensemble Kalman filter the calculation of the error correlations, and hence the filter update, is done based on the ensemble of model evaluations and can therefore be strongly influenced by a few ensemble members with extreme values. With nonlinear state propagation models, extreme values can be a common phenomenon that can be, especially if there are nonlinearities between the observed variable and the modeled states, problematic during the filter update. An illustrative example of this problem is shown using an unsaturated flow model where the modeled states are pressure heads and observations are water content. It is demonstrated that the ensemble Kalman filter can in this case yield a deterioration of state predictions. We discuss the normal score transform and the transform with the retention function applied to the model states in order to mitigate this problem. It is shown that both transforms improve the estimation of the model states and parameters.
KW - Data assimilation
KW - EnKF
KW - Nonlinear models
KW - Normal score transform
KW - State transformation
KW - Unsaturated flow
UR - http://www.scopus.com/inward/record.url?scp=84949666485&partnerID=8YFLogxK
U2 - 10.1016/j.advwatres.2015.09.008
DO - 10.1016/j.advwatres.2015.09.008
M3 - Article
AN - SCOPUS:84949666485
VL - 86
SP - 354
EP - 365
JO - Advances in water resources
JF - Advances in water resources
SN - 0309-1708
ER -