Current browse context:
stat.AP
Change to browse by:
References & Citations
Statistics > Applications
Title: Penalized Ensemble Kalman Filters for High Dimensional Non-linear Systems
(Submitted on 1 Oct 2016 (this version), latest version 11 Mar 2021 (v3))
Abstract: The ensemble Kalman filter (EnKF) is a data assimilation technique that uses an ensemble of models, updated with data, to track the time evolution of a non-linear system. It does so by using an empirical approximation to the well-known Kalman filter. Unfortunately, its performance suffers when the ensemble size is smaller than the state space, as is often the case for computationally burdensome models. This scenario means that the empirical estimate of the state covariance is not full rank and possibly quite noisy. To solve this problem in this high dimensional regime, a computationally fast and easy to implement algorithm called the penalized ensemble Kalman filter (PEnKF) is proposed. Under certain conditions, it can be proved that the PEnKF does not require more ensemble members than state dimensions in order to have good performance. Further, the proposed approach does not require special knowledge of the system such as is used by localization methods. These theoretical results are supported with superior performance in simulations of several non-linear and high dimensional systems.
Submission history
From: Elizabeth Hou [view email][v1] Sat, 1 Oct 2016 21:52:24 GMT (609kb,D)
[v2] Wed, 14 Mar 2018 22:19:38 GMT (617kb,D)
[v3] Thu, 11 Mar 2021 02:58:23 GMT (1289kb,D)
Link back to: arXiv, form interface, contact.