A fast, single-iteration ensemble Kalman smoother for sequential data assimilation
<p>Ensemble variational methods form the basis of the state of the art for nonlinear, scalable data assimilation, yet current designs may not be cost-effective for real-time, short-range forecast systems. We propose a novel estimator in this formalism that is designed for applications in which...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Copernicus Publications
2022-10-01
|
| Series: | Geoscientific Model Development |
| Online Access: | https://gmd.copernicus.org/articles/15/7641/2022/gmd-15-7641-2022.pdf |
| _version_ | 1797985875615809536 |
|---|---|
| author | C. Grudzien C. Grudzien M. Bocquet |
| author_facet | C. Grudzien C. Grudzien M. Bocquet |
| author_sort | C. Grudzien |
| collection | DOAJ |
| description | <p>Ensemble variational methods form the basis of the state of the art for nonlinear, scalable data assimilation, yet current designs may not be cost-effective for real-time, short-range forecast systems. We propose a novel estimator in this formalism that is designed for applications in which forecast error dynamics is weakly nonlinear, such as synoptic-scale meteorology. Our method combines the 3D sequential filter analysis and retrospective reanalysis of the classic ensemble Kalman smoother with an iterative ensemble simulation of 4D smoothers. To rigorously derive and contextualize our method, we review related ensemble smoothers in a Bayesian maximum a posteriori narrative. We then develop and intercompare these schemes in the open-source Julia package DataAssimilationBenchmarks.jl, with pseudo-code provided for their implementations. This numerical framework, supporting our mathematical results, produces extensive benchmarks demonstrating the significant performance advantages of our proposed technique. Particularly, our single-iteration ensemble Kalman smoother (SIEnKS) is shown to improve prediction/analysis accuracy and to simultaneously reduce the leading-order computational cost of iterative smoothing in a variety of test cases relevant for short-range forecasting. This long work presents our novel SIEnKS and provides a theoretical and computational framework for the further development of ensemble variational Kalman filters and smoothers.</p> |
| first_indexed | 2024-04-11T07:23:57Z |
| format | Article |
| id | doaj.art-15d6666e51bd4e61bfb1588a6aae26e2 |
| institution | Directory Open Access Journal |
| issn | 1991-959X 1991-9603 |
| language | English |
| last_indexed | 2024-04-11T07:23:57Z |
| publishDate | 2022-10-01 |
| publisher | Copernicus Publications |
| record_format | Article |
| series | Geoscientific Model Development |
| spelling | doaj.art-15d6666e51bd4e61bfb1588a6aae26e22022-12-22T04:37:07ZengCopernicus PublicationsGeoscientific Model Development1991-959X1991-96032022-10-01157641768110.5194/gmd-15-7641-2022A fast, single-iteration ensemble Kalman smoother for sequential data assimilationC. Grudzien0C. Grudzien1M. Bocquet2Center for Western Weather and Water Extremes (CW3E), Scripps Institution of Oceanography, University of California San Diego, San Diego, CA, USADepartment of Mathematics and Statistics, University of Nevada, Reno, Reno, Nevada, USACEREA, École des Ponts and EDF R&D, Île-de-France, France<p>Ensemble variational methods form the basis of the state of the art for nonlinear, scalable data assimilation, yet current designs may not be cost-effective for real-time, short-range forecast systems. We propose a novel estimator in this formalism that is designed for applications in which forecast error dynamics is weakly nonlinear, such as synoptic-scale meteorology. Our method combines the 3D sequential filter analysis and retrospective reanalysis of the classic ensemble Kalman smoother with an iterative ensemble simulation of 4D smoothers. To rigorously derive and contextualize our method, we review related ensemble smoothers in a Bayesian maximum a posteriori narrative. We then develop and intercompare these schemes in the open-source Julia package DataAssimilationBenchmarks.jl, with pseudo-code provided for their implementations. This numerical framework, supporting our mathematical results, produces extensive benchmarks demonstrating the significant performance advantages of our proposed technique. Particularly, our single-iteration ensemble Kalman smoother (SIEnKS) is shown to improve prediction/analysis accuracy and to simultaneously reduce the leading-order computational cost of iterative smoothing in a variety of test cases relevant for short-range forecasting. This long work presents our novel SIEnKS and provides a theoretical and computational framework for the further development of ensemble variational Kalman filters and smoothers.</p>https://gmd.copernicus.org/articles/15/7641/2022/gmd-15-7641-2022.pdf |
| spellingShingle | C. Grudzien C. Grudzien M. Bocquet A fast, single-iteration ensemble Kalman smoother for sequential data assimilation Geoscientific Model Development |
| title | A fast, single-iteration ensemble Kalman smoother for sequential data assimilation |
| title_full | A fast, single-iteration ensemble Kalman smoother for sequential data assimilation |
| title_fullStr | A fast, single-iteration ensemble Kalman smoother for sequential data assimilation |
| title_full_unstemmed | A fast, single-iteration ensemble Kalman smoother for sequential data assimilation |
| title_short | A fast, single-iteration ensemble Kalman smoother for sequential data assimilation |
| title_sort | fast single iteration ensemble kalman smoother for sequential data assimilation |
| url | https://gmd.copernicus.org/articles/15/7641/2022/gmd-15-7641-2022.pdf |
| work_keys_str_mv | AT cgrudzien afastsingleiterationensemblekalmansmootherforsequentialdataassimilation AT cgrudzien afastsingleiterationensemblekalmansmootherforsequentialdataassimilation AT mbocquet afastsingleiterationensemblekalmansmootherforsequentialdataassimilation AT cgrudzien fastsingleiterationensemblekalmansmootherforsequentialdataassimilation AT cgrudzien fastsingleiterationensemblekalmansmootherforsequentialdataassimilation AT mbocquet fastsingleiterationensemblekalmansmootherforsequentialdataassimilation |