dorsal/arxiv
View SchemaA Local Ensemble Kalman Filter for Atmospheric Data Assimilation
| Authors | Edward Ott, Brian R. Hunt, Istvan Szunyogh, Aleksey V. Zimin, Eric J. Kostelich, Matteo Corazza, Eugenia Kalnay, D. J. Patil, James A. Yorke |
|---|---|
| Categories | |
| ArXiv ID | physics/0203058 |
| URL | https://arxiv.org/abs/physics/0203058 |
Abstract
In this paper, we introduce a new, local formulation of the ensemble Kalman Filter approach for atmospheric data assimilation. Our scheme is based on the hypothesis that, when the Earth's surface is divided up into local regions of moderate size, vectors of the forecast uncertainties in such regions tend to lie in a subspace of much lower dimension than that of the full atmospheric state vector of such a region. Ensemble Kalman Filters, in general, assume that the analysis resulting from the data assimilation lies in the same subspace as the expected forecast error. Under our hypothesis the dimension of this subspace is low. This implies that operations only on relatively low dimensional matrices are required. Thus, the data analysis is done locally in a manner allowing massively parallel computation to be exploited. The local analyses are then used to construct global states for advancement to the next forecast time. The method, its potential advantages, properties, and implementation requirements are illustrated by numerical experiments on the Lorenz-96 model. It is found that accurate analysis can be achieved at a cost which is very modest compared to that of a full global ensemble Kalman Filter.
{
"annotation_id": "b3f475df-cc3c-441b-ad70-2279588c571f",
"date_created": "2026-03-02T18:00:39.032000Z",
"date_modified": "2026-03-02T18:00:39.032000Z",
"file_hash": "99135f8ed8285a010e2a5ca3162308bd3b62e247569a07588f1d02f4a1ffb9ee",
"private": false,
"record": {
"abstract": "In this paper, we introduce a new, local formulation of the ensemble Kalman\nFilter approach for atmospheric data assimilation. Our scheme is based on the\nhypothesis that, when the Earth\u0027s surface is divided up into local regions of\nmoderate size, vectors of the forecast uncertainties in such regions tend to\nlie in a subspace of much lower dimension than that of the full atmospheric\nstate vector of such a region. Ensemble Kalman Filters, in general, assume that\nthe analysis resulting from the data assimilation lies in the same subspace as\nthe expected forecast error. Under our hypothesis the dimension of this\nsubspace is low. This implies that operations only on relatively low\ndimensional matrices are required. Thus, the data analysis is done locally in a\nmanner allowing massively parallel computation to be exploited. The local\nanalyses are then used to construct global states for advancement to the next\nforecast time. The method, its potential advantages, properties, and\nimplementation requirements are illustrated by numerical experiments on the\nLorenz-96 model. It is found that accurate analysis can be achieved at a cost\nwhich is very modest compared to that of a full global ensemble Kalman Filter.",
"arxiv_id": "physics/0203058",
"authors": [
"Edward Ott",
"Brian R. Hunt",
"Istvan Szunyogh",
"Aleksey V. Zimin",
"Eric J. Kostelich",
"Matteo Corazza",
"Eugenia Kalnay",
"D. J. Patil",
"James A. Yorke"
],
"categories": [
"physics.ao-ph",
"nlin.CD",
"physics.data-an",
"physics.geo-ph"
],
"title": "A Local Ensemble Kalman Filter for Atmospheric Data Assimilation",
"url": "https://arxiv.org/abs/physics/0203058"
},
"schema_id": "dorsal/arxiv",
"source": {
"execution_id": "66ea63f2-d5c1-4729-b2aa-7da6c260e99a",
"id": "arXiv Dataset IDs",
"type": "Model",
"variant": "snapshot-2026-03-01",
"version": "0.1.0"
},
"user_id": 1000002
}