Weather regime prediction using statistical learning

Axel Deloncle 1, 2, * R. Berk 3 F. d'Andrea 2 M. Ghil 2, 4, 5
* Corresponding author
1 LadHyX - Laboratoire d'hydrodynamique
2 LMD - Laboratoire de Météorologie Dynamique (UMR 8539)
3 Department of Statistics, University of California Los Angeles, Los Angeles, CA, United States
4 AOS - Department of Atmospheric and Oceanic Sciences [Los Angeles]
5 IGPP - Institute of Geophysics and Planetary Physics [Los Angeles]
Abstract : Two novel statistical methods are applied to the prediction of transitions between weather regimes. The methods are tested using a long, 6000-day simulation of a three-layer, quasigeostrophic (QG3) model on the sphere at T21 resolution. The two methods are the k nearest neighbor classifier and the random forest method. Both methods are widely used in statistical classification and machine learning; they are applied here to forecast the break of a regime and subsequent onset of another one. The QG3 model has been previously shown to possess realistic weather regimes in its northern hemisphere and preferred transitions between these have been determined. The two methods are applied to the three more robust transitions; they both demonstrate a skill of 35%-40% better than random and are thus encouraging for use on real data. Moreover, the random forest method allows one, while keeping the overall skill unchanged, to efficiently adjust the ratio of correctly predicted transitions to false alarms. A long-standing conjecture has associated regime breaks and preferred transitions with distinct directions in the reduced model phase space spanned by a few leading empirical orthogonal functions of its variability. Sensitivity studies for several predictors confirm the crucial influence of the exit angle on a preferred transition path. The present results thus support the paradigm of multiple weather regimes and their association with unstable fixed points of atmospheric dynamics. © 2007 American Meteorological Society.
Document type :
Journal articles
Complete list of metadatas

Cited literature [28 references]  Display  Hide  Download
https://hal-polytechnique.archives-ouvertes.fr/hal-01023347
Contributor : Denis Roura <>
Submitted on : Sunday, July 20, 2014 - 6:54:44 PM
Last modification on : Wednesday, August 7, 2019 - 12:14:40 PM
Long-term archiving on: Thursday, November 20, 2014 - 6:16:20 PM

File

jas39182E1.pdf
Publisher files allowed on an open archive

Identifiers

Collections

CNRS | X-LADHYX | X-LMD | LMD | PARISTECH | INSU | UPMC | X | ENS-PARIS | PSL | ENPC-LMD | ENPC | X-DEP-MECA | SORBONNE-UNIVERSITE | SU-SCIENCES

Citation

Axel Deloncle, R. Berk, F. d'Andrea, M. Ghil. Weather regime prediction using statistical learning. Journal of the Atmospheric Sciences, American Meteorological Society, 2007, 64 (5), pp.1619-1635. ⟨10.1175/jas3918.1⟩. ⟨hal-01023347⟩

Export

BibTeX TEI DC DCterms EndNote

Share

Metrics

Record views

804

Files downloads

244

Contact

Informations

CCSD