Multi-year prediction of the Atlantic Niño: A first approach from ENSEMBLES - CNRS - Centre national de la recherche scientifique Accéder directement au contenu
Article Dans Une Revue Física de la Tierra Année : 2014

Multi-year prediction of the Atlantic Niño: A first approach from ENSEMBLES

Predicción multianual del Niño Atlántico: Una primera aproximación desde ENSEMBLES

Résumé

This work presents for the first time a comprehensive exploration of reliability for decadal predictions upon the Atlantic Nino. Initialized and uninitialized decadal re-forecasts are used to assess multi-year prediction skill of the ATL3 SST index and to evaluate the oceanic regions, over the tropical Atlantic, providing that skill. The analysed decadal re-forecasts were carried out with the UK Met Office Decadal Prediction System (DePreSys) as part of the ENSEMBLES perturbed-parameter ensemble. Re-forecasts started once every year over 1960-2005 are employed; the verification period is 1961-2012, and the summer season considered June throughout September. Model (i.e. HadCM3) SST systematic errors and the impact of the initialization on local SST variability are described. The forecast quality assessment shows that initialized predictions of the ATL3 index yield higher correlation skill and lower root mean square error than the uninitialized counterparts until 2-5 years ahead. The results also indicate that initialization outperforms empirical predictions based on persistence over this forecast range, suggesting that ocean dynamics play a role in this near-term predictability of the Atlantic Nino.

Dates et versions

hal-01855068 , version 1 (07-08-2018)

Identifiants

Citer

Javier García-Serrano, Irene Polo, Francisco Doblas-Reyes, Reindert Haarsma. Multi-year prediction of the Atlantic Niño: A first approach from ENSEMBLES. Física de la Tierra, 2014, 25, ⟨10.5209/rev_FITE.2013.v25.43435⟩. ⟨hal-01855068⟩
29 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More