Premium
Improvement of ENSO prediction using a linear regression model with a southern Indian Ocean sea surface temperature predictor
Author(s) -
Dominiak Sébastien,
Terray Pascal
Publication year - 2005
Publication title -
geophysical research letters
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.007
H-Index - 273
eISSN - 1944-8007
pISSN - 0094-8276
DOI - 10.1029/2005gl023153
Subject(s) - climatology , sea surface temperature , boreal , anomaly (physics) , el niño southern oscillation , environmental science , linear regression , geology , oceanography , paleontology , physics , machine learning , computer science , condensed matter physics
This study presents a detailed comparison between three ENSO precursors which can predict across the spring persistence barrier: the anomalous equatorial Pacific upper ocean heat content, the zonal equatorial wind stress anomaly in the far‐western Pacific and SST anomalies in the South‐East Indian Ocean (SEIO) during the late boreal winter. A new correlation analysis confirms that El Niño (La Niña) onsets are preceded by significant cold (warm) SST anomalies in the SEIO during the late boreal winter after the 1976–77 climate regime shift. Thus, the objective is to examine the respective potential of these three ENSO precursors to predict ENSO events across the boreal spring barrier during recent decades. Surprisingly, in this focus, cross‐validated hindcasts of the linear regression models based on the lagged relationship between Niño3.4 SST and the predictors suggest that SEIO SST anomalies during the late boreal winter is the more robust ENSO predictor.