Role of the southern Indian Ocean in the transitions of the monsoon-ENSO system during recent decades

International audienceThe focus of this study is to document the possible role of the southern subtropical IndianOcean in the transitions of the monsoon-ENSO system during recent decades.Composite analyses of Sea Surface Temperature (SST) fields prior to El Niño-SouthernOscillation (ENSO), Indian Summer Monsoon (ISM), AUstralian Summer Monsoon(AUSM), Tropical Indian Ocean Dipole (TIOD) and Maritime Continent Rainfall (MCR)indices reveal the South East Indian Ocean (SEIO) SSTs during late boreal winter as theunique common SST precursor of these various phenomena after the 1976-1977 regime shift.Weak (strong) ISMs and AUSMs, El Niños (La Niñas) and positive (negative) TIOD eventsare preceded by significant negative (positive) SST anomalies in the SEIO, off Australiaduring boreal winter. These SST anomalies are mainly linked to subtropical Indian Oceandipole events, recently studied by Behera and Yamagata (2001). A wavelet analysis of aFebruary-March SEIO SST time series shows significant spectral peaks at 2 and 4-8 yearstime scales as for ENSO, ISM or AUSM indices. A composite analysis with respect toFebruary-March SEIO SSTs shows that cold (warm) SEIO SST anomalies are highlypersistent and affect the westward translation of the Mascarene high from austral to borealsummer, inducing a weakening (strengthening) of the whole ISM circulation through amodulation of the local Hadley cell during late boreal summer. At the same time, thesesubtropical SST anomalies and the associated SEIO anomalous anticyclone may be a triggerfor both the wind-evaporation-SST and wind-thermocline-SST positive feedbacks betweenAustralia and Sumatra during boreal spring and early summer. These positive feedbacksexplain the extraordinary persistence of the SEIO anomalous anticyclone from boreal springto fall. Meanwhile, the SEIO anomalous anticyclone favors persistent southeasterly windanomalies along the west coast of Sumatra and westerly wind anomalies over the westernPacific, which are well-known key-factors for the evolution of positive TIOD and El Niñoevents, respectively. A correlation analysis supports these results and shows that SEIO SSTsin February-March has higher predictive skill than other well-established ENSO predictors forforecasting Niño3.4 SST at the end of the year. This suggests again that SEIO SST anomaliesexert a fundamental influence on the transitions of the whole monsoon-ENSO system duringrecent decades