Climate Dynamics

, Volume 28, Issue 6, pp 553–580

Impact of southeast Indian Ocean sea surface temperature anomalies on monsoon-ENSO-dipole variability in a coupled ocean–atmosphere model


DOI: 10.1007/s00382-006-0192-y

Cite this article as:
Terray, P., Chauvin, F. & Douville, H. Clim Dyn (2007) 28: 553. doi:10.1007/s00382-006-0192-y


Recent studies show that SouthEast Indian Ocean (SEIO) SSTs are a highly significant precursor of transitions of the whole monsoon-El Niño-Southern Oscillation (ENSO) system during recent decades. However, the reasons for this specific interannual variability have not yet been identified unequivocally from the observations. Among these, the possibility of SEIO SST-driven variability in the monsoon-ENSO system is investigated here by inserting positive/negative SEIO temperature anomalies in the February’s restart files of a state-of-the-art coupled General Circulation Model (GCM) for 49 years of a control simulation. For each year of the control simulation, the model was then integrated for a 1-year period in fully coupled mode. These experiments show that Indian Summer Monsoon (ISM) and tropical Indian Ocean Dipole Mode (IODM) events are significantly influenced by the SEIO temperature perturbations inserted in the mixed layer of the coupled GCM several months before. A warm SEIO perturbation, inserted in late boreal winter, slowly propagates northward during the following seasons, implies enhanced ISM rainfall and finally triggers a negative IODM pattern during boreal fall in agreement with observations. A reversed evolution is simulated for a cold SEIO perturbation. It is shown that the life cycle of the simulated SEIO signal is driven by the positive wind-evaporation-SST, coastal upwelling and wind-thermocline-SST feedbacks. Further diagnosis of the sensitivity experiments suggests that stronger ISM and IODM variabilities are generated by excluding the El Niño years of the control simulation or when the initial background state in the SEIO is warmer. This finding confirms that IODM events may be triggered by multiple factors, other than ENSO, including subtropical SEIO SST anomalies. However, the ENSO mode does not react significantly to the SEIO temperature perturbation in the perturbed runs even though the simulated Pacific pattern agrees with the observations during boreal fall. These discrepancies with the observations may be linked to model biases in the Pacific and to the too strong ENSO simulated by this coupled GCM. These modeling evidences confirm that subtropical Indian Ocean SST anomalies generated by Mascarene high pulses during austral summer are a significant precursor of both ISM and IODM events occuring several months later.

Copyright information

© Springer-Verlag 2006

Authors and Affiliations

  • Pascal Terray
    • 1
    • 3
  • Fabrice Chauvin
    • 2
  • Hervé Douville
    • 2
  1. 1.LOCEANInstitut Pierre Simon Laplace (IPSL) and Université Paris 7ParisFrance
  2. 2.Météo-France CNRM/GMGEC/UDCToulouseFrance
  3. 3.LOCEAN/IPSLUniversité Pierre et Marie CurieParis CEDEX 05France

Personalised recommendations