Impacts of Indian and Atlantic oceans on ENSO in a comprehensive modeling framework

The impact of the Indian and Atlantic oceans variability on El Niño–Southern-Oscillation (ENSO) phenomenon is investigated through sensitivity experiments with the SINTEX-F2 coupled model. For each experiment, we suppressed the sea surface temperature (SST) variability in either the Indian or Atlantic oceans by applying a strong nudging of the SST toward a SST climatology computed either from a control experiment or observations. In the sensitivity experiments where the nudging is done toward a control SST climatology, the Pacific mean state and seasonal cycle are not changed. Conversely, nudging toward an observed SST climatology in the Indian or Atlantic domain significantly improves the mean state and seasonal cycle, not only in the nudged domain, but also in the whole tropics. These experiments also demonstrate that decoupling the Indian or Atlantic variability modifies the phase-locking of ENSO to the annual cycle, influences significantly the timing and processes of ENSO onset and termination stages, and, finally, shifts to lower frequencies the main ENSO periodicities. Overall, these results suggest that both the Indian and Atlantic SSTs have a significant damping effect on ENSO variability and promote a shorter ENSO cycle. The reduction of ENSO amplitude is particularly significant when the Indian Ocean is decoupled, but the shift of ENSO to lower frequencies is more pronounced in the Atlantic decoupled experiments. These changes of ENSO statistical properties are related to stronger Bjerknes and thermocline feedbacks in the nudged experiments. During the mature phase of El Niño events, warm SST anomalies are found over the Indian and Atlantic oceans in observations or the control run. Consistent with previous studies, the nudged experiments demonstrate that these warm SSTs induce easterly surface wind anomalies over the far western equatorial Pacific, which fasten the transition from El Niño to La Niña and promote a shorter ENSO cycle in the control run. These results may be explained by modulations of the Walker circulation induced directly or indirectly by the Indian and Atlantic SSTs. Another interesting result is that decoupling the Atlantic or Indian oceans change the timing of ENSO onset and the relative role of other ENSO atmospheric precursors such as the extra-tropical Pacific Meridional Modes or the Western North Pacific SSTs.