Consistent link between Indian Ocean climate variability and mean state across past and future climates

Future changes in the tropical climate remain largely uncertain because climate models do not agree on whether key modes of variability, such as El Niño-Southern Oscillation (ENSO) or the Indian Ocean Dipole (IOD), will become stronger or weaker. These modes arise from multiple coupled ocean-atmosphere feedbacks that climate models fail to accurately simulate, and thereby propagate errors in the projections. We show, in contrast, that models consistently exhibit increased sea surface temperature (SST) variability in the equatorial eastern Indian ocean (EEIO), a location where SST variations drive large Walker circulation changes. This response is dominated by a robust shoaling of the thermocline in the EEIO, which locally acts to strengthen the upwelling feedback. Analysis of climate model simulations show that this feedback is strengthened because the shoaling of the thermocline increases the vertical temperature gradient at the base of the mixed layer, making upwelling variations more efficient at influencing SSTs. We show that this mechanism also leads to increased variability in simulations of the Last Glacial Maximum (LGM), 21,000 years before present (BP) and the Mid Holocene, 6,000 years BF. Although external forcings associated with each of these climates are different from the global warming case, the forced changes in mean climate exhibit a shallower thermocline in the EEIO and enhanced upwelling feedback, confirming the robustness of this mechanism. We suggest that the EEIO response to similar mean state changes among past and future climates remains consistent because it is dominated by robust changes in the upwelling feedback, compared to the tropical Pacific that involves a subtle balance of multiple feedbacks with different sensitivities. Potential impact of model biases will also be discussed.