Improving predictions of Indian Ocean climate change using paleoclimate data

Abstract:

Models predict a wide range of changes in the climate of the Indian Ocean (IO) and its variability, hindering adaptation strategies in densely populated countries surrounding it. These outcomes depend on whether coupled feedbacks will amplify the changes induced by greenhouse warming or not and could lead to dramatic changes in rainfall patterns and hydrological extremes. The activation of coupled feedbacks has been attributed to model errors, further obscuring the reliability of these predictions. We assessed the viability of these mechanisms in simulations of the Last Glacial Maximum (LGM), a past climatic interval when paleoclimate proxies show large-scale changes in rainfall patterns and oceanographic conditions across the IO. According to the simulations, these patterns can only be explained by the same coupled feedbacks acting in future predictions. Furthermore, both past and future changes show evidence of a mode of climate variability capable of generating unprecedented sea surface temperature and rainfall variability. This mode, which is inhibited under present-day conditions, becomes active in climate states with a shallow thermocline and vigorous upwelling, consistent with the predictions of continued greenhouse warming. These model predictions are supported by modeling and proxy evidence that this mode was active during the LGM. Together these results increase our confidence in the models predicting large climate changes and more extreme variability under continued greenhouse gas emissions.