Pacific origin of the abrupt increase in Indian Ocean heat content during the warming hiatus

Global mean surface warming has stalled since the end of the twentieth century1,2, but the net radiation imbalance at the top of the atmosphere continues to suggest an increasingly warming planet. This apparent contradiction has been reconciled by an anomalous heat flux into the ocean3–8, induced by a shift towards a La Niña-like state with cold sea surface temperatures in the eastern tropical Pacific over the past decade or so. A significant portion of the heat missing fromtheatmosphere is thereforeexpected tobestored in the Pacific Ocean. However, in situ hydrographic records indicate that Pacific Ocean heat content has been decreasing9. Here, we analyse observations along with simulations from a global ocean–sea ice model to track the pathway of heat. We find that the enhanced heat uptake by the Pacific Ocean has been compensated by an increased heat transport from the Pacific Ocean to the Indian Ocean, carried by the Indonesian throughflow. As a result, Indian Ocean heat content has increased abruptly, which accounts for more than 70% of the global ocean heat gain in the upper 700m during the past decade. We conclude that the Indian Ocean has become increasingly important in modulating global climate variability. Several studies have linked the recent pause in the rise of the global mean surface air temperature to a shift towards a more La Niña-like state in the tropical Pacific Ocean, triggered by a series of long-lasting La Niña events since the end of the twentieth century3–8. The resulting La Niña-like state has lifted relatively cold equatorial thermocline water towards the surface, producing persistently cold sea surface temperature (SST) anomalies in the eastern Pacific. Cold SST anomalies are associated with reductions in surface upward longwave radiation and latent heat flux to the atmosphere10. Thus, the net surface heat flux into the tropical Pacific Ocean increased sharply in the 2000s (refs 6,8). In addition, the cold SST anomalies and associated cooling of the atmosphere in the equatorial Pacific could also force extra-tropical stationary waves to remotely enhance heat flux into the Atlantic and Southern oceans11. Two independent studies have suggested that surface heat uptake indeed increased in the Atlantic and Southern oceans during the 2000s, leading to a downward flux of the upper ocean’s heat into the deeper ocean8,12. In agreement with enhanced surface heat uptake in the Pacific, Atlantic and Southern oceans, the global ocean heat content in the upper 700m (OHC700) increased strongly during 2003–2012 (Fig. 1a) at a rate of about 2.9 × 1022 J per decade9,12,13 (see Supplementary Information 1). However, there are significant differences in recent OHC700 changes between the major ocean basins, particularly the Pacific and Indian oceans (Fig. 1b,c; see also Supplementary Information 1 and 2). For the Pacific Ocean, OHC700 decreased during 2003–2012, in spite of the increased surface heat uptake in the eastern Pacific6,8 (Fig. 1c). In sharp contrast, the OHC700 of the Indian Ocean increased abruptly during 2003–2012, at a rate of about 2.1× 1022 J per decade (Fig. 1b), accounting for more than 70% of the global ocean heat gain in the upper 700m during that period. This suggests that a significant portion of the heat missing from the atmosphere now resides in the upper 700m of the Indian Ocean, with little explanation. Given that the OHC700 in the Indian Ocean did not increase during 1971–2000 (Fig. 1b), and that the Indian Ocean (north of 34 S) covers only 12% of the global sea ice-free ocean surface area, the marked increase of the Indian OHC700 is striking. Hence, a significant gap exists in our understanding of the heat missing from the atmosphere and its distribution between the different ocean basins. In particular, the difference between the Pacific and Indian oceans in their recent warming trends needs to be reconciled, which is the main objective of the present study. To do so, we use a series of global ocean–sea ice general circulation model simulations forced with the biascorrected twentieth century reanalysis surface fluxes14 (Methods; see also Supplementary Information 3). Two sets of six-member ensemble experiments are used: a control experiment forced with real-time surface flux fields and a reference experiment forced with climatological surface flux fields (Methods; see also Supplementary Information 4). The results from these ensemble experiments can be summarized as two sets of global OHC700 time series (Fig. 1a). The simulated global OHC700 from the control experiment follows the time variability of in situ observations since the 1950s (ref. 9) reasonably well. Note that there is no apparent drift of the global OHC700 in the reference experiment. The dominant forcing terms of the recent hiatus can be determined by comparing the global ocean heat budget terms averaged for the 1971–2000 period with the 2003–2012 period, with both periods computed as anomalies relative to the reference experiment (Fig. 1d). The heat budget indicates that the simulated global OHC700 increase since 1971 was largely driven by an increased downward longwave radiative heat flux, consistent with the thermodynamic effects of increased anthropogenic greenhouse gases in the atmosphere. This flux has accelerated in the most recent decade (that is, 2003–2012) and is damped by both an increased upward longwave radiative heat flux and latent heat flux. The Indian OHC700 shows very weak to no increase, in contrast to the strong global OHC700 increase during 1971–2000 in both the observational estimates and the control simulation (Fig. 1b). During the 2000s, however, observations show an abrupt increase,