Effect of changing ocean circulation on deep ocean temperature in the last millennium

Abstract

Paleoreconstructions and modern observations provide us with anomalies of surface temperature over the past millennium. The history of deep ocean temperatures is much less well-known and was simulated in a recent study for the past 2000 years under forced surface temperature anomalies and fixed ocean circulation. In this study, we simulate the past 800 years with an illustrative forcing scenario in the Bern3D ocean model, which enables us to assess the impact of changes in ocean circulation on deep ocean temperature. We quantify the effect of changing ocean circulation by comparing transient simulations (where the ocean dynamically adjusts to anomalies in surface temperature – hence density) to simulations with fixed ocean circulation. We decompose temperature, ocean heat content and meridional heat transport into the contributions from changing ocean circulation and changing sea surface temperature (SST). In the deep ocean, the contribution from changing ocean circulation is found to be as important as the changing SST signal itself. Firstly, the small changes in ocean circulation amplify the Little Ice Age signal at around 3 km depth by at least a factor of 2, depending on the basin. Secondly, they fasten the arrival of this atmospheric signal in the Pacific and Southern Ocean at all depths, whereas they delay the arrival in the Atlantic between about 2.5 and 3.5 km by two centuries. This delay is explained by an initial competition between the Little Ice Age cooling and a warming due to an increase in relatively warmer North Atlantic Deep Water at the cost of Antarctic Bottom Water. Under the consecutive Atlantic meridional overturning circulation (AMOC) slowdown, this shift in water masses is inverted and ageing of the water causes a late additional cooling. Our results suggest that small changes in ocean circulation can have a large impact on the amplitude and timing of ocean temperature anomalies below 2 km depth.