Exploring the Physical Causes for Inter-Model Differences in predictions of future THC-related climate change under global warming

Most current coupled ocean/atmosphere climate models simulate a decrease in the oceanic thermohaline circulation in response to anthropogenic global warming. As a result, the models usually simulate a reduction in the northward meridional heat transport in the Atlantic ocean, somewhat mitigating the effects of global warming in the northern hemisphere, while exacerbating them in the southern hemisphere. Large uncertainties remain, however, because the predicted changes can vary greatly from one model to the other, with the possible responses ranging from near stability to an almost complete shutdown of the thermohaline circulation. To understand the physical causes for these inter-model differences, an intercomparison of a coordinated sets of experiments has been undertaken both as an international experiment supplementing the Coupled Model Intercomparison Project and by the UK RAPID programme. In total, about 20 coupled models, comprising both full AOGCMs and EMICs (Earth Model of Intermediate Complexity), are intercompared. The experiments are a CO2 increase experiment and a water-hosing experiment in which freshwater is released at high-latitudes. The first experiment is intended to explore future THC-related climate change with an idealised scenario of increasing anthropogenic GHG emissions, whereas the second one focuses on the effect of freshwater forcing in a sensitive region, and seeks to determine whether the models might possess bistable THC regimes by trying to push them accross their hypothesised stability threshold. This talk will describe the understanding achieved so far. Some of the results are the following: In both kind of experiments, no model shows a rapid, complete, or irreversible collapse. In the CO2 experiments, the models having the strongest overturning in the control climate tend to show the largest THC reductions. In all the models, the THC weakening is caused more by changes in surface heat flux than by changes in surface water flux. No model shows a cooling anywhere, because the greenhouse warming is dominant. In the water-hosing experiments, the reduction of the THC is associated with the suppression of deep convection. Some models are able to stabilise their THC by shifting their deep convection sites north of the freshwater input zone. There is in general a southward shift of the ITCZ, and enhanced evaporation in the tropical/subtropical region which may increase salinity there, which may counteract somewhat the freshening at higher-latitudes.