Thermohaline drivers of the Arctic Ocean circulation

The Arctic Ocean has a significant effect on global ocean circulation because it provides sources of both dense and light waters to the North Atlantic. The processes affecting formation of water masses within the Arctic, however, remain poorly understood because of the sparsity of measurements available for the region. Here, we use data derived from quasi-synoptic hydrographic observations across the main Arctic gateways to diagnose water mass transformations in the Arctic interior. We see a double overturning circulation in density space. The lower cell involves the densification of approximately 1.5 Sv of Atlantic Water (1 Sv ≡ 106 m3 s−1). This is accounted for by surface buoyancy fluxes driven by heat loss on the Barents Shelf, which we quantify using ERA-Interim reanalysis data. In the upper cell, a further 1.8 Sv of inflowing Atlantic Water experiences lightening through turbulent diapycnal mixing with fresher Arctic surface waters. Turbulent diapycnal diffusivities of order 10−5 m2 s−1 are implied by the water mass transformations when averaged over the Arctic Basin. These are an order of magnitude larger than values documented by microstructure observations made around the Siberian shelf. However,observationally-based estimates of tidal energy conversion indicate the existence of highly-localised areas of enhanced turbulence. We find that sufficient energy is converted from the barotropic tide in these locations to account for the mixing inferred for the upper limb of the overturning.We assess the effects on the fresh surface layer of increasing freshwater input, using simulations from a coupled ice-ocean general circulation model. We find that, to the lowest order, the response of ocean freshwater content is linear, with an adjustment timescale of approximately 10 years. However, the details of the ocean response are seen to depend on the source of freshwater input. The response to a change in precipitation is subject to greater complexity than that to increasing river runoff because of more complex interactions with sea ice. The results presented here suggest that future increases in Arctic Ocean freshwater input in the form of precipitation are more likely to be associated with variability in the storage and release of excess freshwater than are increases in freshwater input from river runoff.