Abstract
The climatic influence of the global overturning circulation can be studied using numerical models representing the atmosphere and ocean, as well as sea and continental ice. A variety of models, from the simplest (conceptual) to the most elaborate (3D couplings), have shown that the climate system has several stable states characterized by highly contrasting AMOC fluxes. The transition from one state to another is abrupt when a hydrological threshold is exceeded, but the models do not yet agree on the conditions defining these bifurcation points.
The other way of studying these AMOC-related shifts in the climate system is to reconstruct the intensity of this circulation using ocean sediments. Geochemical indicators (ratios 13C/12C, 14C/12C, 230Th/231Pa, 143Nd/144Nd) and sedimentological indicators (granulometry) give a semi-quantitative idea of AMOC and point to a drastic reduction during the cold transitions of the Tardiglacial : Heinrich event 1 centered around 16 ,000 years BP and Recent Dryas around 12 ,000 years BP.
These cooling events, centered on the North Atlantic and Europe, are the consequences of a reduction in the oceanic heat flux normally transported by the AMOC (1 PW). Modelling suggests that the cause of the tilts is linked to hydrological disturbances in the North Atlantic. Decades of study of marine sediments have enabled us to identify and map the increase in freshwater inflow from the melting ice caps on the periphery of the North Atlantic (Laurentian and Fennoscandian ice caps). This freshwater transport took place, on the one hand, via the postglacial reactivation of river drainage, and on the other hand, directly via the calving of icebergs into the ocean, reducing the density of surface water masses in the convection zones of the AMOC.