Tracking the fate of meltwater from different Northern ice sheet sectors over Heinrich Stadial

The addition of meltwater from continental ice sheets to the North Atlantic is thought to have played a pivotal role in the reorganization of climate and ocean circulation over the last deglaciation as well as during Heinrich events. This is supported by recent analysis of PMIP and CMIP results, which shows that meltwater addition into the North Atlantic can largely alter global climate, and remains a key uncertainty for both reconstruction and climate projections.  To date, most model studies of freshwater “hosing” assume a relatively uniform distribution and apply meltwater to a large portion of the North Atlantic basin. However, AMOC weakening is sensitive to the actual input position of the typically cold and non-saline meltwater perturbation, and, on a centennial-millennial timescale, the resulting temperature and salt anomaly will only partially disperse over the entire North Atlantic surface ocean. In contrast, most proxy data sensitive to meltwater record the signal at a specific location. It is unclear if spatial heterogeneity of the ocean’s distribution of the meltwater anomaly may contribute to disagreements between freshwater proxy records and model simulations with freshwater additions tuned to reproduce the record of past AMOC weakenings. To enhance understanding of the spatial distribution of meltwater anomalies during deglaciations, we present the results of a model sensitivity study using HadCM3 and artificial dye tracers to track the fate of meltwater originating from different Northern ice sheet sectors. We consider different meltwater scenarios consistent with Heinrich Stadial 1 ice sheet reconstructions and compare the results under different AMOC states. The results confirm that, on a centennial timescale, meltwater distribution is not uniform over the North Atlantic Ocean. The emerging patterns expose that the efficiency of a meltwater injection to produce a surface ocean anomaly (in, e.g., salinity or δ18Osw) at a given proxy location differs between different ice sheet sectors by orders of magnitudes. Further, besides the direct effect of meltwater, the sensitivity of climate indicators, such as temperature, to changes in AMOC strength also shows regional discrepancies.