Following observations of a drop in the West Antarctic Ice Sheet (WAIS) mass balance over the last few decades with the possibility to reach a tipping point leading to ineluctable glaciers outlets instability in the region, understanding the driving processes has become a priority. In particular, the Circumpolar Deep Water (CDW) intrusion onto the continental shelf in the Amundsen Sea is, nowadays, in the spotlight, and gathers the attention of both the observers and the modellers. This modelling study presents the analysis of a 1/12° simulation of the Amundsen Sea sector reproducing well the interannual-to-decadal variability of the ice-shelves basal melt rates. The development of a methodology to study the ocean state in the reference frame of the continental shelf break enables us to distinguish and characterize a western fresh shelf zone and an eastern warm shelf zone in the region. Connecting it with the more regional circulation, we try to shed light on the different mechanisms driving the CDW inflow onto the continental shelf in the region. In particular, we draw attention to the sea ice effect in terms of Ekman pumping along the shelf break, and we point out the possible initiation of a southern Antarctic Circumpolar Current (ACC) branch to the south-east of the Ross Gyre, which could control part of the variability along the Amundsen Sea shelf. Finally, we discuss correlations between the ocean variability at the shelf break and the one of the melt activity underneath the ice-shelves.

Understanding the driving processes at stake for the Circumpolar Deep Water (CDW) intrusion onto the Amundsen shelf is crucial. We use a multi-decadal ocean simulation at 1/12° to revisit the ocean dynamics at the Amundsen shelf break, distinguishing a western fresh shelf and an eastern warm shelf. While the prevailing presence of the Antarctic Slope Current - fed to the east of Russel Bay through vortex stretching of an outflow of melted waters - blocks CDW intrusions in the west, the contact of Antarctic Circumpolar Current (ACC) branches along the shelf in the east favors this inflow. Of particular importance is a southern ACC branch initiated to the south-east of the Ross Gyre, which interacts with the topography at the entry of the western Pine Island-Thwaites trough. Then, we link the ocean interannual-to-decadal variability at the shelf break with the ice-shelf basal melting and create a Fresh-Warm Boundary Index (FWBI) to follow the oscillation of the fresh-warm shelves limit through time in Russel Bay, which could be a focal point to understand the low frequency fluctuations of the basal melt. We suggest that not only a wind-induced Ekman pumping could favor the CDW inflow at the shelf break, but also topographic interactions, a bottom Ekman transport, a sea-ice–induced Ekman pumping resulting from strong surface currents, and the baroclinicity of the eastward along-shelf current in the west. Finally, we highlight that El Niño-Southern Oscillation has no strong correlation with the ice-shelf basal melt variability, except for the very recent years.