On the Western Boundary Current System of the Weddell Gyre: a model intercomparison

The Weddell Sea is located in the Southern Ocean, bounded to the south and west by the Antarctic continent and the Antarctic Peninsula, respectively, and to the north by the Antarctic Circumpolar Current. The cyclonic Weddell Gyre stands out as the dominant feature of the basin circulation, driven by wind and thermohaline forcing as well as topographic steering. Importantly, it is the primary source region of Antarctic Bottom Water (AABW), thus becoming one of the key regions for the global thermohaline circulation. Furthermore, the geographical location of the Western Boundary Current System (WBCS) developed in the gyre allows the leakage of near-freezing subsurface waters into the Bransfield Strait. This cold-water pathway has been recently suggested to maintain regionally low rates of glacier retreat.  In this work, we perform the inter-comparison between NEMO-based and HYCOM-based global ocean circulation models at different resolutions over the WBCS domain. To this aim, we analyse the horizontal and vertical structure of the WBCS and its volume transport along the historical ADELIE transect (SOS-Climate II campaign; https://doi.pangaea.de/10.1594/PANGAEA.864578), which extends oceanward from the northernmost tip of the Antarctic Peninsula and across the WBCS. The choice of this transect is not trivial as it captures the hydrodynamic of the WBCS before water masses either leave the basin or recirculate within the gyre.  

Preliminary results support that both eddy-resolving models are in agreement about the major features of the hydrography and dynamic structure of the WBCS as compared to previous modelling studies. Both reproduce the spatial distribution of the Antarctic Coastal Current (CC), the Antarctic Slope Front (ASF) and the Weddell Front (WF), as reported in Thompson and Heywood (2008). Talking about the NEMO-based model at a lower resolution (¼^o), the multi-jet structure of the WBCS is absent, appearing only one major branch. We attribute this mismatch mostly due to a resolution issue. Regarding the volume transport, we find the modelled WBCS displays a seasonal cycle in all cases of study, where minimum values are found in September-December while maximum are in March-July, as also reported Wang et al. [2012]. A major difference occurs towards the interior of the gyre, where the HYCOM-based model exhibits a significantly stronger and wider current branch (~150 km) east of the WF, and whose description is absent in the literature. In previous studies this domain is traditionally excluded and, when volume transport estimates from the NEMO-based model and the HYCOM based model were computed, they both yielded an average transport about 30 Sv, which agrees well with a transport about 24 Sv reported by Wang et al. (2012) and Jullion et al. (2014), also based on modelling estimates across a similar but shorter transect.

These results encourage us to further explore these models in ongoing analyses about the natural variability of the WBCS of the Weddell Gyre and major forcing controlling its variability. We expect that a better understanding of the governing processes will allow us to assess the potential downstream impact of local water masses after their exit from the Gyre.