Variability of the Weddell Sea deep waters in GLORYS12v1 reanalysis

Understanding deep-water variability in the Weddell Sea remains a challenge, given the limited observational coverage and the difficulties ocean models face in representing fine-scale processes, particularly along the Antarctic margin. Recent high-resolution oceanic products offer a valuable opportunity to advance understanding of this region. Nevertheless, it is crucial to critically evaluate their reliability before relying on them for scientific analysis. Here, we assess the representation and variability of Warm Deep Water (WDW), Weddell Sea Deep Water (WSDW), and Weddell Sea Bottom Water (WSBW) in the 1/12° Global Ocean Physics Reanalysis (GLORYS12v1) between 1993 and 2020. GLORYS12v1 reproduces key hydrographic features and vertical stratification below 500 m, despite persistent biases in the upper ocean. WDW shows a spurious freshening trend and a cool bias in Section SR4, diverging from the observed neutral salinity and warming trend. Moreover, WSDW and WSBW exhibit overestimated warming and salinification trends. These discrepancies are primarily linked to (i) a weakening of the Weddell Gyre, which limits inflow and renewal of deep waters; (ii) stronger westerlies enhancing Ekman transport and upwelling; and (iii) changes in sea ice concentration affecting deep convection. A complex maximum covariance analysis reveals strong decadal-scale covariability between the Weddell Gyre barotropic circulation and the thermohaline structure of deep water masses, especially WSDW and WSBW. Although GLORYS12v1 resolves many relevant processes, its overestimation of trends and underrepresentation of coastal dynamics highlight the need for improved vertical coordinate schemes, refined mixing parameterizations, and enhanced observational coverage to better capture the variability of deep waters in polar regions.