Subpolar Southern Ocean Response to Changes in the Surface Momentum, Heat, and Freshwater Fluxes under 2xCO2

Abstract

The Antarctic subpolar Southern Ocean (sSO) has fundamental climate importance. Antarctic BottomWater (AABW) originates in the sSO and supplies the lower limb of the Meridional Overturning Circulation (MOC), occupying 36% of ocean volume. Climate models struggle to represent continental shelf processes that form AABW.We explore sources of persistent model biases by examining response of the sSO to perturbations in surface forcing in a global ocean–sea-ice model (ACCESS-OM2) that forms AABW both on shelf and in open-ocean. The sSO response to individual and combined perturbations of surface heat, freshwater and momentum fluxes follows the WCRP CMIP6 FAFMIP-protocol. Wind perturbations (i.e., a poleward shift and intensification of the Westerlies) is dominant, enhancing AABW formation and accelerating the global MOC. This occurs through upwelling of warm waters and inhibition of sea-ice growth during winter, which triggers large open water polynya (OWP) events with associated deep convection. These events occur in the Weddell and Ross seas and their variability is associated with availability of heat at mid-ocean depths. These OWPs cease when the heat reservoir is depleted. Effects of surface warming and freshening only partially compensate changes from increasing winds on ocean stratification and depletion of AABW formation. These results indicate that overly convective models, such ACCESS-OM2, can respond to CO2-perturbed scenarios by forming too much AABW in OWP, which might not hold in models without OWPs. This might contribute to the large inter-model spread thermosteric sea-level projections, being relevant to the interpretation of future projections by current climate models.

The nonlinear impact of surface forcing changes on bottom water formation and overturning in the Southern Ocean

“…Projected wind stress changes increase bottom water formation and enhance CDW inflow, while projected increases in AIS melting cause a substantially weaker bottom overturning cell and contraction of the AABW volume. A pronounced AABW and lower limb MOC with stronger and/or southwardshifted westerlies have been previously documented in both coupled and ocean-only models using different experimental designs (Bishop et al, 2016;Dias et al, 2021;Hogg et al, 2017;Spence, Sebille, et al, 2014). The simulated change in CM4 and ESM4 agree with these previous findings.…”

“…On the other hand, the response in CM4 to wind perturbations is muted relative to ESM4, where the occurrence of open-ocean deep convection during Stress depends on the amount of heat stored within the CDW at the time of the perturbation initiation (Beadling et al, 2022). The dominance of open-ocean convection in response to wind perturbation has been observed in other climate models, such as ACCESS-OM2 (Dias et al, 2021), but are regarded as unrealistic given that most of AABW in the present day is sourced from DSW on the Antarctic shelf and not in the open ocean (Heuzé, 2021;. However, it is possible that open ocean convection may have been an important process in maintaining lower-limb overturning under the preindustrial climate, and the weaker AABW production during current climate condition is due to the absence of such phenomena in the SO (C. d. de Lavergne et al, 2014).…”

“…However, recent work by Bronselaer et al (2020) and Beadling et al (2022), highlight that both the projected wind stress changes and AIS melting are critical to consider together when investigating the transient response of the SO, given their competing effects on SO ventilation and the thermal response on the Antarctic shelf. Previous studies suggest opposing imprints of wind and AIS melt perturbations on SO water masses and circulation, with contributions that are likely not a simple linear combination of separate perturbation responses (Dias et al, 2021;Beadling et al, 2022). This non-linear response precludes the possibility of attributing changes in the SO MOC to individual contributions from AIS melt and wind stress change in previous studies that have not considered both forcings in tandem.…”

“…Such an agreement suggests that the observed intensification and poleward shift of Southern Hemisphere winds are likely imprinting on the upper limb of the SO MOC, with dynamics consistent with those represented in modeling studies. However, the invigoration of the lower limb of the SO MOC that occurs when observed changes in the surface wind stress field are applied in model simulations (Spence, Sebille, et al, 2014;Hogg et al, 2017;Dias et al, 2021) is inconsistent with the observed weakening of the lower limb of the SO MOC (Rintoul, 2007;Purkey & Johnson, 2010Schmidtko et al, 2014;Anilkumar et al, 2015) and contraction of AABW volume (Purkey & Johnson, 2012;Azaneu et al, 2013;van Wijk & Rintoul, 2014;Anilkumar et al, 2021). Such disagreements may suggest that observed trends in the surface wind stress are not presently imprinting on the SO MOC's lower limb, or competing processes may be acting to limit the sensitivity of the lower limb to wind stress changes.…”

“…The impacts of projected changes in wind stress on the SO MOC's lower limb remain less explored. However, previous modeling studies suggest a strengthening of the lower limb due to greater AABW formation, mostly in connection with enhanced open-ocean polynya activity near the Antarctica margin relative to baseline simulations (Spence, Sebille, et al, 2014;Hogg et al, 2017;Dias et al, 2021). However, numerical models are often limited in their ability to accurately represent AABW formation, due to insufficient horizontal resolution and a lack of representation of critical overflow processes, leading to an underrepresentation of DSW and a bias towards forming AABW via open ocean deep convection (Heuzé et al, 2015;Dufour et al, 2017;Aguiar et al, 2017;Heuzé, 2021;Mohrmann et al, 2021).…”