Luongo, M., Armour, K., and Johnson, G.C. (2025)
Presented at:
AGU Annual Meeting 2025While most coupled climate models predict that the Southern Ocean (SO) should have warmed since 1980, SO sea surface temperatures (SSTs) have in fact cooled over that same period. Many studies have investigated whether this model mismatch might be due to inaccurate surface ocean forcing, misaligned internal variability, or insufficiently fine model resolution. In this work we explore whether i) realistic forcing of an observationally informed ocean-only model can produce both SO temperature and salinity trends that match observations and ii) whether those ocean-only responses have any remote influence on tropical oceans. We initialize ocean-only MITgcm simulations from the 1992 ECCOv4r4 ocean state estimate and spin them up using climatological reanalysis fluxes of heat, freshwater, and momentum to create an ocean mean state that agrees well with observational estimates. We then separately force the surface ocean with instantaneous changes of radiative forcing, freshwater fluxes (from hydrological changes, sea ice melt, and ice sheet melt separately), and a poleward shift and intensification of the surface westerly winds. Using a Green’s function approach, we then convolve these instantaneous responses with observed forcing time series and study the temperature and salinity trends implied by each forcing agent. We show that much of the muted warming and large-scale freshening pattern since 1980 is simply a result of surface radiative forcing without additional changes in freshwater input or winds. We then demonstrate how observed changes in freshwater flux and winds geographically hone this base pattern toward the specific pattern of observed trends. While several forcing mechanisms can create Southern Ocean surface cooling, invoking one physical mechanism at the expense of others can lead to inaccurate salinity trends. We use a detection and attribution analysis to determine the relative contribution of each forcing pattern to the final observed pattern in a linear combination. Lastly, we show that these Southern Ocean forcings do not affect the tropics in this ocean-only set-up, suggesting the importance of atmospheric rather than oceanic pathways for Southern Ocean-Tropical Pacific teleconnections.
