Western Antarctic ocean surface variability during the 20th century in the NCAR-CCSM4

Variability in the Southern Ocean is strongly connected with climate modes, in particular with the Southern Annular Mode (SAM). SAM variability shows strong seasonal asymmetries in the Antarctic Peninsula which could impact ocean-atmosphere dynamics. Thus, the goal of this study is to understand the atmosphere-ocean response to seasonality in the Southern Ocean during the 20th century with a specific focus on the western Antarctic Peninsula. We analyze the seasonal and long-term aspects of the ocean-atmosphere properties in the NCAR-CCSM4 model. Given the high correlation between the CCSM4-SAM index and the significant (at the 95% level) regional long-term trend, both the zonal wind stress (~0.75 and 0.04 N m−2 century−1) and barotropic transport (~0.78 and 12 Sv century−1) in the western domain (155-130W), are associated with the SAM. This is in contrast to the barotropic transport in the north-central domain (130-80W and 55-60S), which shows great seasonality: The changes from the austral winter (JJA) to summer (DJF) are approximately 10 Sv, while its long-term trend is small (7 Sv century−1). On the other hand, the decrease in the surface salinity in the end of the century (~0.06), mainly near Antarctic Peninsula, and the long-term trend of 0.2 century−1 are not seasonal and were not attributed directly to SAM variability. Thus, an increase in precipitation of 8 mm month−1, mostly during the austral summer (DJF) in the late 20th century over the northern Antarctic Peninsula, together with the positive long-term trend of 8 mm month−1 century−1 over the Amundsen-Bellingshausen Seas, contributed, among other factors, to surface desalinization in the late 20th century. SST warmed throughout the 20th century, independent of the season displaying changes that were not directly related with the SAM. Our study shows how different sea-air properties interact with each other and, also, when the SAM plays an important role in controlling long-term variability.