Abstract

As most permanent observations in Antarctica started in the 1950s, understanding Antarctic climate variations throughout the twentieth century remains a challenge. To address this issue, the non-summer multi-decadal variability in pressure reconstructions poleward of 60 degrees S is evaluated and... Show moreAs most permanent observations in Antarctica started in the 1950s, understanding Antarctic climate variations throughout the twentieth century remains a challenge. To address this issue, the non-summer multi-decadal variability in pressure reconstructions poleward of 60 degrees S is evaluated and assessed in conjunction with climate model simulations throughout the twentieth and early twenty-first centuries to understand historical atmospheric circulation variability over Antarctica. Austral autumn and winter seasons show broadly similar patterns, with negative anomalies in the early twentieth century (1905-1934), positive pressure anomalies in the middle twentieth century (1950-1980), and negative pressure anomalies in the most recent period (1984-2013), consistent with concurrent trends in the SAM index. In autumn, the anomalies are significant in the context of estimates of interannual variability and reconstruction uncertainty across most of the Antarctic continent, and the reconstructed patterns agree best with model-generated patterns when the simulation includes the forced response to tropical sea surface temperatures and external radiative forcing. In winter and spring, the reconstructed anomalies are less significant and are consistent with internal atmospheric variability alone. The specific role of tropical SST variability on pressure trends in these seasons is difficult to assess due to low reconstruction skill in the region of strongest tropical teleconnections, the large internal atmospheric variability, and uncertainty in the SST patterns themselves. Indirect estimates of pressure variability, whether through sea ice reconstructions, proxy records, or improved models and data assimilation schemes, will help to further constrain the magnitude of internal variability relative to the forced responses expected from SST trends and external radiative forcing. Show less