An assessment and interpretation of the observed warming of West Antarctica in the austral spring

We synthesize variability and trends in multiple analyses of Antarctic near-surface temperature representing several independent source datasets and spatially complete reconstructions, and place these into the broader context of the behavior of other components of the climate system during the past 30-50 years. Along with an annual-mean trend during the past 50 years of about 0.1°C/decade averaged over Antarctica, there is a distinct seasonality to the trends, with insignificant change (and even some cooling) in austral summer and autumn in East Antarctica, contrasting with warming in austral winter and spring. Apart from the Peninsula, the seasonal warming is largest and most significant in West Antarctica in the austral spring since the late 1970s. Concurrent trends in sea ice are independent evidence of the observed warming over West Antarctic, with the decrease in sea ice area in the Amundsen and Bellingshausen Seas congruent with at least 50% of the inland warming of West Antarctica. Trends in near surface winds and geopotential heights over the high-latitude South Pacific are consistent with a role for atmospheric forcing of the sea ice and air temperature anomalies. Most of the circulation trend pattern projects onto the two Pacific South American (PSA) modes of atmospheric circulation variability, while the Southern Annular Mode lacks a positive trend in spring that would otherwise cause a cooling tendency. The largest circulation trend is associated with the PSA-1 mode, a wave-train extending from the tropics to the high Southern latitudes. The PSA-1 mode is significantly correlated with SSTs in the southwestern tropical and subtropical Pacific. The increased SSTs in this region, together with the observed increase in rainfall, suggest that anomalous deep convection has strengthened or increased the occurrence of the Rossby wave-train associated with PSA-1. This hypothesis is supported by results from two ensembles of SSTforced atmospheric general circulation model simulations. Finally, the implications of the seasonality, timing, and spatial patterns of Antarctic temperature trends with respect to interpreting the relative roles of stratospheric ozone depletion, SSTs and increased atmospheric concentrations of greenhouse gasses are discussed.