Spatial patterns of recent Antarctic surface temperature trends and the importance of natural variability: lessons from multiple reconstructions and the CMIP5 models

Recent studies of Antarctic surface air temperature (SAT) trends over the past several decades have revealed that the Antarctic Peninsula and Western Antarctica are among the most rapidly warming regions on Earth (Turner et al. 2005; Chapman and Walsh 2007; Monaghan and Bromwich 2008; Steig et al. 2009; ODonnell et al. 2011; Bromwich et al. 2013; Nicolas and Bromwich 2014). These regions have also experienced increasing surface mass loss due to glacier retreat (Joughin et al. 2014; Rignot et al. 2014), with potentially severe consequences for global sea level rise. In contrast, East Antarctic SAT has experienced an insignificant, cooling trend (Schneider et al. 2012; Nicolas and Bromwich 2014). In the annual mean, this large spatial asymmetry in SAT trends between West and East Antarctica has become a hallmark of Antarctic climate change: yet, it remains poorly understood. It is important to recall, however, that the annual mean asymmetry in Antarctic SAT trends arises, largely, from opposing trends in austral spring and autumn. Over the past several decades, significantly positive trends have been observed in West Antarctica in spring, and marginally significant negative trends have been observed in East Antarctica in autumn (Schneider et al. 2012). Hence, the causes leading to the annual mean asymmetry are likely to be complex, and might arise for different reasons in different seasons. Abstract The recent annually averaged warming of the Antarctic Peninsula, and of West Antarctica, stands in stark contrast to very small trends over East Antarctica. This asymmetry arises primarily from a highly significant warming of West Antarctica in austral spring and a cooling of East Antarctica in austral autumn. Here we examine whether this East–West asymmetry is a response to anthropogenic climate forcings or a manifestation of natural climate variability. We compare the observed Antarctic surface air temperature trends over two distinct time periods (1960–2005 and 1979–2005), and with those simulated by 40 models participating in Phase 5 of the Coupled Model Intercomparison Project (CMIP5). We find that the observed East–West asymmetry differs substantially between the two periods and, furthermore, that it is completely absent from the forced response seen in the CMIP5 multi-model mean, from which all natural variability is eliminated by the averaging. We also examine the relationship between the Southern Annular mode (SAM) and Antarctic temperature trends, in both models and reanalyses, and again conclude that there is little evidence of anthropogenic SAM-induced driving of the recent temperature trends. These results offer new, compelling evidence pointing to natural climate variability