Much of the climate variability of Antarctica is modulated by the Southern Annular Mode (SAM, Section 14.5.2), the high-latitude atmospheric response to ENSO (Section 14.4) and interactions between the two (Stammerjohn et al., 2008; Fogt et al., 2011; see also Sections 2.7 and 10.3.3). Signatures of the SAM and ENSO in Antarctic temperature, snow accumulation and sea ice have been documented by many observational and modelling studies (Bromwich et al., 2004; Guo et al., 2004; Kaspari et al., 2004; van den Broeke and van Lipzig, 2004; Marshall, 2007).

The positive SAM is associated on average with warmer conditions over the Peninsula and colder conditions over East Antarctica, with a mixed and generally non-significant impact over West Antarctica (Kwok and Comiso, 2002; Thompson and Solomon, 2002; van den Broeke and van Lipzig, 2004). ENSO is associated with circulation anomalies over the southeast Pacific that primarily affect West Antarctica (Bromwich et al., 2004; Guo et al., 2004; Turner, 2004). ENSO variability tends to produce out-of-phase variations between the western and eastern sectors of West Antarctica (Bromwich et al., 2004; Kaspari et al., 2004), in association with the PSA pattern (Section 14.7.1).

The positive summer/autumn trend in the SAM index in recent decades (Section 14.5.2) has been related to the contrasting temperature trend patterns observed in these two seasons, with warming in the east and north of the Antarctic Peninsula and cooling (or no significant temperature change) over much of East Antarctica (Turner et al., 2005; Thompson et al., 2011). The high polarity of the SAM is also consistent with the significant increase in snow accumulation observed in the southern part of the Peninsula (Thomas et al., 2008).

Unlike the eastern Antarctic Peninsula, its western coast shows maximum warming in austral winter (when the SAM does not exhibit any significant trend), which has been attributed to reduced sea ice concentrations in the Bellingshausen Sea. Recent studies have emphasized the role of tropical SST forcing not directly linked to ENSO to explain the prominent spring- and wintertime atmospheric warming in West Antarctica (Ding et al., 2011; Schneider et al., 2012). There is further evidence of tropical SST influence on Antarctic temperatures and precipitation on decadal to inter-decadal time scales (Monaghan and Bromwich, 2008; Okumura et al., 2012).

Modelling of Antarctic climate remains challenging, in part because of the nature of the high-elevation ice sheet in the east Antarctic and its effects on regional climate (Section Moreover, modelling ice properties themselves, for both land ice and sea ice, is an area that is still developing despite improvements in recent years (Vancoppenolle et al., 2009; Picard et al., 2012; Section 9.4.3). Modelling the role of the stratosphere and of ozone recovery is critical for Antarctic climate, as stratospheric change is intimately linked to trends in the SAM (Section 14.5.2).

The projected easing of the positive SAM trend in austral summer (Section 14.5.2) may act to delay future loss of Antarctic sea ice (Bitz and Polvani, 2012; Smith et al., 2012b). It is unclear what effect ENSO will have on future Antarctic climate change as the ENSO response to climate change remains uncertain (see and 14.5.2 for more information). Seasonally, changes in the strength of the circumpolar westerlies are also expected during the 21st century as a result of changes in the semi-annual oscillation caused by alterations in the mid- to high-latitude temperature gradient in the SH. Bracegirdle et al. (2008) considered modelled circulation changes over the Southern Ocean and found a more pronounced strengthening of the autumn peak of the semi-annual oscillation compared with the spring peak. Future changes in surface temperature over Antarctica are likely to be smaller than the global mean, and much smaller than those projected for the Arctic, because of the buffering effect of the southern oceans, and the thermal mass of the east Antarctic ice sheet (Section 12.4.6). Warming is likely to bring increased precipitation on average across Antarctica (Bracegirdle et al., 2008), but the spatial pattern of precipitation change remains uncertain.

In summary, consistency across CMIP5 projections suggests it is very likely that Antarctic temperatures will increase through the rest of the century, but more slowly than the global mean rate of increase (Table 14.1). SSTs of the oceans around Antarctica are likely to rise more slowly than surface air temperature over the Antarctic land mass. As temperatures rise, it is also likely that precipitation will increase (Table 14.1), up to 20% or more over the East Antarctic. However, given known difficulties associated with correctly modelling Antarctic climate, and uncertainties associated with future SAM and ENSO trends and the extent of Antarctic sea ice, precipitation projections have only medium confidence.