On the annual and semi-annual cycles of precipitation across Antarctica

In this study, we examine the importance of the annual and semi-annual cycles of precipitation (H-1(P) and H-2(P), respectively) across Antarctica utilizing the ERA-40 re-analysis from 1980 to 2001. A qualitative comparison between monthly precipitation from ERA-40 and precipitation reports from Antarctic bases reveals that the re-analysis successfully captures the seasonal cycle of precipitation at coastal sites. However, likely excessive summer precipitation on the high Antarctic Plateau in ERA-40 means the re-analysis is uncertain in the continental interior. The spatial variability in the amplitude of both H-1(P) and H-2(P) reveals coastal maxima with a clear wavenumber 3 pattern: unsurprisingly, maximum values are located east of climatological low-pressure centres, where moisture advection into the continent is greatest. This pattern is particularly pronounced for H-1(P), for which the amplitude exceeds 25 mm water equivalent (WE) in coastal west Antarctica. In the Antarctic coastal region H-1(P) often explains greater than 50% - and sometimes more than 80% - of the total variance of the seasonal precipitation cycle. In contrast, the northwest Antarctic Peninsula has a higher amplitude in H-1(P) that is linked to its location within the circumpolar trough (CPT). A marked change in the magnitude of H-1(P) between the 1980s and 1990s in parts of coastal west Antarctica and the west coast of the Antarctic Peninsula was observed in the ERA-40 data and confirmed by similar changes in the seasonal cycle of precipitation reports. The principal modification to H-1(P) results from a decrease in winter precipitation that can be linked to a weaker and smaller winter Amundsen Sea Low, which in turn is known to be influenced by El Nino-Southern Oscillation (ENSO). A simple estimate indicates that the observed change in the seasonal cycle of precipitation Would lead to an apparent warming of similar to 0.6 degrees C in a proxy of mean annual temperature derived from oxygen isotope measurements. Copyright (C) 2009 Royal Meteorological Society

Details