4.6.3 Causes of Changes

4.6.3.1 Changes in Snowfall and Surface Melting

For Greenland, modelling driven by reanalyses and calibrated against surface observations indicates recent increases in temperature, precipitation minus evaporation, surface melt water runoff and net mass loss from the surface of the ice sheet, as well as areal expansion of melting and reduction in albedo (Hanna et al., 2005, 2006; Box et al., 2006). High interannual variability means that many of the trends are not highly significant, but the trends are supported by the consistency between the various component data sets and results from different groups. Estimated net snowfall minus melt water runoff includes an increase in the Greenland contribution to sea level rise of 58 Gt yr^–1 between the 1961 to 1990 and 1998 to 2003 intervals (Hanna et al., 2005), or of 43 Gt yr^–1 from 1998 to 2004 (Box et al., 2006).

For Antarctica, the recent summaries by van de Berg et al. (2006), van den Broeke et al. (2006) and Monaghan et al. (2006) have updated trends in accumulation rate. Contrary to some earlier work, these new studies found no continent-wide significant trends in accumulation over the interval 1980 to 2004 (van de Berg et al., 2006; van den Broeke et al., 2006) or 1985 to 2001 (Monaghan et al., 2006) from atmospheric reanalysis products (National Centers for Environmental Prediction (NCEP), the European Centre for Medium Range Weather Forecasts (ECMWF), Japanese), or from two mesoscale models driven by ECMWF and one by NCEP reanalyses. Strong interannual variability was found, approaching 5% for the continent, and important regional and seasonal trends that fit into larger climatic patterns, including an upward trend in accumulation in the Antarctic Peninsula. Studies of surface temperature (e.g., van den Broeke, 2000; Vaughan et al., 2001; Thompson and Solomon, 2002; Doran et al., 2002; Schneider et al., 2004; Turner et al., 2005) similarly showed regional patterns including strong warming in the Antarctic Peninsula region, and cooling at some other stations. Long-term data are very sparse, precluding confident identification of continent-wide trends.