On the basis of thousands of surface mass balance (SMB) field measurements over the entire Antarctic ice sheet it is currently estimated that more than 2 Gt of ice accumulate each year at the surface of Antarctica. However, these estimates suffer from large uncertainties. Various problems affect Antarctic SMB measurements, in particular, limited or unwarranted spatial and temporal representativeness, measurement inaccuracy, and lack of quality control. We define quality criteria on the basis of (1) an up-to-date review and quality rating of the various SMB measurement methods and (2) essential information (location, dates of measurements, time period covered by the SMB values, and primary data sources) related to each SMB data. We apply these criteria to available SMB values from Queen Mary to Victoria lands (90°-180°E Antarctic sector) from the early 1950s to present. This results in a new set of observed SMB values for the 1950-2005 time period with strong reduction in density and coverage but also expectedly reduced inaccuracies and uncertainties compared to other compilations. The quality-controlled SMB data set also contains new results from recent field campaigns (International Trans-Antarctic Scientific Expedition (ITASE), Russian Antarctic Expedition (RAE), and Australian National Antarctic Research Expeditions (ANARE) projects) which comply with the defined quality criteria. A comparative evaluation of climate model results against the quality-controlled updated SMB data set and other widely used ones illustrates that such Antarctic SMB studies are significantly affected by the quality of field SMB values used as reference. Copyright 2007 by the American Geophysical Union.

cited By 26