The spatial and temporal variability of the surface mass balance in Antarctica: results from a regional climate model

Abstract

A 14 year integration with a regional atmospheric model (RACMO) is used to obtain detailed information on the Antarctic surface mass balance and to understand the mechanisms that are responsible for the spatial and temporal distribution of the surface mass balance. The model (Δx = 55 km) uses the parameterizations of the physical processes from the ECHAM4 general circulation model and is driven from the lateral boundaries by the 15 year re-analyses of the European Centre for Medium-range Weather Forecasts (ERA-15). Sea surface temperature and sea ice extent are prescribed from observations. The model is evaluated with in situ measurements of surface pressure, 2 m temperature, and surface mass balance. Generally, good agreement is found between model output and measurements, although in the interior of the ice sheet the temperatures are slightly too high in summer. The 14 year mean surface mass balance averaged over the grounded Antarctic ice ‹B› is 156 mm water equivalent per year. A statistical relation between precipitation and topographical parameters is derived from model output. Half of the spatial variance in precipitation can be explained by a relation between precipitation and distance to the coast. Locally, the direction of the atmospheric flow is important; in Ellsworth Land and Wilkes Land the surface mass balance is larger than expected on the basis of topography alone, because of flow directed inland. The year-to-year variability in ‹B› is similar to the value found in ERA-15 (standard deviation of annual mean values is 6 to 7% of the 14 year mean) and is determined by the atmospheric circulation and not by variations in temperature or humidity. In the interior of the ice sheet, seasonality of precipitation is mainly determined by temperature, but, near the coast, the dynamics of the flow are important. For example, in Dronning Maud Land the precipitation is highest in autumn, when the upslope component of the wind vector at 500 hPa is largest. Year-to-year variations in seasonality of precipitation are large and might affect proxies for meteorological variables in ice cores.