The prediction of short-term (100 year) changes in the mass balance of ice sheets and longer-term (1000 years) variations in their ice volumes is important for a range of climatic and environmental models. The Antarctic ice sheet contains between 24 M km
3
and 29 M km
3
of ice, equivalent to a eustatic sea level change of between 60m and 72m. The annual surface accumulation is estimated to be of the order of 2200 Gtonnes, equivalent to a sea level change of 6 mm a
-1
. Analysis of the present-day accumulation regime of Antarctica indicates that about 25% (
ca.
500 Gt a
-1
) of snowfall occurs in the Antarctic Peninsula region with an area of only 6.8% of the continent. To date most models have focused upon solving predictive algorithms for the climate-sensitivity of the ice sheet, and assume: (i) surface mass balance is equivalent to accumulation (i.e. no melting, evaporation or deflation); (ii) percentage change in accumulation is proportional to change in saturation mixing ratio above the surface inversion layer; and (iii) there is a linear relation between mean annual surface air tem perature and saturation mixing ratio. For the A ntarctic Peninsula with mountainous terrain containing ice caps, outlet glaciers, valley glaciers and ice shelves, where there can be significant ablation at low levels and distinct climatic regimes, models of the climate response must be more complex. In addition, owing to the high accumulation and flow rates, even short- to medium -term predictions must take account of ice dynamics. Relationships are derived for the mass balance sensitivity and, using a model developed by Hindmarsh, the transient effects of ice dynamics are estimated. It is suggested that for a 2°C rise in mean annual surface tem perature over 40 years, ablation in the A ntarctic Peninsula region would contribute at least 1.0 mm to sea level rise, offsetting the fall of 0.5 mm contributed by increased accum ulation.
Semi-equilibrated global sea-level change projections for the next 10 000 years