# 10.A.4 Changes in Ice Sheet Surface Mass Balance

Quadratic fits are made to the results of Gregory and Huybrechts (2006) (Section 10.6.4.1) for the SMB change of each ice sheet as a function of global average temperature change relative to a steady state, which is taken to be the late 19th century (1865–1894). The spread of results for the various models used by Gregory and Huybrechts represents uncertainty in the patterns of temperature and precipitation change. The Greenland contribution has a further uncertainty of 20% (one standard deviation) from the ablation calculation. The Antarctic SMB projections are similar to those of the TAR, while the Greenland SMB projections are larger by 0.01–0.04 m because of the use of a quadratic fit to temperature change rather than the constant sensitivity of the TAR, which gave an underestimate for larger warming.

## 10.A.5 Changes in Ice Sheet Dynamics

Topographic and dynamic changes that can be simulated by currently available ice flow models are roughly represented as modifying the sea level changes due to SMB change by –5% ± 5% from Antarctica, and 0% ± 10% from Greenland (± one standard deviation) (Section 10.6.4.2).

The contribution from scaled-up ice sheet discharge, given as an illustration of the effect of accelerated ice flow (Section 10.6.5), is calculated as r_{1} × T / T_{1}, with T and T_{1} expressed relative to the 1865 to 1894 average, where r_{1} = 0.32 mm yr^{–1 }is an estimate of the contribution during 1993 to 2003 due to recent acceleration and T_{1} = 0.63°C is the global average temperature during that period.