In the dry snow zone of the Greenland Ice Sheet, accumulation is relatively easy to measure by digging a pit or by drilling to retrieve a core. The span of the sample is measured by counting annual layers. The core or the pit wall is sampled to measure the accumulated mass, which is reported in kg m^-2 a^-1. At lower elevations, melting during summer blurs the distinction between annual layers. The meltwater percolates from the surface down into older layers, but the total accumulation can still be measured. There are between 300 and 400 usable published measurements of this kind, unevenly distributed over the ice sheet and having variable record lengths. At still lower elevations some or even all of the snowfall may be lost by melting and runoff during summer, and there are very few reliable measurements of accumulation.

To estimate accumulation over the whole ice sheet it is necessary to extrapolate from the small number of points with measurements. This exercise is doomed in the ablation zone, where the nearest measurements may be hundreds of kilometres away, unless information is taken from coastal weather stations. At these stations the snowfall is measured regularly, and can be assumed to be proportional to the "accumulation" which would be measured if there were a glacier there. These indirect estimates must, however, be corrected by subtracting an estimate of sublimation.

The dataset being used in Trent's work on accumulation in Greenland is based on the work of earlier compilers, but includes a small number of additional measurements.

The main results to date of work on spatial interpolation of accumulation are illustrated in two maps which show accumulation and its standard error.