Analysis of Error in the Determination of Snow Storage for Small High Arctic Basins

Abstract

Water balance studies in tundra regions require accurate snowfall data but weather station records often underestimate basin snow storage. However, snow storage can be determined by snow surveys conducted prior to the melt period because Arctic snowpacks do not melt during winter. Topography strongly controls snow distribution. A basin can be subdivided into various terrain types and the snow survey then establishes the snow characteristics of each terrain type so that basin snow storage is obtained as their areally weighted mean. Such a survey was carried out in small basins near Resolute, Northwest Territories, traversing different types of terrain. The results confirmed that weather station snowfall grossly underestimated basin snow storage. Since it is also desirable to simplify future snow surveys by reducing the number of transects, an error analysis was performed to determine the error resulting from a grouping of terrain types. It was found that both maximum and mean error increased as the number of terrain types was reduced, but the increase was not substantial when certain terrain types were combined. A mean error of 15% is expected when only four types of terrain (hilltops, flats, gullies-valleys and slopes) are recognized in the survey, but the error quickly increases when further simplification of the terrain is introduced.