STATISTICAL-MODEL OF SPATIALLY DISTRIBUTED SNOWMELT RATES IN A BOREALFOREST BASIN

Spatial variation in snowmelt rates in the boreal forest can be explai ned by differences in canopy density. Canopy density, represented as g ap fractions (GF), controls both the amount of short-wave radiation re aching the snowpack surface and wind speed over the snow surface, whic h in turn regulates sensible and latent heat fluxes, Reductions in sho rtwave, sensible and latent heat fluxes outweigh any increased contrib utions from longwave radiation from the canopy. Differences in the tot al energy available for melt do not translate to equally proportional changes in melt rates under different canopy densities. As available e nergy increases, the melt rate increases with decreasing canopy densit y and the form of the relationship can vary depending on climatic cond itions. A good relationship between ground-based GF measurements and a canopy closure index derived from Landsat TM provides the spatial fab ric for the distribution of snowmelt rates that show comparable patter ns of snow ablation during years of very different climatological cond itions. This physically meaningful method of determining the spatial v ariability of snow ablation and subsequent meltwater delivery to the s oil interface is particularly useful for providing insight to the hete rogeneous active layer development in permafrost regions of the boreal forest and the implications for runoff processes. (C) 1998 John Wiley & Sons, Ltd.