437 Influence of Sub-grid Variability on Snow Deposition and Ablation in North American Mountain Environments: Implications for Upscaling to Meso-scale Representations Danny Marks, Gerald Flerchinger, Mark Seyfried Abstract A measurement and modeling evaluation of how snow distribution and melt are influenced by vegetation and topographic structure at three different experimental catchments in western North America (two in the western U.S. and one in western Canada) has been undertaken. This ongoing research investigates variations in the critical interactions between vegetation, topography and snowcover in different snow-dominated basins, how these variations impact upscaling site- and basin-scale processes for watershed and regional scale analyses, and how these differences are incorporated in transferable methods to account for the effects of sub-grid variability. By uniquely covering a transect of cordilleran research sites from 35 to 61 o N and from 1,000 to 2,700 m asl, it provides a true ‘Western Cordilleran baseline’ for snowmelt runoff prediction for North America that will substantially benefit model development and testing. Keywords: snow, watershed hydrology Introduction Forest snow scaling The influence of forest canopy cover and variable melt energetics on depletion of snowcover was investigated following earlier work in open environments. The results can be stratified into that variability within the Marks is a Research Hydrologist, USDA-ARS, Northwest Watershed Research Center, Boise, ID 83712. E-mail: danny@nwrc.ars.usda.gov. Flerchinger is a Research Hydraulic Engineer and Seyfried is a Soil Scientist, both at the USDA-ARS, Northwest Watershed Research Center, Boise, ID 83712. forest stand and that between forest stands. Within stands, Faria et al. (2000) found the frequency distribution of snow water equivalent (SWE) under boreal forest canopies fit a log-normal distribution. Within-stand covariance between the spatial distributions of snow water equivalent and melt energy promoted an earlier depletion of snowcover than if melt energy were uniform. This covariance was largest for the most heterogeneous stands (usually medium density). Stand scale variability in mean SWE and mean melt energy resulted in more rapid snow covered area (SCA) depletion for stands with lower leaf area. Because of the heterogeneity in the spatial distributions of SWE and melt energy in forest environments, it is necessary that these variations be included in calculations of SCA depletion (Faria et al. 2000). Forest snowmelt energetics Recent studies of energetics of forest snowmelt (Davis et al. 1997, Hardy et al. 1997, Pomeroy and Granger 1997, Link and Marks 1999a, Link and Marks 1999b, Hardy et al. 2000) have focused on sub-canopy radiative exchange. Sub-canopy insolation is roughly one order of magnitude less than that incoming to a mature pine forest during melt. Snow albedo below forest canopies is lower than that of typical open snowfields (Harding and Pomeroy 1996) and is subject to a premelt decay due to deposited leaf litter from forest canopies (Hardy et al. 2000). Melt simulations that include a litter decay algorithm are vastly improved over those with traditional albedo assumptions (Link and Marks 1999b, Hardy et al. 2000). Few studies have considered in detail the contribution of sub-canopy longwave radiation, despite the relatively warm canopy temperatures (10 to 20 °C above that of snow) measured during melt.