Modeling meteorological forcing of snowcover in forests

The architectural properties of a forest are known to modify significantly meteorological forcing of snowcover. Current numerical snow models utilize a wide range of vegetation representations that limit their application to particular biomes or for basic research on specialized problems. Most do not explicitly represent the combined effects of the canopy on processes of mass and energy transfer beneath the canopy.;This project develops forest canopy sub-models that estimate the below-canopy solar and longwave irradiance, wind speed, and accumulation of precipitation, based on meteorological measurements above the canopy and parameters of forest architecture. The wind and solar radiation sub-model predictions were independently compared with meteorological observations at deciduous and coniferous sites in the snowbelt region of northern Michigan. The solar radiation and wind models required adjustments to match sub-canopy measurements.;The primary experiment compared the simulations and measurements of snow depth for eight modified versions of the Utah Energy Balance (UEB) snow model during the 1998--99 snowcover season at the two forest sites and a near-by open site. Independent inclusion of each sub-model and a new stability scheme in the UEB model revealed significant sensitivity of modeled snow depth to stability and each of the four processes estimated by the sub-models. The original UEB model uses a simple forest canopy parameterization that does not consider precipitation interception. Comparison of the original and modified UEB models significantly improved simulations of snow depth at the open and coniferous sites, but performance was slightly worse for a leafless deciduous site. Unlike the modified model, the analysis suggests that the original model produces inconsistent results, which reduces its potential for application to different biomes.;Results suggest that opposing processes of energy and mass exchange tend to moderate meteorological forcing beneath a forest canopy. Each process can substantially affect snow depth, depending on the above-canopy meteorological conditions and architecture of the forest. Future work should consider refinement of the sub-models, testing in different biomes, inclusion of soil substrate processes, and comparison of these results with those of other snow models under similar environmental conditions.