Because of the dominant influence of energy fluxes on snowpack evolution and meltwater generation, efforts have been made to better understand spatial and temporal patterns in fluxes and assess their sensitivity to physical characteristics.
A comprehensive set of data from the Emerald Lake basin in California revealed the changes in relative magnitudes of snow surface energy fluxes throughout the accumulation and snowmelt seasons [ Marks, 1992]. Radiation accounted for 66-90% of the energy available for melt; sensible and latent heat fluxes were roughly of the same magnitude, but of opposite sign and therefore canceled. About 20% (50 cm) of the SWE was lost to sublimation and evaporation over the course of the snow season. The seasonal contribution of soil heat conduction to the energy budget is negligible, but may significantly contribute to midwinter melt from the base of the pack. A 16-year time series of snow depths, air temperatures, and soil temperatures from a mid-continental site indicated that a snow depth of 42.5 cm was required to maintain steady soil temperatures [ Sharratt et al., 1992].
Colbeck [1991] reviewed how layering can have profound effects on the physical properties of snow that are important in both thermal and hydrologic phenomena. Because of the importance of water-vapor diffusion in snow metamorphism, the water-vapor diffusion coefficient is of great interest. Changes in pore geometry during experiments to determine porous-medium parameters of snow make such determinations highly uncertain, and different experiments often lead to contradictory measurements [ Colbeck, 1993]. Sommerfeld et al. [1991] preconditioned the air flow prior to making air-permeability measurements on dry snow by passing the air stream through a column of snow similar to the one being measured, equilibrating the air with the conditions in the snow, thus reducing erosion of the test snow during the experiment. Their results suggested that a simple snow density-permeability relationship fit their data better than dimensionally correct specific-surface-area methods. Hardy and Albert [1993] made concurrent measurements of the permeability and physical properties of snowpack strata, finding that physical properties correlated poorly with permeability measurements and that stereological parameters correlated well for dry snow, but poorly for old snow and ice layers. Vapor transport does not contribute to thermal transport within the snowpack as much as conduction and advection by dry air. The effect of ventilation on the snowpack temperature profile is most likely to be observed when the temperature gradient in the snowpack is not great or when air flow within the pack is relatively high [ Albert and McGilvary, 1992].
Meltwater movement, heat and solute transport, and snow metamorphism are affected by the advance of wetting fronts through the pack. As the meltwater penetrates the pack, the thermal characteristics of the pack change; therefore, the formation of roughly columnar zones, or flow fingers, of wet snow that contain moving meltwater penetrating the snowpack ahead of the main melting front can be observed in situ by monitoring the thermal changes within the pack [ Sturm et al., 1993; Conway and Benedict, 1994].
The penetration of visible light into the snowpack has been suggested as a cause for the presence of a temperature maximum about 10 cm below the surface. This explanation appears less viable when radiative-transfer calculations are carried out with high spectral resolution [ Brandt, 1993]. Wavelengths that penetrate the snowpack (visible) are scattered back to the surface, and wavelengths that are absorbed are absorbed in the first few millimeters, indicating that subsurface absorption is minimal. The conditions under which significant heating of the interior of the snow occur when the scattering coefficient is low (e.g. blue ice) and when the thermal conductivity is low (e.g. near-surface depth hoar).
Heat flow was affected by the presence of trees, because of interception of snow by the foliage resulting in irregular snow depths. For example, heat flow from tree wells can be more than twice that of undisturbed snow [ Sturm, 1992]. Snow ablation in forest openings depends on the size of the opening. Ablation from openings of diameter H (where H is the average height of adjacent uncut trees) was less than from openings of 0, 3 and 5H because net radiation was least in the 1H diameter openings. Evaporation in the canopy-covered case (0H) was slightly greater than the 1H case because of nighttime radiative transfer from the canopy to the snow, which kept the canopy-covered snow warmer than in the openings, and because the canopy prevented nighttime condensation. Furthermore, lateral advective energy transfer between openings and forest canopy appears to be significant [ Berry and Rothwell, 1992; Bernier and Swanson, 1993].