In many northern regions melting of the seasonal snowcover is the most important event of the water year. Water from melting snow recharges surface and groundwater supplies and the partitioning of the water released by melting snow to these components is strongly influenced by the infiltrability of frozen ground. Because the process of infiltration into frozen ground is complex, involving coupled heat and mass flow with phase changes, it is not well understood. This paper uses a numerical simulation of meltwater infiltration into frozen soil to provide a better understanding of the process and to develop parametric relationships for predicting infiltration. The simulation is based on the continuity, energy and momentum equations and well known, accepted procedures to define the transport phenomena and the thermophysical properties of a frozen soil.

Snowmelt infiltration into frozen ground involves two primary flow regimes, a transient regime and a quasi-steady state regime. Transient flow, in which the infiltration rate and the surface heat transfer rate undergo rapid changes, occurs immediately following the application of water at the soil surface. With increasing time flow tends to a quasi- steady state condition in which the time rates of decrease in the infiltration and heat transfer rates are and small and relatively constant Once quasi-steady state flow is reached, the energy used to raise the temperature of a frozen soil at depth is supplied by latent heat released from the freezing of percolating meltwater in the soil layers above. The relative importance of transient and quasi-steady state regimes on infiltration for continuous and interrupted melt sequences is described. Parametric expressions for infiltration into a frozen soil for quasi-steady state flow and a constant supply of meltwater are developed. These dimensionless equations relate infiltration and infiltration rate to initial soil surface saturation, initial soil saturation, initial soil temperature, soil permeability and infiltration opportunity time. The adjustments of these relationships to fit the effects of melt/freeze cycles on water entry are investigated. Reasonable agreement between predicted and measured infiltration is shown.

Litong Zhao
Division of Hydrology
College of Engineering
University of Saskatchewan
Saskatoon, SK, Canada, S7H 5A9
Ph: (306) 966-7837
Fax: (306) 966-7829
E-mail: zhao@dvinci.usask.ca

D.M. Gray
Division of Hydrology
College of Engineering
University of Saskatchewan
Saskatoon, SK, Canada, S7H 5A9
Ph: (306) 966-7828
Fax: (306) 966-7829
E-mail: grayd@sask.usask.ca