H54E-07:
Modelling infiltration processes in frozen soils

Friday, 19 December 2014: 5:30 PM
Andrew M Ireson and Lee S. Barbour, University of Saskatchewan, Saskatoon, SK, Canada
Abstract:
Understanding the hydrological processes in soils subject to significant freeze-thaw is fraught by “experimental vagaries and theoretical imponderables” (Miller 1980, Applications of soil physics). The infiltration of snowmelt water and the subsequent transmission of unfrozen water during thawing, is governed by hydraulic conductivity values which are changing with both ice and unfrozen water content. Water held within pores is subject to capillary forces, which results in a freezing point depression (i.e. water remains in the liquid state slightly below 0°C). As the temperature drops below zero, water freezes first in the larger pores, and then in progressively smaller pores. Since the larger pores also are the first to empty by drainage, these pores may be air filled during freezing, while smaller water filled pores freeze. This explains why an unsaturated, frozen soil may still have a considerable infiltration capacity. Infiltration into frozen soil is a critical phenomena related to the risk of flooding in the Canadian prairies, controlling the partitioning of snowmelt into either infiltration or runoff. We propose a new model, based on conceptualizing the pore space as a bundle of capillary tubes (with significant differences to the capillary bundle model of Wannatabe and Flury, 2008, WRR, doi:10.1029/2008WR007102) which allows any air-filled macropores to contribute to the potential infiltration capacity of the soil. The patterns of infiltration and water movement during freeze-thaw from the model are compared to field observations from the Canadian prairies and Boreal Plains.