https://gwf.usask.ca/projects-facilities/all-projects/p1-frozen-soils.php

This project improves the understanding of soil freeze-thaw processes which critically affect the surface energy and water balance in cold regions. The partitioning of snowmelt into runoff and infiltration is arguably the single most important control on flood risk and water for crops in the Canadian prairies. Understanding of the physical processes involved is fraught with challenges and there remain major gaps. Perhaps the most basic property is the soil freezing characteristic curve, SFC: a relationship between unfrozen water content and soil temperature (below zero degrees Celsius), analogous to the soil moisture characteristic for unfrozen conditions. This represents the phenomenon of freezing-point depression in soils, and controls the hydraulic properties. However, there is no consensus on why this actually happens. Moreover, there is no simple in-situ method to measure this phenomenon directly in the field – the problem being our inability to interpret most soil moisture instrumentation in frozen conditions. From a hydrological perspective, this understanding is critical to being able to predict the fate of snowmelt, and the overall water balance of a watershed or field.