Abstract

Many current models of heat and water flow in frozen soils overestimate the freezing-induced redistribution of water. These models also treat the soil physical properties as constant in time although they are strongly influenced by the frost itself. This study was conducted to determine possible methods to overcome these two problems in an operational hydraulic model. Winter measurements of soil temperature and water content were performed on a clay soil and on a layered loam soil. The data were compared with simulations made with a physically based, one-dimensional model of coupled heat and water flow. Two procedures were tested for the calculation of hydraulic conductivity of partially frozen soil: firstly, an interpolation procedure, taking into account the strong non-linearity of the hydraulic conductivity function close to the freezing front. Secondly, an impedance parameter was used to describe the effect of ice lenses. A spring time modification of the soil moisture characteristic curve was tested to account for the frost-induced changes of the soil structure. Simulated temperatures and water contents agreed well with measurements, both for the clay and the layered loam soil, after introduction of the impedance parameter. The alternative interpolation procedure did not only reduce the hydraulic conductivity of the frozen soil sufficiently to produce a realistic redistribution, but also enabled use of a lower value on the impedance parameter. Changes in water retention properties resulting from frost action during winter caused an overestimation of simulated water content of 15% by volume in the heavy clay soil during spring. This discrepancy was eliminated by increasing the frequency of pore diameters below 0.1mm in the model during spring.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call