Evaluation of the heat pulse probe method for determining frozen soil moisture content

Heat pulse probes (HPP) have been widely utilized to determine soil thermal properties and water content in unfrozen soils; however, their applications in frozen soils are largely restricted by phase change and the presence of unfrozen water. This study explores the possibility of using HPP to determine total water content of frozen soils by (1) establishing the optimum heat applications to limit melting, (2) improving the mathematical representations for frozen conditions, and (3) evaluating the applicability of HPP methods under various temperature and moisture conditions. A custom-built HPP was tested at total moisture levels that varied from full saturation to oven dry and initial soil temperatures from 20 degrees C to -11 degrees C. The applied heat pulse durations varied from 8 to 60 s, with total heat strength varying from 100 to 2000 J m(-1). Comparison of mathematical methods involved two analytical solutions and a one-dimensional finite difference numerical model. While both analytical methods assumed no phase change, the numerical model considered ice melting and unfrozen water. Conclusions include the following: (1) the numerical model with phase change is the only appropriate method to represent the temperature change curve once melting occurs; (2) below -4 degrees C, ice melting could be limited, and measurement errors of total moisture content were within +/-0.05 m(3) m(-3); (3) application of HPP between -2 degrees C and 0 degrees C is difficult because of the retarded response of probe temperature to changing moisture contents and heat applications; and (4) probe spacing is a sensitive parameter requiring calibration once reinstallation of the probe or the thawing and freezing process occurs.