Thermo-TDR probes were installed in a bare field to measure subsurface soil temperatures, thermal properties, and liquid phase water contents in the 3 to 147 mm soil layer in the 2012-2013 winter season. In situ subsurface soil ice content as a function of time and depth were determined at 60-min intervals following the SHB theory. A sensitivity analysis of temperature, heat capacity, thermal conductivity, and time step interval of the SHB method was performed based on numerically produced soil freezing and thawing events.
The results showed that the in-situ ice contents from the SHB method were sometimes unrealistically large or even negative. Thermal conductivity errors and time step interval were the key factors attributing to the errors and instability in the SHB method while temperature and heat capacity showed less influences. To obtain reliable ice content with the SHB method, the required measurement accuracies of soil temperature and thermal conductivity were ±0.02°C and ±5%, respectively. The accuracy of soil heat capacity could be as large as ±50%. The time interval of SHB measurement was another critical factor. A 15 min SHB interval produced more accurate estimation of ice content than that of a 60-min interval. Shorter interval of temperature measurement may help to reduce the impact of thermal conductivity errors. We concluded that accurate thermal conductivity and relatively short time step are required to accurately estimate ice content with the SHB method.