Predicting ground temperature is an important part of the analysis of geothermal resources assessment and use. Thus, we develop and validate one-dimensional numerical model for heat and mass transfer in partially frozen soils. The model is implemented in HVACSIM Plus and used to simulate the thermal regime of soil profile. In addition to modeling heat conduction, model variations also includes moisture transfer, snow accumulation and melting, and soil freezing and thawing. The results are compared against experimental measurements of ground temperature for three locations in Montana, USA. The differences between simulated depth temperature with and without snow cover and freezing and thawing of soil reveal that ground temperatures are predominantly influenced by these two factors. Considering moisture transfer slightly improves temperature predictions, although it increases computational time by one order of magnitude. To balance computational efficiency with prediction accuracy, we propose an equivalent moisture content of 40-60% saturation in predicting ground temperature. (C) 2014 Elsevier Ltd. All rights reserved.