Determination of frozen soil thermal properties by heated needle probe

Determination of the in situ thermal properties of ice‐bearing frozen soil by a two‐prong heated needle probe has traditionally been confounded by the energy consumption and release resulting from the melting and re‐freezing of soil ice. A two‐dimensional soil thermal model that accounts for soil unfrozen water and phase change was used to analyse data from the laboratory and a field site in Spitsbergen. A close correspondence was found between modelled and observed thermal behaviour of the probe. The thermal model highlighted the unsuitability of the probe for determining the thermal properties for ice‐bearing frozen soil at temperatures approximately between −10 °C and 0 °C by standard solver software that does not account for melting of soil ice. The model shows that a heated needle probe can generally be used to determine the soil thermal properties for soil at or below −10 °C. For most soils the thermal properties vary only marginally between about −10 and −2 °C and therefore the thermal properties that are determined for a particular soil at temperature below −10 °C generally apply also to soils that are at temperatures between −10 and −2 °C. Copyright © 2003 John Wiley & Sons, Ltd.