Estimating Ice Content in Unsaturated Frozen Soils from Apparent Heat Capacity during Freeze–Thaw Processes

Abstract This study focuses on the development of a new approach to estimating ice content in frozen soils using apparent heat capacity during freezing and thawing. Existing approaches for ice content have several limitations due to their implementation complexity, unsuitability for in situ measurements, and calibration requirements across different temperatures. In this study, a new approach for estimating volumetric ice content from temperature-dependent apparent heat capacity measured during freeze–thaw processes is presented. Box-scale laboratory experiments were performed on sand, silt, and clay prepared at different dry densities and degrees of saturation. Apparent heat capacity was measured over −15°C to 15°C and used to compute ice content from the latent heat released during the phase change. Because transient heat transfer affects measurements under nonisothermal conditions, an empirical rate-dependent normalization was applied to improve consistency between apparent heat capacity measurements during freezing and thawing. The ice contents calculated using the corrected heat capacities exhibited a strong correlation with the capacitance-derived ice contents for silt and clay, with most values falling within ±0.03 m 3 /m 3 . Deviations were observed in sand due to abrupt freezing over a narrow temperature interval relative to the probe response time. The findings from this research demonstrate that heat capacity provides a reliable indicator of phase change and captures distributed freezing behavior in soils exhibiting freezing characteristic curves. The proposed method offers a low-cost, laboratory-validated alternative for ice content estimation in fine-grained soils exhibiting distributed freezing behavior, with potential for field deployment pending in situ validation. The results support improved thermal modeling of permafrost and frozen ground systems for soils with continuous freezing characteristic curves.