Thermal conductivity of soil: A review on the vast experimental data and predictive models

Abstract

In this article, a comprehensive review is provided that summaries and analyzes the experimental and modeling studies on soil thermal conductivity. The effects of internal and external parameters on soil thermal conductivity are analyzed by extracting data from existing literatures. Generally, soil thermal conductivity increases with the rise of water content, degree of saturation, dry bulk density, quartz content, concentration of contaminants, etc., while it decreases with ratio of clay particles, porosity, concentration of salt solution, temperature below freezing point. Traditional theoretical and experimental models of soil thermal conductivity overcome the time-consuming drawbacks of experimental measurements, but most of them are only available for specific soil types or conditions. Machine learning methods are gradually being applied in recent years, by which models with better accuracy can be established. In future studies, measurement on soil thermal conductivity in specific conditions should be supplemented, such as temperature nearing the freezing point and above the boiling point of water, contamination enrichment, and state nearby the compaction curve, to meet new requirements in engineering. Meanwhile, based on more comprehensive experimental data, various machine learning methods should be applied to training prediction models with improved performance.