Thermal conductivity of soils by weighted average model with transitional air/water inter-phase

This study presents two straightforward weighted average models for estimating the thermal conductivity of unsaturated soils. The first model ( WAM TAWI -1 ) incorporates the weighted average contributions from primary soil constituents (i.e., quartz ) and residual soil minerals , which are surrounded by a single continuous phase ( a transitional air/water inter-phase fluid ) with continuously changing thermal conductivity ( λ ) within its boundary values representing air ( λ a ) and water ( λ w ) . Similarly, the second model ( WAM TAWI -2 ) incorporates the weighted average contributions from all soil minerals and the transitional air/water inter-phase fluid. In contrast to the original model by de Vries in 1963, these models are free of complex latent heat transfer expressions, water/air shape fitting factors, and critical water content for changing continuous medium between air and water; consequently, they are exceptionally easy to use. The WAM TAWI -1 model is recommended because it only requires commonly available data: soil porosity , grain size distribution , and quartz content, while the WAM TAWI -2 model requires full mineral composition of a soil which is rarely available. The WAM TAWI -1 model was effectively validated with respect to λ data of 39 Canadian field soils, three standard sands, and 10 Chinese soils. The average standard deviations ( SD ) were ± 0.093 W⋅m −1 ⋅K −1 for 17 coarse soils, ±0.068 W⋅m −1 ⋅K −1 for 22 fine soils, and ± 0.079 W⋅m −1 ⋅K −1 for all Canadian soils. For the three standard sands, the average SD was ±0.143 W⋅m −1 ⋅K −1 , demonstrating good performance. Among the 10 Chinese soils, the model demonstrated good performance of ±0.142 W⋅m −1 ⋅K −1 for five coarse soils, while delivering superior estimates of ±0.094 W⋅m −1 ⋅K −1 for the remaining five fine soils. The results confirm the outstanding superiority of the model with respect to the original model by de Vries, in terms of its simplicity and λ predictions. • Simplified weighted average models with transitional air/water inter-phase fluid are proposed. • Validation against λ data of 39 Canadian soils and three standard sands. • Application to predict λ for 10 Chinese soils. • Simple model, self-explanatory, and easy to use.