A simplified de Vries‐based model to estimate thermal conductivity of unfrozen and frozen soil. (3rd August 2016)

• Record Type: Journal Article
• Authors: Tian, Z.; Lu, Y.; Horton, R.; Ren, T.
• Is Part Of: European journal of soil science. Volume 67:Number 5(2016:Oct.)
• Page Start: 564
• Page End: 572
• Publication Date: 2016-08-03
• Subjects: Soil science -- Periodicals; 631.4
• DOI: 10.1111/ejss.12366
• Languages: English
• ISSNs: 1351-0754

Abstract

Abstract : Soil thermal conductivity ( λ ) is an important property in soil physics, environmental science, thermal science and engineering disciplines. In situ measurement of λ is complicated and there is a need for a model to predict λ accurately in both unfrozen and frozen soil. In this research, we developed a simplified de Vries‐based model to estimate λ of both unfrozen and frozen soil. The simplified model follows the basic assumptions of the de Vries model, but simplifies or improves the estimates of λ air, λ minerals and g a(minerals), g a(air) and g a(ice) ( g a are the shape factors for various soil components): λ minerals and g a(minerals) are related to soil mineral composition, and g a(air) and g a(ice) are related linearly to the volumetric fractions of air and ice, respectively. The λ and g a values of sand, silt and clay are required for calculating λ minerals and g a(minerals) . The values of λ silt and g a(silt), however, are not available in the literature. In this research, we estimated λ silt and g a(silt) by fitting the simplified model to measured λ values of 17 soil samples with various textures, water contents and bulk densities. The simplified model was validated with measured λ data from an additional 10 unfrozen and 18 frozen soil samples, obtained either from the literature or from our own measurements. The performance of the simplified model was evaluated by comparing estimates of λ with those from other available models. The results showed that the simplified de Vries‐based model provided accurate and consistent data for λ, and it performed better than other de Vries‐based models. Highlights: We developed a simplified de Vries‐based model to estimate thermal conductivity, λ, of unfrozen and frozen soil. We improved methods to calculate thermal conductivities of soil air and minerals and shape factors of minerals, air and ice. Thermal conductivity and shape factor of silt were estimated with inverse modelling. Our new, simplified model gave more accurate predictions of λ than did other de Vries‐based models.