An expert system for estimating soil thermal and transport properties

Prediction of soil thermal conductivity k is particularly difficult at high temperatures T (50-90 °C) and very low moisture content θ; the k value may increase with temperature by a factor of 3-5. This phenomenon is due to water vapor migration resulting in latent heat transfer, which is strongly dependent on soil water characteristics (SWC). In the past, the SWC influence on k prediction was never studied in great detail - mainly due to a lack of reliable SWC experimental data at low θ. Currently, hydraulic properties of soils can be evaluated from numerous predictive models correlated with experimental data. The paper objectives focus on: effects of SWC on k prediction; explanation of nonlinear variation of k at high T with θ ranging over the full degree of saturation; large k over- predictions at low θ and high T. Results obtained show very strong k dependence on the SWC function; therefore, accuracy of SWC estimates cannot be disregarded. The paper explains a nonlinear k behavior at high T as a combined effect of water vapor migration, SWC, and soil air relative humidity. To this end the paper provides also information about Soil Thermal and Transport Properties — Expert System (STTP-ES), collection of the notable predictive models for these properties, for both moist and frozen soils, which have been gleaned from the literature and integrated into one software package. The appendix provides a brief description and recent modifications made to the STTP-ES. Closing discussion concentrates on STTP-ES shortcomings, development of the Windows driven package, Expert System extension to other porous media and the database development.