An Empirical Model for Estimating Soil Thermal Conductivity from Soil Water Content and Porosity

摘要

Soil thermal conductivity lambda is a vital parameter for soil temperature and soil heat flux forecasting in hydrological models. In this study, an empirical model is developed to relate lambda only to soil volumetric water content theta and soil porosity theta(s). Measured lambda values for eight soils are used to establish the empirical model, and data from four other soils are used to evaluate the model. The new model is also evaluated by its performance in the Simple Biosphere Model 2 (SiB2). Results show that the root-mean-square errors (RMSEs; ranging from 0.097 to 0.266 W m(-1) K-1) of the new model estimates of lambda are lower than those (ranging from 0.416 to 1.006 W m(-1) K-1) for an empirical model of similar complexity reported in the literature earlier. Further, with simple inputs and equations, the new model almost has the accuracy of other more complex models (RMSE of lambda ranging from 0.040 to 0.354 W m(-1) K-1) that require additional detailed soil information. The new model can be readily incorporated in large-scale models because of its simplicity as compared to the more complex models. The new model is tested for its effectiveness by incorporating it into SiB2. Compared to the original SiB2 lambda model, the new lambda model provides better estimates of surface effective radiative temperature and soil wetness. Owing to the newly presented empirical model's requirement for simple, available inputs and its accuracy, its usage is recommended within large-scale models for applications where detailed information about soil composition is lacking.