Developing a texture-based soil hydrologic characteristics model and extending this model to predict soil strength characteristics.

摘要

Hydrologic models are used to describe water flow patterns in the soil, but there are not many models to describe the influence of soil water on soil mechanical properties. The purpose of this study was to adapt a hydrologic model to predict soil strength. In order to accomplish this goal, a new soil hydrologic characteristics model was developed using new predictive equations to improve stability and usability of the texture-based hydrologic model. This model, referred to as the Soil Texture and Compaction (STAC) model, was created to predict saturated moisture content, moisture content at 33 and 1500 kPa water tension, water tension at air entry, pore size distribution, and the saturated hydraulic conductivity given the fraction of sand and clay, and an estimate of the soil compaction. The STAC model was created using multi-variable linear regression based on mean values of the six listed hydrologic traits for each soil type. Initial estimates were improved by using density adjustments. When compared with 1717 A horizon samples, the model achieved a coefficient of determination (R2) ranging from 0.02 to 0.99 for the listed hydrologic traits, and the standard error of estimate (Se) ranged from 0.02 to 0.38. The fit is vastly improved with a good estimate of the density adjustment factor. In addition, a method is proposed to extend a soil hydraulic model to predict the soil strength. The soil strength is required to determine the tractive capability of equipment working in the field. Extending a soil hydrologic model will allow the machine tractability to be predicted in simulated surface conditions. The volumetric soil moisture content and level of soil compaction are used as independent variables to predict the cone index. The soil moisture content is determined through a hydrologic model. The soil compaction can be calculated by comparing true and expected bulk densities, or simply estimated by the user. The soil strength predictions were tested using a stochastic weather generator, soil characteristics models, and accepted soil hydrology methods for infiltration, evapotranspiration, and percolation.