Study on the selection of unsaturated flow model for the different types of soil and soft rock.

Abstract

Precise estimation of unsaturated hydraulic properties of porous media is indispensable in various study areas, such as analyzing the moisture flow, the drying process occurring from the surface, and the pollutant migration beneath the ground surface. Although many empirical/theoretical models describing the unsaturated hydraulic properties have been proposed by several previ ous researchers, the best model for the different types of soil/rock may not be identical. Thus, the model selection process and the estimation technique of the parameters included in the models should be developed. In the present study, the inverse technique based on the transient evapo ration change was investigated to select the model and estimate the model parameters. The experimental work was based on a relatively low permeable soft rock and a rela tively high permeable sandy soil (Toyoura standard sand). Experimental equipment was developed to precisely mea sure the evaporation rate for the high permeable sandy soii. The Genetic Algorithm (GA) was adopted in the inverse technique as an optimization tool. In order to simplify the problem, only the drying process from the saturated con dition was considered. It was established that the infor mation concerning the transient evaporation change could be used for the model selection and parameter estimation. Further, the saturation distribution could be used for the selection of the models. The present study provides important information for the development of the model selection process