Experiment Study On Soil-water Characteristic Curve In Unsaturated Zone

In unsaturated zone, a variety of physical, chemical, biological and hydrological processes have been token placed, the flow 0f gas and liquid as well as the transfer and transform processes of various material components also have been happened, coupled with the superposition of human activities and the contaminations, resulting in very complicated process in unsaturated zone. A series of theoretical and practical problems such as Soil erosion , groundwater pollution, groundwater resources evaluation, soil degradation, desertification, irrigation systems, soil pollution, soil improvement, water conservancy and groundsill distortion have closely connection with unsaturated zone. The evaluation of unsaturated zone parameters is right or not have direct impact on the quality and accuracy of evaluation. The water characteristic curve is basis of unsaturated zone parameter. However, the test about the water characteristic curve in unsaturated zone have certain limitation, the relation of moisture content and capillary pressure is very complicated in the water characteristic curve and difficult to theoretically fetch exact relationship currently. Consequently, It is great important of confirming parameter in the water characteristic curve.The object of this paper is to obtain accurate water characteristic curve parameters, based on consulting and borrowing ideas from their predecessors, carefully designed and improved the experimental apparatus, put forward a reasonable plane for test. It was spent one year time to obtain measured data for five kinds of sandy soil. Taking soil water dynamics as a guidance, in virtue of Matlab software , basing on comparison algorithm, choosing six representative models of empirical formula in order to fitting parameter and sensitivity analysis for measured data.Achieving the following results:Designed test equipment can quickly and accurately obtain the measured data of the water characteristic curve.Comparing the nonlinear fitting methods with non-linear regression analysis by Matlab software, the results show that the non-linear regression is higher calculation accuracy.Selecting six models to fit the measured data of the five kind of sandy soil. Mckee and Bumb (1984) model have the smallest deviation of measured and calculated values. The average deviation is 5e-4,and the fitted parameters are well agree with the fact. Therefore Mckee and Bumb (1984) model is suitable for soil and moisture characteristic curve.Making sensitivity analysis for water characteristic curve parameters, in Mckee and Bumb (1984) model,θs andθr (saturated water and residual water content) are linear relation with water content. a(fitting parameter) is the most sensitive parameter for water content ,the parameter b(fitting parameter) is less sensitive.