Hydraulic properties determination of coarse grain soils by evaporation method

Abstract

Coarse-grained soil is the most common and problematic soil of Northeastern Thailand. The knowledge of its hydraulic properties is, therefore, essential for the analysis, design, and management of irrigation, saline soil dissemination, and landslide protection. The two main hydraulic properties are soil water characteristics curve and
hydraulic conductivity function as described by the van Genuchten functions. There are many determination methods for these soil properties available, but most of them are tedious, time consuming, complex, have high cost, and usually provide only one the soil characteristics. The evaporation method is a laboratory method which can be used to determine both properties from a single set of data. This method can deal with both undisturbed soil samples and
repacted soil samples. There are two calculation procedures from the evaporation data, i.e. the traditional so-called “direct method,” and the newly developed method called the “inverse method.” However, for coarse-grained soil, the inverse method can easily result in nonuniqueness, especially when determining the values of the fve parameters (q_r, q_s, a, n and K_s). Therefore, the objective of this study is to test the evaporation method with sandy and loamy sandy soils using the inverse method in order to fnd the correct set of van Genuchten parameters; by utilizing the initial data derived from the direct method for the ftting procedure. During the evaporation experiments with four
sandy and fve loamy sandy soils, series of pressure head and soil water content values were obtained. In the frst stage, the direct method was used to calculate soil hydraulic properties in order to frst get the approximate parameter values. Then, these values were used as the initial guesses for the inverse method. The results showed that parameters (_r, q_s, a and n obtained from the direct method could be used as the initial guesses for the inverse method, making the complete set of parameters (qr, qs, a, n, K(h) and K(q)) more robust and unique.