Determination of Soil Water Characteristic Curve Equation for Estimating Soil Moisture Change at Mae-Sa Watershed, Chiang Mai Province

Abstract

A study to formulate the suitable soil water characteristic curve equations for estimating soil moisture  change at Mae Sa watershed, Chiang Mai province, Northern Thailand. Twenty one samples were collected from seven different places at different soil depths (5, 115 and 215 cm). The objectives were to determine the relationship between soil water content and matric potential and to obtain appropriated equations of that relation considering form Brooks and Corey’s, Campbell’s, and Van Genuchten’s equations. Coefficients from the obtained equation were employed in a soil water movement model in order to estimate change in soil moisture.  Modelling accuracy was verified using a parameter called Model Efficiency (EFF) proposed by Nash and Sutcliffe (1970).  Estimated soil moisture from the model was calibrated using Generalized Reduced Gradient technique.  The results could be summarized as follows.

The relationship between soil water content and matric potential of all samples collected from seven different locations were clear that water content in soil decreases when matric potential increases.  Beside, it is found that at all matric potential level, soil water content at shallower depth of five stations is higher than the deeper.  However, results of  the rest locations are in the opposite direction.  The reason caused remaining water content in shallower soil depth lower than the deeper is due to land cover of these locations are agricultural area which total porosity in soil can be reduced.  To select an appropriated soil water characteristic curve equation for a soil sample, R² can not be used since its value is not much difference. Nevertheless, curves that derived from Van Genuchten’s equation are the most reasonable curve especially when it close to Y axis which is the point that soil is close to saturation point and matric potential is nearly zero.  Based on coefficient from Van Genuchten’s equation, estimated soil moisture change before model calibration of a selected soil sample is lower than observed soil moisture.  However, a very good fit between observed and estimated was found after calibration. Total porosity and saturated hydraulic conductivity which caused the highest EFF (0.9397) are 0.4988 and 0.0001 cm/s respectively.

 Keywords: Soil water content, Matric potential, Soil water characteristic curve equation, Soil moisture change, Mae Sa watershed