ASSESSMENT OF GRAVITY-DRAINAGE UNSATURATED FLOW UNDER FIELD CONDITIONS FOR A SILTY CLAY LOAM SOIL: II. PREDICTING TIME RATE OF CHANGE IN DEPTH OF STORED WATETR IN SOIL PROFILE

Abstract

An 8 by 8m field plot was continuously flooded with water for forty days; the plot was covered then to prevent evaporation after ending water supply. In this study, drainage cycle started when all water infiltrated through soil surface. Gravimetric water content measurements were made from the soil surface to 140cm in 10cm increment during 90 days of drainage. The general flow equation was analytically solved under unit gradient assumptions to predict depth of stored water, and time rate of change in stored water as a function of depth during drainage period using three functions; , , and ; that describe the hydraulic conductivity as a function of water content (where Km is saturated hydraulic conductivity, θ is volumetric water content, and the subscript m, r denote maximum and minimum water content values, and β, α, and n are empirical parameters). When a power function was fitted to describe the log-log relation between depth of stored water and time, straight line trends were obtained with coefficient of determination values ranged from 0.849 to 0.970 for different depths. A close and highly significant 1:1 relationships between measured and predicted values of depth of stored water were obtained during the drainage period with correlation coefficient values of 0.987, 0.989, 0.973 and regression coefficient values of 0.993, 1.126, and 1.124 for the three functions respectively. Measured rate of change in stored water increased with increasing depth of soil profile and ranged from 1.5541 to 17.3855 cm.day-1 during the first time interval and from 0.0001 to 0.037 cm.day-1 during the last time interval for the 10 and 140cm depths respectively. A 1:1 relationships between predicted and measures values of the rate of change in stored water gave regression coefficient values of 0.986, 0.826, 1.966 and coefficient of determination values of 0.889, 0.850, and 0.863 for the three functions respectively. Values of the regression coefficient clearly showed that the first function accurately predicted the rate of change in stored water during drainage period while values of the rate of change in stored water during drainage period were under predicted by the second function and over predicted by the third function.