Spatial Modeling of Soil Properties for Subsurface Drainage Projects

Abstract

Selecting areas for a subsurface drainage system and the proper design of such a system require sufficient and reliable soil data, which are inherently variable in both space and time. This study shows how geostatistical techniques can be applied for both feasibility and design stages of a drainage project to provide decision makers and designers with reliable estimates, a way of selecting areas for a subsurface drainage system and determining its implementation priority, and a way of accurate design of that system. The study is applied to a practical area of 1,533 ha in the Nile Delta of Egypt. Results showed that measured values of water table depth and soil salinity yielded ≈40% underestimation of drainage necessity and 59% overestimation of soil salinity levels, respectively, compared with their kriged values. The effect of using the kriged hydraulic conductivity on the determined lateral drain spacing and the required lateral drain pipe length was demonstrated. Five percent wider drain spacing was obtained with kriged hydraulic conductivity compared with its measured value with most steady- and unsteady-state drainage formulas. This value was approximately doubled using some other unsteady formulas. The minimum sample size for estimating a mean value at a given precision level could be reduced by three to seven times using kriging technique.