Modelling the effect of canal bed elevation on seepage and water table rise in a sand box filled with loamy soil

The HYDRUS 2D finite difference two-dimensional water balance model was experimentally tested for transient and steady state seepage flux, mound height, and piezometric water level from soil surface as a function of time and horizontal distance from the centre of the canal (half width = 45 cm) under different canal bed elevations (20, 0, -40, -80 and -120 cm denoted as experiments D1, D2, D3, D4 and D5, respectively) and constant water head of 5 cm in a sand box (200 cm x 170 cm x 150 cm) filled with Hisar loam soil. Differences of means between measured and predicted values of infiltration flux, seepage flux and mound height as tested by paired t test were not found significant (P = 0.05). Seepage flux and mound height increased with increasing canal bed elevation. Phreatic level depths were everywhere much shallower than the piezometric water level depths in experiments D1, D2 and D3. However, in experiments D4 and D5 both phreatic and piezometric levels were at similar depths. The seepage parameters and mound height increased, and water table depth decreased, linearly with increasing canal bed elevation. Lowering the canal bed to 120 cm below the soil surface reduced the seepage rate to that of lined canals. The projections in a large flow domain also revealed that lowering the canal to -2 and -4 m below soil surface stabilized the water table at 2.5 and 4.5 m below soil surface, respectively. The practical implications are that open drains should be used for irrigation in areas underlain with a brackish groundwater aquifer and gravity canals may be allowed only where groundwater aquifer is of good quality and sub-surface water withdrawal is practiced for irrigation.