Estimating Infiltration Parameters From Surface Irrigation Advance Data in Border Irrigation Systems.

Infiltration is highly spatially and temporally variable hydraulic parameter, which is difficult to measure in-situ. However, it is the most important input parameter in surface irrigation simulation engines. Simulations are essential to optimize parameters to increase the surface irrigation efficiency. Therefore, accurate estimation of real time infiltration parameters is vital. Conventional approach is to measure infiltration parameters once or twice and then use those values for subsequent optimizing simulations. However, field measured parameters are specific to the prevailing conditions and cannot be applied universally. Several methods have been suggested to estimate infiltration parameters from irrigation advance wetting front data. This field study evaluated three different methods of estimating infiltration parameters using irrigation advance and recession data in a border irrigation system to identify the best approach to be used in real time irrigation control.

Advance front data from 19 irrigation events were collected with wetting front sensors. Inflow and outflow rates and volumes were measured using flow monitoring devices. WinSRFR 3.1 was then used to simulate irrigation advance using field measured kostiakov infiltration parameters and kostiakov infiltration parameters estimated by three other methods namely, adjusted parameters, two-point method, and Post irrigation Volume balance method (PIVB).

Simulated advance trajectories were compared with field observed advance trajectories. Out of methods studied, two-point method produced consistently best match to the field observed advance trajectories with 0.0007, 0.0075, 96.77, 0.0105 and 0.9925 of mean values of SSE (Sum of squared error), ME (Mean error), Model efficiency, RMSE (Root of the mean squared error) and Willmott's index of agreement, followed by mean values of 0.0240, -0.0298, 78.27, 0.0371 and 0.9296 for PIVB method, 0.0363, -0.0402, 66.24, 0.0477 and 0.8757 for adjusted method, 0.0512,-0.0482, 60.17, 0.0553 and 0.8337 for field measured parameters. However, in two point method, infiltration parameters substantially varies depending upon the two points selected, out of which middle and end points provided the most reliable estimates.