Estimating the three-parameter Kostiakov infiltration function in furrow irrigation from advance data

Typescript (photocopy).A three-parameter model was developed to evaluate the parameters (K, a, and F) in the extended Kostiakov infiltration equation when continuous flow irrigation is practiced. The model also permits the evaluation of the average cross sectional area (C) of water stored on the soil surface during irrigation and an evaluation of the infiltrated profile shape factor (B). The model predicts K, a, F, C, and B using only measured advance and recession data. The model was verified and its sensitivity tested by comparing results from the three-parameter model with results obtained from a kinematic wave numerical model. Excellent agreement was obtained over a range of soil types, surface roughnesses, field slopes, and furrow flow rates. The three-parameter model was also tested using several sets of field data for furrow irrigation at Flowell (Utah) and Kimberly (Idaho). The infiltration functions obtained using the three-parameter model were compared to the infiltration functions obtained using the recycling furrow infiltrometer method. The three-parameter model and the recycling furrow infiltrometer estimated similar infiltration functions for the furrows studied. The effects of the method chosen to estimate the infiltration function of a soil (recycling furrow infiltrometer, Latortue's (1984) continuous flow two-parameter model, and the three-parameter model) on the simulation of irrigation events were investigated. Results showed no significant effect on the simulation of the advance phase and the simulation of the volume of water infiltrated at any point of the furrow. However, the method chosen to estimate the infiltration function may significantly affect the estimate of the runoff volume. This is due to the fact that the simulated runoff volumes are calculated by integrating the volume of water infiltrated at a point over the length of the furrow. Thus, the longer the furrow, the greater the variation in the simulated runoff