| Infiltration in irrigated furrows is a complex hydraulic phenomenon that varies in both space and time. The effects of spatially variable infiltration on advance and infiltration of water along a furrow must be well understood to improve the management of furrow irrigation systems. Infiltration was measured every 8 m along a 240 m furrow with two types of infiltrometers, blocked furrows with stagnant water (BF-St) and blocked furrows with continuous flowing water (BF-CF), and the data at each location was regressed to fit both the Kostiakov and the modified Kostiakov infiltration functions. The regressed parameters in both equations were analyzed using classical and regionalized statistical theory. The parameters K and 'a' in both the Kostiakov and the modified Kostiakov equations were significantly correlated (K = {dollar}f{dollar}(a)). A first-order Markov autoregressive model was used to generate synthetic realizations of infiltration functions over space which closely resembled the original regressed infiltration data.; A time-solution (steps through space and solves for time) kinematic wave model for continuous flow furrow irrigation was developed and validated. The time-solution model included the advance, runoff, depletion, and recession phases of furrow irrigation for both nonuniform and uniform soil properties. A two-point volume balance procedure with the Kostiakov infiltration equation resulted to be inappropriate for estimating the stochastic properties of spatially variable infiltration in its present form. However, with future modifications and developments, the two-point volume balance procedure may become more appropriate when estimating infiltration under spatially variable soil conditions.; Simulations of furrow irrigation with spatially variable infiltration provided probability distributions of the total advance time, applied and infiltrated volumes, and Christiansen's uniformity coefficient. Significant variability in these parameters occurred due to infiltration variability. In particular, the uniformity coefficient was much lower (45 to 65 percent) than that predicted for a uniform soil (85 to 90 percent).; The time-solution kinematic wave model, combined with the first-order Markov infiltration model, provides a unique tool for analyzing the effects of spatial variability of infiltration on furrow irrigation management. Use of these models may require alterations in presently accepted practices that are based on constant infiltration functions instead of spatially variable infiltration. For example, the two-point volume balance procedure and Christiansen's uniformity coefficient will need modifications for use in a spatially variable infiltration environment. |
