Evaluation of methods for determining infiltration parameters from irrigation advance data

Knowledge of soil infiltration parameters is of the utmost importance for optimum performance and management of surface irrigation. The advance of the surface flow across a field varies according to the infiltration rate, ultimately determining the soil moisture uniformity and water application efficiency. Calculation of the infiltration parameters from irrigation advance data is now the preferred method. If the process is to be included in a real-time control system it must be done accurately, reliably and rapidly, without user intervention, and with a minimum of field data. With this purpose in mind, six infiltration methods were selected for evaluation: the two-point method, the computer model INFILT, the method of Upadhyaya and Raghuwanshi, the Valiantaz one-point method, the Shepard one-point method and a simple linear infiltration function. Field experimental data from 10 furrow irrigation trials were used to evaluate the methods, these data covering a range of soil types and furrow characteristics. The data sets were analyzed to determine infiltration parameters to evaluate the ability of the selected methods for prediction of advance and cumulative infiltration. The results showed INFILT to be the most accurate and reliable method, performing consistently well for all of the data sets. Generally the two-point and linear estimation methods also performed well, although the two-point method is handicapped by its requirement for prior knowledge of the steady or final infiltration rate f(o). None of the methods proved entirely suitable for use in a real-time control system. Copyright (c) 2005 John Wiley & Sons, Ltd.