Development of a regression-based model of border irrigation on cracking soils

Abstract

Earlier studies have shown that the Strelkoff model for simulating border irrigation is the most satisfactory for the field conditions of south-east Australia, and that the model can be employed for practical applications. However, the use of this model for design purposes on cracking soils was found to be tedious and time consuming, particularly if the number of simulation runs required are excessive (say > 10). To overcome these difficulties, a “ regression-based model of border irrigation” is developed by performing multiple regression analysis on the input and output data of the Strelkoff model. The model so developed is an approximation of the Strelkoff model, but due to its simplicity it can be used for simulating an irrigation event in a fraction of time required for the Strelkoff model. The model developed in this study was evaluated for its performance using four irrigation events monitored in south-east Australia, and its prediction accuracy was found comparable to that of the Strelkoff model.