ABSTRACT
The problems of on-farm water management and irrigation system design parameters which are often chosen with limited or no analysis of a unique local condition leads to the poor performance of most irrigation systems today. In this study, three furrow irrigation water advance models, referred to as Walker and Skogerboe model, Michael's advance model and Philip and Farrell model were verified and evaluated for their prediction of water advance distances under specific field conditions at the Galma Irrigation Scheme, kilometer 16 along Zaria - Jos road. Fifteen irrigation events were monitored. Several parameters were measured including the furrow geometry, slope, furrow length, furrow width, advance and recession time, and infiltration characteristics. The models were applied to all the events and the results were compared to those obtained from the field measurements. Thereafter, the predictions by the different models were compared with one another. The indices used in evaluating the performance of the models include slope of the best-fit line ?, standard error of estimate ?, coefficient of determination R2 and the absolute error E (%). This study shows that the Walker and Skogerboe model performed best in predicting advance distances with average indices values of ? = 1.0381, ? = 0.42, R2 = 0.9985 and E = 4.40%. It is followed by Michael's advance model which performed reasonably well with average indices values of ? = 1.0175, ? = 1.40, R2 = 0.9837 and E = 9.67% while the performance of Philip and Farrell was poor with average indices values of ? = 0.3508, ? = 7.43, R2 = 0.9499 and E = 37.08%. Considering the overall suitability of the models to the field condition encountered in this study, both the Walker and Skogerboe model, and Michael's model are suitable for predicting water advance distance.