Comparison of methods for predicting discharge in compound open channels

Abstract

The relative accuracies of four conventional stage-discharge prediction methods, for flow in compound channels, were compared in a laboratory study. Discharge measurements in a physical model comprising a deep central channel flanked by wide shallow berms (flood plains), as well as similar observations by others, provided the basis for comparison. System discharge was either undertestimated (‘single-channel’ method) or overestimated (various ‘separate channels’ methods) when depth on the flood plain zones was small, i.e. when the velocity differential between the deep and shallow zones was large. These discrepancies can be attributed to questionable assumptions made in the different methods tested regarding the nature of flow in compound channels. A new method is proposed which accounts for momentum transfer between adjacent deep and shallow zones of compound sections. This transfer mechanism and its effect on system conveyance are usually ignored in conventional methods. The new method gives satisfactory results for the conditions examined. Its accuracy can be further improved with the development of more accurate friction factor relationships for compound channels.