Prediction of overland flow resistance and its components based on flow characteristics using support vector machine

Abstract

Abstract Due to the significance of overland flow resistance (f) in hillslope hydrology and models of erosion, the present study peruses the capability of non-linear approaches to estimate the overland flow resistance and its components. For this purpose, numerous support vector machine (SVM) models were developed and tested using four series of flume experimental data sets. This study was divided into two parts; the first section aimed to model the total overland flow resistance and investigates the effect of the different parameters on the resistance. In the second section, the resistance was linearly divided into different types of resistance (namely, grain resistance (fg), form resistance (ff), wave resistance (fw), sediment transport resistance (fs)). Then the separated components (fw and fs) were estimated by SVM. The results revealed the importance of the Froude number (Fr) values as an input data in most of the estimated models. Also, it was concluded that the slope gradient has a stronger impact on the on the sediment transport resistance over plane beds than the other hydraulic properties do. The outcome of the models approved the capability of the proposed models based on SVM. Also, SVM outperformed the empirical approaches.