Kiyoumars Roushangar; Saman Shahnazi
Abstract
In most water resources studies, the bed load transport rate is considered as a constant proportion of total load due to the difficulty and costs associated with measuring of it, which is not reasonable due to the high variability of this ratio. In this study, data collected from 19 coarse-grained rivers ...
Read More
In most water resources studies, the bed load transport rate is considered as a constant proportion of total load due to the difficulty and costs associated with measuring of it, which is not reasonable due to the high variability of this ratio. In this study, data collected from 19 coarse-grained rivers in the United States were employed to predict bed load, total load transport rates and the ratio of bed to total sediment load using Support Vector Machine which is a branch of intelligent methods. Next, the results were compared and evaluated with classical methods. Results showed that this method has a very high performance compared to the classical methods and performance criteria in predicting the bed to total sediment load ratio has acceptable results. In addition, the modeling showed that the ratio of average velocity to shear flow velocity and the Froude number is the most effective parameters in predicting bed load, total load and the ratio of these.
