Stochastic Modeling of Local Scour Depth Prediction on the Basis of Prediction of River Morphological Changes in Braided River

آب و خاک

Article 2, Volume 25, Issue 2, May 2011   

Document Type: Research Article

DOI: 10.22067/jsw.v0i0.9217

Authors

S. Porbakhshian¹; M.R. Majdzadeh Tabatabaei^{* 2}; S.S. Mousavi²; Sh. Mansouri²

¹Power and Water University of Technology

²Water and Environmental Engineering Department, Power and Water University of Technology

Abstract

Abstract
Morphological river models are designed to provide physical insight into the morphological response and to assist river engineers and managers in the design, operation and maintenance of river systems. Here deterministic modeling weak for a dynamic and stochastic of nature river environment. Specially, these could not predict the exact shape of the river bed, Specially e.g. for braided river because the bed level variability and variations in cross-sectional. Since a stochastic model approach copes with the variability of system behavior of the time, therefore need for Stochastic modeling on the location of morphological changes in rivers and variations in river bed seems necessary. Many large rivers in the world have recently undergone through a great deal of morphological changes, Which has led to the development of local scouring, therefore, it has become an important problem for the river engineering. The change of river morphology is evaluated by braid parameter in braided rivers. A decrease in braid parameter results in a braided channel changes to meandering. As a result, local scouring process is accelerated. Since Process of the changes in river cross section are usually caused by change in water and sediment discharges or by river works. Moreover, river gradient plays a key role in channel morphological changes therefore In this research, local scouring relationship with river morphologic changes are investigated by stochastic modeling in braided rivers based upon for parameters such as maximum water, sediment discharges, river bed gradient river and bed elevation. The model was then tested by data obtained from Yahagi river in Japan. That the month Maximum Stream flow data is predicted by time series models (ARIMA) and three sediment transport equation were used to calculate the bedload such as Bagnold, Meyer-peter and Einstien Brown. predicted results show If calculate the bedload with the Bagnold equation, this model could predict significantly in cross-sectional and local scour depth, predict river morphological changes.

Keywords: Braided river, Local scouring, Stochastic modeling, ARIMA, Non linear variant regression