Using artificial neural network with clustering techniques to predict the suspended sediment load
by Abdelghafour Dellal; Abdelouahab Lefkir; Yamina Elmeddahi; Samir Bengherifa
International Journal of Hydrology Science and Technology (IJHST), Vol. 19, No. 2, 2025

Abstract: Rivers are natural water channels that are influenced by a variety of factors, including erosion and sedimentation, which have a detrimental impact on the ecosystem's health and water quality. Recently, researchers resorted to using an artificial neural network (ANN) to model the suspended sediment load. This study addressed the application of a multi-layer ANN model. Feed-forward with a backpropagation algorithm based on five different collection methods for the input data. To model the daily suspended sediment load in the Sacramento River, California, USA, current and delayed Ql flow discharge data and solid flow Qs data were used. The accuracy of the five methods was compared in 10 different input groups based on proficiency criteria: standard deviation ratio RSR, coefficient of determination R², percentage bias (PBIAS), and Nash-Sutcliffe efficacy (NSE). The ANN model with the k-mean clustering technique provides the best results. The RSR values varied between 0.30 to 0.42, and the R² values ranged from 0.82 to 0.91, while the range of NSE values was from 0.79 to 0.90.

Online publication date: Mon, 03-Feb-2025

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Hydrology Science and Technology (IJHST):
Login with your Inderscience username and password:

Username:        Password:          Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com