Journal of Wetland Ecobiology
Volume 17, Issue 3 (Autumn 2025)
2025, 17(3): 1-13 |
Back to browse issues page
Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
Sakiani M A, elmizadeh H, Zoratipour A. Estimation of river suspended sediments with ANNs and ANFIS methods by modeling in MATLAB (case study: Dez river in Khuzestan). Wetland Ecobiology 2025; 17 (3) : 1
URL: http://jweb.ahvaz.iau.ir/article-1-1114-en.html
URL: http://jweb.ahvaz.iau.ir/article-1-1114-en.html
Abstract: (327 Views)
In recent years, the use of intelligent systems in order to increase the accuracy of estimating the amount of river sediments has become common. The studied area is the end part of the Dez River basin, which is one of the largest basins of Karoon with an area of 6288 square kilometers, which passes through the agricultural areas of Dezful and Shush. In this research, using MATLAB software and in order to evaluate the model estimation results, the data related to the measurements of concentration or sediment discharge with their corresponding flow rate during the period of time were transferred to a logarithmic axis and created through the relationship between river discharge, sediment discharge and the generalization of this relationship to river flow statistics, the best fitting line was passed through the points using the least squares method and was extracted. Next, in order to evaluate the results of artificial neural network models (ANNs), ANFIS and compare them with regression estimates, these models are simulated. Finally, the criteria of coefficient of determination (R2), root mean square error (RMSE), Nash-Sutcliffe coefficient (NS), standard deviation (STDVE) and mean absolute value of relative error (MAE) were used to find the best model. According to the results, it can be concluded that neuro-fuzzy inference system (ANFIS) is a suitable choice. The results of forecasting and estimation of suspended sediments in Dez River indicate the superiority of FIS over artificial neural network (ANNs). In this regard, models based on neural network are more effective in the field of point forecasts and do not have the ability to reflect and consider the stochastic behavior of hydrological variables and cannot reflect the uncertainty of the forecast in the output.
Article number: 1
Type of Study: Research |
Subject:
Special
Received: 2025/12/20 | Accepted: 2026/01/31 | Published: 2026/01/31
Received: 2025/12/20 | Accepted: 2026/01/31 | Published: 2026/01/31
Send email to the article author
| Rights and permissions | |
 |
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. |