A hybrid method for forecasting river-suspended sediments in Iran

• Published in: International journal of river basin management Vol. 15; no. 4; pp. 453 - 460
• Main Authors: Tavakoli Targhi, Alireza, Abbaszadeh, Sadegh, Arabasadi, Zeinab
• Format: Journal Article
• Language: English
• Published: Abingdon Taylor & Francis 02.10.2017; Taylor & Francis Ltd
• Subjects: artificial neural network; Artificial neural networks; Correlation coefficient; Correlation coefficients; cuckoo optimization algorithm; discharge; Fluvial sediments; Forecasting; Hydrology; Learning algorithms; Mathematical models; Methods; Neural networks; Optimization; Rivers; Sediment; Sediment discharge; Sediments; Suspended sediments
• ISSN: 1571-5124; 1814-2060
• DOI: 10.1080/15715124.2017.1315815

Estimation of sediment mass carried by rivers is an important issue in Hydrological Sciences. The main purpose of this research was to find an appropriate method to compute sediment discharge. Some machine-learning approaches have been used to forecast river-suspended sediments, correctly. One of the most effective and traditional approaches for forecasting events is to use artificial neural networks (ANNs). So, we are going to improve the performance of ANNs in estimation of suspended sediments, upon a data of Baba Aman basin in Iran. We first apply a typical neural network and obtain the root-mean-square-error of and the correlation coefficient of . Then, to improve the prediction ability of ANNs, we hybridize this method with cuckoo optimization algorithm (COA). Combination of ANNs with COA causes reduction in root-mean-square-error to , increasing in correlation coefficient to and also proposing a better model.