Daily Rainfall – Runoff Modeling of Darreh-Rud River in Ardabil Province, Iran

• Main Authors: Mohammad Reza Nikpour, Hadi Sanikhani, Sajad Mahmodi Babelan, Saeid Nastarani Amuqin
• Format: Article
• Language: fas
• Published: Shahid Chamran University of Ahvaz 2018-12-01
• Series: علوم و مهندسی آبیاری
• Subjects: gene expression programming; estimation of streamflow; least-squared support vector machines; artificial neural networks; hybrid of wavelet-ann
• Online Access: http://jise.scu.ac.ir/article_13922_cde1fce58524dc774576b43c6db55e58.pdf
• ISSN: 2588-5952; 2588-5960

Rainfall- runoff is one of the most complicated hydrological processes that is affected by various physical and hydrological variables. Therefore, understanding and predicting runoff formation processes and their transfer to the outlet point of the watershed is one of the most important issues of hydrological sciences (Salajegheh et al., 2009). Botsis et al., (2011) simulated daily rainfall-runoff of a catchment in the California, USA. They compared the performance of SVM with three types of kernel function with ANN. Finally, the SVM had a more accurate simulation of rainfall-runoff. Nourani et al., (2009) applied the hybrid of wavelet-ANN to model rainfall-runoff of Lighvan-Chay basin in Iran. The results showed that the proposed model is capable to predict long-term and short-term rainfall events due to the use of the time series with multiple scales as input layer of ANN. Darreh-rud river as the most important branch of Aras border region in Iran, is one of the main rivers of Ardabil province and the main source of water supply in different parts of the province. On the other hand, the Emarat reservoir dam is under construction on the river. Therefore, ANN, WNN, GEP and LS-SVM models were evaluated for estimating the inflow of Emarat dam (located on Darreh-rud river, Ardabil province).