Journal of Watershed Management Research

Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:

---

The most important issues of watershed management, is predicting hydrological processes. Using new models in this field can help to management and proper planning. In addition, predicting of river flow, especially in flood conditions, will allow the authorities to reduce flood damage with the preparation. One of the ways which have recently been used to predict and estimate the flow rate of rivers is neuro-fuzzy model. Neuro-fuzzy with review and determine the relationships between inputs and output, estimate the desired output deals. In this study, the three years values of the daily rainfall and discharge of different stations in Taleghan watershed were used as input to the neuro-fuzzy model and according to the statistical coefficients (RMSE, R² and E), the best structure and inputs composition of neuro-fuzzy to predict the river flow was determined. Results demonstrated that the best estimates were of the Gaussian fuzzfier. Although different input modes, provided acceptable results, Best estimates with coefficients RMSE and R² (training data 0.02 and 0.98-checking data 0.06 and 0.87), was discharge of Mehran and Joestan and previous day discharge of Garab and Dehdar. The results indicated that neuro-fuzzy can predict the daily flow with high accuracy and can be used in watershed management and flood control.
 
 

 

Keywords: Daily discharge, Neuro-fuzzy, Flood forecasting, Taleghan watershed

Full-Text [PDF 1429 kb]   (2671 Downloads)    

Add your comments about this article : Your username or Email:

---