Accurate forecasting of river flow is one of the most important factors in surface water resources management especially during flood and drought periods. Because of the importance of river flow forecasting, in this study, daily flow of Barandozchay river in two stations (Bibakran and Dizaj) for a period of 20 years using Wavelet Neural Network (WNN) which is a combination of wavelet analysis and Artificial Neural Network (ANN) has been predicted. The results of ANN model have been compared with WNN model. Data of the years 1990-2005 and 2006-2009 were used for training and verification of the networks, respectively. The performance of the two models was evaluated by statistics: r, RMSE and MAE. The results showed that the WNN model with a correlation coefficient of 0.972 and 0.976 (for stations of Dizaj and Bibkran, respectively) was able to forecast daily river flows better than the ANN model. Therefore, the results indicated that the proposed WNN method performed quite well compared to Artificial Neural Network method and could be applicable for river flow forecasting.
Marofi, S., Amirmoradi, K., & Parsafar, N. (2013). River flow prediction using Artificial Neural Network and Wavelet Neural Network models (Case study: Barandozchay River). Water and Soil Science, 23(3), 93-103.
MLA
Safar Marofi; Kimyia Amirmoradi; Nasreddin Parsafar. "River flow prediction using Artificial Neural Network and Wavelet Neural Network models (Case study: Barandozchay River)". Water and Soil Science, 23, 3, 2013, 93-103.
HARVARD
Marofi, S., Amirmoradi, K., Parsafar, N. (2013). 'River flow prediction using Artificial Neural Network and Wavelet Neural Network models (Case study: Barandozchay River)', Water and Soil Science, 23(3), pp. 93-103.
VANCOUVER
Marofi, S., Amirmoradi, K., Parsafar, N. River flow prediction using Artificial Neural Network and Wavelet Neural Network models (Case study: Barandozchay River). Water and Soil Science, 2013; 23(3): 93-103.
)