Application of Bayesian Networks and Genetic Programming for Predicting Daily River Flow (Case Study: Barandoozchay River)

Document Type : Research Paper



     Accurate estimation of river discharge is an important issue in forecasting of drought and floods, designing of water structures, dam reservoir operation and sediment control. So far, several methods such as time series models, Artificial Neural Networks, Fuzzy models and Genetic programming have been used for accurate predicting of river flow. In this study, Genetic Programming and Bayesian Networks methods were used to forecast daily discharge of Barandoozchay River. The daily discharge data of Barandoozchay River measured at the Dizaj hydrometric station during 2007 to 2011 was used for modeling, which 80% of the data used for training and remaining 20% used for testing of models. For assessing the role of memory in increasing or reducing of model accuracy, we tested different combinations of input variables. The results showed that at first, the accuracy of models increase with increasing of memory, as the most accuracy obtained in third combination of input variables in both of methods. After that with increasing of memory the accuracy of models decreased. Comparing the performance of GP and BNs models indicated that the accuracy of the GP method with the R=0.978 and RMSE=1.66 (m3/s) was slightly more than BNs method with R=0.964 and RMSE=1.96 (m3/s). In addition, the performance of GP method was better than BNs method in predicting minimum and average discharges.


Volume 39, Issue 4
December 2017
Pages 213-223
  • Receive Date: 07 June 2014
  • Revise Date: 27 December 2016
  • Accept Date: 19 January 2016
  • First Publish Date: 21 December 2016