Adaptive Neuro-Fuzzy Inference System for Long-term Streamflow Forecasts Using K-fold Cross-validation: Taleghan basin, Iran

Streamflow forecasting is an important issue in water resource management. In this paper, the application of Adaptive Neuro-fuzzy Inference System (ANFIS) is investigated in modeling monthly and seasonal streamflow forecasts. Moreover, K-fold as the cross-validation method is used to evaluate test-training data in the model. Results are compared with those of the typical method (i.e., using 75% of data for training and the remaining 25% for testing the validity of the trained model). Study area is Taleghan basin located at northwestern Tehran, Iran. The data used in this research consists of 19 years of monthly streamflow, precipitation and temperature records. To apply temperature and precipitation data in the model, the whole basin was divided into sub-basins and average values of each parameter for each sub-basin were allocated as model input. Finally, results are compared with those of the ANN model. It was found that the forecasting models using K-fold are more reliable. In addition, the ANFIS model shows better performance than the ANN model in predicting peak flows and other model evaluation indices including the Nash-Sutcliffe Efficiency Index and Scatter Index

Keywords:

Streamflow forecasting , Tehran , ANFIS , K-fold , ANN