Prediction of daily streamflow using adaptive neuro-fuzzy inference systems and group method of data handling approaches: a case study of kone river, binh dinh province

Abstract

ement, and planning. This study evaluates the performance of the Group Method of Data
Handling (GMDH) and Adaptive Neuro-Fuzzy Inference System (ANFIS) models in predicting daily
streamflow. Rainfall and streamflow data, collected from rain gauges and hydrological stations in
the upstream area of the Kone River in Binh Dinh Province, were used as inputs for the models and
tested across various input scenarios. Model performance was assessed using three statistical
metrics: the coefficient of determination (R²), root mean square error (RMSE), and mean absolute
error (MAE). The results revealed that the ANFIS model consistently outperformed the GMDH
model, achieving the highest R² value of 0.94 and the lowest RMSE (64.6 m³/s) and MAE (14.2 m³/s).
Additionally, Scenario 1 demonstrated the best predictive performance across both models. This
study successfully developed reliable approaches for daily streamflow forecasting and provided
valuable insights into the influence of input variables on prediction accuracy.