Quality simulation of dam reservoir using GP model (case study: ZayandehRoud dam reservoir)

Abstract

In this research, temperature and nitrate values at the outlet of the ZayandehRoud dam reservoir are simulated and predicted using a genetic programming (GP) model. Initially, the CE-QUAL-W2 model is applied to simulate the water quality condition of the dam reservoir from autumn 2016 to the end of summer 2021 for calibration and from autumn 2021 to the end of summer 2022 for the validation process. In addition, the genetic programming (GP) model is also used to reduce the simulation computation costs, and the results are presented and compared with the artificial neural network (ANN) model. For temperature simulation using GP, the best RMSE values of the training and validation process are 1.800 and 1.925 °C, respectively, compared to related values of 1.405 and 1.932 °C using ANN. Furthermore, for nitrate simulation of GP, the best RMSE values of the training and validation process are 0.383 and 0.536 mgL−1, respectively, compared to related values of 0.362 and 0.711 mgL−1, respectively, using the ANN model. The results demonstrate the GP model’s good performance in simulating the dam reservoir’s quality conditions. © The Author(s) under exclusive licence to Iranian Society of Environmentalists (IRSEN) and Science and Research Branch, Islamic Azad University 2025.