Calibration and Verification of QUAL2E Using Genetic Algorithm Optimization

In this study, the QUAL2E model is linked with a genetic algorithm (GA) in order to conduct the calibration and verification of the model. The efficacy of the optimization model was tested for different observation data quality represented by the perfect and noisy data assumptions. Four cases were studied. In the first (base) case, calibration was conducted for the reach-variable reaeration coefficient ( K2 ) and the sediment oxygen demand rate ( K4 ) . In the second case, the number of sampling points was increased. The third case investigated the impact of objective function formulation itself on the optimization model performance. Finally, a high number of parameters were calibrated using the weighted and unweighted objective functions. In general, it was not possible to reach the exact values of the parameters at all reaches. However, the resulting water quality profiles were satisfactory and close to the goal. The performance of the optimization model improved for better observation data quality and increased number of sampling points. Furthermore, it was observed that the objective function formulation impacted the performance of optimization, depending on the observation data quality. Results of the fourth case suggested that weighted objective function minimized the domination of water quality constituents with higher orders of magnitude in concentration values in search for the best solution.