A fast and effective parameterization of water quality models

• Published in: Environmental modelling & software : with environment data news 2022-03, Vol.149, p.105331, Article 105331
• Main Authors: Khorashadi Zadeh, Farkhondeh, Nossent, Jiri, Woldegiorgis, Befekadu Taddesse, Bauwens, Willy, Van Griensven, Ann
• Format: Article
• Language: English
• Subjects: Bayesian analysis; Calibration; DREAM(ZS); Error model; Identification methods; Mathematical models; Parameter identification; Parameterization; Screening; Sensitivity analysis; Uncertainty analysis; Water quality
• ISSN: 1364-8152; 1873-6726
• DOI: 10.1016/j.envsoft.2022.105331

Water quality (WQ) models parameterization remains a challenging task, as these models are typically characterized by a high number of parameters. The objective of this study was to present a solution to the WQ parameterization problem by the use of a fast sensitivity analysis (SA) method and a manual calibration. For this purpose, we applied the simple screening LH-OAT method to the conceptual WQ model of the River Dender, Belgium. To evaluate the effectiveness of LH-OAT to identify the influential parameters, the advanced PAWN method was applied. A manual calibration was done using the influential parameters. LH-OAT provided a parameter ranking that was very similar to the one of PAWN but in a much more efficient way. The Bayesian uncertainty assessment showed the effectiveness of the LH-OAT results. To conclude, a fast screening method is preferred over an advanced SA method to identify the influential parameters for the calibration. •Water quality models parameterization is still a challenging task.•The use of a fast sensitivity analysis (SA) method and a manual calibration is proposed as a solution to the parameterization problem.•LH-OAT provides a parameter ranking that is very similar to the one of PAWN but in a much more efficient way.•For parameter screening, a simple method is preferred over an advanced method.