Evaluation of the explainability and reliability of two surrogate models for predicting outflow of combined sewer systems
Water Science & Technology · 01.07.2026 · 94 (1) · 1-13
DOI: 10.2166/wst.2026.296 ↗Kurz gefasst
Fachartikel in der Zeitschrift Water Science & Technology, online erschienen am 18.06.2026. Verfasst von Martin Oberascher und 8 weiteren Autor:innen. Beteiligte Einrichtung: Unit of Environmental Engineering, Department of Infrastructure Engineering, University of Innsbruck, Technikerstrasse 13, Innsbruck 6020, Austria. Hauptthemen laut MEDLINE: Models, Theoretical, Sewage, Drainage, Sanitary. Die Arbeit nennt 39 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Austrian Science Fund.
Abstract (Originalsprache)
Urban drainage networks (UDNs) are designed to reliably discharge wastewater and stormwater, with machine learning (ML) increasingly leveraged for a wide range of applications. A legal assessment of the European ‘Artificial Intelligence Act’ (AI Act) indicates that, while UDNs are considered part of critical infrastructure, ML applications in this field are generally not classified as high-risk AI systems, being subject to specific implementation requirements. Nevertheless, aspects such as model explainability are important to align with best practices in ML model development. In this context, two ML models (i.e., purely data-driven and physics-leveraged) used herein as surrogates for hydrodynamic modelling to predict combined sewer flow, were adapted from the authors' previous work and enhanced with various explainable AI techniques such as SHapley Additive exPlanations and robustness scores for a comprehensive understanding of the models' decision-making processes. The analysis revealed that the engineered classification into dry or wet flow is the most important feature for both ML models, followed by historical rainfall features for the purely data-driven ML model and hydrodynamic model outputs for the physics-leveraged ML model. Additionally, the achievable accuracy of both utilised ML models was highly sensitive to deviations in rainfall data, further limiting their reliability and trustworthiness.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Substanzen
Autor:innen und Institutionen
- Unit of Environmental Engineering, Department of Infrastructure Engineering, University of Innsbruck, Technikerstrasse 13, Innsbruck 6020, Austria.
- Department of Urban Water Management, RPTU University of Kaiserslautern-Landau, Paul-Ehrlich-Straße 14, Kaiserslautern, Germany.
- Department for Public Law, Constitutional and Administrative Theory, University of Innsbruck, Innrain 52d, 6020 Innsbruck, Austria.
- Independent Researcher, Reading, RG1 1ED, UK.
- Unit of Environmental Engineering, Department of Infrastructure Engineering, University of Innsbruck, Technikerstrasse 13, Innsbruck 6020, Austria E-mail: robert.sitzenfrei@uibk.ac.at.
ORCID-Profile: Martin Oberascher, Amin E Bakhshipour, Karim Sedki, Arnold Autengruber, Arno Kahl, Sina Hesarkazzazi, Ali Haghighi, Ulrich Dittmer, Robert Sitzenfrei
Förderung
- Austrian Science Fund
Zitieren
APA
Oberascher, M., Bakhshipour, A. E., Sedki, K., Autengruber, A., Kahl, A., Hesarkazzazi, S., Haghighi, A., Dittmer, U., & Sitzenfrei, R. (2026). Evaluation of the explainability and reliability of two surrogate models for predicting outflow of combined sewer systems. Water Science & Technology, 94(1), 1-13. https://doi.org/10.2166/wst.2026.296
Vancouver
Oberascher M, Bakhshipour AE, Sedki K, Autengruber A, Kahl A, Hesarkazzazi S, et al. Evaluation of the explainability and reliability of two surrogate models for predicting outflow of combined sewer systems. Water Science & Technology. 2026;94(1):1-13. doi:10.2166/wst.2026.296
BibTeX
@article{Oberascher2026,
author = {Oberascher, Martin and Bakhshipour, Amin E and Sedki, Karim and Autengruber, Arnold and Kahl, Arno and Hesarkazzazi, Sina and Haghighi, Ali and Dittmer, Ulrich and Sitzenfrei, Robert},
title = {Evaluation of the explainability and reliability of two surrogate models for predicting outflow of combined sewer systems},
journal = {Water Science & Technology},
year = {2026},
volume = {94},
number = {1},
pages = {1--13},
doi = {10.2166/wst.2026.296}
}
EUVIMED