Making waves: Knowledge and data fusion in urban water modelling.

Area of Science:

• Environmental Engineering
• Computer Science
• Systems Science

Background:

• Mechanistic models are foundational for urban water systems (UWS) but face challenges with complexity and computational demands.
• Data-driven models offer potential for UWS but are limited by interpretability and data availability.
• Current UWS modeling struggles to balance accuracy, efficiency, and interpretability.

Purpose of the Study:

• To advocate for a paradigm shift towards knowledge-informed data-driven modeling in UWS.
• To present a conceptual framework for integrating mechanistic knowledge into data-driven UWS models.
• To highlight the benefits of this integrated approach for UWS management.

Main Methods:

• Leveraging abundant sensor data with advanced data-driven techniques.
• Integrating existing mechanistic knowledge into data-driven model frameworks.
• Developing a conceptual framework for knowledge-informed modeling in UWS.

Main Results:

• Knowledge-informed models reduce data requirements and improve interpretability.
• The integrated approach offers a more accurate representation of UWS dynamics.
• This methodology enhances efficiency and addresses uncertainties in UWS modeling.

Conclusions:

• A knowledge-informed data-driven modeling approach is crucial for advancing UWS management.
• This paradigm shift enables more efficient, interpretable, and accurate UWS modeling.
• The proposed framework supports improved decision-making in the digital era of UWS operations.