Exploring drivers and barriers to urban stormwater management towards enhanced water quality

Stormwater represents a significant threat to the environment and public health, as recently highlighted by the revised Directive 2024/3019. The context in which this issue develops is rapidly evolving: population growth, urbanization and technological innovation, together with the impacts of climate change, all suggest that the effects of stormwater on surface waters are likely to worsen in the coming years. The thesis investigates methodological barriers that currently limit improvements in stormwater quality and management, with a specific focus on data management and hydraulic and water quality modelling approaches. By identifying critical gaps in current practices, the research proposes strategies to improve the use of data and models on the road towards enhanced water quality. In particular, the thesis discusses the current absence of standardized formats for managing water quality measurements, which limits data reusability both within and beyond academia. A comprehensive dataset of rainfall-runoff and water quality measurements has been generated and described using metadata aligned with FAIR principles. The research also explores the concept of model structural uncertainty, and the importance of accounting for it prior to parameter calibration, in order to build reliable models with an explanatory power: a methodology for detecting structural inadequacies based on model ensemble disagreement has been developed. Furthermore, a risk-based management framework for diffuse pollution has been applied with the aim of identifying areas more likely to release pollutants during storm events, using proxies pollutant data. Finally, the thesis bridges the gap between research and practice through the development of a storyline focused on mitigating stormwater pollution via the planning and implementation of green infrastructure in urban areas. This approach illustrates how the main achievements of the thesis are intrinsically connected and contribute to building reliable and meaningful scenarios, thereby supporting practitioners in their decision-making processes.