Climate change is expected to lead to more intense and severe rainfall events in the future and thereby significantly increase the risk of urban flooding. This change, characterized by spatial and temporal changes in precipitation patterns, challenges the capacity of existing urban drainage systems, which may be exceeded by higher runoff flows than originally anticipated. Relying solely on enlarging stormwater infrastructure to address this issue is costly and may change flood risk downstream rather than effectively resolving it. Furthermore, climate change may result in prolonged dry periods, potentially causing soil compaction and reduced infiltration capacity. In cold‑climate regions, rising temperatures are projected to increase the frequency of rain‑on‑snow events and mid‑winter snowmelt, leading to more rapid runoff and elevated risks of surface flooding. These processes can also contribute to the formation of impermeable frozen soil layers, reduce infiltration, and increase the likelihood of ice‑lens formation. Given these challenges, urban drainage systems must be both adaptable and space‑efficient, capable of managing not only increased rainfall volumes and intensities but also the hydrological effects associated with snowmelt and freeze–thaw dynamics under a changing climate.

As awareness of the hydrological and environmental impacts of urbanization on catchments grows, there has been a paradigm shift towards adopting green infrastructure solutions. These approaches diverge from traditional end‑of‑pipe strategies, by emphasizing more holistic and sustainable methods. This thesis reports on a combination of modelling and field experimentation aimed at providing a better understanding of the influence of local climate (e.g., humid continental, oceanic, and subarctic climates), soil permeability (e.g., spatial and temporal variability), and initial moisture content (degree of saturation) on the water retention capacity of a grass swale and a biofilter cell. These factors were evaluated on a long‑term basis using historical meteorological time series and on an event basis, using design storms to represent a range of rainfall intensities from 1‑ to 50‑year return periods. This analysis made it possible to identify conditions that lead to the occurrence of overflow events and to test the implementation of an outflow control structure that aims to increase retention capacities of grass swales.

Results showed that robust estimates of infiltration rates and a clear understanding of the local climate and its influence on soil moisture dynamics are prerequisites for designing well-functioning infiltration-based stormwater facilities. Design strategies should include a trade-off between selecting permeable soils for locations characterized by evenly distributed rainfall and prioritizing surface storage for areas with intense, short-duration events. Results from field irrigation experiments demonstrated that retrofitting an existing grass swale with a controlled outflow control release can enhance its retention capacity and reduce peak flows in downstream urban areas. Additionally, a characterization of grass swale infiltration rates revealed long-term changes in infiltration—driven by sedimentation and vegetation density. While no strong statistical correlation was found between the depth of the unsaturated zone and swale infiltration capacity, lower saturated hydraulic conductivity values were observed in shallower water table conditions. The study underscores the trade-off between selecting or modifying permeable swale soils to reduce runoff and protect groundwater quality. Finally, results from snowmelt modeling in a location with a humid continental climate showed that overflow dynamics in a biofilter cell varied significantly by the event type. Rain-on-snow events were the most frequent triggers of overflows but produced the lowest overflow volumes per event, suggesting that the snowpack buffered and delayed runoff.