Restoring catchment low flow hydrology by infiltration-based stormwater source-control systems

Urbanisation has long been recognised as an important stressor to stream ecosystems, with the alterations of flow regime and water quality affecting the health of many urban streams worldwide. Recognition of these effects has recently led to new practices in stormwater management. In line with the “natural flow paradigm”, they aim to protect aquatic ecosystems by maintaining or restoring flow regimes close to their pre-development state. Source-control technologies, featuring various types of stormwater infiltration or harvesting systems, have been increasingly used to help restore a near-natural hydrology. They are thought to mitigate the impacts of urbanisation by: i) decreasing surface runoff, which is increased by the sealing of surfaces in urban areas; ii) increasing subsurface flows; and iii) reducing total runoff through evapotranspiration from vegetated systems or reuse of harvested water. However, the extent of these alterations and their effects on the hydrograph is still uncertain, particularly at low flow. There is no systematic approach yet developed to assess urbanisation impacts on subsurface flow processes, meaning that the strategies to adopt for stormwater management are still unclear. Given the importance of low flows in urban water management policy, in a context of water scarcity and ecological degradation, integration of low flow objectives in management strategies appears as an area of critical interest. To advance knowledge in the field, this research aimed to better characterise the effects of infiltration-based source-control techniques on low flows. It comprised two areas of research, which can be characterised by their scale, namely the site and catchment scale. At the site scale, the hydrological behaviour of infiltration bioretention systems was characterised through experimental and modelling work. Bioretention systems are popular source-control infiltration systems, particularly in new urban developments. Experimental work involved long-term monitoring of two bioretention systems, and the development of an automated system to measure evapotranspiration in the field. The data analyses determined the water balance of the bioretention systems, partitioning between seepage and evapotranspiration from the systems and their nearby environment. They showed that the water balance was largely dominated by seepage, thus informing the integration of such systems in catchment-scale models. Numerical modelling suggested that the experimental results may be extrapolated to various soil and climatic conditions. The second part of the research aimed to evaluate the effects on the baseflow regime of a catchment-scale implementation of source-control techniques. The selection of metrics to quantify baseflow alterations by urbanisation, and subsequent effect of stormwater management strategies, is poorly defined in the literature. A regional study relying on principal component analysis was thus conducted to identify the most relevant metrics to use in the assessment of baseflow alterations. Based on these metrics, scenario studies were conducted to evaluate the role of source-control techniques, namely bioretention systems and harvesting tanks, on the flow regime. The results suggested that infiltration systems designed with a focus on emulating the natural flow regime helped mitigate the effects of urbanisation; however, in the study catchment, they failed to completely restore the baseflow regime due to limitation of their storage capacity. The results also illustrate the variability in metrics response to stormwater management strategies, confirming the importance of metrics selection in the assessment of stormwater strategies. The scenario studies were conducted with a model built using the software MUSIC, where subsurface flow processes were represented through a standard linear reservoir structure. While common in the literature, this modelling approach is limited by the lack of validation of the physical processes represented by the model. The final part of the thesis thus explores the question of the suitability of conceptual models for stormwater management. The analyses demonstrate the sensibility of the predictions of baseflow restoration by source-control techniques to the model structure, thus suggesting work towards more physical approach to urban catchment modelling.