The challenges of hydrograph separation and catchment mean transit time estimation in a mesoscale, urbanised, snowmelt-influenced catchment

• Hydrograph separation is done using major ion measurements over a ten-year period. • Calibration of a mathematical baseflow filter using tracers. • Groundwater controls both availability and quality of surface water. • Dissolved silica contents are indicators of water transit time in the watershed. • Water transit times are maximum during winter and minimum after the snowmelt period. Surface water extracted from rivers is the main source of drinking water of many major cities worldwide. In wet, temperate climates, groundwater often contributes substantially to river discharge, yet quantifying these inputs remains challenging – particularly in heterogeneous, urbanised, mesoscale catchments. This study focuses on the Saint-Charles River catchment (344 km 2 ), Quebec City, Canada, a partially urbanised catchment which supplies drinking water to over 300 000 people from surface water. Long-term stream monitoring of major ions and trace elements (2003 – 2023) was used to identify water quality end-members and their temporal variability. Hydrograph separation based on major ions concentrations, enabled the estimation of monthly groundwater inflows to the river between 2013 and 2023. A mathematical filter was calibrated against the tracer-based approach results to determine the daily groundwater contribution between 2013 and 2023. Results show that groundwater is the major contributor to the river annually with a BFI between 0.59 and 0.63. In addition, up to 87 % of the year is classified as baseflow-dominated. The highest monthly average contribution are found in summer (>80 % of total river flow from July to September), while the lowest are found in spring (35–45 % of total river flow in April). Groundwater also controls the river total mineralization. While electrical conductivity proved unreliable due to anthropogenic inputs, alkalinity emerged as a robust tracer of groundwater contribution. River silica dynamics were modelled using a lumped parameter approach, yielding estimates of mean transit times ranging from a few days during high flows to several months or years during low flows. These findings improve the understanding of river–aquifer interactions and have significant implications for the vulnerability assessment of surface water intakes under climate change and urbanization in similar hydrological settings.