Understanding low-flow periods based on river and aquifer recessions using a sequential groundwater-surface water modelling approach

Low-flow periods are a seasonal component of a river regime characterized by a reduction in discharge. This recession phenomenon is associated with water shortages and quality problems that are detrimental to communities and ecosystems that rely on groundwater-fed rivers. This paper presents a methodology to understand the dynamics of low-flow periods by utilizing the information contained in the response of the river-aquifer system during recessions. The methodology is developed and applied to the 5000 km² Yamaska River watershed in Quebec (Canada), where critical low-flow conditions are frequently observed in winter and summer, and where the heterogeneous nature of the geology can lead to complex interactions between the rivers and the aquifer. Multiple water table and streamflow recession events recorded over a period of 20-50 years at 16 monitoring wells and 22 gauging stations were combined to obtain an averaged recession response at each location, referred to as the master recession curve (MRC). An MRC, which is minimally influenced by precipitation and evapotranspiration processes, contains important information about the flow and storage characteristics of an aquifer and its connection to rivers. Moreover, MRCs from wells and gauging stations provide complementary information. The recession-based analysis provided a tenable framework to disregard the surface modelling component at this stage since, during the depletion periods, the system is minimally influenced by atmospheric processes. A sequential modelling approach was devised to construct an integrated hydrological model using the HydroGeoSphere simulator to capture the groundwater-surface water interactions during low flows. First, the hydraulic characteristics of the subsurface were derived from the MRCs by history matching with the model in fully saturated mode. With the subsurface domain characterized, the rest of the processes were parameterized to capture the observed groundwater and surface water hydrographs using the fully integrated model. Beyond elucidating the low-flow dynamics, this methodology showcases efficiency due to its sequential strategy, alleviating the inherent computational burdens of setting up integrated models. This communication presents the outcomes from conceptual and numerical analyses, contributing to understanding hydrologic systems under low-flow conditions.