Marie Larocque

Groundwater discharge to headwater streams plays a critical role in maintaining flow during dry periods and buffering stream temperatures, yet groundwater‐stream interactions remain poorly understood in low‐ to moderate‐relief headwater catchments. In this study, we used paired analysis of air, stream and groundwater temperature across 53 site‐years collected over 2 years to assess the relative importance and depth of groundwater contributions in four moderate‐relief headwater catchments in a cold, humid region (eastern Canada). Groundwater contributions were characterized using thermal sensitivity (regression slope between air and stream temperatures) as well as amplitude ratios () and phase shifts () between annual air and stream temperature signals. Thermal sensitivity was strongly correlated with amplitude ratios, which captured the relative importance of groundwater contributions, whereas phase shifts captured information on groundwater flow depth and together these metrics were used to define groundwater signatures. Across site‐years, 51% displayed an atmospheric signature, 42% a deep groundwater signature and 7% a shallow groundwater signature. The relative contribution of groundwater to streamflow declined with increasing drainage area, with no groundwater‐driven signatures observed in catchments larger than 10 km ² over the study period. Across catchments, shallow groundwater signatures occurred only in the catchment containing permeable sand sediments near its outlet, suggesting a role of surficial geology in shaping groundwater contributions. Air‐stream temperature relationships varied between the two study years, with several sites shifting from deep groundwater to atmospheric signatures, highlighting interannual variability in groundwater contributions. Overall, results emphasized spatial variability within catchments (driven by drainage area), across catchments (driven by surficial geology) as well as temporal variability (interannual changes) in groundwater‐stream interactions. These findings underscore the importance of considering both spatial and temporal variability when using paired air‐stream temperature analysis to assess groundwater contributions to moderate‐relief headwater streams.

Aims: Boreal ombrotrophic peatlands provide essential ecosystem services but are increasingly intersected by permanent roads. Roads can modify hydrology and water chemistry, but their impacts on plant communities remain poorly documented. The study compared plant communities near roads with those in undisturbed areas to estimate the extent of road impacts and identify the species, plant groups, and road‐altered environmental variables best explaining vegetation changes. Location: Fourteen ombrotrophic peatlands, crossed by roads, perpendicular to groundwater flow, in Minganie and Lower North Shore regions, Quebec, Canada. Methods: Peatlands were surveyed along transects perpendicular to the road with vegetation plots (1 m ² ) sampled at systematic distances up to 100 m on both sides of the road. Plant cover and environmental variables were measured. Redundancy analysis (RDA), indicator species analysis, and nonparametric tests were used to evaluate vegetation responses and identify disturbance‐associated species. Results: Vegetation‐composition shifted within approximately 25 m of the road, and up to 75 m for certain species. Sphagnum mosses and lichens declined sharply near roads, whereas deciduous shrubs increased. Species richness decreased within 5–10 m of the road, mainly due to the loss of ombrotrophic species. Peat water pH and electrical conductivity (EC) were elevated near roads, especially downgradient. Mineral enrichment, likely from runoff and groundwater seepage of granular construction materials and road maintenance salts, appeared to be the main driver of vegetation change. Conclusions: Permanent roads transform ombrotrophic peatland ecology by altering nutrient and mineral availability, reducing specialized taxa, and promoting shrub‐dominated vegetation. These changes may weaken their long‐term carbon‐sequestration capacity. Mitigation should focus on reducing mineral inputs and stabilizing road embankments with native vegetation.

In North America, the Great Lakes contain approximately 20% of the available surface fresh water in the world. As a result, the Great Lakes Basin (GLB) is a well-known region for its extensive agricultural and food production activities. Such agricultural activities are considered one of the most significant non-point sources of nutrient transport, particularly nitrogen and phosphorus, to surface water and groundwater. This is mainly because of the application of synthetic fertilizers and manure for enhanced crop productivity and soil fertility. Such elevated nutrient concentrations can disrupt aquatic ecosystems, degrade surface and groundwater quality, and harm both human and aquatic life. However, quantification of nutrient concentrations in agricultural watersheds is challenging because it is influenced by different process parameters including soil type, climate, and land use conditions. These parameters are highly non-linear and uncertain which hinders the applicability of typical mathematical models in nutrient transport applications in surface water and groundwater quality. Therefore, data-driven models using machine learning (ML) algorithms have been extensively applied to unravel the complexities of nutrient transport in surface water and groundwater, tackling the main challenges associated with the mathematical models. This is mainly because ML algorithms can deal with complex datasets with high uncertainty and non-linearity while considering the interdependence between the process parameters. By leveraging historical datasets, ML algorithms can model the explain the cause-result and intricate interdependencies between process parameters, making them well-suited for simulating nutrient transport processes in surface and sub-surface water applications. In the current study, different ML algorithms were adopted to predict nutrient concentrations in surface water and groundwater in a sand plain agricultural watershed within the GLB in Ontario, Canada. These ML algorithms included regression (e.g., artificial neural network) and classification (e.g., decision trees) techniques to better simulate nutrient concentrations in surface water and groundwater. The ML input variables involved meteorological (e.g., precipitation), hydrogeological (e.g., groundwater levels), and water physico-chemical (e.g., pH) conditions. The performance of these ML algorithms was evaluated using different evaluation metrics such as root-mean squared error and F1-score for regression and classification models, respectively. The optimal ML models were selected according to the outcomes of these evaluation metrics. In addition, the interdependence between the involved process parameters (e.g., land use and precipitation) and nutrient concentrations was interpreted to determine the governing parameters on the nutrient transport process in surface and sub-surface water. The main outcomes of this study can help decision-makers in assessing the most effective management efforts to protect and improve surface water and groundwater quality in agricultural watersheds. In addition, these insights enable the interpolation of nutrient concentrations from discrete sampling points, facilitating predictions at unmonitored locations across the watersheds.

Wetlands, particularly peatlands, have historically been used for agricultural production, as exemplified by the Lanoraie peatland complex in the St. Lawrence Valley (Quebec, Canada). In this region, unlined artificial ponds located at the interface between the peat and the surrounding sandy substrate are used for agricultural irrigation. However, low water levels in these ponds, as well as in neighboring rivers, have led to irrigation deficits, especially during summer low-flow periods when water demand is at its peak. This situation poses the risk of water use conflicts and draining the peatland could irreversibly harm its ecological functions. A recent project assessed the impact of agricultural ponds on the hydrology of the peatland-river-aquifer system to support sustainable water management. A comprehensive monitoring program has successfully collected essential environmental data, including information on geology, river flows, and groundwater levels. Using these data, a groundwater flow model was developed for a small area of the peatland complex. The results showed that pumping from the ponds could partially dewater the peatland, thereby endangering its ecological integrity. Building on these findings, a new project aims to evaluate the hydrological and hydrogeological dynamics of the peatland, to assess the impacts of vegetation, water use, and climate changes on its hydrology, to develop indicators to guide sustainable water allocation, and to explore potential Nature-based solutions to mitigate the effects of pumping. Methodological advancements are planned to develop a modelling framework allowing to incorporate the impact of peatland afforestation while accounting for the high sensitivity of peat deposits to groundwater level fluctuations. The knowledge generated will directly support integrated water resource management in the region.

Ephemeral ponds (EPs) are seasonally flooded isolated wetlands that provide a variety of hydroecological benefits, including the provision of breeding habitat for several amphibian and invertebrate species. However, the lack of their explicit representation in hydrological models limits a comprehensive understanding of their interaction with surrounding landscapes and their vulnerability in the context of human interventions and climate change. The purpose of this research was to improve the isolated wetland module of the Soil Water Assessment Tool (SWAT) to better represent EP hydrology. The changes include (1) representation of groundwater and hypodermic flow as the only inflows from the pond drainage surface, due to the intermittent and negligible presence of inflow from surface runoff in forested ponds, (2) revision of how evapotranspiration within EPs is represented and (3) implementation of distinct volume‐area‐depth relationships for ponds based on their geometrical shape. The accuracy of these improvements was assessed against that of a previous isolated wetland formulation in replicating water depth observations of 10 EPs of a portion of the Kenauk forest (68 km²) in the Canadian Shield of the Outaouais region (Québec, Canada). The comparison results show that the revised SWAT model presented here significantly improves the distinct filling and drying water cycle of EPs (average root mean square error of 0.1 m of the revised model vs. 0.23 m for the original model). Besides, the new module allowed to identify that hypodermic flow, evapotranspiration and seepage to the underlying soil are the main EP source and sinks. The new module also allowed to explicitly quantify the differences in filling/drying pattern of the EPs of the Kenauk forest and unlike the original model structure, the new module was able to closely replicate the interannual variation of spring and annual hydroperiod duration.

Ephemeral ponds in cold and humid forested regions are often vulnerable because of their small size and ephemeral nature. The aim of this study was to identify the geomorphological and meteorological drivers of ephemeral pond hydrology in the forests of cold and humid climates. A multi‐year study of 40 ponds was conducted in the Kenauk forest located in the Canadian Shield of the Outaouais region (Quebec, Canada). Pond bathymetry, mineral and organic sediment thickness, watershed area, slope, canopy cover and maximum depth were measured in‐situ. The spring hydroperiod (the period during which there is water in the ponds between April and October) as well as the surface and groundwater recession rates were measured throughout the study period, from 2016 to 2020. Data from this study show that ponds located at lower altitudes in the landscape are larger, receive water from a larger area, and have longer hydroperiods than ponds located at higher altitudes. The results demonstrated a connection between the ponds and the surrounding aquifer. Spring and early summer precipitation was shown to affect hydroperiods the most, while summer precipitation influenced the presence of water in the ponds from April to October. Winter precipitation appear determinant for pond recession rates. Simple multiple regression models were able to simulate hydroperiods and hydroperiod indices relatively well, but pond recession rates were not well represented by the models. This study brings original multi‐year and multi‐site data identifying parameters and variables that determine ephemeral pond hydrology in cold and humid climates. The results provide new insights into their resilience on the landscape and bring new arguments for their long‐term protection.

Large scale and long-term estimates of groundwater recharge (GWR) are strategic for assessing the relative impacts of climate change and land cover (LC) change on groundwater resources. This is especially true in cold and humid climates where global change has a high disrupting potential. Therefore, this work aims to determine the driving processes of long-term and large-scale GWR in cold and humid climates. Using a parsimonious model, GWR was simulated in the cold and humid region of southern Quebec, Canada (35 800 km2) over the past decades (1961-2017) and for potential future conditions (12 scenarios, 1951-2100). Constant and time-variant LC were used, with a monthly time step and a 500 m x 500 m spatial resolution. The datasets and model are open source. The simulated past and future results showed the importance of seasonality for GWR and the key role of annual temperature in the spatial distribution of GWR rates. They highlighted the high responsiveness of the cold and humid region hydrology to long-term interannual climate variability and the importance of simulating the snow and freezing processes when estimating GWR in these climates. In the future, warmer temperatures during the cold months (less precipitation as snow, earlier snowpack melting) led to more liquid water available when the vegetation was dormant, leading to higher GWR. Warmer temperatures during the rest of the year extended the growing period and increased plant water uptake, directly decreasing the water available for GWR. The direction of the annual changes in GWR thus depended on whether the increase during the cold months could offset the decrease during the rest of the year. The sensitivity of GWR to climate change increased with the increase in LC change intensity. The spatial distribution of LC changes was identified as another driver of GWR change as afforestation took place on agricultural areas, located on the flattest and clayey areas, thus reducing the GWR rates for forested area over time. This work called for a more systematic inclusion of LC changes in long-term groundwater resource simulation and proposes a computationally-affordable methodology to tackle it.

Reliable predictions of water systems’ response to external pressures and ongoing changes are highly important to ensure informed decision-making to support sustainable water resources management for human use and the functioning of healthy ecosystems. Recent strong development of numerical models offers a potential to understand and forecast water systems under anthropogenic and climatic influences to provide information for decision-making, process understanding of the ‘unseen’ part of the water cycle and hazard risk analysis. However, the reliability of numerical model predictions is strongly influenced by various sources of uncertainties, data qualities and assumptions, and often lacks stakeholders' point-of-view. A new, improved approach is needed and in this paper, we present six basic principles to improve the reliability and accuracy of numerical water model predictions considering explicitly stakeholders' needs and, thereby, better serving the society. Six highlighted principles are: (i) clearly defining the objectives and the purpose of the model, sustaining them during the entire modelling process; (ii) incorporating expert and local community knowledge through stakeholders' feedback; (iii) implementing a multi-model approach in which a range of conceptualizations are explored ; (iv) considering and representing the uncertainties arising from model inputs, parameters, conceptual model structure and measurement/information error; (v) translating the results to concrete and understandable strategies that policymakers can use for their informed decision-making; and (vi) long term capacity building and monitoring data collection to reduce knowledge gaps, test and improve predictions. We argue that implementing these six principles reduces uncertainties, improves the predictive capacity of the numerical water models, and ensures informed decision-making to support sustainable water resources management and thereby serve society better.