Tidal, hydrological and meteorological contributions to high-water level events in the Saint Lawrence River Estuary: Local responses to regional drivers

Extreme water levels in estuaries exhibit considerable variability on multiple spatial and temporal scales, driven by complex responses to coastal, hydrological, meteorological, geological, and anthropogenic factors. These factors complicate the prediction of high-water level events and their characteristics. Although extensive literature exists on hydroclimatic and coastal extremes, few studies conduct comprehensive statistical analyses that jointly consider multiple drivers and the spatiotemporal variability of high-water-level events in estuaries. This study addresses these gaps by developing a robust statistical framework to identify, characterize and attribute local and regional high-water-level events in the St. Lawrence River Estuary (Canada) over the period 1980–2023. Building on existing approaches, a Generalized Additive Model (GAM) formulation of non-stationary harmonic regression is used to decompose the water level variability at 11 gauging stations, with automated selection of covariates and tidal constituents. This decomposition quantifies the relative contributions of the underlying processes, enabling event classification according to their dominant coastal, tidal, or fluvial drivers. Distinct spatial patterns of tidal and non-tidal processes are observed along the fluvial estuary, and a specific transition zone is identified between dominant tidal and fluvial influences. The results indicate that non-stationary dynamics are distinct in different tidal and non-tidal frequency bands. The resulting event classification also demonstrated a different landward shift from coastal to hydrological drivers for tidal and non-tidal non-stationarity. These findings guide the statistical analysis and configuration of the hydrodynamic models used in meteorological forecasting, climate projections, and other specific applications. • Framework to identify high-water-level events in the St. Lawrence River Estuary. • GAM to attribute events to tidal and non-tidal drivers. • Results reveal distinct spatial patterns and detect a tidal-fluvial transition zone. • Non-stationary dynamics are characteristic of each tidal and non-tidal frequency band. • Focus shift from peaks to whole event dynamics and underlying processes.