Statistical properties of water level extremes along the St. Lawrence fluvial estuary

• DOI: 10.5194/egusphere-egu24-11761
• Publication Date: 2024-03-08T22:29:45Z
• Authors: Silvia Innocenti; Pascal Matte; Remi Gosselin; Mouna Doghri; Caroline Sevigny; Olivier Champoux; Jean Morin

The governmental Flood Hazard Identification and Mapping Program (FHIMP) seeks to update standards for flood mapping risk area definition in Canada. Within this initiative, Environment Climate Change Canada (ECCC) has been mandated provide 2D simulations of water levels the St. Lawrence fluvial estuary estimate return periods extreme under historical future conditions. Long-term fine-scale hydrodynamic are necessary reproduce accurately complex interplay hydrological, meteorological tidal processes responsible system. However, substantial computational resources time needed run numerical models constrain feasibility producing numerous long-term with a wide range potential flood-generating Consequently, study considers complementary statistical framework assess characteristics drivers from data prepare input scenarios climatic projections.  Event-based analyses level records conducted at 18 stations across system using univariate multivariate techniques characterize observed dynamics events. Specifically, frequency analysis is applied each station quantify local based on approximately 400 events Estuary between 1972 2022. Multivariate investigations non-stationary harmonic regression tool (NS Tide) then used involved major reconstruct series set meteorological, astronomical covariates. Finally, spatial performed identified NS Tide continuous reconstructions. goal high water-level (e.g., duration, seasonality, probability distribution) regional scales.