Answers are grounded in this paper and cite their sources.
Large-scale flood models such as those developed by Verisk require regional dominant flood generation mechanisms to be realistically represented by the model. Some of the unique challenges for modelling floods in cold regions found in Canada are: (a) frozen ground runoff generation process because of below-freezing temperatures for the major part of the year, (b) the dynamics of snow accumulation and subsequent melting and (c) highly uncertain and complex processes of ice jam formation that cause significant flooding under relatively low flow conditions. To address this, we used a conceptual hydrological model with components enhanced for addressing the above challenges, coupled with a physically-based hydraulic model with a machine learning component to account for ice jam events. The model was calibrated and validated with the observed datasets from the HYDAT agency (Hydro climatological Data Retrieval Agency) and the historical Canadian River Ice Database (CRID).
No takes yet. Share an insight, caveat, or question.
Kondapalli et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: