Modeling Hydrodynamic Behavior of the Ottawa River: Harnessing the Power of Numerical Simulation and Machine Learning for Enhanced Predictability

The Ottawa River Watershed is a vast area that stretches across Ontario and Quebec holds great importance for Canada’s people, economy, collective history, both in the present future. river has faced numerous floods recent years due to climate change. most significant flood occurred 2019, surpassing 100-year event, serves as stark reminder of how change impacts our environment. Considering limitations machine learning (ML) models, which heavily rely on historical data used during training, they may struggle accurately predict such “non-experienced” or “unseen” were not encountered training process. To tackle this challenge, study utilized combination numerical modeling ML create an integrated methodology. Indeed, comprehensive dataset flow discharge was generated using model, encompassing wide range potential future floods. This significantly improved process generalize accuracy results. Utilizing dataset, novel model called Expanded Framework Group Method Data Handling (EFGMDH) been developed. Its purpose provide decision-makers with explicit equations estimating three crucial hydrodynamic characteristics River: floodplain width, velocity, depth. These predictions various inputs, including location desired cross-section, slope, Manning roughness coefficient at different sections (right, left, middle), discharge. establish practical models each aforementioned River, input combinations tested identify optimal ones. EFGMDH demonstrated high throughout testing stages, achieving R2 value exceeding 0.99. proposed model’s exceptional performance demonstrates its reliability applications area.