Applied ML Systems · Climate Resilience & Infrastructure
Urban Flood Forecasting
Conventional flood models need soil moisture readings, calibrated gauges and physics-based solvers, inputs that take hours to assemble when a storm is already arriving. This system predicts flood extent and depth across Calgary's Bow and Elbow watersheds from rainfall alone. A CNN reads spatial rainfall against topography, an LSTM models accumulation over time, and a custom synchronization layer resolves the lag between storm and inundation.
Runs as an automated TFX pipeline on GCP Vertex AI, promoting a new model only when it beats the current one.
- Nash–Sutcliffe Efficiency
- Average Depth Error
- 2100
- Climate Horizon, RCP 8.5
Built WithCNN + LSTM · TensorFlow · TFX · Vertex AI · GIS