Development Churchill River Flood Forecasting System

4DM as part of consortium with KGS Group and Global Institute of Water Security was awarded a project on behalf of the Government of Newfoundland and Labrador to conduct a flood risk assessment and forecasting system. 4DM’s role in the project is to conduct land cover mapping using remote sensing and implement the flood forecasting system using the HydrologiX platform. In this project, 4DM designed and developed an automated web-based forecasting system for predicting hydrological flows, ice formation and ice breakup. Key elements of the development included the integration of hydrometeorological data feeds, tidal data, ice measurements, developing a hydrological model runner (HEC-HMS), a hydraulic ice model runner (RIVICE), as well as open water hydraulic model (HEC-RAS) to generate a daily 3-day water level forecast along the lower Churchill River.

The remote sensing work involved using Sentinel-2 data with supervised training machine learning approach to conduct the land classification and convert the data to Curve Numbers for hydrological Modeling. Data used in the project was Atmospherically corrected using Sen2Cor and mosaic into block regions for classification using Support Vector Machine (SVM).