About
CEFNet is an experimental AI model for ensemble forecasts of coastal water levels at tide gauges across the CONUS coastline. It uses a multimodal transformer architecture to combine gridded atmospheric forcing, recent station-level water-level observations, and harmonic tide predictions. The current forecast product uses GEFS ensemble forecasts and recent gauge observations to provide 3-day water-level forecasts with raw ensemble spread. CEFNet is being developed in the Department of Civil and Environmental Engineering at Florida International University.
AI-driven Coastal Ensemble Forecast