Understanding migratory waterfowl spatiotemporal distributions is important because, in addition to their economic and cultural value, wild waterfowl can be infectious reservoirs of highly pathogenic avian influenza virus (HPAIV). Waterfowl migration has been implicated in regional and intercontinental HPAIV dispersal, and predictive capabilities of where and when HPAIV may be introduced to susceptible spillover hosts would facilitate biosecurity and mitigation efforts. To develop forecasts for HPAIV dispersal, an improved understanding of how individual birds interact with their environment and move on a landscape scale is required. Using an agent-based modeling approach, we integrated individual-scale energetics, species-specific morphology and behavior, and landscape-scale weather and habitat data in a mechanistic stochastic framework to simulate Mallard (Anas platyrhynchos) and Northern Pintail (Anas acuta) annual migration across the Northern Hemisphere. Our model recreated biologically realistic migratory patterns using a first principles approach to waterfowl ecology, behavior, and physiology. Conducting a limited structural sensitivity analysis comparing reduced models to eBird Status and Trends in reference to the full model, we identified density dependence as the main factor influencing spring migration and breeding distributions, and wind as the main factor influencing fall migration and overwintering distributions. We show evidence of weather patterns in Northeast Asia causing significant intercontinental pintail migration to North America. By linking individual energetics to landscape-scale processes, we identify key drivers of waterfowl migration while developing a predictive model responsive to daily weather patterns. This model paves the way for future waterfowl migration research predicting HPAIV transmission, climate change impacts, and oil spill effects.