Integrating and evaluating AI forecasts for Bureau of Reclamation

Building on our ongoing work with the U.S. Bureau of Reclamation, Dr. Laura Read recently presented a technical poster at the 2025 AGU Annual Meeting detailing our collaborative efforts to implement and evaluate operational machine learning inflow forecasts across basins in the Western U.S. This partnership, which stems from Reclamation’s Facilitated Adoption Program, focuses on expanding forecasting for critical operations like flood control (FIRO operations), power generation, and environmental flow management. By integrating HydroForecast’s AI-driven model, we are providing Reclamation regions with high-resolution, reliable forecast data to inform vital water infrastructure management in the face of increasingly complex hydrologic conditions.  

New case study: Early forecasts for managing flood risks during extreme atmospheric rivers and historic flows

A standout case study from this collaboration occurred during the recent historic "weather whiplash" in the Pacific Northwest in December 2025. Following a record-breaking drought, a relentless sequence of atmospheric rivers (ARs) brought warm temperatures and massive runoff to the region.

HydroForecast identified record-breaking flows days in advance of the peak. At key Reclamation sites like the Bumping and Rimrock reservoirs, our forecasts provided critical lead time—showing inflows that would exceed storage capacity a full week before the event—allowing operators to proactively manage spillways and coordinate with downstream partners to protect public safety.  

To learn more about how our forecasts are informing water management in the West, get in touch with our team.