Precipitation is a key component of the hydrological cycle, where frequency of occurrence and intensity shape environmental and societal impacts from short-range to seasonal timescales

The National Oceanic and Atmospheric Administration awarded a cooperative agreement to the University of Maryland, College Park for $300,000 on September 1, 2026, under the Cooperative Institute (Inter-Agency Funded Activities) program (CFDA 11.405).

The recipient will develop artificial intelligence and machine learning techniques leveraging low-earth orbit and geostationary orbit satellite observations to improve 6-to-12-hour quantitative precipitation forecasts. The AI/ML-based QPF products will provide National Weather Service forecasters with new information to better predict short-term flash flood events and increase public awareness of potential high-impact weather threats, addressing uncertainty in timing, location, intensity, and duration of extreme rainfall events including those associated with atmospheric rivers, tropical cyclones, and mesoscale convective systems.

Work is performed at College Park, Maryland, with a period of performance from September 1, 2026, through August 31, 2028. The Cooperative Institute program supports research and education in environmental sciences through cooperation between government, university, and nonprofit research institutions, with awards made on the basis of proposal evaluations, technical and administrative reviews, and external reviews.

The NOAA quantitative precipitation estimates (QPE) and quantitative precipitation forecasts (QPF) are critical for communicating impacts or potential impacts from extreme rainfall events to the public, emergency responders, and a variety of other stakeholders. The advancement of numerical weather prediction (NWP) models and the global observing system has improved performance and confidence in QPE/QPF products. However, even at short lead times, uncertainty remains in QPE/QPF in terms of timing, location, intensity, and duration of extreme rainfall events, including those associated with atmospheric rivers, tropical cyclones, and mesoscale convective systems. This project will leverage AI/machine learning (ML) techniques and low-earth orbit (LEO)/geostationary orbit (GEO) satellite observations to improve model 6-12-hour QPF. The AI/ML-based QPF products would provide a plethora of new information for national weather service (NWS) forecasters to better predict short-term flash flood events and increase public awareness of potential high-impact threats. Societal impact statement: the AI/ML-based QPF products would provide a plethora of new information for NWS forecasters to better predict short-term flash flood events and increase public awareness of potential high-impact threats.