Our researchers are working to better understand and predict the formation and evolution of these storms with the goal of producing earlier and more accurate warnings.
The weather models developed at NSF NCAR with our university collaborators provide some of the best short-term, storm-scale forecast products in the world and are used by researchers, government agencies, nonprofits, and private companies across the nation.
NSF NCAR scientists are employing artificial intelligence to improve the model guidance to forecasters about the risk of tornadoes, hail, and damaging winds up to a week in advance.
An innovative model developed at NSF NCAR, the Model for Prediction Across Scales (MPAS), can simulate the severe storms that impact local communities and the global atmospheric conditions that affect those storms simultaneously. This technology is the foundation for The Weather Company’s high-resolution forecast system, the first hourly updating model that could predict an event as small as a thunderstorm virtually anywhere on the planet.
UCP’s Joint Center for Satellite Data Assimilation has produced a groundbreaking software tool that provides the highest quality analysis of the current state of the atmosphere available to kick off model simulations. The Weather Company became the first operational user of this technology in early 2025.
UCP’s NSF Unidata program provides a number of software packages that allow forecasters and other researchers to visualize weather data, shedding light on severe storms.
