$424K grant to better predict weather, climate through machine learning, AI
Improved weather and climate forecasting using machine learning and artificial intelligence is the focus of a new University of Hawaiʻi at Mānoa project. Results are expected to have a major impact in Hawaiʻi and other tropical climate areas around the world.
Associate Professor Peter Sadowski from the Information and Computer Sciences Department in the College of Natural Sciences earned a five-year, $424,293 CAREER grant from the National Science Foundation (NSF). CAREER grants are designed to support early-career faculty to serve as academic role models in research and education.
“One of the risks of climate change for Hawaiʻi is extreme weather events, and current scientific models are poor at estimating these risks,” Sadowski said. “This project will provide a completely new approach modeling these risks, using the latest advancements in AI (artificial intelligence).”
Sadowski’s project will develop machine-learning methods to predict the risk of adverse weather and climate events. AI will be used to develop new data-driven computational methods for modeling risk and apply these methods to weather applications.
In particular, these models will be applied to forecasting solar irradiance and precipitation, two areas that are particularly important for tropical islands such as the Hawaiian Islands. Estimating the risk of rapid changes in solar power generation is necessary for managing energy grids that are seeing a rapid increase in variable renewable sources, and floods claim hundreds of lives and billions in property damage each year in the U.S. alone.
Artificial intelligence methods have greatly improved translating text into predictions using images and video. A key development is the ability to learn probabilistic models of images and video. The research will leverage existing data from numerical simulations of atmospheric variables, observations from satellites and ground-based weather station data from the NSF-funded CHANGE-HI project. The machine-learning methods developed by this project will complement existing physics-based weather prediction models by providing location-specific forecasts with increased speed, higher resolution and probabilistic accuracy.
Fostering the next generation
This research will be paired with an educational outreach program that includes a summer data science course for high school students and a workshop to share data science teaching materials with Hawaiʻi’s K–12 teachers.