Improving weather and climate prediction using machine learning and artificial intelligence is the focus of the new university Hawaii at the Manoa Project.The results are expected to have a major impact on the following areas: Hawaii and other tropical climate regions around the world.
Peter Sadowski, associate professor of information and computer science at the University of Natural Sciences, won $424,293 in prize money over five years. career Grant from the National Science Foundation (NSFMore). career The grant is intended to support junior faculty members who serve as academic role models in research and teaching.
“One of the risks of climate change is Hawaii “This is an extreme weather event, and current scientific models are poor at estimating these risks,” Sadowski said. “This project leverages the latest advances to provide a completely new approach to modeling these risks. AI (artificial intelligence). “
Sadowsky’s project will develop machine-learning methods to predict the risks of severe weather and climate change. AI It is used to develop new data-driven computational techniques for modeling risk and to apply these techniques to weather applications.
In particular, these models are applied to forecasting solar radiation and precipitation, two areas of particular importance for tropical islands such as the Hawaiian Islands. Estimating the risk of rapid changes in solar power is necessary to manage a rapidly growing energy grid with fluctuating renewable sources, and in the United States alone, flooding kills hundreds of people and causes billions of property damage each year.
Artificial intelligence techniques have greatly improved the conversion of text into predictions using images and videos. A key advancement is the ability to learn probabilistic models for images and videos. This work makes use of existing data obtained from numerical simulations of atmospheric variables, observations from satellites, and data from ground-based weather stations. NSFMore-funded change high plan. The machine learning method developed by this project complements existing physics-based weather forecasting models by providing location-specific forecasts with high speed, high resolution, and probabilistic accuracy.
Nurturing the next generation
This research will be conducted in conjunction with educational outreach programs such as a data science summer course for high school students and workshops to share data science teaching materials. Hawaiikindergarten through high school teachers.
