In the clouds

WashU team uses AI, statistics to improve predictions for rainfall, extreme weather

Talia Ogliore 10.08.2026

Marine low clouds cover vast areas of the world’s oceans and act almost like a giant reflector, sending incoming sunlight back to space and helping cool the planet. But these clouds change when they are exposed to outside influences. Under a new grant from the NSF, a WashU team will use AI and statistics to study the impact of aerosols on marine low clouds and improve predictions for rainfall and extreme weather. (Image: Shutterstock)

Facebook Twitter Linkedin Email

The National Science Foundation (NSF) has awarded researchers at Washington University in St. Louis a grant worth $1 million to develop an artificial intelligence (AI)-powered approach to improve predictions of aerosol and cloud properties. The new tools that researchers will develop can help isolate and quantify the effects of aerosols on clouds that play an important role in global climate.

Marine low clouds cover vast areas of the world’s oceans and act almost like a giant reflector, sending incoming sunlight back to space and helping cool the planet. But these clouds change when they are exposed to outside influences. Scientists know that aerosols — tiny particles from sources such as wildfire smoke — can influence how bright these clouds are and how long they persist, but the strength of these effects remains poorly understood.

“Small changes in these clouds can have a surprisingly large effect on Earth’s energy balance and climate,” said Jian Wang, professor of energy, environmental & chemical engineering in the McKelvey School of Engineering and director of the Center for Aerosol Science and Engineering (CASE) at WashU. “Yet we still do not fully understand how aerosols change marine low clouds.”

“Even for people living far from the ocean, including here in the Midwest, understanding these processes matters because they affect our ability to predict future warming, precipitation and other climate changes,” said Wang, the principal investigator for the project.

Cloud science becomes clear

The new project requires an unusual combination of expertise in aerosol and cloud science, artificial intelligence and environmental statistics. Wang brings deep expertise in atmospheric aerosols, clouds and field observations, while co-principal investigators Bo Li and Maxine Yu — both in the Department of Statistics and Data Science in WashU Arts & Sciences — together contribute complementary expertise in AI, statistical learning, environmental and spatial statistics, causal inference and uncertainty quantification.

“Bringing these perspectives together within WashU allows us to develop AI methods that are not only technically advanced, but also physically meaningful and statistically rigorous,” Li said.

The researchers will train their new models on actual satellite observations and field measurements that capture episodic smoke plumes from North American wildfires. Importantly, instead of looking only at conditions at one location and time, the models will follow the history of an air mass as it travels across the atmosphere, incorporating emissions and meteorological conditions along its path.

WashU researchers will use this data to develop three physics-informed generative AI models. The first two models will predict scenarios of aerosol concentrations and cloud properties if there were no wildfire influences, while the third AI model will quantify the probability distributions of aerosol concentration and cloud properties that are influenced by wildfire smoke.

“The AI creates the counterfactual: what would the aerosols and clouds likely have looked like if the wildfire had not occurred?” Yu said. “This allows us to separate the aerosol effect from the many meteorological factors that also influence clouds.”

The scientists said that the AI and statistical framework they are developing will be adaptable to address challenges beyond wildfire impacts, and it could help researchers separate the effects of particular environmental drivers from background weather and climate variability.

Workforce implications

The project also will train students and researchers in cutting-edge AI methods, strengthening the nation’s AI-ready scientific workforce — a priority for the NSF.

“Training is an important part of the project, because the next generation of scientists increasingly needs to understand both the physical sciences and modern AI and statistics,” Li said.

The WashU team’s goal is to give students hands-on experience using advanced AI to address consequential real-world scientific problems.

“We also plan to develop a human-in-the-loop agentic AI framework that can help scientists automate parts of the process — from preparing and screening data to training models and analyzing results — while keeping scientists involved in the critical decisions,” Wang said. “Our goal is therefore not just one scientific result, but a reusable tool for tackling similar problems in earth science.”

The McKelvey School of Engineering at Washington University in St. Louis promotes independent inquiry and education with an emphasis on scientific excellence, innovation and collaboration without boundaries. McKelvey Engineering has top-ranked research and graduate programs across departments, particularly in biomedical engineering, environmental engineering and computing, and has one of the most selective undergraduate programs in the country. With 170 full-time faculty, 1,492 undergraduate students, 1,284 graduate students and 22,000 living alumni, we are working to solve some of society’s greatest challenges; to prepare students to become leaders and innovate throughout their careers; and to be a catalyst of economic development for the St. Louis region and beyond.