Atmospheric Sciences Research Center Partners with Tomorrow.io to Build New Forecasting Tools

Two researchers review weather data on a laptop from inside the XCITE Lab at UAlbany.

Researchers review weather data from inside the XCITE Lab at UAlbany. (Photo by Patrick Dodson)

ALBANY, N.Y. (April 29, 2025) — The University at Albany today announced a new partnership with Tomorrow.io, a leading weather technology company based in Boston, to collaborate on developing new AI-driven wind and extreme weather forecasting models.

The project, led at UAlbany by researchers at the Atmospheric Sciences Research Center (ASRC), will use a combination of artificial intelligence (AI) modeling techniques, observational weather data from the New York State Mesonet operated by UAlbany, and Tomorrow.io’s proprietary satellite data.

The NYS Mesonet is the nation’s most advanced and largest early warning weather detection network. It features 127 standard weather stations, located an average of about 19 miles apart. The Mesonet’s weather data is collected in real-time every five minutes, feeding weather prediction models and decision-support tools for users across New York.

“The future of weather forecasting is poised to be fundamentally reshaped, and our collaboration with the University at Albany marks a significant milestone in predictive intelligence,” said Cole Swain, VP of Strategy at Tomorrow.io. “By uniquely training AI models on combined real-time and historical observation datasets — including satellite-derived and Mesonet-grounded data — we are setting the stage for dramatically enhanced precision and practical utility of weather predictions, driving the development of next-generation derivative models for critical decision-making.”

“Wind and extreme weather events are notably complex elements in forecasting,” said UAlbany ASRC Director Chris Thorncroft. “By integrating the New York State Mesonet’s detailed observational data with Tomorrow.io’s satellite constellation data, we are aiming to develop new tools that could represent a significant leap forward in atmospheric prediction capabilities for industry applications.”

The new AI tools developed from this partnership will aim to provide hyperlocal, real-time weather information and predictive insights that can support a multitude of industries, such as aviation, along with other forms of transportation, renewable energy generation and emergency management.

The project will also build upon existing research at UAlbany’s Center of Excellence in Weather & Climate Analytics to develop novel methods of training AI models directly from observations.

“UAlbany’s Center of Excellence in Weather & Climate Analytics is focused on developing AI forecasting models that use weather observations to help industry decision-makers reduce the risks to their businesses and optimize efficiency. We are thrilled to start this new partnership,” said COE’s director of operations, Jan Woodcock.

“Our cutting-edge deep learning model, 'deepExWind,' will be the first of its kind to integrate both surface and upper-air observations from the New York State Mesonet and Tomorrow.io’s sensors, enabling short-term extreme wind and weather forecasting with unprecedented accuracy," said ASRC Innovation Professor Sukanta Basu, who is leading the collaboration for UAlbany with postdoctoral researcher Harish Baki.

The joint initiative aligns with Tomorrow.io’s commitment to advancing AI-driven weather prediction. The company already offers a suite of weather tools, including a free Weather API, hyper-local weather data, and a consumer app, all powered by AI and proprietary technologies.

Researchers at ASRC are also contributing to a number of projects that are focused on using artificial intelligence to develop smart weather solutions. This includes a $20 million National Science Foundation institute exploring the use of artificial intelligence and machine learning technologies to improve our understanding of weather and climate.

Through the institute, ASRC researchers are leading the development of technologies that can better monitor and predict winter weather, such as a partnership with the New York State Department of Transportation to build machine learning models that help monitor road surface conditions across the state.