Clickability tracking pixel

University at Albany Uses AI to Forecast Winter Storms

A group of researchers at the University at Albany would like residents of the state of New York to start thinking of winter snow storm forecasting when they hear the term artificial intelligence.

by Rick Karlin, Times Union / September 8, 2020

(TNS) — Sometimes, when people mention artificial intelligence they may think of facial recognition used by police, or speech recognition that your Alexis or Siri device miraculously - sometimes annoyingly - uses to order a pizza or call up a song.

AI has other uses as well, including financial in which stock traders try to divine which way the market or a given stock is heading.

Now, a group of researchers at the University at Albany would like New Yorkers to think of winter snow storms when they hear the term AI.

The school’s Atmospheric Sciences Research Center just received a $2.4 million, five-year National Science Foundation grant to explore and develop ways that AI can better track and predict winter storms.

The overall $20 million project, funded by the National Science Foundation, includes a total of seven schools and is titled the Institute for Research on Trustworthy AI in Weather, Climate and Coastal Oceanography.

In short, they want to harness AI in which computer programs can “teach” themselves to make calculations and predictions based on the feedback or additional information that is constantly being fed to them.

The idea is to be able to tailor winter weather predictions to people who can make immediate use of them, particularly transportation and utility services, said Chris Thorncroft, who directs UAlbany’s atmospheric center and is co-leading the winter research.

UAlbany’s focus on winter weather makes sense given the harsh winters and large population of the state, which means snowstorms are highly impactful.

Certainly we get plenty of snow.

Rochester, for example, has more snow than any other large U.S. city with nearly 100 inches per year, according to UAlbany.

And more than 60 percent of the state’s landmass gets over 70 inches of snow annually.

The snow, along with ice storms, frequently creates havoc with traffic delays and utility outages as well as road and school closures.

With that in mind, UAlbany’s researchers will explore ways of harnessing AI to produce usable information for transportation and utility companies including the state Department of Transportation and various power providers.

“This is a huge national effort to develop and exploit AI for societal good as well as economic development,” Thorncroft said. “We are looking forward to combining resources with institutions across multiple sectors.”

Furthermore, New York has its Mesonet system which covers the entire state, taking in real-time weather observations on a granular level. That’s the kind of data that is perfect for AI since it would be overly time consuming for humans to make sense of data to the extent that a computer can.

“Individual scientists, we can’t sit down and sift through that data in an efficient way,” said Kara Sulia, a UAlbany researcher and co-investigator in the study.

One example: using the countless temperature, barometer, wind and other readings that come in from around the state Mesonet stations to build better models of how snowstorms develop and travel over time.

Nationally, the project is led by Oklahoma University.

The term “trustworthy” also speaks to another aim of this project:  letting the public know that AI shouldn’t be scary. For some, it conjures images of mass government surveillance, sophisticated computer hacking and other potential threats.

To address that, the meteorological researchers are working with social scientists to help determine how AI can be helpful when it comes to making weather-based decisions. “We are building in the social science and human dimension of this,” said Thorncroft.

By enhancing forecasts, for example, AI can help decision makers decide where to put their resources such as snowplows or utility crews during a looming storm.

“We’re not just trying to improve weather information and forecasts in its own right. We’ll be working with people that have to make decisions based on weather,” said Thorncroft. “I like to use the word ‘tailoring’ “ when it comes to what they will do with their findings.

Other universities are looking at tropical cyclones, tornadoes and coastal oceanography including a study of sea turtles.

In addition to UAlbany the other participating schools include Oklahoma, Colorado State University,  University of Washington, North Carolina State University, Texas A&M University-Corpus Christi, and Del Mar College (Corpus Christi).

Also participating is the National Center for Atmospheric Research and private industry partners including Google, IBM, NVIDIA and Disaster Tech.

©2020 the Times Union (Albany, N.Y.). Distributed by Tribune Content Agency, LLC.

Looking for the latest gov tech news as it happens? Subscribe to GT newsletters.

E.REPUBLIC Platforms & Programs