CSU hopes artificial intelligence can teach us more about the atmosphere

FORT COLLINS, Colorado – The science of weather forecasting is improving every year, but there are still so many mysteries to solve.

Elizabeth Barnes, a professor at Colorado State University (CSU), thinks some of those answers may come from artificial intelligence (AI), also known as machine learning. This is basically when a computer program makes a prediction based on patterns it finds in massive amounts of data.

“It can sort data much faster than we can, and in most cases it can do it better too,” Barnes said. “And sometimes it can even find relationships that we didn’t know existed. We can learn new science.”

Barnes said she was often impressed with the accuracy of an AI-based climate forecast, but was more interested in how the machine got that answer in the first place.

What Barnes and his collages are working on at CSU is called Explainable Artificial Intelligence (XAI). Barnes said it’s like breaking the lid on the so-called black box that seals the methods behind the machine.

“We take that forecast or that prediction, and the idea is that you feed that information back into your machine learning model,” Barnes said. “And that gives you a map of what was important to him making his decision. What ingredients were used.

RELATED: Yes, the Farmer’s Almanac says Colorado is going to have a snowy winter. No, it doesn’t mean anything.

Barnes said the information roadmap has already led to a new understanding of the impact of ocean conditions on long-term weather more than a month in advance.

“It also helps us learn more about our climate patterns,” Barnes said. “Inside the models, parts are actually replaced by machine learning algorithms to do a better job.”

Barnes said one of the beauties of machine learning is that you can keep the rules very simple and can use almost any kind of data, even maps, words and pictures instead of just numbers. and statistics.

This is a straightforward approach to data-driven prediction modeling; AI doesn’t need equations to find a solution. Unlike numerical weather prediction models which are a more physical approach. These models use things like Newtonian and thermodynamic equations to make a weather forecast.

“Machine learning tools allow us to be creative about how we do science,” Barnes said. “It allowed me to think about how I ask questions and what types of questions I ask, without barriers in the way that I think a lot of climate science had in the past.

RELATED: Controversial AI-Generated Artwork Wins First Place at Colorado State Fair

SSUGGESTED VIDEOS: Latest news from 9NEWS

Watch more 9NEWS on the free 9NEWS+ app for Year and Television fire.

9NEWS+ offers several daily live shows, including 9NEWS Mornings, Next with Kyle Clark, and 9NEWS+ Daily, an original streaming program. 9NEWS+ is where you can watch the latest live news, weather updates and press conferences. You can also replay recent news reports and find on-demand videos of our top stories, local politics, investigations, and Colorado-specific features.

To download 9NEWS+ of the year search for KUSA.

To download 9NEWS+ on Fire TV search for 9NEWS.

RELATED: Watch 9NEWS for free on ROKU, Apple TV, Fire TV