Here's how AI could help astronomers unlock new secrets about meteors

• Researchers used AI to analyze over 28,000 meteor events based on 13 different properties.
• The new method led to a classification system that ranks a meteoroid's hardness.
• This "Hclass" scale reveals more about a meteoroid's composition and origin from asteroids or comets.

Artificial intelligence is now helping astronomers gain more insights into one of the most well-known sights in the night sky.

Streaking meteors burning up in the atmosphere regularly dazzle skywatchers across the world, especially during the peak activity of famous showers like the Perseids and Geminids. Now, modern data collection techniques relying on AI are revealing more information about the physical nature and origin of the phenomena when they're still meteoroids shed from asteroids and comets in space.

“Meteors have been observed for centuries, but only recently have we had datasets large and detailed enough to apply modern machine-learning methods,” Sam Hemmelgarn, a researcher at the Lowell Observatory in Arizona who led a new study, said in a statement. “This allows us to extract physical information that was previously hidden in the data.”

Here's what to know about how AI technology is powering a surge of insights into meteors.

More than 28,000 meteor events inform new classification method

To validate their new method, the researchers drew on more than 28,000 meteor events, including meteor showers, recorded in 2023 at the Lowell Observatory in Arizona.

Each meteor was classified based on 13 properties, including its speed, brightness and its height. That's well beyond traditional classification methods, which the researchers said typically only use a handful of parameters.

“Modern meteor networks capture a wealth of observational information, and we wanted a framework that could fully take advantage of that,” said Lowell Observatory astronomer Nick Moskovitz, who co-authored the study, published in May in the journal Science Direct.

AI may help astronomers classify meteors

The researchers then turned to artificial intelligence, using a combination of machine learning algorithms to identify natural groupings in the data they collected.

The analysis led the team to settle on three main factors that determine what happens to meteors when they enter Earth's atmosphere:

• how the meteoroid is moving through the atmosphere, including its speed
• how easily it begins to heat up and glow
• and how its size and shape influence how it breaks up.

Those factors further led to a new classification system that ranks a meteoroid on a hardness scale that the researchers refer to as "Hclass." That nuanced scale reveals more about a meteoroid's composition, which may range from dense, iron-rich material originating mostly from asteroids, to more fragile, porous debris that often sheds from a comet.

What are meteors? Difference from meteorites, meteoroids

Rocks in space are known as meteoroids. If those space rocks enter Earth's atmosphere, they become meteors that streak across the sky in events colloquially referred to as "shooting stars."

Meteors – or fragments of them – that survive their atmospheric trip and land on the surface without burning up become meteorites, according to NASA.

What causes meteor showers?

Named after star constellations, meteor showers occur when Earth passes through dusty debris trails left by comets and other space objects as they orbit the sun.

The debris – space rocks known as meteoroids – collides with Earth's atmosphere at high speed and disintegrates, creating fiery and colorful streaks in the sky, according to NASA. Those resulting fireballs, colloquially known as "shooting stars," are meteors no larger than the size of a pea that burn up in the atmosphere, NASA explains.

Meteor showers occur on a predictable schedule each year, with some lasting for mere days and others stretching on for weeks. But a meteor shower is at its best when the Earth passes through the densest part of the associated cosmic debris, otherwise known as the shower's peak activity.

For the recent study, the researchers applied their machine-learning method to several well-known meteor showers to confirm that its results accurately reflected what we already know about the origins of the associated asteroids and comets.

With the approach validated, the researchers were confident in saying their method can be applied widely to both individual meteors and larger events like meteor showers.

Eric Lagatta is the Space Connect reporter for the USA TODAY Network. Reach him at [email protected]