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

A team of astronomers at Lowell Observatory in Arizona has developed an AI-driven method to classify meteors, unlocking new details about their origins and composition. Using data from more than 28,000 meteor events recorded in 2023, researchers led by Sam Hemmelgarn and Nick Moskovitz analyzed 13 properties, including speed, brightness, and atmospheric height, far exceeding traditional classification techniques. The study, published in May in *Science Direct*, applied machine learning algorithms to identify patterns in the data. The findings revealed three key factors influencing a meteoroid’s behavior upon entering Earth’s atmosphere: its speed, heat response, and structural breakdown. These insights led to a new classification system called 'Hclass,' which ranks meteoroids on a hardness scale, differentiating between dense, iron-rich asteroid fragments and fragile comet debris. Traditional methods relied on limited parameters, but modern meteor networks now capture vast observational data. The AI framework allows researchers to extract hidden physical information, providing a clearer understanding of meteoroid origins. This approach could improve predictions about meteor showers, such as the Perseids and Geminids, which captivate skywatchers worldwide. The study highlights how AI is transforming astronomy by processing large datasets to reveal subtle details about celestial phenomena. By refining classification methods, researchers aim to better track meteoroids before they become meteors or meteorites, enhancing both scientific knowledge and public awareness of space events.