A novel asteroid discovery algorithm, HelioLinc3D, spotted its first potentially harmful asteroid (PHA), a 600-foot (183 meters)-long asteroid that has been named 2022 SF289 and that had been previously overlooked. The algorithm was developed by a team from the University of Washington’s DiRAC Institute and will be used with data from the Vera C. Rubin Observatory in the Chilean Andes, during its decade-long survey of the night sky. This new asteroid was observed during testing during the ATLAS survey in Hawaii, and hints at the wealth of data that the research team expects to collect over the next decade. 

On Potentially Harmful Asteroids

The majority of asteroids are far from the Earth’s orbit, but there are others, known as “near-Earth objects” (NEOs), that orbit within 5 million miles (8 million kilometers) of Earth’s orbit, which is around 20 times the distance to the Moon. These asteroids have the designation of PHA, which refers to how close to the Earth an asteroid gets. Given the risks posed by these asteroids, they are systematically studied, and this is often done using specialized telescope systems such as the ATLAS survey at the University of Hawaii’s Institute for Astronomy. These systems take multiple images of segments of the sky, four or more times during the night. When light is observed moving in a straight line across a series of images, that indicates that a new asteroid has been found. According to the University of Washington there are 2,350 that have been discovered in this way. 

According to NASA, PHAs have an average width of 140 meters and orbit within 7.5 million kilometers of Earth. It is highly improbable that any of these PHAs will strike our planet within the next century, although, as HelioLinc3D’s discovery highlights, there are still many unknown PHAs, and, furthermore, the orbits of many PHAs have become more unpredictable. With this discovery, it is clear that HelioLinc3D will be able to fulfill its mission of identifying near-Earth asteroids using fewer and more sporadic observations than today’s methods allow. 

The Algorithm

While it is true that the algorithm’s mission in using the generative AI in software testing is to discover near-Earth asteroids, with the Rubin Observatory predicting that the algorithm will help them discover “millions of new asteroids”, it is also expected to advance astronomy as a whole. In the University of Washington press release, the Rubin scientist Ari Heinze said, “By demonstrating the real-world effectiveness of the software that Rubin will use to look for thousands of yet-unknown potentially hazardous asteroids, the discovery of 2022 SF289 makes us all safer”. 

The Rubin Observatory will launch its survey in 2025. The survey, funded by the U.S. National Science Foundation and the U.S. Department of Energy, will exponentially improve the speed of asteroid discovery. Whereas traditional methods require at least four image-taking segments during the night, the Rubin Observatory will require just two times to take images with its 8.4-meter mirror and 3,200-megapixel camera. This cadence required a new algorithm so that there would be no blind spots during observations. 

Heinze, the principal developer of the algorithm, worked with Smithsonian senior astrophysicist and Harvard University lecturer, Matthew Holman, and University of Illinois at Urbana-Champaign assistant professor, Siegfried Eggl, to develop HelioLinc3D. The critical question was whether the algorithm could discover new asteroids with existing data, given that, using traditional methods, Rubin’s cadence provides too few observations for new discoveries to be made with conventional algorithms. Thanks to John Tonry and Larry Dennueau, the lead astronomers at ATLAS, the group was given access to ATLAS’ data for testing. It was during testing that the group discovered 2022 SF289 on September 19, 2022, which was 13 million miles (20.0 kilometers) from Earth.

Source: University of Washington

Although ATLAS observed the asteroid on three occasions across four different nights, this fell below the threshold of four observations in one night for an asteroid to be classified as an NEO. However, as Denneau pointed out, 

“Any survey will have difficulty discovering objects like 2022 SF289 that are near its sensitivity limit, but HelioLinc3D shows that it is possible to recover these faint objects as long as they are visible over several nights. This in effect gives us a ‘bigger, better’ telescope.”

This is borne out by the number of surveys that had missed the asteroid. Since ATLAS discovery, the Pan-STARRS and Catalina Sky Survey have also observed the asteroid. This is very encouraging for the Rubin Observatory. When its survey comes online in 2025, HelioLinc3D will be up every night, catching fragments of data and using it to discover asteroids that would otherwise be left undiscovered. This will not only make our planet safer, it will advance our understanding of astronomical science. In a year that has been defined by artificial intelligence, it is apt that this algorithm is introducing a new era in the search for PHAs.