AI Discovers ‘Potentially Hazardous’ Near Earth Asteroid Astronomers Missed

In a significant breakthrough for space exploration and safety, an asteroid discovery algorithm designed for the Vera C. Rubin Observatory's has successfully identified its first "potentially hazardous" asteroid.

The newfound asteroid, designated 2022 SF289, was spotted during a test run of the algorithm with the ATLAS survey in Hawaii, Its discovery has been officially announced in the International Astronomical Union's Minor Planet Electronic Circular MPEC 2023-O26.

Discovery images from the ATLAS survey, with 2022 SF289 visible in the red boxes. Credit: ATLAS | University of Hawaii Institute for Astronomy | NASA

Measuring approximately 600 feet in length, 2022 SF289 falls into the category of "potentially hazardous asteroids," which are celestial bodies in proximity to Earth's orbit and are closely monitored by scientists to mitigate potential collision risks. However, the newly discovered asteroid poses no immediate threat to Earth.

2022 SF289 is categorized as an Apollo-type NEO, with its closest approach bringing it within 140,000 miles of Earth's orbit, even closer than the moon. 

The algorithm, known as HelioLinc3D, offers a more efficient and accurate method of uncovering near-Earth asteroids than current techniques.

"This discovery demonstrates the real-world efficacy of the software that the Rubin Observatory will employ to identify thousands of as-yet unknown potentially hazardous asteroids, ultimately enhancing our safety," - Ari Heinze, a Rubin Scientist, Principal Developer of HelioLinc3D, Researcher at University of Washington.

The solar system comprises a multitude of rocky bodies, ranging from small asteroids to dwarf planets, remnants from the early formation of our planetary system over four billion years ago. While most of these objects exist at a distance, a subset known as near-Earth objects (NEOs) orbits relatively close to our planet. Among these NEOs, those with trajectories approaching within approximately 5 million miles of Earth's orbit warrant special attention due to their potential risk. These potentially hazardous asteroids (PHAs) are systematically searched for and monitored to prevent any potential future collisions.

Current methods of identifying PHAs involve specialized telescope systems, such as the NASA-funded ATLAS survey. This system captures images of sections of the night sky at least four times each night, with a discovery being made when a point of light is observed moving consistently in a straight line across the image series. Although around 2,350 PHAs have been discovered using this technique, scientists believe that many more await detection.

The Vera C. Rubin Observatory, perched atop the Chilean Andes, is poised to join the search for these objects starting in early 2025.

Supported primarily by the U.S. National Science Foundation and the U.S. Department of Energy, Rubin's observations will vastly accelerate the pace of PHA discovery.

 Equipped with an 8.4-meter mirror and a massive 3,200-megapixel camera, Rubin will scan the sky with unprecedented speed, visiting each region twice nightly instead of the current four times. This innovative observing strategy necessitates a novel discovery algorithm to effectively identify potential threats.

Scientists at the University of Washington's DiRAC Institute, part of Rubin's solar system software team, have been diligently developing such algorithms. Collaborating with Matthew Holman, a senior astrophysicist at the Smithsonian and a Harvard University lecturer, Heinze and Siegfried Eggl, an assistant professor at the University of Illinois at Urbana-Champaign and a former University of Washington researcher, crafted HelioLinc3D.

This new algorithm was designed to identify asteroids within Rubin's dataset and was put to the test using existing ATLAS survey data.

During this testing phase, HelioLinc3D combed through ATLAS data and successfully pinpointed 2022 SF289 on July 18, 2023. This particular asteroid had been observed by ATLAS three times on different nights, falling short of the required four observations on a single night for identification as a new NEO. HelioLinc3D excelled in combining fragments of data from multiple nights, culminating in the asteroid's discovery.

Additional observations of 2022 SF289 captured by the Zwicky Transient Facility survey  Credit: Joachim Moeyens | University of Washington | B612 Asteroid Institute

Larry Denneau, one of the lead ATLAS astronomers, emphasized that while surveys might struggle to identify objects like 2022 SF289 due to their faintness, HelioLinc3D's ability to recover such faint objects across several nights is groundbreaking. Denneau likened this to having a "bigger, better" telescope.

Although other surveys initially missed 2022 SF289 due to its trajectory passing in front of dense starfields in the Milky Way, the algorithm's success provided crucial information for follow-up observations.

Confirmatory data from the Pan-STARRS and Catalina Sky Survey, along with contributions from the B612 Asteroid Institute's ADAM platform and the NSF-supported Zwicky Transient Facility telescope, verified the discovery.

While scientists are currently aware of 2,350 PHAs, it is believed that over 3,000 more remain to be discovered.