The remarkable Hubble Space Telescope, a continuous source of astronomical revelations for over three decades, has inadvertently made a new discovery: a trove of over 1,000 asteroids. This discovery is attributed to an international collaboration involving citizen scientists, astronomers from the European Space Agency (ESA), and innovative machine learning techniques, representing a groundbreaking approach to identifying objects within vast historical data archives.
Led by Pablo García-Martín, a researcher at the Autonomous University of Madrid’s Department of Theoretical Physics, the research team comprised experts from ESA, NASA’s Jet Propulsion Laboratory (JPL), and various institutions across Europe. Their findings were published in a recent paper titled “Hubble Asteroid Hunter III. Physical properties of newly found asteroids” in the esteemed journal Astronomy & Astrophysics.
A fundamental aspect of astronomy involves studying asteroids, remnants of the Solar System’s formation approximately 4.5 billion years ago. These celestial objects exhibit a wide range of sizes and shapes, from tiny rocks to massive planetoids. Observing asteroids presents challenges due to their faintness and constant motion around the Sun. However, Hubble’s unique geocentric orbit allows it to capture these wandering objects through their distinctive curved trails in its exposures.
As Hubble orbits Earth, its vantage point changes, enabling scientists to observe asteroids and map their trajectories. By analyzing the curvature and position of these trails, astronomers can determine the asteroids’ distances and infer the characteristics of their orbits. This approach has proven valuable for testing theories about the formation and evolution of the Main Asteroid Belt.
According to one prevalent model, smaller asteroids are fragments of larger ones, a result of countless collisions and grinding over billions of years. A contrasting theory proposes that these small bodies formed in their current state long ago and have remained largely unchanged. However, astronomers have yet to identify a plausible mechanism explaining why these smaller asteroids would not have accumulated circumstellar disk dust from the Sun’s surroundings, from which the planets originated.
In 2019, astronomers from the European Science and Technology Centre (ESTEC) and the European Space Astronomy Center’s Science Data Center (ESDC) joined forces with the immensely popular citizen science platform Zooniverse and Google to launch Hubble Asteroid Hunter (HAH). This ambitious project invited volunteers to scour archival Hubble data for asteroids.
The enthusiastic response resulted in 11,482 citizen scientists contributing nearly two million identifications. Their efforts generated a training set for an automated algorithm that could identify asteroids using machine learning. This yielded a total of 1,701 asteroid trails, with 1,031 corresponding to previously uncatalogued asteroids, around 400 of which measured less than 1 km in size.
The potential applications of this groundbreaking approach extend beyond Hubble data. It can be employed to analyze datasets from other asteroid-hunting observatories, such as NASA’s Spitzer Space Telescope and Stratospheric Observatory for Infrared Astronomy (SOFIA). Once the James Webb Space Telescope (JWST) accumulates a substantial dataset, the same method could be applied, unlocking further discoveries.
The HAH project is poised to embark on a new phase, meticulously examining the trails of these newfound asteroids to characterize their orbits, rotation periods, and other properties, contributing to our understanding of the Solar System’s formation and evolution.
