Asteroid hunt by Hubble telescope using AI
- Over 1,000 new small asteroids have been found in Hubble Space Telescope images. They lie in the main asteroid belt between Mars and Jupiter.
- Researchers, along with over 11,000 volunteer citizen scientists, found the asteroids in older images from Hubble. The asteroid hunt relied on machine learning, a type of artificial intelligence (AI), to find the faint trails, or streaks, left by asteroids in the images.
- The results provide more clues about how the asteroid belt originated and evolved.
Astronomers have found over a million asteroids, mostly in the main asteroid belt between Mars and Jupiter. But many smaller asteroids, from the size of boulders to pebbles, are still waiting to be discovered. Recently, astronomers and citizen scientists have used archived images from the Hubble Space Telescope to search for small asteroids. They used machine learning and AI in their quest to identify the asteroids. And this asteroid hunt was a big success! Researchers from the U.S. and Europe said on April 18, 2024, that they discovered over 1,000 previously uncatalogued asteroids. About 400 of the objects were less than 1/2 mile (one km) in size.
The research team first published their peer-reviewed results in the journal Astronomy & Astrophysics on March 15, 2024.
Hubble’s asteroid hunt
Hubble has provided incredible views of distant space objects, from planets to galaxies. But it can also see relatively nearby small objects, such as small asteroids in our own solar system.
The researchers, along with volunteer citizen scientists, searched through some 37,000 images, which spanned 19 years of Hubble observations. And the asteroid hunt paid off. They found 1,701 asteroid trails in the images. Asteroid trails appear as curved streaks in some Hubble images. They appear because the moving asteroids are closer to us than many of the objects being studied by Hubble.
Of the 1,701 asteroids detected, 1,031 were from small asteroids that were previously unknown and uncatalogued. Lead author Pablo García Martín at the Autonomous University of Madrid, Spain, said:
We are getting deeper into seeing the smaller population of main belt asteroids. We were surprised with seeing such a large number of candidate objects. There was some hint of this population existing, but now we are confirming it with a random asteroid population sample obtained using the whole Hubble archive.
This is important for providing insights into the evolutionary models of our solar system.
Citizen scientists
Hubble looked for smaller asteroids in the main asteroid belt, between Mars and Jupiter. These are ones that astronomers hadn’t yet detected. Most of the ones Hubble found are indeed in the asteroid belt. And they are extremely faint, about one forty-millionth the brightness of the faintest star that the human eye can see.
But this was a huge task, and the researchers needed help. So over 11,000 volunteer citizen scientists – 11,482 to be exact – also took part in the endeavor. As Martín noted:
Asteroid positions change with time, and therefore you cannot find them just by entering coordinates, because at different times, they might not be there. As astronomers we don’t have time to go looking through all the asteroid images. So we got the idea to collaborate with over 10,000 citizen-science volunteers to peruse the huge Hubble archives.
Using AI in the asteroid hunt
The effort also used a form of artificial intelligence (AI) called machine learning to more effectively go through the thousands of archived Hubble images. The citizen scientists used an automated algorithm to identify the very faint asteroid trails in the images. AI has distinct advantages for this kind of work, as the paper explained:
One of the advantages of applying machine learning to find solar system objects in complete astronomical archives is the large number of potential results obtained. This allows us to apply purposely strict filtering conditions to improve accuracy and still keep a large enough sample to obtain statistically meaningful results.
This work began in 2019, when an international group of astronomers launched the Hubble Asteroid Hunter project.
Researchers and engineers at the European Space Research and Technology Centre (ESTEC) and the ESAC Science Data Centre developed the initiative. They did so in collaboration with Zooniverse, a popular citizen-science platform, and Google.
Next, the researchers will study the asteroids to characterize their orbits and physical properties. However, most of them are in image taken years ago, so it is not possible to track them now in real time.
History of the asteroid belt
So why are these discoveries important? They provide valuable clues about the origin and evolution of the asteroid belt. One possibility that astronomers have favored is that the asteroid belt is like smashed pottery. The smaller asteroids are the remaining pieces of larger ones that have collided and were destroyed over billions of years. The asteroids we have now are the leftover debris.
The other primary theory is that the smaller asteroids formed the way they are. But scientists say if that is so, then they should have conglomerated together as they accumulated dust from the planet-forming disk of the early solar system, forming larger bodies. However, astronomers don’t know of any mechanism that would prevent them from doing so. Co-author Bruno Merín at the European Space Astronomy Centre (ESAC), in Madrid, Spain, said:
Collisions would have a certain signature that we can use to test the current main belt population.
Asteroids have also previously photobombed images of galaxy clusters, as EarthSky reported back in 2019. You can clearly see the tell-tale curved streaks in those images, too.
Bottom line: Astronomers used the Hubble Space Telescope to go on an asteroid hunt. They found over 1,000 previously unknown small asteroids in the asteroid belt.
Source: Hubble Asteroid Hunter: III. Physical properties of newly found asteroids
Read more: How big are asteroids? Compare sizes in this video
Read more: Asteroids photobombed deep Hubble images
The post Asteroid hunt by Hubble telescope using AI first appeared on EarthSky.