FEATURES
CITIZEN SCIENCE WITH ESA SCIENCE DATA
THE HUBBLE ASTEROID HUNTER PROJECT
l Sandor Kruk and Bruno Merín – DOI :
https :// doi . org / 10.1051 / epn / 2023206 l European Space Agency ( ESA ), European Space Astronomy Centre ( ESAC ), Camino Bajo del Castillo s / n 28692Villanueva de la Cañada , Madrid , Spain
The vast amount of data in astronomy archives presents an opportunity for new discoveries . Deep learning combined with crowdsourcing provides an efficient way to explore this data using the intuition of the human brain and the processing power of machines .
▲ Faint trail of main-belt asteroid
2002 SE101 , discovered by the ground-based LINEAR survey in 2001 , crossing the famous Crab Nebula , as imaged by the Hubble
Space Telescope in 2005 . Discovery and colour composition by citizen scientists
Melina Thévenot . Credit : ESA / Hubble
& NASA , M . Thévenot
In the Hubble Asteroid Hunter project , we used citizen science on the Zooniverse platform and a deep learning algorithm on Google Cloud , to explore two decades of Hubble Space Telescope observations from the ESA Hubble archives for objects not targeted by the Hubble observations . The project , which was set up as a collaboration between Zooniverse , ESAC Science Data Center and Google , led to the detection of 1701 asteroids , including 1031 previously unknown ones , 198 new strong gravitational lenses and to quantifying the impact artificial satellites have on Hubble Space Telescope observations . This study is a proof of concept and shows what can be achieved by using new tools to explore the extensive astronomical archives . The archives of data held by the ESA Science Data Center are a treasure trove of information about our Universe , containing over 800 terabytes of data . But as more missions like Gaia , the Hubble Space Telescope and the James Webb Space Telescope continue to add new data every day , we need new tools and techniques to process and analyse this vast amount of information . And future missions like Euclid or Roman , will deliver over 30 petabytes over the lifetimes of the missions . Fortunately , the emergence of machine learning has provided us with the ability to process large amounts of data quickly .
28 EPN 54 / 2