A new citizen science project launched by the European Space Agency (ESA) and the Euclid Consortium, in which he collaborates Canary Islands Astrophysical Institute (IAC) and the Zooniverse platform, will allow that volunteers around the world to identify the shapes of millions of galaxies in images taken by ESA’s Euclid space telescope.
The objective of the initiative is to train deep neural networks of artificial intelligence that will allow the creation of the largest catalogue of galactic morphology to date, according to the IAC.
(See: Diamonds on Mercury? Here’s what we know.)
“In November 2023 and May 2024, the world got a taste of the quality of ESA’s Euclid mission when its first images were made public, showing a wide variety of sources, from nearby nebulae to distant galaxy clusters. And this is just the beginning, as Euclid, in its mission to map the Universe, will continue to take images of hundreds of thousands of distant galaxies.“, details the IAC.
Over the next six years the mission is expected to send about 100 GB of data per day to Earth and Euclid will publish its first data catalogues for the scientific community starting in 2025, In the meantime, any volunteer for the Galaxy Zoo project can take a look at previously unpublished images from the telescope.
(See: NASA discovers the coldest exoplanet ever seen and similar to Jupiter).
The first dataset, containing tens of thousands of galaxies selected from over 800,000 images, is now available on the platform awaiting classification.
“If you take part in the project, you could be the first to see the latest images from Euclid. And not only that, you could also be the first human to see the galaxy in the image.”, he points out Marc Huertas-Company, researcher at the Institute of Astrophysics of the Canary Islands (IAC) and responsible for the scientific exploitation of the satellite in the Euclid Consortium for the characterization of the structure of galaxies.
The image shows 1,000 galaxies belonging to the Perseus cluster and more than 100,000 more distant galaxies in the background, each containing hundreds of billions of stars.
Many of these faint galaxies have never been seen before and some are so far away that Its light has taken 10 billion years to reach Earth.
(See: New origin theory suggests the Universe would be someone’s chemistry project).
Galaxies.
The IAC says that these classifications are not only useful for their immediate scientific potential, but also as a training set for artificial intelligence (AI) algorithms.
“If humans don’t teach AI what to look for, algorithms have a hard time classifying galaxies. But together, humans and AI can accurately identify an unlimited number of galaxies.”, warns Huertas-Company.
(See: Day length may be increasing due to climate change, study says).
At Zooniverse, the team developed an AI algorithm called ZooBot, which examines Euclid images and labels the “easiest” ones, of which there are already many examples in previous galaxy surveys, while when ZooBot is not sure of the classification of a galaxy, it shows it to users to get their classifications, that help the algorithm learn more.
The IAC recalls that Euclid was launched in July 2023 and began its routine scientific observations on 14 February 2024 with the aim of revealing the hidden influence of dark matter and energy on the visible Universe.
Thus, over a period of six years, Euclid will observe the shapes, distances and motions of billions of galaxies up to 10 billion light-years away.
EFE
Add Comment