AI Reveals Unknown Objects in Our Dust-Cloaked Galaxy (Weekend Feature)


Milky Way Galaxy


“The amount of data we are receiving from missions such as the European Space Agency GAIA telescope is staggering. We develop software to characterize the one  billion stars, but finding the unexpected becomes an impossible needle in a haystack, replied astronomer Albert Zijlstra with the of the University of Manchester and the EXPLORE project in response to an email from The Daily Galaxy asking  what new discoveries have been revealed in your current analysis of Gaia data? “Weird stars such as the recently discovered remnant of the supernova of 1181AD are hard enough to find with older data.  With the new telescopes, only machine learning will be able to find the needles that don’t look like other stars.”

An animated dive into the dusty Milky Way in the video below reveals the outlines of our galaxy taking shape as we look out further and further from Earth. Based on new data from an interactive tool that exploits data from the European Space Agency’s Gaia mission and other space science data sets, astronomers have created an animation to model dust in the Milky Way. The work was presented this week at the National Astronomy Meeting (NAM 2022) at the University of Warwick.

Dust clouds are related to the formation and death of stars

The video shows the cumulative build-up of dust looking from Earth’s local neighborhood to ~13000 lightyears towards the galactic center – around 10% of the overall distance across the Milky Way. Close by, dust swirls all around but, further out, the concentration of dust along the galactic plane becomes clear. Two ‘windows’, one above and one below the galactic plane, are revealed.

“Dust clouds are related to the formation and death of stars, so their distribution tells a story of how structures formed in the galaxy and how the galaxy evolves,” said Nick Cox, coordinator of the EXPLORE project which is developing the tools. “The maps are also important for cosmologists in revealing regions where there is no dust and we can have a clear, unobstructed view out of the Milky Way to study the Universe beyond, such as to make Deep Field observations with Hubble or the new James Webb Space Telescope.”

State-of-the-art machine learning and visual analytics have the power to greatly enhance scientific return and discovery

The tools used to create the animation combine data from the Gaia mission and the 2MASS All Sky Survey. The tools are part of a suite of applications designed to support studies of stars and galaxies, as well as lunar exploration, and have been developed through funding from the European Union’s Horizon 2020 Program.

“State-of-the-art machine learning and visual analytics have the power to greatly enhance scientific return and discovery, said Zijlstrat.  “With a constant stream of new data, such as the recent third release of Gaia data in June 2022, we have an increasing wealth of information to mine – beyond the scope of what humans could process in a lifetime. We need tools like the ones we are developing for EXPLORE to support scientific discovery, such as by helping us to characterize properties within the data, or to pick out the most interesting or unusual features and structures.”

Albert Zijlstra  and Eurekalert/RAS

Image credit: Shutterstock License

Leave a Reply

Your email address will not be published. Required fields are marked *