He largest catalog of 3D astronomical images of stars, galaxies and quasars has been created by a group of astronomers from the University of Hawaii at the Manoa Institute for Astronomy (IfA).
The team used data from the Panoramic Telescope and the UH Rapid Response System or Pan-STARRS1 (PS1) to decipher which of the 3 billion objects are stars, galaxies or quasars using new computational tools.
New computational tools
Previously, the largest map of the universe was created by Sloan Digital Sky Survey (SDSS), which covers only a third of the sky. To achieve this, astronomers took publicly available spectroscopic measurements that provide definitive object classifications and distances, and They sent them to an artificial intelligence algorithm.
This artificial intelligence or machine learning approach with a “feedback neural network” achieved an overall classification accuracy of 98.1% for galaxies, 97.8% for stars, and 96.6% for quasars. The galaxy's distance estimates are accurate to nearly 3%.
It is approximately 300 GB in size and scientific users can query the catalog via the MAST CasJobs SQL interface or download the entire collection as a readable table. According to the main author of the study, Robert Beck, former cosmology postdoctoral fellow at IfA:
Using a state-of-the-art optimization algorithm, we leverage the spectroscopic training set of nearly 4 million light sources to teach the neural network to predict source types and galaxy distances, while at the same time correcting for extinction of light by dust in the Milky Way.
–
The news
It occupies 300 GB and is a 3D galaxy catalog that covers three quarters of the sky
was originally published in
Xataka Science
by
Sergio Parra
.