Madrid. An international team of astronomers is compiling the largest catalog of galaxy morphological classification. The material contains 27 million galaxies and combines high-quality images with artificial intelligence mechanisms.
The catalog, which was shared by researchers from the Institute of Physics of Cantabria (IFCA-CSIC-UC) and the Institute of Space Sciences (ICE-CSIC), has been published in the Journal of Monthly Notices of the Royal Astronomical Society (MNRAS).
To do so, the researchers used data from the Dark Energy Survey (DES) – which cataloged hundreds of millions of distant galaxies over a six-year period – and a machine learning algorithm with up to 97% accuracy for learning. To classify galaxies into two types of galaxies. Morphology, even the weakest and most distant galaxies.
The morphology of galaxies is closely related to the type of stars that make them up and the mechanisms of their formation.
Essentially, this catalog classifies galaxies into two types of shapes: spiral galaxies, which have a rotating disk where new stars are born, and elliptical galaxies, which are the largest massive galaxies in the universe, and are made up of old stars that perform random movements, CSIC explains it is a statement.
Although the two types of galaxies are easy to distinguish with the naked eye, researchers have found two main problems: on the one hand, the large number of galaxies to be classified, and on the other hand, the fact that galaxies are located further away. They appear weaker and smaller, so the collages were a lot of noise.
For the work, the team used convolutional neural networks, a method that uses a machine learning (deep learning) algorithm that can receive an input image and assign a label to different aspects of that image and distinguish them from one another.
This method made it possible to produce the largest morphological catalog of galaxies published to date and assign a classification to 27 million galaxies, some of them as far as 8 gigawears (Ga), that is, 8000 million years.
This catalog allows to get an approximate view of what the galaxies were like when the universe was half its current age, and to study how the galaxies changed in the last 8 Ga and note how these structural changes related to the evolutionary paths of the galaxies.