Several years of observations and theoretical models have helped astrophysicists better understand how gas collapses to form new stars in galaxies. But not all galaxies actively form stars. These galaxies have an abundant population of “resting” objects, including stars, at significantly lower rates.
What prevents the formation of stars in galaxies? For two decades, scientists have been exploring the answer to this question.
A new study that sought to find out seems to find what could be responsible.
The study involves using machine learning and three state-of-the-art simulations to back up the results of a large sky survey. These three simulations were EAGLE, Illustris and IllustrisTNG. The simulations help scientists determine what we would expect to see in the real universe, as observed by the SDSS when different physical processes stopped the birth of stars in massive galaxies.
The machine learning algorithm was then applied to classify galaxies into star formation and at rest, asking which of the three parameters: the mass of supermassive black holes in the center of galaxies, the total mass of galaxy stars or the mass of dark matter halo around galaxies, better predict how galaxies turn out.
These parameters allowed scientists to determine what physical process is working and force galaxies to retreat.
According to the simulations, the supermassive mass of the black hole is the most important factor in slowing down star formation. Most importantly, the results matched the observations of the local Universe.
Joanna Piotrowska, Ph.D. Cambridge University student said: “It’s really exciting to see how the simulations predict exactly what we see in the real universe. Supermassive black holes (objects with masses equivalent to millions or even billions of suns) really have a big effect on their environment. These monster objects force their host galaxies to kind of withdraw from star formation.
The work will be presented today at the National Astronomy Meeting (NAM 2021).