Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain
the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in
Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles
and JavaScript.
Galaxies and clusters are the ensembles of stars and other astronomical objects bound together by gravitational forces. Galaxies are ordered into galaxy groups, which are in turn organised into clusters. The Earth is in a galaxy called the Milky Way, which is in the Local Group in the Virgo Supercluster.
Machine learning provides an opportunity to probe dark matter in massive galaxy clusters, more precisely and hundreds of times faster than current methods.
The spin of Sagittarius A* from Event Horizon Telescope observations is compatible with a historical merger with a 4:1 mass ratio, according to a model. This finding supports the idea that supermassive black holes grow through hierarchical mergers.
A deep map of the diffuse gas surrounding a nearby galaxy reveals changes in the physical conditions: a mark of the boundary between the edge of the galaxy and the onset of the circumgalactic medium, the dominant reservoir of normal matter.
Quiescent galaxies have similar amount of cool gas to star forming galaxies, yet why galaxies stop forming stars remains an open question. The authors investigate why passive galaxies remain quiescent using a gravitationally lensed background galaxy to probe the faint, diffuse cool gas around a massive quiescent galaxy, and use lensing configuration to constrain the total mass and geometry of this gas reservoir.
By extracting non-Gaussian cosmological information on galaxy clustering at nonlinear scales, a framework for cosmic inference (SimBIG) provides more precise constraints for testing cosmological models.
Supermassive black holes regulate the amount of atomic hydrogen in galaxies and the atomic hydrogen gas mass to stellar masses ratio is more strongly correlated with black hole masses.
A probabilistic machine learning method trained on cosmological simulations is used to determine whether stars in 10,000 nearby galaxies formed internally or were accreted from other galaxies during merging events. The model predicts that only 20% of the stellar mass in present day galaxies is the result of past mergers.
Giant shock waves at the physical boundaries of the most massive structures in the Universe could be used as an accurate tool to measure the total mass of clusters of galaxies.
Computer simulations based on the prevailing cosmological model, ΛCDM, reproduce many observed properties of our Universe. But a study of coherent satellite motions in galaxy clusters yields discrepancies that challenge the definition of ‘today’.
