The Astrophysical Journal Letters had just published opit’s article titled “The CALIFA and HIPASS velocity function for all morphological galaxy types” (paywalled, available for download at the arXiv).

The velocity function (VF) of galaxies is one of a few fundamental properties of the galaxies in our Universe: it is simply a normalised distribution of galaxies rotating at each velocity. It is important because the rotation velocity of a galaxy is the best indicator of its mass, which includes the mass of the dark matter. It lets us know more about the composition and formation of the Universe. Also, knowing the total masses and internal mass distributions of galaxies lets us create better models and theories of their evolution, among other things.

Of course, we cannot know the total number of galaxies in the Universe, nor observe them all, so we do a lot of statistics. All of it starts with the main principle claiming that the properties of the Universe are the same if viewed on a sufficiently large scale – that the same physical laws are valid everywhere, that the properties of galaxies and stars are identical, etc. All this means that if we have a sufficiently large, well-defined sample of galaxies, we can calculate the velocity function for the whole Universe – the number of galaxies rotating with a given velocity per cubic Megaparsec (the preferred volume unit in cosmology).

In the paper, opit used the star rotation velocity measurements described in her previous article and combined them with a published radio observation-based VF estimate in order to produce the first VF that is valid for both the massive ellipticals (which are missed by radio surveys) and for smaller galaxies which are not observed by the CALIFA survey. That’s how it looks (Fig. 3 from the manuscript, the thick teal line):


That’s not all, though. Together with our collaborators we compared our velocity function with simulations of the universe. We used both the dark matter-only simulations yielding dark universes (i.e. only gravity is important, no processes such as gas cooling, star formation and explosions, etc. take place) and a large realistic, full-physics simulation for a comparison. What we found is that the simulations using our current model of cosmology not only over-predict the number of dwarf galaxies (see this at the lower panel, showing the ratio of simulated and observed VFs), but underpredicts the number of galaxies with intermediate and high rotation velocities (above ~150 km/s). This is very exciting – we may infer how a forming galaxy impacts the dark matter halo around it, what are the structures of the haloes surrounding each galaxy, where simulations should be improved and where theory might fail from checks like that.