The new semi-analytic code GalICS 2.0 – reproducing the galaxy stellar mass function and the Tully–Fisher relation simultaneously

Abstract

GalICS 2.0 is a new semianalytic code to model the formation and evolution of galaxies in a cosmological context. N-body simulations based on a Planck cosmology are used to construct halo merger trees, track subhaloes, compute spins and measure concentrations. The accretion of gas onto galaxies and the morphological evolution of galaxies are modelled with prescriptions derived from hydrodynamic simulations. Star formation and stellar feedback are described with phenomenological models (as in other semianalytic codes). GalICS 2.0 computes rotation speeds from the gravitational potential of the dark matter, the disc and the central bulge. As the rotation speed depends not only on the virial velocity but also on the ratio of baryons to dark matter within a galaxy, our calculation predicts a different Tully-Fisher relation from models in which v rot ∝ v vir . This is why GalICS 2.0 is able to reproduce the galaxy stellar mass function and the Tully-Fisher relation simultaneously. Our results are also in agreement with halo masses from weak lensing and satellite kinematics, gas fractions, the relation between star formation rate (SFR) and stellar mass, the evolution of the cosmic SFR density, bulge-to-disc ratios, disc sizes and the Faber-Jackson relation.

Three problems about multi-scale modelling in cosmology

“…15 In the GALAXY module again assumptions enter in calculating disc radii and rotation speeds, because this SAM like others compute galaxy disc sizes by assuming that baryons and dark matter have the same angular momentum distribution and that angular momentum is also conserved, while knowing from cosmological hydrodynamical simulations that this is in fact not the case and angular momentum is not conserved. However, even if mistaken at the level of individual galaxies, the assumption is retained because of its statistical validity and agreement with observations (Cattaneo et al, 2017(Cattaneo et al, , p. 1405. 14 It is worth stressing in the words of Oman et al (2015) that "the rotation curves of many galaxies, dwarf included, are actually consistent with LCDM predictions.…”

“…On the SAMs side, Di Cintio and Lelli (2016), using semiempirical models have been able to reproduce the BTF relation but with a scatter higher than the observed one (unless parameters are tuned accordingly). But the real turning-point in this literature has been a very recent article by Cattaneo et al (2017) that has attracted significant attention in the community, with headlines along the lines of "dark matter theory triumphs" because of the unprecedented ability to retrieve both BTF relation and the Faber-Jackson relation (which is the equivalent of the BTF relation but not for spiral galaxies but for elliptical galaxies, in linking luminosity with stellar velocity dispersion). Let us take a closer look at how Cattaneo et al's new SAMdcalled GalICS 2.0dhas been able to achieve such a remarkable result.…”

“…I thank an anonymous reviewer for this journal for stressing this point. 15 Cattaneo et al (2017Cattaneo et al ( , 1402 use the following analogy to explain the independence between modules: "The relation between HALO and GALAXY can be compared to that between a mill and a baker. There is an exchange of matter both ways (inflows and outflows, flour and money) but the baker does not need to know whether the flour has been ground with a water or a wind mill.…”

“…the rate at which cold gas is removed from galaxies. Any feedback mechanism that regulates star formation without removing gas from galaxies is already incorporated phenomenologically into our star formation efficiency" (Cattaneo et al, 2017(Cattaneo et al, , 1409.…”

“…Behind this unprecedented result, there is Cattaneo's definition of the rotational velocity of galaxies (v rot ) as being the same as the circular velocity (v c ) at the optical radius r opt ¼ 3.2 r d "which contains 83% of the mass of an exponential disk" (Cattaneo et al, 2017(Cattaneo et al, , p. 1418. This is because "measuring v rot at the outer edge of the disk makes our results less sensitive to the real form of the dark matter density profile which is likely to differ at the center from the NFW model assumed in GalICS2.0" (Cattaneo et al, 2017). This assumption, however, seems to contrast with the way McGaugh (2012) data sets are gathered (the blue data point in Fig.…”