Warm Dark Matter Galaxy Formation

Abstract

Numerous hypothetical particles have been predicted which might possibly make up the dark matter content of the Universe. One class of these particle candidates includes warm dark matter (WDM) particles, which have large early-time thermal velocities that serve to erase small-scale perturbations. This creates a cutoff in the linear power spectrum - the scale of which depends on the mass of the WDM particle - and results in a suppression in the numbers of low mass halos. Since the number of satellite galaxies around Milky Way-mass host galaxies is sensitive to this cutoff, we can use the number of satellites actually observed around our own galaxy as a test of different WDM models (such as sterile neutrinos).
First, we explore the simplest case of a thermal relic WDM particle (and alter- natively a sterile neutrino produced via non-resonant oscillations). We use the galform semi-analytic model of galaxy formation to compare predicted satel- lite luminosity functions to Milky Way data and determine a lower bound on the WDM particle mass. This depends strongly on the Milky Way halo mass, and to some extent, on the baryonic physics assumed. For our fiducial model we find that for a thermal relic particle mass of 3.3 keV (the 2σ lower limit from an anal- ysis of the Lyman-α forest by Viel et al.) the Milky Way halo mass is required to be > 1.4 × 1012 M⊙. For this same fiducial model, we also find that all WDM particle masses are ruled out (at 95% confidence) if the Milky Way halo mass is smaller than 1.0 × 1012 M⊙, while if the mass of the Galactic halo is less than 1.8 ×1012 M⊙, only WDM relic particle masses larger than 2 keV are allowed.
Next, we consider models in which some of the WDM particles are resonantly produced sterile neutrinos, which behave “colder” than the non-resonantly pro- duced population also being generated. This model of sterile neutrino darkmatter is well-motivated theoretically, and is also in less conflict with current Lyman-α bounds. This scenario then becomes a two-parameter problem involving both the particle mass and the resonant fraction. We repeat the satellite abundance test applied to this new problem to rule out parts of the parameter space for different Milky Way halo masses. Focusing on a 7 keV sterile neutrino particle which may have been hinted at by recent observations, we find that if the Milky Way halo mass is 2 × 1012 M⊙ then most cases are allowed, but if the mass is 1 × 1012 M⊙ then this particle is likely ruled out.