Variation of the stellar initial mass function in semi-analytical models III: testing the cosmic ray regulated integrated galaxy-wide initial mass function

Abstract

In our previous work, we derive the CR-IGIMF: a new scenario for a variable stellar initial mass function (IMF), which combines numerical results on the role played by cosmic rays in setting the thermal state of star forming gas, with the analytical approach of the integrated galaxy-wide IMF. In this work, we study the implications of this scenario for the properties of local Early-Type galaxies (ETG), as inferred from dynamical, photometric and spectroscopic studies. We implement a library of CR-IGIMF shapes in the framework of the Galaxy Evolution and Assembly (GAEA) model. Our realization includes a derivation of synthetic spectral energy distribution for each model galaxy, allowing a direct derivation of the mass fraction in the mean IMF of low-mass stars (i.e. the dwarf-to-giant ratio - $\rm f_{dg}$) and a comparison with IMF sensitive spectral features. The predictions of the GAEA model implementing the CR-IGIMF confirm our previous findings: it correctly reproduces both the observed excess of z$\sim$0 dynamical mass (mass-to-light ratios) with respect to spectroscopic (photometric) estimates assuming a universal, MW-like, IMF, and the observed increase of [$\alpha$/Fe] ratios with stellar mass in spheroidal galaxies. Moreover, this realization reproduces the increasing trends of $\rm f_{dg}$, and IMF-sensitive line-strengths with velocity dispersion, although the predicted relations are significantly shallower than the observed ones. Our results show that the CR-IGIMF is a promising scenario that reproduces at the same time dynamical, photometric and spectroscopic indications of a varying IMF in local ETGs. The shallow relations found for spectral indices suggest that either a stronger variability as a function of galaxy properties or additional dependences (e.g. as a function of star forming gas metallicity) might be required to match the strength of the observed trends.