SatGen: a semi-analytic satellite galaxy generator

Halo merger trees

  • We compare different algorithms (based on the extended-Press-Schechter formalism) for generating halo merger trees (Jiang & van den Bosch 2014) and show that the algorithm by Parkinson+ (2008) yields merger statistics in best agreement with those from cosmological simulations.

  • We characterize halo ``progenitor-mass functions" for halos in different environments (i.e., different large-scale density), and re-calibrate the EPS merger tree algorithm for different environments (Jiang+ in prep).

Statistics of dark matter subhalos

  • We characterize the statistics of dark matter halo substructure using the SatGen model and N-body cosmological simulations (Jiang & van den Bosch 2015;  16;  17; van den Bosch & Jiang 2016). We show that ~1% of Milky-Way sized halos are free from the too-big-to-fail problem due to the dramatic halo-to-halo variance of the formation histories. We also show that the halo-to-halo variance of substructure abundance is important in multiple contexts, including subhalo detection by gravitational lensing, and halo models. The model is now being upgraded to model satellite galaxies, allowing for different halo profiles, subhalo response to dynamical and baryonic processes etc (Jiang, van den Bosch, Greene, in prep.). 


Galaxy-halo connection

Is dark matter halo spin a predictor of galaxy spin and size?

  • We find with different hydro-cosmological simulations that galaxy spin and size are barely correlated with the spin of host dark matter halos; instead, galaxy size correlates with the halo concentration parameter, : R_eff ~ c^-0.7 R_vir (Jiang, Dekel+2018)

Dwarf galaxy structure

  • We characterize the formation of ultra-diffuse galaxies (UDGs) in cosmological simulations. We show that field-UDGs dwell in cored dark matter halos with normal spin, and that normal dwarf satellites can be transformed into red UDGs at orbital pericenters due to tidal heating and ram pressure stripping (Jiang, Dekel, Freundlich+2019).

  • Current cosmological simulations produce a plethora of low-surface brightness dwarf galaxies but almost all fail to produce compact dwarf galaxies, in the mass regime of M_star≈10^8 M_sun (or M_halo ≈ 10^11 M_sun). The gas density threshold for star formation (n_th), which is largely arbitrary in simulations, has an impact on the morphology of the galaxies and the structure of their host halos. Notably, high-n_th results in bursty star formation histories and thus repeated impulsive supernovae outflows, which make galaxies puffy and host halos having flat cores. In order to reproduce observations, which reveal both dark-matter cusps and cores (and both compact and diffuse galaxies) in the aforementioned mass regime, we resort to a virial parameter-based star formation threshold, which gives rise to a range of n_th, and therefore may help with the dwarf structural diversity (Jiang, Bournaud, and Dekel, in prep).

Cusp-core transformation

  • We present a simple analytic model that predicts the change of halo density profile upon mass increments or decrements from the center (which mimic gas inflows and outflows). The model makes use of a 3-parameter density profile that has good analytic properties and the flexibility in the inner slope, and assumes energy conservation (Freundlich, Dekel, Jiang+2019, arXiv:1907.11726). 


Galaxy morphology

Low-surface brightness

  • Deep imaging reveals the low-surface brightness components of massive ellipticals out to several R_eff (e.g., Jiang, Huang, Gu 2011); and with S. Huang and A. Leauthaud et al., we are trying to infer the triaxial 3d stellar distribution from deep imaging data.

Red nugget

  • We examine a nearby galaxy that resembles a high-z red nugget, and reveal a significant stellar disk in the center. Such structures would not be detected in z~2 early-type galaxies if they are there, due to limited spatial resolution of telescopes, while they would naturally account for the observed ellipticities of high-z red-nuggets (Jiang, van Dokkum, Bezanson, Franx 2012).​