Structures of Dwarf Satellites of Milky Way-like Galaxies: Morphology, Scaling Relations, and Intrinsic Shapes

Abstract

The structure of a dwarf galaxy is an important probe of the effects of stellar feedback and environment. Using an unprecedented sample of 223 low-mass satellites from the ongoing Exploration of Local Volume Satellites survey, we explore the structures of dwarf satellites in the mass range 105.5 < M ⋆ < 108.5 M ⊙. We survey satellites around 80% of the massive, M K < − 22.4 mag, hosts in the Local Volume (LV). Our sample of dwarf satellites is complete to luminosities of M V <−9 mag and surface brightness μ 0,V < 26.5 mag arcsec−2 within at least ∼200 projected kpc of the hosts. For this sample, we find a median satellite luminosity of M V = −12.4 mag, median size of r e = 560 pc, median ellipticity of ϵ = 0.30, and median Sérsic index of n = 0.72. We separate the satellites into late- and early-type (29.6% and 70.4%, respectively). The mass–size relations are very similar between them within ∼5%, which indicates that the quenching and transformation of a late-type dwarf into an early-type one involves only very mild size evolution. Considering the distribution of apparent ellipticities, we infer the intrinsic shapes of the early- and late-type samples. Combining with literature samples, we find that both types of dwarfs are described roughly as oblate spheroids that get more spherical at fainter luminosities, but early-types are always rounder at fixed luminosity. Finally, we compare the LV satellites with dwarf samples from the cores of the Virgo and Fornax clusters. We find that the cluster satellites show similar scaling relations to the LV early-type dwarfs but are roughly 10% larger at fixed mass, which we interpret as being due to tidal heating in the cluster environments. The dwarf structure results presented here are a useful reference for simulations of dwarf galaxy formation and the transformation of dwarf irregulars into spheroidals.

Quenching of satellite galaxies of Milky Way analogues: reconciling theory and observations

“…This suggests the existence of a strong correlation between the star-forming properties and morphologies of dwarf galaxies. This correlation is also discussed in Carlsten et al ( 2021a ) in the context of the observational LV sample, although the observed star formation rates are currently tentative. Further comparisons between the morphologies of dwarf galaxies in the simulations versus observations will shed more light on this correspondence, but this is beyond the scope of this study.…”

“…Note that applying the corrected factor to the LV data does not change the conclusion that the SAGA sample appears to have a surface brightness limit of μ e , r 25 mag arcsec −2 and actually slightly impro v es the agreement of ARTEMIS with LV with regards to satellite abundances in Font et al ( 2021 ). Carlsten et al ( 2021a ) for splitting the observed satellite galaxies by morphological type: early-type galaxies are abo v e this line and late-type below. This fit also separates well the quenched from the starforming satellites in our simulations.…”

“…Notably, this ∼1.5 mag difference in surface brightness between star-forming and quenched satellites is about the same level as found observationally between late-type and early-type dwarfs in the LV surv e y (see fig. 2 of Carlsten et al 2021a ).…”

“…The two populations could be easily delineated by a simple linear fit to the simulations. Ho we ver, we use instead the observational fit derived by Carlsten et al ( 2021a ), who used an extended set of observations of dwarf satellites of MW analogues in the LV. We choose this fit for consistency with the LV observations, although we note that the simulated fit separating the two populations is very similar and adopting it would not change the following results.…”

“…The two trends predicted from the simulations agree very well, confirming that the colour cut is an appropriate criterion for designating quenched satellites. We also present a third case for the simulations (in purple), which shows the effect of adding observational noise to the simulated colours at a level of 0.07 dex (assuming a Gaussian distribution), which is approximately the median statistical error on the colour estimates in Carlsten et al ( 2021a ). Convolving the raw g − i simulated colours with this noise has the effect of fanning out the sharp red sequence in the top panel, resulting in some satellites at intermediate luminosities being assigned active status rather than their true quenched status.…”