The dwarf stellar mass function in different environments and the lack of a generic missing dwarfs problem in ΛCDM

TL;DR

This study investigates how the dwarf stellar mass function, for $M_\star < 10^{9.5} M_\odot$, depends on environment defined by distance to the cosmic web, and whether a generic missing dwarfs problem exists in $\\Lambda$CDM when all environments are included. It combines deep COSMOS and XMM-LSS photometry with high-resolution hydrodynamical simulations (NewHorizon, TNG50, FIREbox), using the DisPerSE framework to connect environment to the observed mass function down to $M_\star \sim 10^{7} M_\odot$ out to $z \sim 0.4$. The results show strong environmental modulation: proximity to filaments increases the knee mass $M^*$ and flattens the dwarf slope $\\alpha_1$, boosting the massive-to-dwarf ratio, with qualitative agreement across fields and simulations, though the simulations differ by up to ~0.5 dex in the dwarf regime. Importantly, there is no evidence for a generic missing dwarfs problem when all environments are considered; subtle discrepancies likely reflect differences in baryonic physics and resolution, and a modest mass-recalibration (~0.2 dex) could further align observations with theory. Overall, the work highlights the critical role of environmental bias in shaping the dwarf population and supports the view that baryonic processes mitigate the classic missing satellites tension in $\\Lambda$CDM.

Abstract

We combine deep photometric data in the COSMOS and XMM-LSS fields with high-resolution cosmological hydrodynamical simulations to explore two key questions: (1) how does the galaxy stellar mass function, particularly in the dwarf (Mstar < 10^9.5 MSun ) regime, vary with environment, defined as distance from the large-scale structure (LSS) traced by nodes and filaments in the cosmic web? (2) is there a generic 'missing dwarfs' problem in LambdaCDM predictions when all environments - and not just satellites around Milky Way like galaxies - are considered? The depth of the observational data used here enables us to construct complete, unbiased samples of galaxies, down to Mstar ~ 10^7 MSun and out to z ~ 0.4. Strong environmental differences are found for the galaxy stellar mass function when considering distance from LSS. As we move closer to LSS, the dwarf mass function becomes progressively flatter and the knee of the mass function shifts to larger stellar masses, both of which result in a higher ratio of massive to dwarf galaxies. While the stellar mass functions from the three simulations (NewHorizon, TNG50 and FIREbox) considered here do not completely agree across the dwarf regime, there is no evidence of a generic missing dwarfs problem in the context of LambdaCDM, akin to the results of recent work that demonstrates that there is no missing satellites problem around Galactic analogues.

Figures (10)

• Figure 1: An example showing a number density map in the COSMOS field at $0.210<z<0.235$ created using DisPerSE. The colours indicate the local density (see colour bar), while the solid black lines show the locations of the filaments.
• Figure 2: Total magnitudes (left) and effective surface brightnesses (right) vs rest-frame $(u-r)$ colours of COSMOS2020 galaxies in the stellar mass and redshift ranges of 10$^7$ M$_{\odot}$ < $M_\star$ < 10$^{7.25}$ M$_{\odot}$ and $0.15 < z < 0.18$ respectively. The black points indicate the COSMOS2020 galaxies. The grey shaded region indicates the location of a purely old SSP that forms at $z=2$ (assuming half solar metallicity). The upper and lower limits of the grey shaded region indicate the rest-frame $(u-r)$ colour of this SSP with and without a reddening of $E(B-V)$ = 0.1 respectively The solid grey vertical line indicates the mid-point of the grey shaded region. The red and blue horizontal lines indicate the detection limits of the Deep and Ultradeep layers of the HSC-SSP imaging.
• Figure 3: The galaxy stellar mass function for the COSMOS and XMM-LSS fields in different environments. Galaxies at small distances from LSS (i.e. nearest to LSS) are defined as those that reside in the lowest tercile (0$^{\rm th}$ -- 33$^{\rm rd}$ percentile values) of both the distances to nodes and the distances to filaments. Galaxies at intermediate and large distances from LSS are defined as those that reside in the middle (33$^{\rm rd}$ -- 66$^{\rm th}$ percentile values) and upper (66$^{\rm th}$ -- 100$^{\rm th}$ percentile values) terciles of both the distances to nodes and the distances to filaments respectively.
• Figure 4: Best-fit parameter values and their uncertainties for M$^\star$ and $\alpha_1$ from the double Schechter fits to the mass functions in the COSMOS and XMM-LSS fields in different environments. Galaxies at small distances from LSS (i.e. nearest to LSS) are defined as those that reside in the lowest tercile (0$^{\rm th}$ -- 33$^{\rm rd}$ percentile values) of both the distances to nodes and the distances to filaments. Galaxies at intermediate and large distances from LSS are defined as those that reside in the middle (33$^{\rm rd}$ -- 66$^{\rm th}$ percentile values) and upper (66$^{\rm th}$ -- 100$^{\rm th}$ percentile values) terciles of both the distances to nodes and the distances to filaments respectively.
• Figure 5: The distribution of projected distances from the nearest filaments for galaxies in the COSMOS and XMM-LSS fields. The median distance is shown using the dashed blue vertical line.