Halo Properties from Observable Measures of Environment: II. Central versus Satellite Classification

TL;DR

This work develops a neural-network framework to classify halos into centrals vs satellites using observable environmental measures and to infer interaction history (historical centrals vs satellites) and current orbital status (infalling vs orbiting). It trains on the SMDPL simulation and validates on Bolshoi-Planck, using UniverseMachine galaxies to connect halo properties to observable stellar masses; a baseline optimal isolation is outperformed by a kNN-based neural network, achieving ~89–90% accuracy for present central–satellite classification and ~86–89% for history and orbiting classifications. Projection effects are identified as the dominant source of misclassification, while full 3D phase-space information dramatically reduces errors (to ~4.1% misclassification) for the orbiting/infalling task. The method offers a practical, observable-pathway to quantify environmental influence on galaxy evolution, with potential applications to local surveys like GAMA and DESI BGS and room for enhancements with velocity information. Overall, observable environment encodes substantial information about halo dynamics and histories, enabling new studies of environment-driven galaxy formation.

Abstract

A physical understanding of galaxy formation and evolution benefits from an understanding of the connections between galaxies, their host dark matter halos, and their environments. In particular, interactions with more-massive neighbors can leave lasting imprints on both galaxies and their hosts. Distinguishing between populations of galaxies with differing environments and interaction histories is therefore essential for isolating the role of environment in shaping galaxy properties. We present a novel neural-network based method, which takes advantage of observable measures of a galaxy and its environment to recover whether it (1) is a central or a satellite, (2) has experienced an interaction with a more massive neighbor, and (3) is currently orbiting or infalling onto such a neighbor. We find that projected distances to, redshift separations of, and relative stellar masses with respect to a galaxy's 25 nearest neighbors are sufficient to distinguish central from satellite halos in $> 90\%$ of cases, with projection effects accounting for most classification errors. Our method also achieves high accuracy in recovering interaction history and orbital status, though the network struggles to distinguish between splashback and infalling systems in some cases due to the lack of velocity information. With careful treatment of the uncertainties introduced by projection and other observational limitations, this method offers a new avenue for studying the role of environment in galaxy formation and evolution.

Figures (18)

• Figure 1: Here we show a 2D projection of four objects, each centered at a point with its virial radius indicated by a shaded circle surrounding that point and a dashed line representing 2R$_{\text{vir}}$ enclosing both. The values shown along the connecting lines between objects represent the halo-centric distance, i.e., the distance between the two halo scaled by the virial radius of the more-massive halo. By this definition, the orange halo clearly falls within 1 $R_{vir}$ of the black halo (it has a halo-centric distance $< 1$ with respect to the black halo), with no other nearby-larger objects, and thus is a satellite of the black halo. On the other hand, the blue and green halos do not meet this criteria. Additionally, the blue halo, while physically closer to the green halo, has a smaller halo-centric distance relative to the black halo, due to its much larger size, making the black halo the object exerting the largest tidal force on the blue halo.
• Figure 2: Halos from the Bolshoi-Planck simulation are plotted according to their halo-centric distance (x-axis) and relative radial velocity (y-axis) to the neighbor exerting the largest tidal influence. Satellites are represented by blue points and centrals by pink. The dividing line between the two classes is the dotted vertical line, which is set by the point when the distance between a halo and its more-massive neighbor is less than virial radius of said neighbor. Moving outwards, contours contain 15%, 30%, 60%, and 90% of the population.
• Figure 3: Halos from the Bolshoi-Planck simulation are plotted according to their halo-centric distance (x-axis) and relative radial velocity (y-axis) to the neighbor exerting the largest tidal influence. Historical satellites are represented by blue points and centrals by pink. Moving outwards, contours contain 15%, 30%, 60%, and 90% of the population. In contrast to Figure \ref{['fig:def1']}, historical satellites can be found for R/R$_{\text{vir}}$ > 1, i.e., to the right of dotted vertical line. However, all halos within the radial cut remain classified as satellites.
• Figure 4: The number density of halos with a halo-centric distance of less than one is shown in relative velocity and accretion-time space. Only halos falling in the top-left box, which have a recent accretion scale ($a_\text{acc} > 0.87$) and a negative radial velocity with respect to their future host are considered infalling. All other objects with halo-centric distance less than one are considered orbiting.
• Figure 5: Halos from the Bolshoi-Planck simulation are plotted according to their halo-centric distance (x-axis) and relative radial velocity (y-axis) for the more-massive neighbor exerting the largest tidal influence. Orbiting subhalos are represented by blue points and infalling by pink. Moving outwards, contours contain 15%, 30%, 60%, and 90% of the population. The orbiting halos fall within a triangle to the left of the figure. Infalling halos primarily fall outside this region with the exception of some objects that meet the orbiting velocity criteria but fell into their host less than half a dynamical time ago. The vertical black dotted line shows 1 $R_\text{vir}$, while the dashed curve shows the trajectory of a particle released at 2 $R_\text{vir}$ falling into a halo (see text for details).