A giant galaxy’s stellar layers preserve clues to its dark halo’s past

A massive galaxy does not grow alone. It sits within a much larger dark matter halo, alongside satellite galaxies that can eventually merge with it or lose their stars to its outskirts. Those processes change both the galaxy and its surroundings. But how much of their history remains readable afterward?

Shuo Xu and colleagues investigate this question in an arXiv preprint. Their computational analysis finds that the stellar mass in different radial regions of massive central galaxies retains information about halo mass at different earlier epochs. The supplied record does not establish peer-reviewed publication; these are simulation results, not an observational detection.

Reading a simulated history

The researchers used IllustrisTNG’s TNG300-1 cosmological hydrodynamical simulation. Their starting sample contained 3,388 halos more massive than ten trillion solar masses at redshift 0.4. Redshift labels cosmic epochs, with higher values indicating earlier times. Crucially, the paper’s reference epoch—sometimes called “present-day” in its analysis—is redshift 0.4, not today’s Universe.

They projected each central galaxy’s stars into a two-dimensional mass map, excluding stars still bound to satellites. They measured stellar mass inside elliptical apertures extending to 10, 30, 50, 100 and 150 kiloparsecs, and in the intervening annuli. Merger trees then connected those systems to their main progenitors back to redshift 3. Eight halos were excluded because their usable histories did not extend that far.

Two diagnostics asked different questions. The “fossil correlation” measured how strongly a galaxy property at the reference epoch correlated, across the sample, with its progenitor halo’s mass at each earlier epoch. The “rank retention fraction” measured how many halos selected by a reference-epoch property had progenitors among the most massive halos at an earlier epoch.

The first describes a statistical relationship; the second tests its consequences for selecting a population.

Different radii, different epochs

The central stellar components did not simply correlate ever more strongly with halo mass toward the reference epoch. Instead, their correlations peaked earlier, with a systematic radial ordering.

Stellar mass within 10 kiloparsecs correlated most strongly with halo mass around redshift 1. The outermost annulus, between 100 and 150 kiloparsecs, peaked later, around redshift 0.7. The annular peak positions remained broadly stable as the researchers changed the minimum halo mass from group to cluster scales, although correlation strengths varied.

To interpret those peaks, the team also examined subsequent halo growth while statistically controlling for the mass already assembled. Growth stopped adding substantial information about the reference-epoch stellar mass earlier for inner regions and later for outer regions. The authors interpret each correlation peak as marking the end of that region’s main sensitivity to halo accretion—not necessarily the moment when its stars formed or its growth ceased.

This is therefore not a set of precisely dated stellar layers. It is a population-level relationship between spatial structure and cumulative halo history.

The contribution of disrupted galaxies

The researchers separated stars formed within the central galaxy from stars formed elsewhere and later accreted. The accreted stars retained the radial ordering of correlation peaks, identifying them as the main carriers of the stratified signal. Locally formed stars contributed especially to the innermost region’s memory of early halo assembly.

Comparisons with three-dimensional stellar measurements produced nearly identical fossil correlations, arguing against the pattern being merely a projection artifact within this simulation.

The outer stellar halo also correlated with the past satellite population. At a cosmic age of about 6.5 billion years, its correlation with historic satellite richness reached 0.80, compared with 0.60 for reference-epoch richness. Diffuse accreted stars can preserve information about disrupted galaxies that a census of surviving satellites misses.

A promising tracer, not a reconstruction yet

For epochs earlier than roughly redshift 0.7, stellar-mass selections beyond 10 kiloparsecs retained more of the most massive progenitors than satellite-richness selections did. That suggests a potential way to reduce bias when connecting populations across time.

Important limits remain. The richness measurements used true three-dimensional membership, avoiding the projection contamination faced by observations. Historic and reference-epoch satellite counts also used different stellar-mass thresholds. The innermost component had poorly constrained peaks in the highest-mass samples, and the authors call for tests with other simulations.

The result supports using stellar profiles to constrain halo growth statistically. It does not yet provide a uniquely recoverable history for an individual observed galaxy.