Mass Proxy Quality of Halo Properties in IllustrisTNG and FLAMINGO Simulations: II. Galaxies, Stars, and Black Holes

arXiv:2610.10901 · astro-ph.GA, astro-ph.CO · Submitted 2026-10-07 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Astrophysics Radio. Generated commentary on the latest astrophysics papers.

Vera: Today's paper: "Mass Proxy Quality of Halo Properties in IllustrisTNG and FLAMINGO Simulations".

Jocelyn: Galaxies, Stars, and Black Holes." The information is dense, focusing on characterizing integral properties within group- and cluster-scale halos across different cosmological simulations (illustrisTNG, TNGcluster, and flamingo).

Vera: First, who's behind it and why it matters.

Paper summary: Vera: So, wrapping up this talk on Mass Proxy Quality of Halo Properties in IllustrisTNG and FLAMINGO Simulations: II. Galaxies, Stars, and Black Holes by Aljamal et al., the authors are essentially telling us that total stellar mass is our best bet for estimating the true halo mass at cluster scales.

Jocelyn: And they back that up by saying that number has a scatter of seven to fourteen percent when you condition on it, which is pretty tight for something this big. They also say that the number of satellite galaxies and total black hole mass are good secondary proxies, but they don't give them the same precision as total stellar mass.

Subrahmanyan: The overall implication for the field is that this statistical analysis provides a sort of ground truth for population studies. It tells us how to weigh different observational probes when we try to map out the cosmic web using galaxy clusters.

Vera: It’s about moving beyond just looking at one property and understanding how those properties are connected, like the correlations between stellar, satellite, and gas properties mentioned in the abstract. That connection structure is really important for building better models of how these halos evolve over time.

Jocelyn: And they point out that while total stellar mass is stable in its proxy quality across redshift for some halo masses, other properties like the brightest central galaxy stellar mass and intracluster light mass show more evolution.

Subrahmanyan: So when you think about future work, it’s about developing observable measures that account for things like projection effects or confusion from non-cluster sources so we can really match these simulation findings with what telescopes are going to actually see out there.

Vera: Exactly. We need to make sure the real world data we collect has those same kinds of statistical handles so we can trust our measurements of the universe's structure.

Conclusion: Vera: So we’re wrapping up this look at how reliable different galaxy and black hole measurements are for telling us about the underlying dark matter halos in these large simulations.

Jocelyn: The paper is called "Mass Proxy Quality of Halo Properties in IllustrisTNG and FLAMINGO Simulations: II. Galaxies, Stars, and Black Holes." The authors are looking at how well we can use things like total stellar mass or satellite counts to guess the actual mass of a halo when we only see its galaxies.

Subrahmanyan: It’s about testing those relationships across different simulations—IllustrisTNG, TNG-Cluster, and FLAMINGO—at various redshifts. They are checking scaling relations and how much scatter there is when you try to use one property to predict the true halo mass.

Vera: What they found is that total stellar mass turns out to be the best way to estimate the halo mass when you're looking at cluster scales. The scatter there, when we condition on it, is pretty tight, between seven and fourteen percent.

Jocelyn: And they also looked at things like the number of satellite galaxies and black hole mass as other candidates for proxies. They found those are good too, but the precision isn't quite as high as total stellar mass.

Subrahmanyan: The big picture here is that this gives us a way to really test our models of how structure grows in the universe. If we can reliably map out these relationships, it helps us understand how galaxies and black holes live inside those invisible dark matter scaffolds.

Vera: It shows that while there are some complexities, like redshift changes affecting the accuracy, we have a solid statistical tool for picking what works best when looking at these simulations.

Jocelyn: It really tells us which observational properties we should focus on if we want to use galaxy surveys to measure the underlying dark matter distribution accurately.

Subrahmanyan: And knowing that total stellar mass is so consistent across the samples is a strong indicator for future observational studies aiming to probe this cosmic structure directly.

Department of Physics and Leinweber Institute for Theoretical Physics, University of Michigan · Department of Statistics and Data Science, The University of Texas at Austin · Max-Planck-Institut für Astrophysik

astro-ph.GA, astro-ph.CO

Submitted: 2026-10-07

Updated: 2026-10-09

Comments: 21 pages, 12 figures

Code: https://github.com/illustristng/illustris_python

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Importance score: 82/100

The gist: Galaxies, Stars, and Black Holes." The information is dense, focusing on characterizing integral properties within group- and cluster-scale halos across different cosmological simulations

Key concepts

Mass Proxy Quality (MPQ)
This measures how well an observable property from a simulation can predict the true underlying mass of a dark matter halo. A high MPQ means the property is very reliable for estimating the halo's mass, which is crucial for using simulations to understand real galaxies.
Scaling Relations
These describe how different properties of halos change systematically as their total mass increases. The study examines these relations (like how stellar mass scales with halo size) to see if they are consistent across different simulations and redshifts.
Conditional Likelihoods
This refers to the probability distribution of a property given a specific halo mass. For instance, it tells us how likely a certain stellar mass is for a halo of a specific size. Many properties show lognormal likelihoods, indicating predictable statistical behavior.
Intracluster Light (ICL)
ICL refers to the diffuse stellar material found within galaxy clusters that is not gravitationally bound to any single galaxy. The study investigates how ICL mass correlates with other stellar and gas properties, suggesting it's a useful tracer for cluster evolution.

Terminology

Summary

Galaxies, Stars, and Black Holes." The information is dense, focusing on characterizing integral properties within group- and cluster-scale halos across different cosmological simulations (illustrisTNG, TNGcluster, and flamingo).

Here is a detailed synthesis of the findings:

Overview of Study Scope

The primary objective of this study is to characterize the mass-conditioned scaling relations, covariance structures, conditional likelihoods, and Mass Proxy Quality (MPQ) for several intrinsic halo properties within three major cosmological hydrodynamical simulations: illustrisTNG, TNGcluster, and flamingo. The analysis focuses on properties measured within an R500 c aperture at redshifts z in 0, 0.5, 1, 2.

Key Findings on Scaling Relations and Correlations

The study reveals nuanced behaviors across the measured properties:

  • General Consistency: Many properties exhibit remarkably similar mass-localized slopes and intrinsic scatters across halo samples and redshifts. Furthermore, correlations among stellar, satellite, and gas properties are generally consistent in sign and their dependence on halo mass and redshift.

  • Variations in Specific Properties: Correlations involving the Black Hole Mass (MBH) and Satellite Occupation (Nsat) show greater variation compared to those involving stellar properties.

  • Stellar Property Correlations: The correlation structure of property deviations about their mean behaviors is complex but reveals several common expectations:

  • Intracluster Light (ICL) shows correlations with other stellar properties.

  • There is a strong anti-correlation between the Bright Central Galaxy (BCG) stellar mass and the total stellar mass and counts of satellite galaxies.

  • Stellar mass measures exhibit anti-correlated behavior with hot gas mass in clusters.

Mass Proxy Quality (MPQ) Assessment

A central theme is identifying which properties serve as reliable proxies for the underlying true halo mass (M halo).

  • Best Proxies: Across cluster scales, the total stellar mass (M, tot), the number of satellite galaxies (Nsat), and the total black hole mass (MBH) are identified as the best proxies for total halo mass.

  • Precision: The precision (scatter) in true halo mass conditioned on these properties is estimated using KLLR-derived MPRs, with halo mass precisions generally ranging between 7% and 15%.

  • Exceptional Proxies: The total stellar mass (M, tot) and the total BH mass (MBH) are highlighted as the best MPQs across most of the mass range, although they exhibit some quantitative disagreement between TNG and flamingo expectations.

Conditional Likelihoods and Skewness

The conditional likelihoods for many properties follow a log-normal shape. Specifically:

  • M, tot, Nsat, and the 1D satellite velocity dispersion (sigma sat, 1 D) are nearly lognormal.

  • M, BCG, and M, ICL show a negative skew.

  • The largest sample of 94,000 halos from FLAMINGO at z=0 often displays a mildly negative skew and positive kurtosis.

Simulation-Specific Differences and Normalization Issues

A critical finding relates to the normalization differences between simulations:

  • Stellar vs. Gas Properties: The normalizations of stellar properties, such as total and BCG stellar mass, differ more dramatically across simulations than the normalizations of gas properties. This is evidenced by observed differences in overall baryon fractions ranging from 10 to 20% in clusters (Figure 3).

  • Gas Properties: In contrast, the mass-dependent slopes, property variance, and implied mass proxy qualities for gas properties show generally good qualitative and quantitative agreement across the simulated samples.

Redshift Dependence of MPQ

Appendix F provides a detailed look at how the MPQ changes with redshift:

  • Stable Properties: Properties with strong mass dependence in their MPQ (M, tot, M, sat, and Nsat) exhibit relatively weak redshift evolution.

  • Evolving Properties: Conversely, properties whose MPQ is nearly independent of halo mass—specifically ** M, BCG stellar mass, ICL mass, MBH, and M 14 **—show more noticeable redshift evolution.

  • Scatter Evolution: For the total stellar mass (M, tot), the implied mass scatter decreases by approximately 5% between successive redshift bins for halos with M 500 c 10 13.5 M, while remaining nearly unchanged at higher masses. The underlying causes of this evolution differ: in TNG300, it is driven by a steepening of the M, tot scaling relation, whereas in FLAMINGO, it results mainly from a decrease in intrinsic scatter.

  • ** M, BCG Evolution:** The MPQ of M, BCG exhibits stronger redshift evolution than other discussed properties. In both TNG300-1 and FLAMINGO, the implied mass scatter increases primarily between z=0 and z=0.5, with comparatively little evolution at higher redshifts, though this evolution is strongest for halos with M 500 c 10 14 M.

Conclusion and Future Directions

The intrinsic findings derived from these synthetic observables provide a ground truth for population analysis. The study concludes that total stellar mass, number of satellite galaxies, and total black hole mass are the most reliable proxies for total halo mass at cluster scales. Future work is suggested to focus on developing equivalent observable measures that incorporate real-world observational challenges such as projection effects, non-cluster sources of confusion, and instrumental characteristics.

Improvements for AI systems

  1. Improved mass-conditioned likelihood estimation by using KLLR to extract continuous, mass-conditioned scaling relation parameters and assessing their consistency with a commonly assumed log-normal form. This allows for more precise inference of true halo masses by accounting for the scale dependence of properties, as noted in Section 6.

  2. Enhanced discrimination between simulation outcomes by using the calculated mass proxy quality (MPQ) to guide observational data analysis. For example, researchers can use the finding that total stellar mass is the best mass proxy and its associated scatter of 7–14% to prioritize which observables are most reliable for constraining cosmological parameters.

  3. More robust modeling of galaxy evolution by incorporating simulation-informed priors into forward modeling approaches for simulation-based inference. This leverages the consistent behavior found across simulations, such as how the local slopes and scatter remain consistent despite normalization differences, to improve the accuracy of predicting baryonic physics in new halo samples.

  4. Refined understanding of satellite galaxy assembly by utilizing the derived scaling relations for satellite occupation and velocity dispersion to probe how astrophysical processes orchestrate the arrangement of hot gas, stars, black holes, and diffuse ICL. This helps distinguish between simulation behaviors like the joint shift toward higher bound stellar mass and v c, dm indicates that the TNG-Cluster satellites are, on average, associated with more massive or more centrally concentrated dark-matter subhalos.

Sources

Related papers