🔭 astrophysics

Learning the Universe: Constrained simulations of the Coma galaxy cluster -- I. Radial X-ray and Compton-y signatures

This paper presents a suite of 50 high-fidelity constrained simulations of Coma cluster analogues using the IllustrisTNG model to successfully reproduce observed X-ray and thermal Sunyaev-Zel'dovich profiles, thereby providing a robust statistical framework for understanding how feedback, mergers, and environmental factors shape massive galaxy clusters.

Original authors: Ulrich P. Steinwandel, Stuart McAlpine, Richard Stiskalek, Rüdiger Pakmor, Volker Springel, Eugene Churazov, Ildar Khabibullin, Jens Jasche, Guilhem Lavaux, Greg L. Bryan

Published 2026-06-10

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Ulrich P. Steinwandel, Stuart McAlpine, Richard Stiskalek, Rüdiger Pakmor, Volker Springel, Eugene Churazov, Ildar Khabibullin, Jens Jasche, Guilhem Lavaux, Greg L. Bryan

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine trying to understand a specific, massive city on Earth (like New York or Tokyo) by looking at a single, blurry photograph taken from space. You can see the general shape, but you can't tell if the traffic jams are caused by a construction accident, a parade, or just the usual rush hour. Now, imagine you could generate 50 different, hyper-realistic "what-if" scenarios of that city's history, all starting from the same basic neighborhood layout but evolving differently over time. You could then compare each scenario to the actual photo to see which history matches the reality we see today.

That is essentially what this paper does, but instead of cities, the authors are studying the Coma Galaxy Cluster—a massive, crowded neighborhood of thousands of galaxies located about 300 million light-years away.

Here is a breakdown of their work in simple terms:

1. The Problem: One Snapshot, Many Stories

Astronomers have powerful telescopes (like eROSITA and Planck) that take pictures of the Coma cluster. These pictures show us how hot the gas is and how bright the cluster is. However, a single picture is just a snapshot in time. It doesn't tell us why the cluster looks that way. Did it form slowly and peacefully? Was it recently smashed together by a giant collision?

Standard computer simulations usually create random universes. They might produce a cluster that looks like Coma, but it's just a lucky guess. It's like trying to find a specific person in a crowd by generating random faces; you might get a match, but you won't know if it's the right person or just someone who looks similar.

2. The Solution: "Constrained" Simulations

The authors used a clever trick called constrained initial conditions. Think of this as working backward from the crime scene.

They took the actual map of the universe around Coma (the "neighborhood" of galaxies and dark matter).
They used a sophisticated statistical method (called BORG/Manticore) to reconstruct the exact starting conditions that would lead to Coma existing exactly where it is today.
Instead of running just one simulation, they ran 50 different versions.

Imagine you have a specific recipe for a cake. You know the ingredients (the initial conditions) must be very specific to get a cake that looks like Coma. But, you don't know the exact temperature of the oven or the precise stirring speed for every single batch. So, they baked 50 cakes using the same base ingredients but with slight variations in the "stirring" (the assembly history).

3. The Ingredients: The "IllustrisTNG" Model

To bake these cosmic cakes, they used a very advanced "recipe" called the IllustrisTNG model. This model includes all the physics needed to make a realistic universe:

Gravity: Pulling matter together.
Gas Physics: How hot gas cools, heats up, and flows.
Star Formation: How stars are born from gas clouds.
Black Holes: How supermassive black holes at the centers of galaxies blow out energy and regulate the growth of the galaxy (like a thermostat).

4. The Results: Finding the Best Match

After baking their 50 virtual Coma clusters, they compared them to the real photos taken by telescopes.

The "Best Fits": They found that about 10 of their 50 simulations looked almost exactly like the real Coma cluster in terms of brightness and heat distribution.
The "Scatter": The other 40 simulations looked a bit different. Some were too bright in the center; others were too dim. This variation wasn't a mistake; it was a feature. It showed that even if you start with the same neighborhood, the final result depends heavily on the cluster's history.

5. The Big Discovery: History Matters More Than Mass

The most important finding is about why some simulations matched better than others.

The "Cool Core" Mystery: Some clusters have a very hot, bright center (a "cool core" in astronomical terms, which is counter-intuitive but refers to low entropy). The simulations showed that clusters with these bright centers were the ones that formed earlier and grew calmly.
The "Violent" Clusters: Clusters that formed later or were constantly being smashed by smaller clusters (mergers) ended up with messy, less bright centers.
The Analogy: Think of it like a family tree. Two families might have the same number of people today (mass), but one family grew slowly and peacefully over generations, while the other had many sudden, chaotic events (like big arguments or sudden arrivals). The paper found that the "peaceful" families (early formers) looked more like the real Coma cluster in the center, while the "chaotic" ones looked different.

6. What This Means for Science

The paper concludes that to truly understand a galaxy cluster like Coma, you can't just look at its mass or size. You have to look at its history.

By running 50 different "what-if" scenarios, they proved that the way a cluster looks today is a direct record of how it was built.
They identified that the "best" matches to the real Coma cluster are those that grew mostly through small, gentle mergers rather than giant, violent collisions.

Summary

In short, the authors built 50 digital twins of the Coma galaxy cluster. By comparing them to real telescope data, they proved that the cluster's current appearance is a fingerprint of its past. They found that the real Coma cluster likely grew up "early" and "calmly," avoiding the kind of violent collisions that would have scrambled its center. This method allows astronomers to stop guessing and start reading the history books of the universe with much greater precision.

Technical Summary: Learning the Universe: Constrained simulations of the Coma galaxy cluster – I. Radial X-ray and Compton-𝑦 signatures

Problem Statement
While large-volume cosmological simulations (e.g., IllustrisTNG, MillenniumTNG) have successfully reproduced statistical properties of galaxy populations and cluster scaling relations, they sample only a limited volume of the universe. Consequently, they contain very few instances of rare, massive systems like the Coma cluster, making them ill-suited for detailed, object-by-object comparisons with high-fidelity observations of specific local clusters. Conversely, previous targeted "zoom-in" simulations of local clusters often relied on single realisations or employed galaxy formation physics that differed from state-of-the-art models like IllustrisTNG. There is a need for a statistically robust framework that combines constrained initial conditions (to match the local large-scale environment) with high-resolution hydrodynamical evolution using modern galaxy formation physics to interpret multi-wavelength observations of specific clusters.

Methodology
The authors present a suite of 50 high-fidelity zoom-in simulations of Coma cluster analogues. The methodology integrates three distinct components:

Constrained Initial Conditions (ICs): The simulations utilize posterior samples from the BORG/Manticore Bayesian inference of the local Universe (McAlpine et al. 2025), which reconstructs the local large-scale structure using the 2M++ galaxy catalogue. This ensures that the large-scale environment, filamentary structure, and mass of the Coma analogue are consistent with observational constraints.
Ensemble Approach: Rather than selecting a single "best" realisation, the authors evolve 50 independent posterior samples. This ensemble captures both cosmic variance and uncertainties in the local initial conditions, providing a statistical distribution of possible assembly histories and dynamical states for a Coma-mass cluster.
Hydrodynamical Evolution: Each of the 50 realisations is evolved using the Arepo code with the IllustrisTNG (TNG) galaxy formation model. The simulations employ a multi-level refinement scheme, achieving a dark matter particle mass of $6.0 \times 10^7 M_\odot$ and a gas cell mass of $1.2 \times 10^7 M_\odot$ in the high-resolution region. The physics includes magnetohydrodynamics (MHD), metal-line cooling, stellar feedback, and a dual-mode AGN feedback scheme.

The simulations are oriented to match the observed sky frame of Coma, allowing for direct projection of observables. The authors compute projected maps and azimuthally averaged radial profiles for the thermal Sunyaev–Zel'dovich (tSZ) effect (Compton-𝑦) and X-ray surface brightness (0.4–2 keV), comparing them directly with eROSITA and Planck satellite data.

Key Contributions

First TNG-based Constrained Ensemble: This work provides the first ensemble of Coma analogues evolved with the full IllustrisTNG physics suite derived from BORG/Manticore constrained initial conditions.
Statistical Framework for Local Clusters: By running 50 realisations, the study quantifies the intrinsic scatter in cluster properties (mass, gas fraction, thermodynamic profiles) that arises from variations in assembly history and merger activity, even when the large-scale environment is fixed.
Direct Multi-wavelength Comparison: The paper establishes a rigorous pipeline for comparing simulated radial profiles of Compton-𝑦 and X-ray surface brightness against specific observational datasets (Planck and eROSITA), identifying a subset of "best-fitting" realisations.

Results

Scaling Relations: The ensemble of 50 Coma analogues reproduces standard galaxy cluster scaling relations (gas fraction, $Y_{SZ}$ –mass, $L_X$ –mass, stellar mass–halo mass, and black hole scaling relations) consistent with the TNG-Cluster suite and observational compilations. A systematic, mild excess in the shell-integrated $Y_{SZ,500}$ and gas fraction ( $f_{gas}$ ) relative to Planck normalisation is noted, consistent with known TNG characteristics at the cluster scale.
Radial Profile Scatter: The 50 realisations exhibit significant scatter in the inner radial profiles of Compton-𝑦 and X-ray surface brightness. While the ensemble mean brackets the observed Coma profiles, individual realisations vary widely.
Best-Fitting Realisations: Ten realisations were identified as the best fits to the observed radial profiles. These best-fitting clusters span a wide range of total halo masses ( $M_{200c} \approx 0.87\text{--}1.45 \times 10^{15} M_\odot$ ) and exhibit diverse two-dimensional morphologies (from relaxed to disturbed), demonstrating that a good match to a 1D radial profile does not uniquely determine the 2D structure or total mass.
Origin of Central Scatter: The scatter in central brightness is primarily driven by the central entropy of the intracluster medium (ICM).
- Low-entropy (cool-core) clusters (16/50 realisations) exhibit higher central X-ray and Compton-𝑦 brightness.
- High-entropy clusters (34/50) are fainter in the center.
- This central entropy state is not a transient feature driven by short-timescale AGN bursts or recent major mergers (which are rare at $z < 0.3$ in the sample). Instead, it is determined by the long-term assembly history.
Assembly History Correlation: Low-entropy clusters are "early formers" (median formation redshift $z_f(10\%) \approx 1.59$ ) compared to high-entropy clusters ( $z_f(10\%) \approx 1.15$ ). They accreted mass predominantly through "soft" channels (mini-mergers and smooth accretion) rather than major mergers. This "soft-grown" history allows the central ICM to develop and maintain a low-entropy state. The central black hole growth tracks this history, with low-entropy BCGs having more massive black holes at $z=1$ and slower subsequent growth.
Time Variability: For the vast majority of clusters, the radial profiles are stable over the last 500 Myr. A notable exception is the top-ranked best-fitting realisation (Coma_042), which undergoes rapid profile reshaping due to late-stage minor-merger activity, highlighting that even the best fits can be transient.

Significance and Claims
The paper claims that combining constrained cosmological initial conditions with state-of-the-art galaxy formation physics is an effective strategy for generating targeted, observation-driven analogues of specific clusters. The resulting dataset serves as a valuable resource for:

Testing ICM Physics: Providing a controlled environment to study how feedback, mergers, and large-scale environments shape observable cluster properties.
Calibrating Scaling Relations: Offering a statistical baseline for interpreting the scatter in cluster observables.
Interpreting Observations: Enabling the interpretation of upcoming joint X-ray and Sunyaev–Zel'dovich observations (e.g., with XRISM) by providing a forward-model emulator based on the posterior distribution of the local Universe.

The authors emphasize that the scatter observed in the ensemble is a physical posterior scatter reflecting the residual freedom in the local Universe data, rather than a calibration uncertainty. They suggest that future work could extend the constrained inference likelihood to include cluster-scale thermodynamic observables (Compton-𝑦 and X-ray profiles) to further tighten constraints on the assembly history of the Coma cluster.