🔭 astrophysics

Turbulence in Simulated Local Cluster Analogs: One-to-one comparisons between SLOW and XRISM/Hitomi

This study utilizes SLOW constrained simulations of the Coma, Virgo, and Perseus clusters to perform direct one-to-one comparisons with XRISM/Hitomi observations, revealing that the simulated turbulent velocities and pressure support align closely with observed values and highlighting the critical role of selection effects in previous statistical studies.

Original authors: Frederick Groth, Milena Valentini, Benjamin A. Seidel, Stephan Vladutescu-Zopp, Veronica Biffi, Klaus Dolag, Jenny G. Sorce

Published 2026-03-16

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Frederick Groth, Milena Valentini, Benjamin A. Seidel, Stephan Vladutescu-Zopp, Veronica Biffi, Klaus Dolag, Jenny G. Sorce

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 the space between galaxies in a giant cluster isn't empty. It's filled with a super-hot, invisible gas called the Intracluster Medium (ICM). Think of this gas like a giant, cosmic ocean.

For a long time, scientists have been trying to understand how "rough" this ocean is. Is it a calm, glassy lake, or is it a stormy sea with massive waves and swirling eddies? These "waves" are actually turbulence—gas moving chaotically at hundreds of kilometers per second.

This paper is like a cosmic detective story where the authors try to solve a mystery: Why do our computer simulations of these galaxy clusters predict much wilder, stormier gas than what we actually see with our new, super-precise telescopes?

Here is the breakdown of their investigation, explained simply:

1. The New Telescope vs. The Old Guesses

A few years ago, a telescope called Hitomi took a peek at the Perseus cluster and found the gas was surprisingly calm. Now, a new, even sharper telescope called XRISM is looking at three famous clusters: Coma, Perseus, and Virgo.

The Problem: When scientists run computer simulations to predict how these clusters should behave, the simulations usually say, "It's a hurricane in there!" They predict high turbulence and lots of energy.
The Reality: XRISM looks and says, "Actually, it's mostly a gentle breeze." The gas is moving, but not as wildly as the computers predicted.

2. The "One-to-One" Match-Up

Usually, scientists compare their computer models to observations by looking at many random clusters. It's like trying to guess the average height of a basketball team by measuring 1,000 random people and hoping some of them are on the team. This introduces a lot of "selection bias" (picking the wrong people).

The authors' clever trick: Instead of guessing, they used a special simulation called SLOW.

Think of SLOW as a Cosmic Time Machine that didn't just simulate any universe, but simulated our specific local neighborhood.
They built digital twins of the exact Coma, Virgo, and Perseus clusters we see in the sky.
Because they are looking at the exact same clusters (one-to-one comparison), they didn't have to worry about picking the wrong ones. They could say, "Here is the digital twin of Coma; let's see how it compares to the real Coma."

3. The Results: Who is the Stormy One?

The team ran their simulation and checked the "wind speeds" (turbulence) of their digital twins.

Coma (The Stormy Teenager): This cluster is very active. It's like a teenager who just had a huge fight; it's still shaking with energy. The simulation showed it has the most turbulence, with gas moving at about 600 km/s.
Perseus (The Calmer Adult): This one is more settled but still has some "sloshing" motions, like water in a bucket that was just shaken. It's in the middle, with speeds around 500 km/s.
Virgo (The Zen Master): This cluster is very relaxed. It's like a sleeping cat. The gas is moving very slowly, only about 100–200 km/s.

The Big Surprise: Even for the stormiest cluster (Coma), the amount of "push" the turbulence gives the gas (turbulent pressure) is tiny—only about 8%. For the calm ones, it's even less (1%).

This matches what XRISM is seeing! The simulations finally agreed with the telescope: The gas is much calmer than we thought.

4. Why Was There a Discrepancy Before?

So, why did older simulations predict such wild storms?

The "Selection" Trap: Previous studies looked at random clusters. They might have accidentally picked the most violent, merging clusters (the "stormy teenagers") and assumed all clusters were like that.
The "Zoom" Effect: The authors also found a tricky physics problem. When you look at a distant cluster, your telescope sees a bigger chunk of space at once. It's like looking at a crowd from a skyscraper; you can't see individual people walking, so the whole crowd looks like a blur of motion. The telescope might mistake big, slow movements (bulk motion) for fast, chaotic turbulence. The authors showed that if you don't account for this "blur," you overestimate the turbulence.

5. What's Missing?

The authors admit their simulation is a bit "bare-bones." They only included gravity and gas physics, but they left out cooling (gas getting cold) and feedback (energy blasts from black holes in the center).

Imagine the simulation is a movie without the special effects. The black holes (AGN) are like giant fans in the center of the cluster. In reality, these fans might be stirring the gas, adding a little bit of turbulence that the simulation missed.
However, even without these fans, the simulation matched the telescope data surprisingly well, suggesting that the "calmness" is real and not just a missing ingredient in the code.

The Takeaway

This paper is a victory for precision. By building a digital twin of the exact clusters we observe, the authors proved that:

Galaxy clusters are generally much calmer than we used to think.
The "storm" we see in older simulations was partly an illusion caused by looking at the wrong clusters or not accounting for how telescopes "blur" distant objects.
Our new telescopes (XRISM) and our new, smarter simulations are finally on the same page, helping us understand the invisible "ocean" between galaxies with much greater clarity.

In short: The cosmic ocean is calmer than we feared, and we finally have the right map to navigate it.

1. Problem Statement

The Intracluster Medium (ICM) of galaxy clusters is shaped by gas motions ranging from large-scale bulk flows to small-scale turbulence. These motions act as non-thermal pressure, causing deviations from hydrostatic equilibrium and introducing mass biases in cluster measurements.

While the Hitomi and XRISM missions have provided precise measurements of gas velocity dispersions and turbulent pressure fractions in specific clusters (e.g., Perseus, Coma, Virgo), a significant tension exists between these observations and theoretical simulations:

Observations: XRISM/Hitomi consistently find low turbulent pressure fractions (typically a few percent, e.g., ~4% in Perseus).
Simulations: Previous statistical comparisons often predict higher non-thermal pressure fractions (e.g., median ~8% in TNG-Cluster).

The discrepancy is likely driven by selection effects (simulations often compare to the "average" cluster, whereas observations target specific, often relaxed or specific merger states) and a lack of one-to-one comparisons with specific observed clusters. Furthermore, the role of physical viscosity and the impact of specific evolutionary histories on turbulence remain unclear.

2. Methodology

The authors address this gap using constrained simulations to create direct, one-to-one analogs of observed clusters, eliminating statistical selection biases.

Simulation Framework:
- Dataset: The SLOW (Simulating the LOcal Web) simulations, based on the CLONES simulation set. These are constrained initial conditions that reproduce the local universe's large-scale structure, positions, and masses of observed clusters.
- Target Clusters: Coma, Virgo, and Perseus. These were chosen because they have high-quality constraints in the simulation and have been recently observed by XRISM/Hitomi.
- Code: OpenGadget3, a cosmological SPH/MFM TreePM code.
- Physics: The study uses non-radiative simulations (gravity + hydrodynamics only). Cooling, star formation, and AGN feedback are excluded to isolate the cosmological contributions (mergers, accretion, substructures) to turbulence.
- Resolution: High-resolution zoom-in regions with a background cosmology consistent with Planck 2014.
Analysis Techniques:
- Dynamical State Classification: Used both simulation-based criteria (center-of-mass offset, substructure mass fraction) and observation-based criteria (X-ray surface brightness asymmetry $\alpha$ and profile parameter $\kappa$ ) to classify clusters as "relaxed" or "active."
- Turbulence Measurement:
  - Calculated multi-scale filtered velocities using the vortex-p code to separate bulk motions from turbulence.
  - Applied volume-weighting (rather than mass-weighting) to better match X-ray observations.
  - Excluded cold gas ( $T < 10^5$ K) from calculations.
- Mock Observations:
  - Generated mock X-ray spectra using the Phox simulator (APEC model, $Z=0.3 Z_\odot$ ).
  - Simulated XRISM/Hitomi energy resolution (4 eV) and exposure times to produce synthetic spectra of Fe XXV and Fe XXVI lines.
  - Performed XSPEC fitting to derive line-of-sight velocity dispersions ( $\sigma_z$ ).

3. Key Contributions

First One-to-One Comparison: This study provides the first direct comparison between constrained simulations and XRISM/Hitomi data for specific clusters (Coma, Virgo, Perseus), bypassing the need for statistical averaging.
Validation of Selection Effects: It demonstrates that the low turbulent pressure fractions observed by XRISM are consistent with simulations when specific clusters are matched to their specific dynamical states, rather than comparing to a statistical mean.
Filtering Length Analysis: The paper highlights the critical importance of physical filtering length. It shows that comparing clusters at different distances without accounting for the physical scale of the observation can lead to misinterpreting bulk motions as turbulence, potentially overestimating turbulent velocities.
Mock Spectra Generation: Provides realistic mock X-ray spectra that reproduce the line broadening and shifts observed by XRISM, validating the simulation's ability to reproduce spectral features.

4. Key Results

Dynamical States:
- Coma: The most active cluster, showing traces of a recent merger, large substructures, and two Brightest Cluster Galaxies (BCGs).
- Perseus: Shows a cool core but sloshing motions; classified as "active" but close to relaxed.
- Virgo: Very relaxed, consistent with its regular structure and cool core, despite mild sloshing.
- Note: In the non-radiative simulations, Virgo appeared less relaxed in observation-based metrics (due to missing cooling/feedback effects on the core density), but simulation-based criteria confirmed its relaxed nature.
Turbulent Velocities and Pressure:
- Ranking: Turbulent velocity follows the dynamical state: Coma > Perseus > Virgo.
- Magnitudes:
  - Coma: Filtered turbulent velocities $\approx 600$ km s $^{-1}$ ; Turbulent pressure support $\approx 8\%$ (within central 1–2% $R_{200}$ ).
  - Perseus: Filtered velocities $\approx 500$ km s $^{-1}$ ; Pressure support $\approx 6\%$ .
  - Virgo: Filtered velocities $\approx 100\text{--}200$ km s $^{-1}$ ; Pressure support $\approx 1\%$ .
- Comparison to XRISM: The simulated values are lower than previous statistical simulation averages and align well with XRISM/Hitomi findings. This suggests that the "low turbulence" seen by XRISM is not a failure of simulations but a result of observing specific, often relaxed, clusters.
Spectral Analysis:
- Coma: Mock spectra showed significant line broadening (several 10 eV) and required a two-temperature model to fit, reflecting the complex bimodal velocity structure of the merging system.
- Virgo & Perseus: Showed narrower lines. Simulated $\sigma_z$ values were slightly lower than observations ( $\sim 73$ km s $^{-1}$ for Virgo vs. observed $\sim 262$ km s $^{-1}$ ; $\sim 106$ km s $^{-1}$ for Perseus vs. observed $\sim 144$ km s $^{-1}$ ).
- Interpretation: The slight underestimation in Virgo and Perseus may be due to missing AGN kinetic feedback in the non-radiative simulations, which could inject additional turbulence in the cluster cores.
Distance/Filtering Effect:
- When simulating clusters as if they were at a farther distance (larger physical filtering scale), Coma's turbulent velocity estimate increased significantly (by factor $\gtrsim 2$ ) due to the mixing of bulk and turbulent motions. This confirms that distance-dependent resolution effects can artificially inflate turbulence estimates.

5. Significance and Future Outlook

Resolving the Tension: The study resolves the tension between high-pressure simulations and low-pressure XRISM observations by proving that selection effects are a primary driver. The specific evolutionary history of a cluster dictates its turbulence level.
Plasma Physics Constraints: The agreement between constrained simulations and XRISM data (even without feedback) suggests that the ICM turbulence is largely driven by cosmological accretion and mergers. However, the slight discrepancies in Virgo/Perseus imply that AGN feedback and physical viscosity are necessary to fully reproduce observations.
Methodological Shift: The paper advocates for constrained simulations as the gold standard for comparing with specific astronomical observations, moving away from purely statistical ensemble comparisons.
Future Work: The authors plan to include radiative cooling and feedback in future runs to better model the cluster cores and AGN effects. They also aim to analyze a larger sample of SLOW clusters to establish statistical trends and further constrain plasma properties like viscosity.

In conclusion, this work validates the SLOW simulation framework as a powerful tool for interpreting XRISM data, confirming that the low non-thermal pressure fractions observed are consistent with the specific dynamical states of the observed clusters, while highlighting the need for future simulations to incorporate feedback mechanisms for a complete picture.