🔭 astrophysics

Substructure in redMaPPer clusters and its impact on X-ray morphology and scaling relations

By leveraging DECaLS and eROSITA data, this study reveals that substructure is prevalent in approximately 40% of redMaPPer clusters and significantly drives disturbed X-ray morphologies and enhanced scatter in the luminosity-richness scaling relation, particularly at low redshifts where mergers boost X-ray luminosity and at lower richness where AGN feedback increasingly influences morphology.

Original authors: R. Tuomainen, A. Finoguenov, J. Comparat, L. Doubrawa

Published 2026-04-29

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: R. Tuomainen, A. Finoguenov, J. Comparat, L. Doubrawa

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 universe as a giant, ever-expanding city. In this city, galaxy clusters are the massive metropolitan areas where thousands of galaxies (the "people" of the universe) live together. For a long time, astronomers have tried to use these cities as "rulers" to measure the history and expansion of the universe. To do this, they rely on a simple rule: bigger cities (more galaxies) should be brighter in X-rays (hot gas).

However, this rule isn't perfect. Sometimes, a city looks bright not because it's huge, but because it's in the middle of a chaotic construction project or a massive traffic jam. This paper investigates those "construction projects"—the substructures—and explains why they mess up our measurements.

Here is a breakdown of the paper's findings using everyday analogies:

1. The Problem: The "Moving Truck" Effect

Galaxy clusters aren't static; they are constantly growing by swallowing smaller groups of galaxies. Think of a galaxy cluster as a large family reunion. Sometimes, a smaller family unit (a substructure) arrives late, bringing their own furniture and food.

The Issue: When astronomers look at the cluster, they see the main family plus the late arrivals. This makes the cluster look "richer" (more members) and "brighter" (more hot gas) than it really is if you only counted the original family.
The Consequence: If you don't account for these late arrivals, your "ruler" for measuring the universe becomes blurry. You might think a cluster is bigger or older than it actually is.

2. The Tool: A Smart "Party Guest" Detector

The authors used a new, smart algorithm called HDBSCAN to look at a massive list of galaxy clusters (the redMaPPer catalog).

The Analogy: Imagine you are at a crowded party. Some people are standing in tight, chatty groups, while others are wandering alone. HDBSCAN is like a super-observant host who can instantly spot the tight-knit groups within the crowd, even if they are mixed in with the general noise.
The Result: They found that about 40% of these galaxy clusters have these "late-arriving groups" (substructure) inside them. In fact, for about a quarter of the clusters, these late arrivals make up more than 35% of the total "wealth" (richness) of the cluster.

3. The Connection: Chaos in the Gas

The researchers then compared these optical "party groups" with X-ray images of the same clusters. X-rays show the hot gas that fills the space between galaxies.

The Finding: When a cluster has these substructures (the late arrivals), the hot gas is usually a mess. It's like throwing a rock into a calm pond; the water ripples and swirls.
The Evidence: Clusters with substructure had much "flatter" and more disturbed gas clouds. They weren't the neat, round, calm bubbles you'd expect from a peaceful cluster. The paper found a very strong link: If you see a messy gas cloud, there's almost certainly a substructure crashing the party.

4. The Twist: The "Low-Redshift" Surprise

The most interesting discovery happened when they looked at clusters at different distances (which corresponds to different times in the universe's history).

The Analogy: Think of "low redshift" as looking at the universe's "recent past" (closer to us) and "high redshift" as looking at the "distant past."
The Discovery: At low redshifts (closer to us), clusters with substructure were surprisingly brighter in X-rays than expected.
- Why? The authors suggest that in the recent past, these clusters often have "cool cores" (dense, cold centers) that are very efficient at glowing. When a substructure merges, it might bring another cool core, or the collision might temporarily boost the brightness. It's like two powerful generators crashing together and creating a massive, temporary surge of electricity.
- The Result: This creates a lot of "noise" (scatter) in the data. Some clusters become super bright, while others get disrupted and dim. This makes it very hard to use these clusters as precise rulers for the universe right now.

5. The Conclusion: Cleaning Up the Data

The paper concludes that ignoring these substructures is a major mistake for cosmology.

The Fix: By using their new method to identify and separate these "sub-groups," the researchers were able to clean up the data. They found that once they accounted for the substructure, the relationship between the size of the cluster and its brightness became much clearer and more reliable.
The Takeaway: To get the most precise measurements of the universe's expansion, we must stop treating galaxy clusters as single, smooth objects. We have to recognize that they are often messy, merging families, and we need to count the "late arrivals" separately to get the true picture.

In short: Galaxy clusters are often messy mergers. If you don't spot the "sub-groups" crashing the party, your measurements of the universe's size and history will be off. This paper provides a new, better way to spot those sub-groups and fix the measurements.

1. Problem Statement

Galaxy clusters are vital cosmological probes for understanding the growth of structure in the Universe. However, their utility is limited by significant intrinsic scatter in mass-observable scaling relations (e.g., X-ray luminosity $L_X$ vs. optical richness $\lambda$ ). A primary driver of this scatter is the dynamical state of clusters, specifically the presence of substructure resulting from ongoing mergers and accretion.

Current studies face a systematic bias where cosmological constraints derived from X-ray luminosity differ from those derived from optical richness. The authors hypothesize that this discrepancy is driven by the correlation between optical richness and the large-scale environment, as well as the failure to account for substructure, which disrupts the thermal equilibrium of the intracluster medium (ICM) and alters X-ray properties.

2. Methodology

The study leverages a multiwavelength approach combining optical data from the DESI Legacy Imaging Surveys (via the redMaPPer cluster catalog) and X-ray data from the SRG/eROSITA all-sky survey (specifically the eRASS1 catalog).

A. Substructure Identification (Optical)

Algorithm: The authors applied HDBSCAN (Hierarchical Density-Based Spatial Clustering of Applications with Noise), a density-based clustering algorithm, to the redMaPPer catalog.
Input Data: Only the 2D sky positions of member galaxies were used; color, magnitude, and redshift information were excluded to avoid bias.
Two-Step Procedure: To mitigate the algorithm's inherent bias against identifying single-component clusters (which would classify all relaxed clusters as "noise" or single clusters), a two-step process was implemented:
1. Step 1: HDBSCAN runs with a configuration prohibiting single-component clusters to identify clear substructures.
2. Step 2: For clusters not identified as having substructure in Step 1, the algorithm is re-run with parameters adjusted to allow single-component identification.
Refinement: Galaxies classified as "noise" by HDBSCAN are assigned a membership probability of zero. The cluster richness ( $\lambda$ ) is recalculated, and a new metric, infalling richness ( $\lambda_{infalling}$ ), is defined for the substructure component.
Validation: The method was validated against mock catalogs derived from the Uchuu simulations, achieving a 77.1% correct grouping rate for member galaxies.

B. X-ray Morphology Analysis

Cross-Matching: The refined optical sample was cross-matched with the eRASS1 X-ray cluster catalog (within $R_{500}$ and $\Delta z < 0.02(1+z)$ ).
Parameters: The study utilized the largest set of uniformly derived X-ray morphological parameters to date (from Sanders et al. 2025), including:
- Shape parameters: Ellipticity, centroid shift, fit-peak offset.
- Density parameters: Concentration ( $c_{80-800}$ ), inner density slope ( $\alpha$ ).
- Composite Scores: $D_{shape}$ (shape-based disturbance) and $D_{comb}$ (combining shape and concentration).

C. Scaling Relation Analysis

The authors analyzed the $L_X - \lambda$ scaling relation using Bayesian linear regression (via the linmix routine) to account for measurement errors in both variables.
The analysis was stratified by richness ( $\lambda \ge 20$ and $\lambda \ge 40$ ) and redshift ( $z < 0.2$ vs. $z \ge 0.2$ ) to disentangle evolutionary trends.

3. Key Contributions

Robust Substructure Catalog: Created a refined catalog of 6,421 matched optical/X-ray clusters, identifying 2,159 clusters with substructure and 4,262 without, using a novel two-step HDBSCAN approach that minimizes classification bias.
Quantification of Substructure Properties: Defined and measured the mass ratio ( $\mu = M_{infalling}/M_{host}$ ) and infalling richness, establishing completeness limits ( $\mu > 0.4$ for $\lambda \ge 20$ ; $\mu > 0.25$ for $\lambda \ge 40$ ).
Multiwavelength Correlation: Directly linked optical substructure identification to a comprehensive suite of X-ray morphological parameters, moving beyond simple visual classification.
Systematic Bias Correction: Demonstrated that updating richness estimates based on substructure removal reduces the scatter in the $L_X - \lambda$ relation by ~7%.

4. Key Results

A. Prevalence of Substructure

Substructure is a common feature, present in approximately 40% of the sample.
About 25% of the full sample exhibits a fractional contribution to richness from substructure exceeding 35%.
Detection is strongly dependent on richness; substructure is more common in high-richness systems.

B. Impact on X-ray Morphology

Disturbance Correlation: There is a highly significant correlation between optical substructure and disturbed X-ray morphologies.
Key Parameters: Clusters with substructure exhibit:
- Flatter surface brightness profiles: Significantly lower concentration ( $c_{80-800}$ ).
- Higher centroid shifts ( $\omega$ ): Indicating spatial variability in emission.
- Larger multipole moments ( $M_4$ ): Suggesting pronounced large-scale asymmetries.
Composite Scores: The combined disturbance score $D_{comb}$ (which includes concentration) shows a much stronger distinction between substructure and relaxed clusters than the shape-only score ( $D_{shape}$ ).
Richness Dependence: The morphological differences are most pronounced in high-richness clusters ( $\lambda > 40$ ). In low-richness bins, the distinction is less clear, suggesting Active Galactic Nucleus (AGN) feedback may play a larger role in morphology for lower-mass systems.

C. Impact on Scaling Relations ( $L_X - \lambda$ )

Low-Redshift Evolution ( $z < 0.2$ ):
- Clusters with substructure show a systematically higher X-ray luminosity at fixed richness compared to relaxed systems.
- They exhibit significantly higher intrinsic scatter in the scaling relation.
High-Redshift ( $z \ge 0.2$ ): The differences in normalization and scatter between substructure and relaxed clusters diminish, likely due to the lower prevalence of stable cool cores at higher redshifts.
Physical Interpretation: The excess luminosity and scatter at low redshifts are attributed to the survival of cool cores from infalling subclusters. These cool cores boost X-ray luminosity but are subject to disruption (leading to under-luminous systems) or survival (leading to over-luminous systems), creating a bimodal distribution that increases scatter. The authors suggest magnetic draping may suppress disruptive mixing, allowing these cool cores to survive longer at low redshifts.

5. Significance and Conclusions

Cosmological Precision: The study demonstrates that identifying and characterizing substructure is essential for next-generation cluster cosmology. Ignoring substructure introduces systematic biases that limit the precision of cosmological constraints.
Methodological Viability: The HDBSCAN-based approach using probabilistic membership is a viable and effective method for cleaning cluster catalogs and refining richness estimates.
Physical Insight: The results highlight that the dynamical state of a cluster (specifically the presence of substructure) is a primary driver of the scatter in mass-observable relations, particularly at low redshifts where cool cores are prevalent.
Future Directions: The authors note that future work incorporating spectroscopic redshifts for 3D clustering and larger samples from upcoming surveys will further refine the understanding of substructure evolution and its role in cluster assembly.

In summary, this paper provides a rigorous statistical framework for linking optical substructure to X-ray morphology, proving that substructure is a dominant source of scatter in scaling relations and a critical factor to correct for in precision cosmology.