🔭 astrophysics

The size of the HI disk across different environments: isolated, compact groups, clusters, and pairs

This study reveals that galaxies in Hickson Compact Groups exhibit the most severe truncation of their neutral hydrogen (HI) disks relative to their optical sizes—reaching at least 71% smaller than expected for isolated galaxies—placing them at the extreme end of environmental impact alongside Virgo cluster members, with truncation severity increasing monotonically along the group's evolutionary sequence.

Original authors: R. Ianjamasimanana, L. Verdes-Montenegro, K. M. Hess, P. Kamphuis, M. G. Jones, J. Garrido, S. H. A. Rajohnson, A. Sorgho, B. Namumba, S. Sánchez-Expósito, M. Korsaga

Published 2026-07-21

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: R. Ianjamasimanana, L. Verdes-Montenegro, K. M. Hess, P. Kamphuis, M. G. Jones, J. Garrido, S. H. A. Rajohnson, A. Sorgho, B. Namumba, S. Sánchez-Expósito, M. Korsaga

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 vast, cosmic neighborhood where galaxies are like houses. Some houses stand alone in the middle of a quiet, empty field, while others are packed tightly together in a bustling, crowded city. In astronomy, scientists have long known that these "neighborhoods" change how a house looks. A house in a crowded city might get its paint stripped off by the wind, or its garden trampled by neighbors, while a house in the middle of nowhere stays exactly as it was built.

To understand this, astronomers look at two specific parts of a galaxy: its "optical disk" (the bright, star-filled center you can see with a telescope, like the main living room) and its "HI disk" (a giant, invisible halo of cold hydrogen gas that surrounds the stars, like a massive, fluffy cloud of fog extending far beyond the walls). Usually, this gas cloud is much bigger than the starry center. But in crowded neighborhoods, like galaxy clusters or tight groups, the gas gets stripped away, leaving the galaxy looking smaller and more "bald" than it should be. The big question scientists have been asking is: just how much does the environment shrink this gas cloud, and does the type of neighborhood matter?

This paper takes a deep dive into that question by comparing two very different types of galactic neighborhoods. On one side, they look at "Hickson Compact Groups" (HCGs), which are like tiny, super-dense apartment complexes where four to ten galaxies are squeezed so close together that they are constantly bumping into each other. On the other side, they study the "AMIGA" sample, which are galaxies that are so isolated they are practically alone in the middle of a cosmic desert, with no neighbors to bother them for billions of years. The researchers wanted to see if the "apartment dwellers" have lost their gas clouds compared to the "desert hermits."

The team developed a clever new way to measure this. Instead of just taking a simple ratio of the gas size to the star size (which can be misleading because galaxies come in different shapes and sizes), they built a "baseline" using the isolated galaxies. Think of it like measuring the height of a tree: if you know how tall a tree should be based on the size of its trunk, you can tell if it's been stunted by poor soil. They used the isolated galaxies to define what a "normal" gas cloud size looks like for a given star size. Then, they measured the compact group galaxies against this standard.

The results are striking. The galaxies in the crowded compact groups have lost a massive amount of their gas. On average, their hydrogen disks are at least 71% smaller than what you would expect for an isolated galaxy of the same star size. In fact, about 90% of the galaxies in these compact groups fall below the "normal" line, meaning their gas clouds have been severely trimmed. The study also found that this gas stripping gets worse as the group evolves. In the early stages of a group's life, the gas is still mostly there, but as the galaxies interact more over time, the gas gets stripped away until, in the most evolved groups, the galaxies are almost completely bare.

When the researchers compared these compact groups to other environments, like loose groups of galaxies or massive galaxy clusters (like the famous Virgo Cluster), they found that the compact groups are at the extreme end of the spectrum. They are just as "bald" as the galaxies in the densest clusters. Interestingly, the study also looked at pairs of galaxies that are close together but not in a tight group. Unlike the crowded compact groups, these pairs actually kept their gas, and some even seemed to have more gas than the isolated ones, suggesting that simply being near another galaxy isn't enough to strip the gas; you need the intense, chaotic environment of a compact group or a massive cluster to do the job.

The paper concludes that the environment is a powerful sculptor. While the stars in the center of these galaxies remain largely untouched, the outer gas clouds are incredibly sensitive to their surroundings. In the tightest, most crowded cosmic neighborhoods, the gas is stripped away efficiently and early, leaving galaxies with tiny, truncated gas disks compared to their lonely, isolated cousins who get to keep their fluffy, extended gas clouds intact.

Technical Summary: The Size of the HI Disk Across Different Environments

Problem Statement
The evolution of galaxies is heavily influenced by their environment, particularly in dense regions like clusters and compact groups where interactions and mergers alter gas content and structure. While the relationship between a galaxy's stellar mass and its atomic hydrogen (H i) mass is known to be tight across environments, the relative extent of the H i disk compared to the optical disk ( $D_{HI}$ vs. $D_{25}$ ) is a sensitive tracer of environmental processing. Previous studies have utilized the ratio $D_{HI}/D_{25}$ to quantify gas removal, but this metric carries a systematic size-dependent bias because the relationship between H i and optical diameters is non-linear. Furthermore, systematic comparisons between the extreme ends of the environmental spectrum—specifically, the most isolated galaxies versus those in the densest compact groups—have been scarce. This paper addresses the need for a bias-free quantification of how the compact-group environment truncates H i disks relative to an equivalent isolated galaxy baseline.

Methodology
The authors utilize two primary samples:

AMIGA (Analysis of the interstellar Medium in Isolated GAlaxies): A sample of 791 strictly isolated galaxies, serving as the control group for secular evolution with minimal external influence. The study focuses on 35 galaxies with resolved H i data and extends the sample to 407 by including single-dish measurements where H i diameters are inferred via a size-mass relation.
Hickson Compact Groups (HCGs): A sample of 30 groups (124 members) covering the full evolutionary sequence of HCGs (Phases 1 through 3c), representing the densest local environments.

Key Technical Approaches:

H i Diameter Measurement: $D_{HI}$ is defined as the diameter at the $1 M_{\odot} \text{pc}^{-2}$ iso-density contour. The authors fit ellipses directly to the contour vertices extracted from column density maps using a numerically stable direct least-squares method (Halir & Flusser 1998), rather than fitting Gaussians to the surface density distribution. Beam effects are corrected by subtracting the beam area in quadrature.
Baseline Definition: Instead of using the biased $D_{HI}/D_{25}$ ratio, the authors establish a Bayesian linear regression baseline for the $\log(D_{HI})$ – $\log(D_{25})$ relation using the AMIGA sample. They tested multiple fitting estimators (OLS, Bisector, ODR, etc.) and selected the Bayesian emcee fit because it showed no residual trend with optical diameter, ensuring the baseline is free of size-dependent bias.
Truncation Quantification: Truncation is measured as the residual ( $\Delta \log D_{HI}$ ) between the observed H i diameter and the diameter predicted by the AMIGA baseline for a given $D_{25}$ . Negative residuals indicate truncation.
Statistical Analysis: The study employs survival analysis (Kaplan–Meier estimators and Gehan generalized Wilcoxon tests) to handle H i non-detections in HCGs, treating them as upper limits. Non-parametric tests (Mann–Whitney U, Kolmogorov–Smirnov) are used to compare residual distributions.
Reproducibility: The analysis is distributed as a fully reproducible Python package and Snakemake workflow, adhering to FAIR principles.

Key Results

AMIGA Baseline: The isolated galaxy sample follows a sub-linear scaling relation: $D_{HI} \propto D_{25}^{0.70}$ . This confirms that a fixed ratio is inappropriate and that the AMIGA sample provides a distinct, shallower baseline compared to H i-selected compilations (which often show near-linear relations).
HCG Truncation: HCG galaxies lie systematically below the AMIGA baseline. When including non-detections as upper limits, the Kaplan–Meier median indicates that HCG H i disks are at least 71% smaller than expected for isolated galaxies of the same optical diameter. Approximately 90% of HCG members fall below the baseline.
Evolutionary Sequence: H i truncation increases monotonically with the HCG evolutionary phase:
- Phase 1: Already shows significant truncation (median deficit ~23%), suggesting outer gas removal begins early.
- Phase 2: Shows the largest spread in truncation, indicating a continuous stripping process.
- Phase 3a/3c: The most evolved phases show the strongest truncation. Phase 3c detections (where one member retains gas) are statistically indistinguishable from isolated galaxies, suggesting these are recently accreted systems, while the overall group population is severely truncated.
Environmental Comparison: When compared to literature samples (Virgo cluster/VIVA, Hydra I, Ursa Major, and galaxy pairs), HCGs occupy the most-truncated end of the environmental spectrum. They are statistically indistinguishable from the Virgo cluster (VIVA) sample, which also exhibits severe truncation. In contrast, galaxy pairs (Bok et al. 2020) show no measurable truncation and are statistically distinct from all other samples, likely due to selection biases favoring gas-rich systems.

Significance and Claims
The paper claims to provide the first systematic, bias-free comparison of H i disk sizes between the most isolated and most compact environments. By moving away from the $D_{HI}/D_{25}$ ratio to a residual-based approach, the study demonstrates that the compact-group environment is as effective at truncating H i disks as the dense cores of galaxy clusters. The results indicate that environmental processing in HCGs is efficient enough to strip low-density outer gas early in the group's evolution (Phase 1), with truncation becoming more severe as the group evolves. The study concludes that HCGs and the Virgo cluster represent the most extreme environments for H i disk truncation, while isolated galaxies (AMIGA) and small pairs represent the least truncated populations. The authors emphasize that their framework allows for the rigorous quantification of these environmental effects without the systematic errors inherent in previous ratio-based metrics.

More like this