The Gas-Phase Mass-Metallicity Relation of Dwarf Galaxies Across Large-Scale Environments Using the CAVITY Parent Sample

Source: arXiv:2605.25557 · Published 2026-05-25 · By Bahar Bidaran, Salvador Duarte Puertas, Isabel Pérez, Almudena Zurita, Daniel Espada, María Argudo-Fernández et al.

TL;DR

This study investigates the gas-phase mass-metallicity relation (MZR) in dwarf galaxies (stellar mass 8.9 < log(M*/M☉) < 9.5) across different large-scale cosmic environments: voids, filaments, and clusters, using the CAVITY parent sample and SDSS-DR7 spectroscopic data. The authors fit emission lines with pyPipe3D to derive aperture-corrected gas-phase metallicities and star formation rates (SFR) for 353 void dwarfs, 311 filament dwarfs, and 22 cluster dwarfs. They find that the MZR slope varies systematically with environment, being steepest in voids (0.28±0.03) and flattest in clusters (0.17±0.08), indicating environmental influence on chemical enrichment. They also analyze local environment within voids and filaments, finding isolated dwarfs show similar MZR slopes regardless of large-scale structure, while non-isolated filament dwarfs have flatter MZR slopes comparable to cluster dwarfs.

Key findings

• MZR slope is steepest in voids at 0.28 ± 0.03, intermediate in filaments at 0.25 ± 0.04, and flattest in clusters at 0.16 ± 0.09 (Table 2).
• Average gas-phase metallicity normalization (12+log(O/H)) shifts from 8.45 ± 0.01 in voids and filaments to 8.55 ± 0.02 in clusters, showing cluster dwarfs are more metal-rich at fixed mass.
• Within voids and filaments, isolated dwarfs exhibit steeper MZR slopes (void isolated 0.30 ± 0.04; filament isolated 0.28 ± 0.06) than non-isolated dwarfs (void non-isolated 0.24 ± 0.08; filament non-isolated 0.19 ± 0.07).
• Non-isolated dwarfs in filaments have MZR slopes similar to cluster dwarfs, suggesting local processes within filaments shape chemical enrichment, consistent with pre-processing scenarios.
• Dwarf galaxies in clusters show higher average metallicities, indicating environmental mechanisms (e.g., ram pressure stripping) impacting gas content and enrichment.
• No significant metallicity difference between isolated and non-isolated dwarfs in voids, suggesting local density effects are minor in underdense regions.
• The SFR-M* relation normalization and slope also show environmental dependence, with void isolated dwarfs having highest normalization (-0.57 ± 0.02) and slope (~0.98), while cluster dwarfs demonstrate lower and more uncertain values.
• Aperture corrections and multiple metallicity diagnostics confirm robustness of trends across environment, though results for cluster dwarfs are based on a smaller sample (N=22).

Methodology — deep read

1. Threat Model & Assumptions: The study is observational and astrophysical rather than adversarial, so no traditional threat model applies. The implicit goal is to distinguish whether environment affects dwarf galaxy metallicity and its scaling with stellar mass and SFR. Assumptions include SDSS spectral data quality, sample completeness for the chosen stellar mass range, and validity of emission line metallicity calibrations.

2. Data: The base data come from three samples: 1754 void dwarf galaxies from the CAVITY parent sample (SDSS-DR7 footprint, z=0.005–0.05), 4473 filament dwarfs, and 161 cluster dwarfs from Domínguez-Gómez et al. (2023b). After applying quality cuts (signal-to-noise >3 on key lines, BPT diagram-based AGN removal, morphological inspections), and removing outliers by SFR offset, the final sample is 353 void, 311 filament, and 22 cluster dwarf galaxies. Stellar masses are from the MPA-JHU catalog. Local environment classification (isolated vs non-isolated) uses criteria of no neighbors brighter than Mr ~ -17 mag within 1.5 Mpc projected radius and ±500 km/s line-of-sight velocity.

3. Architecture/Algorithm: Emission line fluxes are measured using the pyPipe3D full spectral fitting pipeline, which fits stellar continuum with SSP models (1272 Vazdekis+ models) before extracting gas emission lines, correcting for Galactic and internal extinction. Metallicity is traced by oxygen abundance 12+log(O/H) estimated primarily via the N2 indicator ([NII]6584/Hα) using Pettini & Pagel (2004) linear calibration. Star formation rates are derived from aperture-corrected Hα luminosities following Kennicutt et al. (2009) with aperture corrections from Duarte Puertas et al. (2017).

4. Training Regime: Not applicable as this is empirical data analysis.

5. Evaluation Protocol: The MZR is modeled by linear fits (metallicity vs log stellar mass) within the dwarf mass range, separately for each environment and local density category. Normalizations and slopes with uncertainties are estimated via regression. SFR-M* relations are similarly fit. Comparisons to previous large SDSS galaxy samples (Duarte Puertas et al. 2017) are presented as baselines. Statistical significance assessed via errors on fit parameters. No mention of cross-validation or simulations of distribution shift.

6. Reproducibility: Data and metallicity catalogues come from SDSS-DR7 public data and established catalogs (MPA-JHU, NASA-Sloan Atlas, CAVITY). The pyPipe3D code is public. The sample selection cuts and criteria are detailed to enable replication. However, no direct code or frozen weights release is mentioned given the observational nature.

Example: For a given void dwarf galaxy with emission line spectra, pyPipe3D fits the continuum and extracts emission line fluxes, corrects for reddening using Balmer decrement. The N2 index is computed from [NII]6584 and Hα lines, converted to metallicity using the Pettini & Pagel relation. The aperture-corrected Hα flux is used to compute SFR. This galaxy's M* and derived metallicity and SFR contribute to linear regression fits to measure the MZR slope and normalization in voids. This process repeats for each environment and subpopulation.

Technical innovations

• Use of a large, carefully curated dwarf galaxy sample spanning voids, filaments, and clusters within SDSS to robustly probe environmental effects on the low-mass MZR, improving upon previous smaller or biased samples.
• Systematic classification of local environment (isolated vs non-isolated) within cosmic web components to disentangle large-scale versus local environmental influences on dwarf galaxy metallicities.
• Application of aperture corrections calibrated from previous SDSS analyses to derive more accurate global metallicities and SFRs from SDSS fiber spectra in low-mass galaxies.
• Demonstration that the MZR slope variation with environment is driven primarily by local environment effects within filaments, consistent with galaxy pre-processing, a nuance not fully established previously.

Datasets

• CAVITY Parent Sample — ~4866 nearby void galaxies (subset used: 1754 dwarfs 8.9 < log(M*) < 9.5) — public SDSS-DR7 footprint
• Domínguez-Gómez et al. (2023b) Filament catalogue — 15000 galaxies (subset dwarfs: 4473) — SDSS-DR7
• Domínguez-Gómez et al. (2023b) Cluster catalogue — 6189 galaxies (subset dwarfs: 161) — SDSS-DR7
• MPA-JHU stellar mass catalog — mass estimates for SDSS galaxies

Baselines vs proposed

• Duarte Puertas et al. (2017) SDSS large galaxy sample: MZR slope ~0.26, normalization ~8.44 (N2 method) vs combined dwarf sample slope 0.26 ± 0.03, normalization 8.45 ± 0.01
• Void dwarfs (isolated): MZR slope = 0.30 ± 0.04 vs non-isolated void dwarfs: 0.24 ± 0.08
• Filament dwarfs (isolated): slope = 0.28 ± 0.06 vs filament non-isolated dwarfs: 0.19 ± 0.07
• Cluster dwarfs MZR slope flatter at 0.16 ± 0.09 and metallicity normalized higher at 8.55 ± 0.02 vs void and filament dwarfs ~8.45
• SFR-M* slope ~0.98 ± 0.09 for void isolated dwarfs vs ~0.85 ± 0.30 for cluster dwarfs

Limitations

• Small sample size for cluster dwarf galaxies (N=22) limits robustness of results in densest environments and increases uncertainty in MZR slope and normalization estimates.
• Reliance on strong-line metallicity calibrations (N2, O3N2, R23) without direct method [OIII]4363 detections can introduce systematic uncertainties in metallicity estimates.
• SDSS fiber aperture covers only central ~0.3-1.6 kpc of galaxies; although aperture corrections are applied, they may not perfectly recover global metallicities and SFRs, especially in dwarfs with irregular morphologies.
• Environmental classification depends on redshift-space projections and velocity cuts, which may misclassify some interacting or satellite galaxies, especially in filaments.
• The study is observational and cross-sectional, unable to directly probe causality or time evolution of environmental effects on chemical enrichment.
• No explicit accounting for potential biases from galaxy inclination, internal extinction variations, or AGN/shock contamination beyond BPT cuts.

Open questions / follow-ons

• What physical mechanisms within filaments cause the flattening of the MZR slope in non-isolated dwarf galaxies—ram pressure, tidal pre-processing, or gas starvation—and how do these operate quantitatively?
• How does the chemical enrichment history and metal retention efficiency of dwarf galaxies evolve over cosmic time across different large-scale environments?
• Can future integral field surveys with deeper spectroscopy confirm these trends and measure spatial metallicity gradients in dwarfs to link internal processes to environmental impact?
• What role do gas inflows and outflows play in driving the environment-dependent scatter in the MZR specifically in low-mass galaxies?

Why it matters for bot defense

For bot-defense engineers and CAPTCHA practitioners, this paper offers an astrophysical analog of how environment and local context shape behavior, here of dwarf galaxies' metallicity scaling. While not directly related to security, the methodology exemplifies rigorous multi-scale data analysis to isolate subtle environmental effects amid potentially confounding factors. By analogy, bot detection systems could draw inspiration from disentangling local vs global context influences on behavior metrics similarly. Understanding heterogeneity and environment-driven scatter within populations can aid in modeling diverse threat profiles or user characteristics. No direct CAPTCHAs or bot-defense techniques arise, but the study is a case study in environmental dependency of modeled relations across scales.

Cite

@article{arxiv2605_25557,
  title={ The Gas-Phase Mass-Metallicity Relation of Dwarf Galaxies Across Large-Scale Environments Using the CAVITY Parent Sample },
  author={ Bahar Bidaran and Salvador Duarte Puertas and Isabel Pérez and Almudena Zurita and Daniel Espada and María Argudo-Fernández and Rubén García-Benito and Laura Sánchez-Menguiano and Simon Verley and Sebastián F. Sánchez and Jesús Falcón-Barroso and Anna Ferré-Mateu and Pedro Villalba-Gonzalez and Andoni Jiménez and Reynier F. Peletier and Tomás Ruiz-Lara },
  journal={arXiv preprint arXiv:2605.25557},
  year={ 2026 },
  url={https://arxiv.org/abs/2605.25557}
}

Read the full paper

• arXiv:2605.25557 (abstract)
• arXiv:2605.25557 (PDF)