J-PAS & FLAMINGO: Cosmic voids and void galaxies in the gravitational landscape of photometric surveys

Technical Summary: J-PAS & FLAMINGO: Cosmic voids and void galaxies in the gravitational landscape of photometric surveys

Problem Statement
Photometric surveys, such as the Javalambre Physics of the Accelerating Universe Astrophysical Survey (J-PAS), offer a powerful mechanism for mapping the large-scale structure of the Universe. However, the identification of cosmic voids and the characterization of their resident galaxies are significantly hindered by photometric redshift (photo-z) errors. These errors smear the line-of-sight positions of galaxies, distorting the underlying cosmic web and complicating the extraction of an unbiased inventory of void properties and the environmental effects on galaxy formation. While voids are essential components of the Cosmic Web and serve as fundamental cosmological probes, their detection in photo-z surveys remains a challenge due to the blurring of density fields.

Methodology
The authors employ a two-tier approach using galaxy mocks derived from the FLAMINGO hydrodynamical simulation at redshift $z=0.3$ with an apparent magnitude limit of $m_i < 20$. Two distinct samples are constructed:

1. FBI (FLAMINGO-based Ideal): A reference sample with original, unaltered galaxy positions.
2. JP (FLAMINGO-based J-PAS): A sample where galaxy coordinates are displaced along the line-of-sight to emulate realistic J-PAS photo-z errors, modeled using the TOPz pipeline and J-PAS Early Data Release (EDR2024) statistics.

To mitigate the impact of these redshift errors, the study moves away from direct density field analysis. Instead, it computes a quasi-gravitational potential field derived from the logarithm of the galaxy number density ($\ln(1+\delta)$). This approach acts as a low-pass filter, suppressing small-scale noise while preserving large-scale dynamical features.

The identification of voids is performed using a watershed algorithm applied to this quasi-potential field. Specifically, the algorithm identifies local minima in the negative quasi-potential, which correspond to the expansion centers of dynamically dominant "super-voids." This method isolates voids that are gravitationally repulsive ($\Phi > 0$) and expanding, effectively filtering out smaller sub-voids or those dominated by surrounding overdensities (the "void-in-cloud" phenomenon).

Following void identification, the authors define samples of void core galaxies (residing within spherical regions enclosing 50% of the void's mean density and satisfying a local density threshold) and a comparison sample of galaxies in high-density regions. The study focuses on massive galaxies ($M_* \ge 10^{10} M_\odot$).

Key Contributions and Results
The study evaluates the efficacy of the quasi-potential method by comparing the void populations and galaxy properties in the FBI and JP samples.

• Void Identification and Recovery:

  • The quasi-potential approach successfully identifies dynamically relevant voids in the JP sample despite photo-z errors. While the JP sample shows a slightly lower void abundance and a marginal shift toward larger, less spherical shapes, the overall size and ellipticity distributions show strong correspondence with the FBI sample.
  • Kolmogorov-Smirnov tests indicate no statistically significant difference in the ellipticity distributions between the two mocks.
  • Individual void recovery is robust: approximately 425 voids (out of 1074 in FBI and 817 in JP) are recovered with an Intersection over Union (IoU) $> 0.5$. These recovered voids occupy $\sim 63\%$ of the thresholded quasi-potential volume in the FBI sample and exhibit excellent agreement in size and shape distributions with their FBI counterparts.

• Density Profiles:

  • The primary impact of photo-z errors manifests as a contamination of void interiors. In the JP sample, void density profiles show an artificial increase in central density and a reduction in the contrast of high-density boundaries. This is attributed to galaxies from high-density regions being scattered into the void interiors due to redshift uncertainties.
  • Despite this, the identified voids in both samples predominantly exhibit R-type density profiles (smoothly rising from the center), indicating they are deep, expanding underdensities rather than collapsing "void-in-cloud" systems.

• Void Galaxy Properties:

  • The study confirms that the expected environmental trends are preserved in the JP mock. Compared to galaxies in high-density regions, void core galaxies in both samples exhibit:
    • Lower stellar masses (with a distinct lack of massive galaxies $M_* > 10^{11.5} M_\odot$ in the FBI sample).
    • Bluer colors ($g-r$).
    • Enhanced star formation activity (higher SFR and sSFR).
  • While the JP sample shows a slight contamination of massive galaxies into voids (shifting the Stellar Mass Function at $M_* \gtrsim 10^{11} M_\odot$), the fundamental trend of void galaxies being lower mass, bluer, and more actively star-forming remains detectable.

Significance and Limitations
The paper claims that a quasi-gravitational potential field serves as an effective tool for mitigating photo-z errors of the order expected for the J-PAS survey. By leveraging the smoothing properties of the quasi-potential, the method allows for the robust identification of dynamically dominant underdensities that are less sensitive to small-scale noise and redshift smearing. Consequently, the framework enables the detection of voids and the preservation of massive void galaxy trends in forthcoming photometric survey data.

The authors note specific limitations:

• The study is restricted to massive galaxies ($M_* \ge 10^{10} M_\odot$) due to the resolution limits of the FLAMINGO simulation; the properties of lower-mass void galaxies are not covered.
• The current work is an application of the quasi-potential method to J-PAS mocks rather than a full methodological validation; a dedicated companion paper is planned for quantitative calibration of parameters.
• The analysis does not yet include full survey systematics such as angular masks, selection functions, or redshift-space distortions, which will be addressed in future work applying this method to real data and larger surveys like Euclid.