1FLAT: A <tt>Firmamento</tt> -based Catalog of Active Galactic Nuclei in Fermi-LAT High-Galactic-latitude γ -Ray Sources

P. Giommi, M. Doro, M. Gouvêa, L. Fronte, F. Metruccio, F. Arneodo, U. Barres de Almeida, Simonetta Di Pippo, Tobias Kerscher, Andrea V. Macciò, B. Mazzon, M. Morrone, E. Prandini, Á. Rodríguez, Arshia Ruina, N. Sahakyan, Leandro Silveri, Durgesh TripathiThe Astrophysical Journal Supplement Series 2025 · DOI

TL;DR

Astronomers have created a new catalog, 1FLAT, by re-evaluating over 5,000 high-energy gamma-ray sources detected by the Fermi-LAT telescope. This independent analysis, using a novel web-based platform called Firmamento, has identified 421 new blazar counterparts, significantly reducing the number of mysterious gamma-ray sources and providing valuable data for future research.

What Problem Does This Paper Address?

Active Galactic Nuclei (AGN), particularly blazars, are powerful cosmic objects that emit radiation across the entire electromagnetic spectrum, making them prominent sources in gamma-ray surveys. When their relativistic jets are pointed towards Earth, they become even brighter and are classified as blazars. Understanding these blazars is crucial for comprehending high-energy astrophysical processes. However, current gamma-ray catalogs, like the Fermi-LAT's 4FGL, often have a significant fraction of "unassociated" sources – meaning their counterparts at other wavelengths (radio, X-ray, optical) are not clearly identified. This lack of identification hinders our ability to study these objects and complete our census of the gamma-ray sky. The problem is to independently verify and improve these associations, especially for sources at high galactic latitudes, and to provide a more complete catalog of blazar counterparts.

Methods

The core of this work is the use of Firmamento, a web-based platform designed for discovering and analyzing multi-frequency astrophysical sources. Firmamento employs a two-step process:

  1. Automated Search and Initial Candidate Identification: The platform uses an enhanced version of the Error Region Counterpart Identifier (ERCI) tool. ERCI retrieves multi-frequency survey data and other relevant information for a given gamma-ray source's error region. It then uses sophisticated data-handling capabilities, including machine learning models, to identify potential counterparts. This process is primarily spectromorphological, meaning it considers both the shape of the source's spectral energy distribution (SED) and its spatial morphology.
  2. Human Supervision and Refinement: A crucial aspect of Firmamento is the integration of human oversight. The platform provides tools for users to visually inspect SEDs, source morphology, and flux variability. This manual validation is used to confirm or reject candidate associations, and to refine classifications. For blazar candidates, Firmamento also uses specialized tools like BLAST and W-Peak to estimate the synchrotron peak frequency and flux.

This approach contrasts with traditional catalog-based methods that might rely more heavily on spatial coincidence and statistical probabilities.

Datasets

The study utilizes data from a vast array of publicly accessible catalogs and surveys, totaling 90 across various wavelengths (Radio/Microwave, Infrared, Optical/Ultraviolet, X-rays, Gamma-rays). Key datasets include:

  • Fermi-LAT 4FGL-DR4 catalog: The primary catalog of gamma-ray sources being investigated.
  • 4LAC-DR3 catalog: A dedicated catalog of blazars from Fermi-LAT data, used for comparison.
  • Numerous multi-wavelength survey data: These include radio surveys (e.g., FIRST, NVSS), infrared surveys (e.g., WISE, 2MASS), optical surveys (e.g., SDSS, PanSTARRS), and X-ray surveys (e.g., XMM-Newton, eROSITA).

The paper does not introduce a new benchmark dataset but rather uses existing data to create a new, independently validated catalog (1FLAT).

Key Insights and Ideas

  • The power of human-in-the-loop analysis: While automated tools are essential, the paper highlights that human supervision of SEDs and multi-wavelength data significantly improves the reliability of source associations, especially for complex or ambiguous cases.
  • Firmamento as a versatile research and education tool: The platform's design, with its intuitive interface and integrated tools, not only facilitates cutting-edge research but also actively engages students, providing them with hands-on experience in astrophysical data analysis.
  • High-Galactic latitude sources are rich in blazars: The focus on high-Galactic latitude sources, often considered more challenging to associate, has yielded a substantial number of new blazar identifications, suggesting these regions are not just "empty" space but contain a significant population of active emitters.
  • Discrepancies reveal data limitations: Differences between the Firmamento catalog and previous catalogs often point to areas where improved or more recent data (especially X-ray) can significantly alter source associations.

Key Results

  • High agreement with existing catalogs: Firmamento confirmed associations for 88.8% of the 5,062 high-Galactic latitude sources from the 4FGL catalog.
  • Significant reduction in unassociated sources: The study identified 421 new blazar counterparts among previously unassociated sources, reducing the fraction of unidentified extragalactic Fermi-LAT sources from 25% to 17%.
  • Alternative associations and non-confirmations: In 64 cases, alternative blazar associations were found, and in 49 instances, the 4FGL-DR4 association was not confirmed.
  • Homogeneous SED parameter estimates: The catalog provides consistent estimates of synchrotron peak frequency and flux for confirmed blazar counterparts, using machine learning and template-based methods.
  • Publicly available catalog: The primary outcome is the First Firmamento LAT AGN Table (1FLAT), made publicly available through the Firmamento platform, including all related multi-wavelength data and images.

Limitations and Future Work

  • Challenges with sparse or poor-quality data: Firmamento's automated identification can be challenging in crowded fields or when multi-wavelength data is sparse or of low quality.
  • Subjectivity in visual inspection: While valuable, visual inspection of SEDs can introduce a degree of subjectivity.
  • Potential for biased representation: The method might be more sensitive to blazars with specific multi-wavelength properties, potentially leading to a biased representation of the overall unassociated source population.
  • Future work: The authors suggest incorporating quantitative variability measures, developing more sophisticated machine learning for SED classification, and leveraging upcoming data releases (like eROSITA X-ray survey data) to further validate and expand the catalog. A comprehensive redshift study is also deferred to future work.

Summary updated

Abstract

Abstract We present a systematic reassessment of 5062 high-Galactic-latitude γ -ray sources from the Fermi-Large Area Telescope (LAT) 4FGL-DR4 catalog using Firmamento , a web-based platform for multifrequency source discovery and analysis. Our goal is to provide an independent evaluation of LAT γ -ray source associations using alternative spectral and spatial methods that integrate both recent and legacy survey data. The evaluation is further refined through human supervision of spectral energy distributions, source morphology, flux variability, and template-based comparisons. Firmamento confirms the 4FGL-DR4 and 4LAC-DR3 counterparts or unassociated sources in 4493 cases (88.8%), demonstrating the robustness of both approaches. Beyond this general agreement, we identify 421 new blazar counterparts among previously unassociated sources, thereby reducing the fraction of unidentified extragalactic Fermi-LAT sources from 25% to 17%. In addition, in 64 cases, we find alternative blazar associations, while in 49 instances, we do not confirm the 4FGL-DR4 association. For all confirmed blazar counterparts, we provide homogeneous estimates of synchrotron peak frequency and peak flux using machine learning and template-based methods. The results agree with 4LAC-DR3 values in most cases, though significant discrepancies appear for a few dozen sources, often due to improved X-ray coverage. The primary outcome of this work is the First Firmamento LAT AGN Table, made publicly available through the Firmamento platform ( https://firmamento.nyuad.nyu.edu ), where all related multiwavelength data and images are available. The project involved extensive manual validation and benefited from the active participation of graduate and undergraduate students, highlighting the platform’s value for both research and education.
BibTeX
@article{giommi2025flat,
  author = {P. Giommi and M. Doro and M. Gouvêa and L. Fronte and F. Metruccio and F. Arneodo and U. Barres de Almeida and Simonetta Di Pippo and Tobias Kerscher and Andrea V. Macciò and B. Mazzon and M. Morrone and E. Prandini and Á. Rodríguez and Arshia Ruina and N. Sahakyan and Leandro Silveri and Durgesh Tripathi},
  title = {{1FLAT: A <tt>Firmamento</tt> -based Catalog of Active Galactic Nuclei in Fermi-LAT High-Galactic-latitude <i>γ</i> -Ray Sources}},
  journal = {The Astrophysical Journal Supplement Series},
  year = {2025},
  doi = {10.3847/1538-4365/ae146e},
}
Figure