A New Sample of Closely Separated Dual and Binary AGN Candidates Revealed with the Radio Fundamental Catalog

The Radio Fundamental Catalog (RFC) has been used to identify and characterize a new sample of closely separated dual and binary Active Galactic Nuclei (AGN) candidates, providing crucial insights into parsec-scale SMBH pairing and the astrometric precision challenges facing the International Celestial Reference Frame.

How it works

The research combines the Big Multi-AGN Catalog (BigMAC), which is a literature-complete compilation of confirmed and candidate multi-AGN systems, with archival data from the Radio Fundamental Catalog (RFC). This cross-matching process yielded 241 candidate and confirmed dual, binary, and recoiling AGN with available RFC images. The primary goal was to present the first detailed analysis of ten systems exhibiting multiple compact components to assess whether each system is more consistent with dual/binary AGN or intrinsic jet structure.

The analysis involved several steps:

1. Identifying sources in the BigMAC-RFC sample with two (or more) cores in the RFC image, noting that Ten of the 241 BigMAC-RFC targets (∼4%) were identified as possible multi-AGN, based on the presence of two (or more) cores in the RFC image.

2. Measuring radio parameters such as flux densities, peak flux densities, and calculating compactness and brightness temperature to distinguish between AGN activity and star formation.

3. Analyzing radio spectral shape by fitting standard power law or curved power law models to determine if synchrotron emission is present, which helps distinguish non-thermal AGN from other emission mechanisms.

4. Measuring angular separations from the highest available resolution observations, calculating projected physical separations for the BigMAC-RFC sources.

Radio Parameters and Morphology Assessment

The ten systems were categorized based on their radio morphology and derived parameters:

-58% were found to exhibit extended emission, in many cases significant enough to be attributable to jet activity (e.g., radio lobes, collimated jet emission, etc.).

-38% of the BigMAC-RFC targets were identified as point sources, in which the RFC source appears compact and Gaussian in nature.

Ten sources were identified as possible multi-AGN based on two or more cores. The analysis utilized compactness (C) and brightness temperature (Tb) to evaluate these candidates. For instance, for J0216-0105, the North component had a compactness of 1.57 and a brightness temperature well in excess of the 105 K limit for star formation, while the South component had a compactness of 1.09 and a brightness temperature exceeding this limit as well.

Classification into Dual AGN vs. Jet Activity

The radio parameters were used to tentatively assign designations:

-Six systems are most consistent with candidate dual or binary AGN.

-Four remaining systems are more plausibly explained by compact jet activity.

The analysis considered various setups, including two peaks, each with C ∼ 1 (indicative of dual AGN), "two peaks, one with C ∼ 1, one with C > 2 (indicative of jet activity), or multiple peaks with C > 2." For example, J0843+4537 was classified as likely jet activity because its hotspot component had a steeper spectral index and a higher compactness value.

Projected Separations and Scale Discrepancies

A key finding is the disparity between initial selection criteria and VLBI measurements:

-Several candidate multi-AGN systems exhibit projected separations of only a few to tens of parsecs.

-In many cases, the parsec-scale radio separations are orders of magnitude smaller than the separations inferred from the observations that originally motivated their selection, emphasizing the importance of VLBI followup.

For example, J1305-1033 exhibits a projected separation of approximately 6.3 pc and if confirmed, would become the smallest known binary AGN.

Implications for Astrometry and ICRF Realization

The study highlights that all ten analyzed systems are ICRF sources exhibiting either intrinsic source multiplicity or significant compact radio structure capable of introducing astrometric centroid shifts.

-It is recommended that all ten BigMAC-RFC sources analyzed here be removed from future iterations of the ICRF.

This work establishes the RFC as an archival resource for characterizing the closest SMBH pairs while improving future ICRF realizations, and demonstrates that compact radio structures can introduce apparent position shifts which degrade the stability of the reference frame as astrometric precision approaches the microarcsecond regime.

Future Work

The paper concludes by emphasizing that confirmation requires dedicated follow-up observations capable of demonstrating that detected radio components remain compact at higher frequencies. Furthermore, it suggests that combining VLBI with high-resolution optical, infrared, and X-ray observations is necessary to fully characterize the BigMAC sample and to address the radio-optical offsets observed in literature.

This research provides valuable data for improving various AI systems, particularly those involved in astrophysics, astronomical data processing, and modeling of black hole dynamics.

Here are specific improvements and capabilities for enhanced AI systems based on this paper:

)1. Enhanced Multi-Scale Source Characterization and Classification (for Image Processing & CNNs):

The paper provides a detailed framework for distinguishing between compact AGN cores, extended jet structures, and dual/binary AGN based on multi-frequency radio morphology, compactness parameters (C), and spectral indices.

• AI System Improvement: Develop Convolutional Neural Networks (CNNs) or Graph Neural Networks (GNNs) trained on the VLBA images and derived radio parameters.

• Improved Capability: The AI can automatically classify a new radio source image not just as point source or extended, but with high confidence, distinguishing between single compact cores, dual cores with distinct brightness temperatures (suggesting dual AGN), and sources dominated by jet activity (high C values and steep spectral indices). This moves beyond simple morphology to physical state classification.

)2. Improved Astrometric Error Modeling for ICRF Realizations (for Reference Frame AI):

The paper explicitly notes that intrinsic source multiplicity and compact radio structure introduce astrometric centroid shifts, making these sources critical contaminants for the International Celestial Reference Frame (ICRF).

• AI System Improvement: Create a machine learning model that predicts the expected positional error/bias for a given source based on its radio properties (e.g., compactness, spectral index) and its known multi-scale structure.

• Improved Capability: This AI can be integrated into astrometric pipelines to proactively flag sources with high predicted centroid jitter or systematic errors, allowing researchers to prioritize them for exclusion from the ICRF realization process before they introduce bias into the reference frame.

)3. Automated Selection of Parsec-Scale Candidates (for Literature Mining & Catalog Searching):

The study demonstrates that current selection techniques (like those used in BigMAC) may miss compact radio structures on parsec scales, as measured by VLBI.

• AI System Improvement: Implement a deep learning system capable of analyzing multiwavelength catalogs (optical, IR, radio) and cross-matching them with archival VLBI data (like RFC).

• Improved Capability: The AI can search for hidden or compact binary/dual AGN systems that were missed by traditional optical selection methods because their radio cores are too compact to be resolved by lower-resolution instruments. It can specifically flag targets where the VLBI separation is orders of magnitude smaller than the separation inferred from initial literature, highlighting potential overlooked binary systems.

)4. Automated Spectral Modeling and Physical Mechanism Identification (for Data Analysis Pipelines):

The paper details how spectral fitting (standard vs. curved power law) helps distinguish between synchrotron emission and potential absorption mechanisms (SSA or free-free absorption).

• AI System Improvement: Build a supervised learning model that takes multi-frequency radio flux data as input and outputs the most probable physical emission mechanism (e.g., standard synchrotron, curved power law indicating SSA, or jet activity signature).

• Improved Capability: This system can rapidly screen large datasets of radio spectra to identify sources where the spectral shape is indicative of either pure AGN accretion or jet dominance, significantly speeding up the vetting process for candidate systems.

)5. Cross-Frame Astrometric Consistency Checking (for Radio-Optical Offset Analysis):

The study highlights discrepancies between separations measured by VLBI and those inferred from radio-optical offsets (e.g., Orosz & Frey 2013).

• AI System Improvement: Develop a system that takes source positions across different reference frames (ICRF, Gaia DR3) and compares them against the measured physical separation derived from VLBI imaging.

• Improved Capability: This AI can quantify the nature of radio-optical offsets—determining whether they are likely due to intrinsic source multiplicity (as suggested by compact radio cores) or larger-scale jet emission. This provides a quantitative tool for interpreting observational discrepancies in AGN studies.