Infrared-Selected Active Galactic Nuclei in the Kepler Fields

arXiv:2505.02253 · astro-ph.GA, astro-ph.CO · Submitted 2025-05-04 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Astrophysics Radio. Generated commentary on the latest astrophysics papers.

Vera: Today's paper: "Infrared-Selected Active Galactic Nuclei in the Kepler Fields".

Jocelyn: The gist: This study identifies and characterizes Active Galactic Nuclei (AGN) candidates monitored by the Kepler missions using infrared selection techniques and subsequent ground-based spectroscopic follow-up,

Vera: First, who's behind it and why it matters.

Paper summary: Vera: So we're looking at this paper called "Infrared-Selected Active Galactic Nuclei in the Kepler Fields". Basically, they used infrared selection techniques combined with follow-up ground-based spectroscopy to find and characterize AGN candidates that were monitored by the Kepler missions.

Jocelyn: Right, so the main point here is confirming what those infrared selections actually turned out to be—whether they were real active galactic nuclei or just random objects. It’s about testing the selection methods used on Kepler prime and K2.

Subrahmanyan: From a theoretical side, this work is important because it validates a way to find these types of objects that might otherwise be missed by purely optical surveys, which is crucial for understanding the population of black holes in galaxies.

Vera: They started by combining known AGN catalogs like the V´eron-Cetty and V´eron catalog and the SDSS survey with other methods like Edelson and Malkan and Stern et al. to create a preliminary list of candidates before even looking at Kepler data.

Jocelyn: Then they did these miscellaneous checks on those targets before proposing them for Kepler observations, which sounds like a careful filtering process to reduce noise.

Subrahmanyan: And the real test, as you read in page three was the follow-up spectroscopy at Lick, Palomar, and Keck to confirm whether these candidates were actual AGN > <ref:2505.02253#pg1>

Vera: They ended up with Kepler observing a total of one thousand one hundred fifty-five AGN and candidates proposed by their group after all that testing > <ref:2505.02253#pg3,Kepler observing a total of 1155 AGN and>

Jocelyn: That brings us to the ground-based observations, where they confirmed one hundred eighty-six of those candidates using the Kast spectrograph on Lick, the Double Spectrograph on Palomar, and LRIS on Keck > <ref:2505.02253#pg3,candidates using the Kast spectrograph on>

Subrahmanyan: The observational data spanned from two thousand eleven to two thousand sixteen with integration times ranging from three hundred to one thousand eight hundred seconds, which gives them a good amount of data for studying the light and emission lines > <ref:2505.02253#pg3,spanned from 2011 to 2016>

Vera: They also talked about how they reduced their images by subtracting bias level and flat-fielding using a uniformly illuminated dome ceiling to get clean data >

Jocelyn: And for calibration, they used Ne arcs for red spectra and a mix of Hg, He, Ar, and Cd arcs for blue spectra to make sure their wavelength measurements were accurate >

Subrahmanyan: Flux calibration was done using the spectroscopy of stars like Feige thirty-four and BD+twenty-five three thousand nine hundred forty-one to establish a flux scale for all their measurements >

Vera: Now, looking at the reliability of those selection techniques, they tested the Edelson and Malkan sample and found that forty-nine AGN candidates from the W2R list were spectroscopically identified as Type one AGN and three as Type two AGN, which gave them a reliability of ninety-six percent > <ref:2505.02253#pg1>

Jocelyn: And for the WISE color cut from Stern et al. that they used, they found a reliability of ninety percent in identifying those candidates >

Subrahmanyan: The full Kepler AGN sample showed that out of one hundred eighty-six objects, one hundred forty were AGN, which means seventy-five percent were AGN, and twenty-five percent of the AGN were Type two >

Vera: That suggests the main identification techniques they used actually yielded significantly higher reliability than some other methods they tested >

Jocelyn: When you look at the sample properties themselves, they examined these through the Baldwin Phillips Terlevich diagram for Type one AGN and Spectral Energy Distributions for Type one AGN > <ref:2505.02253#pg1>

Subrahmanyan: In the BPT diagram, four out of sixteen galaxies turned out to be composite galaxies with prominent narrow-line emissions from HII regions, while the rest of the galaxies above the Ke01 line are in the AGN region >

Vera: They also looked at power-law relationships in their SEDs showing a bump around three µm or log(νrest) = fourteen and another starting around fourteen point seven five >

Jocelyn: And they determined the Balmer decrements for Hα/Hβ were between four and four point five, which gives them a measure of the broad line region size >

Subrahmanyan: The conclusion of this paper is that the overall line and continuum properties of these newly discovered AGN do not differ markedly from those of previous samples >

Vera: They concluded that this sample yielded the same AGN population as those found using other methods based on optical spectroscopy and color selection >

Jocelyn: So, in short, they’ve shown that selecting AGN via infrared colors gives the same population and properties as what you'd find using traditional optical spectroscopy >

Subrahmanyan: This confirms consistent trends with existing literature and shows that the method they used successfully captured the same population as those identified through other means >

Conclusion: Vera: So we've just been digging into how they found these active galactic nuclei candidates using infrared colors from Kepler data, and now we’re wrapping up what this whole study actually means for our understanding of these objects.

Jocelyn: Yeah, looking at the title itself, "Infrared-Selected Active Galactic Nuclei in the Kepler Fields," it tells you right away that they aren't just picking things out randomly; they used those infrared colors as a filter to narrow down a huge amount of data from Kepler.

Subrahmanyan: Exactly. It suggests that this infrared selection technique isn't just some random color match, but it’s actually filtering for something physically specific about the AGN itself, which is important for what we know about black hole fueling in galaxies.

Vera: It brings us back to those initial methods they used to find these targets—the catalogs and preliminary lists—and this study confirms that when you apply that infrared color selection, you end up with a population that lines up with what optical spectroscopy would find anyway.

Jocelyn: So, the main implication here is consistency; it shows that we don't need *only* optical data to get a good picture of these AGN populations; the infrared approach gives us the same results as traditional methods.

Subrahmanyan: That’s significant because it means we can use these infrared selections to find and study more AGN, which helps build a bigger picture of how galaxies host them across cosmic time.

Vera: It’s like they found a reliable shortcut that works across different ways of looking at the same sky data.

Jocelyn: Right, and they also confirmed that the properties—like the line ratios and energy distribution—of these newly discovered AGN are just as consistent with what we already know from other samples.

Subrahmanyan: So, to sum up, this paper’s conclusion is that this infrared selection method successfully captured the same population and physical characteristics as those found through optical spectroscopy and color selection.

Vera: It’s a solid piece of work because it bridges two different observational approaches to get the same result.

Jocelyn: And they've laid a foundation showing that these IR-selected samples are representative of the larger AGN population in the Kepler fields.

Physics Division, Lawrence Berkeley National Laboratory · Department of Physics and Astronomy, University of California Los Angeles Department of Physics and Astronomy at Texas A & M University, Jet Propulsion Laboratory, California Institute of Technology Center for Advanced Computing Research

astro-ph.GA, astro-ph.CO

Submitted: 2025-05-04

Updated: 2025-05-04

Comments: Accepted by ApJS. 12 pages, 14 figures. Figure 2 is a figure set in the HTML edition, and it has 186 spectra of AGN candidates

Journal ref: Astrophysical Journal Supplement Series 279, 22 (2025)

DOI: 10.3847/1538-4365/addbdf

Code: https://github.com/KeplerGO/K2FootprintFiles

License: http://creativecommons.org/licenses/by/4.0/

Importance score: 68/100

The gist: The gist: This study identifies and characterizes Active Galactic Nuclei (AGN) candidates monitored by the Kepler missions using infrared selection techniques and subsequent ground-based

Key concepts

Active Galactic Nuclei (AGN)
These are extremely luminous centers of galaxies powered by supermassive black holes. They emit intense radiation across the electromagnetic spectrum, which is what makes them visible and a major focus for this study.
Kepler Missions
The Kepler space telescopes were used to monitor large areas of the sky for long periods. The researchers specifically looked at data from these missions to find potential AGN candidates using infrared observations.
BPT Diagram
This is a diagnostic tool used by astronomers to classify galaxies. By plotting the ratio of specific emission lines, scientists can determine if a galaxy's activity is dominated by star formation or by an AGN.

Terminology

Summary

The gist: This study identifies and characterizes Active Galactic Nuclei (AGN) candidates monitored by the Kepler missions using infrared selection techniques and subsequent ground-based spectroscopic follow-up, confirming their nature and examining their physical properties.

Candidate Selection Methods

The researchers utilized several methods to identify AGN candidates that were monitored during the Kepler prime and K2 missions First, they searched for known AGN within a certain degree of the expected Kepler boresight center using catalogs like the V´eron-Cetty & V´eron (2010) AGN catalog and the SDSS survey. Second, they identified highly-likely AGN candidates using methods such as Edelson & Malkan (2012) and Stern et al. (2012), combining these lists to form a preliminary target list for each field. Third, they performed miscellaneous checks of the observability of these targets before proposing them for Kepler observations. Finally, they performed follow-up spectroscopy at the Lick, Palomar, and Keck observatories to test whether the AGN candidates were actual AGN. This approach led to Kepler observing a total of 1155 AGN and AGN candidates proposed by their group.

Ground-Based Observations and Data Reduction

The ground-based observations involved confirming 186 AGN candidates using the Kast spectrograph on the Lick 3.0-meter reflector, the Double Spectrograph (DBSP) on the Palomar 5-meter reflector, and the Low-Resolution Imaging Spectrograph (LRIS) on Keck 10-m reflector. The spectroscopic observations spanned from 2011 to 2016, with integration times ranging from 300 to 1800 seconds. Data reduction involved subtracting the bias level and flat-fielding the images using a uniformly illuminated dome ceiling to remove sensitivity variation. Wavelength calibration was obtained using Ne arcs for red spectra and a combination of Hg, He, Ar, and Cd arcs for blue spectra. Flux calibration was performed using spectroscopy of stars Feige 34, Feige 15, and BD+25 3941 to generate a flux scale.

Reliability Testing of Selection Techniques

The reliability of the two main techniques used to find AGN candidates was tested. For the Edelson & Malkan (EM12) sample, forty-nine AGN candidates from the W2R list were spectroscopically identified as Type 1 AGN and three as Type 2 AGN, resulting in a reliability of 96%. For the WISE color cut from Stern et al. (2012), they found a 90% reliability. The full Kepler AGN sample showed that 140/186=75% were AGN, and 35/140=25% of the AGN were Type 2 AGN. This indicated that the main identification techniques yielded significantly higher reliability than other methods did.

AGN Sample Properties and Classification

The sample AGN properties were examined through the Baldwin, Phillips & Terlevich (BPT) diagram and Spectral Energy Distributions (SEDs) for Type 1 AGN. In the BPT diagram, four out of 16 galaxies turned out to be composite galaxies with prominent narrow-line emissions from HII regions, while the rest of the galaxies above the Ke01 line are in the AGN region. The SEDs demonstrated a power-law relationship with a “bump” at around 3 µm or log(νrest) = 14 and another one starting around 14.75 in Figure 10. The Balmer decrements of Hα/Hβ were determined to be 4 (broad) to 4.5.

Conclusion

The study concluded that the overall line and continuum properties of the newly discovered AGN do not differ markedly from those of previous samples. They determined that the sample yielded the same AGN population as those that would have been found using other methods based on optical spectroscopy and color selection. The analysis of emission line properties confirmed consistent trends with existing literature. The study successfully demonstrated that AGN selected via the IR colors yield the same population and properties as those identified through optical spectroscopy. The overall line and continuum properties of the newly discovered AGN do not differ markedly from those of previous samples. The sample yielded the same AGN population as those that would have been found using other methods based on optical spectroscopy and color selection >.

Improvements for AI systems

  1. Start candidate identification with high-confidence infrared criteria to filter large initial samples effectively, as suggested by using W2R candidates which showed a 96% reliability in confirmation. This allows AI systems to prioritize targets based on robust multi-wavelength data rather than relying solely on less reliable color cuts.

  2. Implement a multi-stage validation pipeline for AGN candidates by first applying selection methods like "Edelson & Malkan (2012, EM12) and then cross-validating with other criteria, such as the WISE color criterion W1 − W2 ≥ 0.8." This mirrors the paper's approach of combining lists to form a preliminary target list before spectroscopic follow-up.

  3. Develop a classifier that distinguishes between Type 1 and Type 2 AGN based on line ratios from the BPT diagram, using the established boundaries such as "log([OIII]/Hβ) > 0.61 log([NII]/Hα) − 0.47 + 1.19 (2). This allows AI to classify spectroscopic results with a specified reliability related to the paper's findings of 87% [OIII] and 19% Type 2 AGN" for one sample.

  4. Integrate spectral feature analysis to estimate physical parameters, specifically by fitting emission lines like Hβ and Hα using Gaussian profiles, as described in Section 4.1 (We measured the emission lines using the IRAF routine splot). This enables AI to calculate line fluxes, equivalent widths, and FWHM for subsequent mass and accretion rate estimations.

  5. Train a model to analyze Spectral Energy Distributions (SEDs) by fitting observed photometry from WISE/2MASS, GALEX, and SDSS against established templates like the R06 line or AGN + Stars templates. This allows the AI to determine if an AGN's SED exhibits the characteristic features such as a “bump” at around 3 µm or a power-law relationship with a specific slope.

Abstract

We utilized the Edelson and Malkan (2012) and Stern et al. (2012) selection techniques and other methods to identify AGN candidates that were monitored during the Kepler prime and K2 missions. Subsequent to those observations, we obtained 125 long-slit optical spectra with the Lick 3-m telescope, 58 spectra with the Palomar 5-m telescope, and three with the Keck 10-m telescope to test these identifications. Of these 186 AGN candidates, 105 were confirmed as Type 1 AGN and 35 as Type 2 AGN, while the remaining 46 were found to have other identifications (e.g., stars and normal galaxies). This indicated an overall reliability of about 75%, while the two main methods had much higher reliability, 87%-96%. The spectra indicated redshifts out to z = 3.4. Then, we examined the AGN sample properties through the Baldwin, Phillips & Terlevich diagram and compared the AGN's spectral energy distributions (SEDs) with those from the literature. We found that our sample yielded the same AGN population as those identified through other methods, such as optical spectroscopy.

Related papers