Solid Identification of Extragalactic Gamma-Ray Source Using High-Resolution Radio Interferometric Observation

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The authors aimed to explore the origin of the γ-ray emission in a source from the Fermi-LAT Fourth Source Catalog, 4FGL 0959.6+4606. In the first place, this γ-ray object has been associated with a radio galaxy at z=0.148 while, later on, with a narrow-line Seyfert 1 (NLS1) galaxy at  z=0.399. To disantangle these two possible scenarios, the authors carried out high-resolution observations with the European very long baseline interferometry (VLBI) network (EVN), of both the hypothesised counterparts of this γ-ray source. By means of radio maps, they searched for signatures of milliarcsecond-scale compact jets in both radio and NLS1 galaxies at different redshifts. They found the NLS1 galaxy hosting a compact core and a fainter jet, while no jet features were detected for the radio galaxy. Their result validates the NLS1 galaxy as the counterpart of the γ-ray source 4FGL 0959.6+4606.

 

The introduction is the only section that could be improved in my opinion, by adding more references to recent papers on this topic. The observational procedure followed by the authors has proven to be correct for the purpose of the work. The methodology is clear and well described and the results are in agreement with previous findings (at other wavelengths) from the literature. The confirmation of a new γ-ray NLS1 galaxy should have more relevance while giving the conclusions. For these reasons, I recommend the publication after some minor revisions.

I list here some of the points for consideration by the authors:

1. The percentage of blazars found in the 4FGL should be revised. Two years after 'reference [1]', Foschini et al. (2022), carefuly checked the 4FGL catalogue reclassifying 2980 sources by means of optical spectra. They found 63% of blazars in the whole catalogue. The same percentage is shown in the paper cited as ref [12] in the text.
2. It must be considered that NLS1s have been identified as the third class of AGN that can produce γ-ray emission, after BL Lacs and flat-spectrum radio quasars (FSRQ). This can be also linked to the fact that they lie in the low-mass tail of the FSRQ distribution (Abdo et al. 2009a,c, Foschini et al. 2015, Berton et al. 2015b). This aspect needs to be highlighted in the introduction. 
3. The authors should check for the most updated number of γ-ray NLS1s in the literature. Foschini et al. 2022 found 24 NLS1s for example and other ambiguous sources that could have been lately confirmed as NLS1s. 
4. Radio-loudness ambiguity has been largely discussed in the literature, especially in the case of NLS1s (see Ho&Peng et al. 2001, Gu et al. 2015, Caccianiga et al. 2015, Lähteenmäkiet al. 2018 Järvelä et al. 2022 as main references). The presence of significant star formation activity and the role of variability can be considered to affect  the radio loudness parameter. For this reason, the radio-loud NLS1s should be referred to as NLS1s for which the radio emission is brighter than the optical one.
5. It should be clarified from the beginning (abstract), that 4FGL 0959.6+4606 was originally associated with a radio galaxy.
6. After pointing out that the counterpart of γ-ray source 4FGL 0959.6+4606 has been identified as an NLS1 galaxy, this finding should be put into a more extended scenario. The authors should highlight in the conclusions why the association of a γ-ray source with an NLS1 object is an important finding. For several decades this class of AGN has not been believed to be capable of even launching powerful relativistic jets.
7. Line 1: 'at a small angle' or 'at small angles'
8. Line 60: a reference to the used cosmological model shoud be added 

Author Response

Reply to Reviewer 1

We thank you for the careful reading, useful suggestions and references. We made the proposed changes in the text. In the revised version of the manuscript, we highlighted with boldface the changes performed following the suggestions of all three referees. Below we give our replies to your comments, after the * mark.

--------------

"The percentage of blazars found in the 4FGL should be revised. Two years after 'reference [1]', Foschini et al. (2022), carefuly checked the 4FGL catalogue reclassifying 2980 sources by means of optical spectra. They found 63% of blazars in the whole catalogue. The same percentage is shown in the paper cited as ref [12] in the text. 
It must be considered that NLS1s have been identified as the third class of AGN that can produce γ-ray emission, after BL Lacs and flat-spectrum radio quasars (FSRQ). This can be also linked to the fact that they lie in the low-mass tail of the FSRQ distribution (Abdo et al. 2009a,c, Foschini et al. 2015, Berton et al. 2015b). This aspect needs to be highlighted in the introduction. 
The authors should check for the most updated number of γ-ray NLS1s in the literature. Foschini et al. 2022 found 24 NLS1s for example and other ambiguous sources that could have been lately confirmed as NLS1s." 

* We rewrote the introduction and included more recent publications to revise the blazar fraction, and a few others to better reflect the most up-to-date status of the population of the known gamma-ray-emitting NLS1.
We slightly rearranged the text to provide an individual paragraph for the black hole mass issue and the similarities, connections of NLS1s to FSRQs.

"Radio-loudness ambiguity has been largely discussed in the literature, especially in the case of NLS1s (see Ho&Peng et al. 2001, Gu et al. 2015, Caccianiga et al. 2015, Lähteenmäkiet al. 2018 Järvelä et al. 2022 as main references). The presence of significant star formation activity and the role of variability can be considered to affect  the radio loudness parameter. For this reason, the radio-loud NLS1s should be referred to as NLS1s for which the radio emission is brighter than the optical one."

* We included a reference to the diverse nature of the radio emission from NLS1s. 

"It should be clarified from the beginning (abstract), that 4FGL 0959.6+4606 was originally associated with a radio galaxy."

* Added "radio" (galaxy) in the abstract, and in other parts of the text.

"After pointing out that the counterpart of γ-ray source 4FGL 0959.6+4606 has been identified as an NLS1 galaxy, this finding should be put into a more extended scenario. The authors should highlight in the conclusions why the association of a γ-ray source with an NLS1 object is an important finding. For several decades this class of AGN has not been believed to be capable of even launching powerful relativistic jets."

* Thank you for pointing this out. We included another paragraph in the Conclusion section to highlight these points.

"Line 1: 'at a small angle' or 'at small angles'"

* corrected

"Line 60: a reference to the used cosmological model shoud be added"

* A reference has been added to the ESA Planck satellite mission whose rounded-up values of cosmological parameters were used, and to the cosmological calculator.

Reviewer 2 Report

Comments and Suggestions for Authors

In their manuscript "Solid Identification of an Extragalactic Gamma-ray Source Using High-Resolution Radio Interferometric Observation" the authors investigate the correct cross-identification of gamma-ray source detected by the Fermi LAT. Two sources, a radio galaxy and a narrow-line Seyfert 1 (NLS1) galaxy, fall within the gamma-ray localisation region when all Fermi data is considered. Although, localisation during a gamma-ray flare excludes the radio galaxy as a possible origin. The authors perform VLBI observations of these sources at 5 GHz, finding that the radio galaxy is non-detection or barely detectable, but the NLS1 nucleus shows a tentative core-jet structure, a brightness temperature exceeding equipartition, indicating mild relativistic boosting, and a flat radio spectrum. Based on this they identify the NLS1 as the counterpart for the gamma-ray source. The paper is well-written, concise, and clear. It highlights the known issue of incorrect cross-identifications in the Fermi catalogues, especially when it comes to less understood gamma-ray-emitters, such as NLS1 galaxies. Overall, the manuscript is in good shape. Below I give my detailed report, including some suggestions for improvement. I hope to be able to recommend the publication of this manuscript after a minor revision.

Line-by-line comments:

ABSTRACT

line 2: "-- at small angle to the line sight --" -> "-- at a small angle to the line of sight --"

lines 2-3: " -- there are a few NLS1s --"
The number of known gamma-ray NLS1s is certainly more than a few these days (see my comment later). I would use a word that better reflects the current situation.

line 6: "-- associated with a galaxy."
I would say "radio galaxy" here, as gamma-ray emission from a radio galaxy is much more likely than from a normal, inactive galaxy.

1. INTRODUCTION

line 17: "-- dominated by blazars (98%) --"
A recent reclassification of 4FGL sources indicates that the fraction of blazars is smaller, around 65% (Foschini et al., 2022: https://ui.adsabs.harvard.edu/abs/2022Univ....8..587F/abstract). Although they do find that about 30% of gamma-ray sources lack reliable classification.

lines 19-21: "-- there are 9 narrow-line Seyfert 1 galaxies. Additionally, 8 of the currently unassociated γ-ray sources are also hypothesized to originate from NLS1 galaxies."
These numbers should be updated. Romano et al. 2018 (https://ui.adsabs.harvard.edu/abs/2018MNRAS.481.5046R/abstract) already lists 15 gamma-ray NLS1s, and the reclassification by Foschini et al. (2022) discovers 24 new gamma-ray NLS1 candidates.

lines 33-34: "It was shown that radio-loud NLS1s with blazar-like characteristics often experience enhanced variability in the infrared, a significant fraction also on intraday timescales."
For consistency, I suggest adding a brief mention about the variability of gamma-ray NLS1s also in other bands (optical, X-rays etc.). Does the jet contribute to their multiwavelength variability the same way it does in case of blazars?

lines 58-59: A reference for the used cosmology is missing.

2. OBSERVATION AND DATA REDUCTION

No comments.

3. RESULTS

lines 107-108: A brief qualitative interpretation of the radio map could be added here or in the Discussion. Just noting that it looks like it has a pc-scale core-jet structure, also seen in other gamma-ray NLS1s.

lines 114-115: "The peak intensity of the map created from the phase-referenced data was only marginally smaller, 36.1 mJy/beam --"
The peak intensity of the fringe-fitted data is only mentioned in the Figure 1 caption, so it would be useful to mention it here for comparison.

4. DISCUSSION

lines 176-177: "For J0959+4600, due to the apparent flux density variability, we did not calculate the spectral index."
It is at least worth mentioning that the 1.4-3 GHz spectral index seems flat or even inverted. (Of course, with the caveat that FIRST and VLASS are 20+ years apart.) Do these sources fall in the RACS-mid field?

5. CONCLUSIONS

No comments.

Author Response

Reply to Reviewer 2

We thank for the careful reading and useful suggestions of the reviewer. We made the proposed changes. In the revised version of the manuscript, we highlighted with boldface the changes performed following the suggestions of all three referees. Below we give our replies to all your comments, after the * mark.

-------------------

"ABSTRACT
line 2: "-- at small angle to the line sight --" -> "-- at a small angle to the line of sight --""

* changed 

"lines 2-3: " -- there are a few NLS1s --"
The number of known gamma-ray NLS1s is certainly more than a few these days (see my comment later). I would use a word that better reflects the current situation."

* Abstract is rewritten to better indicate the current status, and added the word dozens

"line 6: "-- associated with a galaxy."
I would say "radio galaxy" here, as gamma-ray emission from a radio galaxy is much more likely than from a normal, inactive galaxy.""

* Corrected. Also whenever we are referring to 2MASX J09591976+4603515 in the text, we refer to it as radio galaxy.

"1. INTRODUCTION
line 17: "-- dominated by blazars (98%) --"
A recent reclassification of 4FGL sources indicates that the fraction of blazars is smaller, around 65% (Foschini et al., 2022: https://ui.adsabs.harvard.edu/abs/2022Univ....8..587F/abstract). Although they do find that about 30% of gamma-ray sources lack reliable classification."

* Revised the blazar fraction with the help of the Foschini et al. 2022 paper.

"lines 19-21: "-- there are 9 narrow-line Seyfert 1 galaxies. Additionally, 8 of the currently unassociated γ-ray sources are also hypothesized to originate from NLS1 galaxies."
These numbers should be updated. Romano et al. 2018 (https://ui.adsabs.harvard.edu/abs/2018MNRAS.481.5046R/abstract) already lists 15 gamma-ray NLS1s, and the reclassification by Foschini et al. (2022) discovers 24 new gamma-ray NLS1 candidates."

* We revised the quoted number of known gamma-ray emitting NLS1 sources, using the Foschini et al. 2022, the Romano et al. 2018 papers, and additional literature search.

"lines 33-34: "It was shown that radio-loud NLS1s with blazar-like characteristics often experience enhanced variability in the infrared, a significant fraction also on intraday timescales."
For consistency, I suggest adding a brief mention about the variability of gamma-ray NLS1s also in other bands (optical, X-rays etc.). Does the jet contribute to their multiwavelength variability the same way it does in case of blazars?'"

* We added some text describing the optical and X-ray variability of radio-emitter NLS1 objects.

"lines 58-59: A reference for the used cosmology is missing."

*  A reference has been added to the ESA Planck satellite mission which rounded-up values of cosmological parameters were used, and to the cosmological calculator.

"3. RESULTS
lines 107-108: A brief qualitative interpretation of the radio map could be added here or in the Discussion. Just noting that it looks like it has a pc-scale core-jet structure, also seen in other gamma-ray NLS1s."

* This is added to the beginning of the Discussion section.

"lines 114-115: "The peak intensity of the map created from the phase-referenced data was only marginally smaller, 36.1 mJy/beam --"
The peak intensity of the fringe-fitted data is only mentioned in the Figure 1 caption, so it would be useful to mention it here for comparison."

* It is added as suggested.

"4. DISCUSSION
lines 176-177: "For J0959+4600, due to the apparent flux density variability, we did not calculate the spectral index."
It is at least worth mentioning that the 1.4-3 GHz spectral index seems flat or even inverted. (Of course, with the caveat that FIRST and VLASS are 20+ years apart.) Do these sources fall in the RACS-mid field?"

* The 1.4-3GHz spectral shape of J0959+4600 added in the Discussion. Unfortunately, the sources are at too high declinations to be included in the RACS.

Reviewer 3 Report

Comments and Suggestions for Authors

The paper reports an EVN observation of the potential counterparts of the Gamma-ray source 4FGL 0959.6+4606. The result significantly supports the NLS1 SDSS J095909.51+460014.3 is a better counterpart than the radio galaxy 2MASX J09591976+4603515. It warrants publication.

I have a series of comments that I would like the authors to address before recommending the manuscript. A general issue is that the paper lacks a discussion on the radio properties of the sources.

1. It will be good if the Gaia position is also plotted in Figure 1. What are the separations between the Gaia position and the C1 and C2 components? Which component is the AGN central core? Though C1 is much brighter than C2, it is possible that C1 is the lobe and C2 is the core, if the object is lobe-dominated.

2. In Table 1, it will be good to also list the peak flux density (in mJy/beam). The ratio of the peak to total flux can tell us the compactness of the source, which also provides information on which component is the core and the lobe. Is the FWHM size in Table 1 the major or the minor axis? Maybe list both? I noted that the image beam size is 5 mas x 1 mas. So the source size is smaller than the resolution, and is generally unresolved.

3. Though the 1-3 GHz slope of the NLS1 SDSS J095909.51+460014.3 is biased by the variability, the variability implies a compact source with a flat or inverted slope, which is consistent with the slope measured by the non-simultaneous 1 and 3 GHz fluxes.

4. It will be good to also discuss the radio emission in the galaxy 2MASX J09591976+4603515. Its emission is resolved out on mas scales, and has a 1-3 GHz slope of -0.8 on arcsec scales. Is it possible that the radio emission is from star formation?

5. What is the radio loudness of the NLS1 and the galaxy?

6. It is easy to get confused between the names J0959+4603 and J0959+4600. It will be better to use names which are easier to identify, for example, 2MASX J0959+4603 and SDSS J0959+4600, or just simply A and B, or any other else.

Author Response

Reply to Reviewer 3

We thank for the constructive comments and suggestions. We made changes in the manuscript following those. In the revised version of the manuscript, we highlighted with boldface the changes performed following the suggestions of all three referees. Below, we give our replies to your comments, after * mark.

---------------------

"1. It will be good if the Gaia position is also plotted in Figure 1. What are the separations between the Gaia position and the C1 and C2 components? Which component is the AGN central core? Though C1 is much brighter than C2, it is possible that C1 is the lobe and C2 is the core, if the object is lobe-dominated."

* Because we are presenting the image originating from fringe-fitting the data thus losing absolute positional information, we think that adding Gaia coordinates would be confusing. However, the difference between the Gaia and the VLBI coordinates are so small (less than half mas) that it would not be illustrative. 
Since the optical position is less than a mas different from C1, its distance to C2 is basically the same as the distance between C1 and C2.
Nevertheless, we added more detailed description of the coordinates in the text as suggested.

"2. In Table 1, it will be good to also list the peak flux density (in mJy/beam). The ratio of the peak to total flux can tell us the compactness of the source, which also provides information on which component is the core and the lobe. Is the FWHM size in Table 1 the major or the minor axis? Maybe list both? I noted that the image beam size is 5 mas x 1 mas. So the source size is smaller than the resolution, and is generally unresolved."

* We found that fitting circular Gaussian components provided a good and stable fit to the visibilities, there was no need to introduce the ellipticity of the components. We clarified this in the caption of Table 1.
We added the peak intensity values to Table 1.
Concerning the sizes of the components, towards the end of Sect. 3.1, we compared the theoretical resolution limit of an interferometric array as given in Kovalev et al. (2005) (reference [44] in the text) with the sizes we obtained. The size of the more compact component is close to but slightly larger than this value, thus it is not completely point-like/unresolved.

"3. Though the 1-3 GHz slope of the NLS1 SDSS J095909.51+460014.3 is biased by the variability, the variability implies a compact source with a flat or inverted slope, which is consistent with the slope measured by the non-simultaneous 1 and 3 GHz fluxes."

* We thank you for pointing this out, we added this information to the spectral index discussion Sect. 4.

"4. It will be good to also discuss the radio emission in the galaxy 2MASX J09591976+4603515. Its emission is resolved out on mas scales, and has a 1-3 GHz slope of -0.8 on arcsec scales. Is it possible that the radio emission is from star formation?"

* We added a few sentences to discuss this possibility. 

"5. What is the radio loudness of the NLS1 and the galaxy?"

* This was calculated by Jiang et al. in their previous paper. We added this information to the Introduction section.

"6. It is easy to get confused between the names J0959+4603 and J0959+4600. It will be better to use names which are easier to identify, for example, 2MASX J0959+4603 and SDSS J0959+4600, or just simply A and B, or any other else."

* The name of the objects are changed as suggested throughout the text.