Draft version October 21, 2024

Typeset using LATEX twocolumn style in AASTeX631

PKS 2254+074: A Blazar in Likely Association with the Neutrino Event IceCube-190619A
Shunhao Ji1 and Zhongxiang Wang1, 2
1 Department of Astronomy, School of Physics and Astronomy, Yunnan University, Kunming 650091, China; [email protected];
[email protected]
2 Shanghai Astronomical Observatory, Chinese Academy of Sciences, 80 Nandan Road, Shanghai 200030, China
arXiv:2410.14079v1 [astro-ph.HE] 17 Oct 2024

ABSTRACT
We report our study of the field of a ≃0.2 PeV neutrino event IC-190619A. This neutrino be-
longs to Gold events, which more likely have an astrophysical origin. Among the two γ-ray sources
within the neutrino’s positional uncertainty region, we find that one of them, the BL-Lac–type blazar
PKS 2254+074, had a γ-ray flare at the arrival time of the neutrino. The flare is determined to have
lasted ∼2.5 yr in a 180-day binned light curve, constructed from the data collected with the Large Area
Telescope (LAT) onboard the Fermi Gamma-ray Space Telescope (Fermi). Accompanying the flare,
optical and mid-infrared brightening is also seen. In addition, ≥10 GeV high energy photons from the
source have been detected, suggesting a hardening of the emission during the flare. Given both the
positional and temporal coincidence of PKS 2254+074 with IC-190619A, we suggest that this blazar
is likely another member of a few recently identified (candidate) neutrino-emitting blazars.

Keywords: Blazars (164); Gamma-ray sources (633); Neutrino astronomy (1100)

1. INTRODUCTION ergy of >0.1 GeV and the accompanying upto 400 GeV
The IceCube South Pole neutrino observatory very-high-energy (VHE) emission was also observed
(Aartsen et al. 2017) has been detecting neutrino (IceCube Collaboration et al. 2018a).
events with energy from ∼30 TeV to PeV since 2010 Following this identification, studies for connect-
(IceCube Collaboration 2013). These high energy neu- ing IceCube-detected neutrinos to blazars have been
trinos have a high probability of arising from extrater- extensively carried out (e.g., Giommi et al. 2020a;
restrial sources. While the origin of such events had Franckowiak et al. 2020; Stathopoulos et al. 2022;
been under intense investigation, it was not until 2017 Plavin et al. 2023; Rodrigues et al. 2024). Among
that the association of such an event, IceCube-170922A, them, one is to find a flaring blazar that is posi-
and an extra-galactic source, the blazar TXS 0506+056, tionally coincident with a reported neutrino event,
was established (IceCube Collaboration et al. 2018a,b). which would strongly suggest the association be-
In addition, more recently the detection of the tween the two and thus enable the identifica-
neutrino emission from the nearby Seyfert galaxy tion of the blazar as being a neutrino source.
NGC 1068 at a signficance of 4.2σ and that from Thus far, the reported cases are PKS B1424−418
the Galactic plane at a 4.5σ significance level have (Kadler et al. 2016), GB6 J1040+0617 (Garrappa et al.
also been reported (IceCube Collaboration et al. 2022; 2019), MG3 J225517+2409 (Franckowiak et al. 2020),
Icecube Collaboration et al. 2023). GB6 J2113+1121 (Liao et al. 2022), PKS 0735+178
Blazars, the subclass of Active Galactic Nuclei (Sahakyan et al. 2023), and NVSS J171822+423948
(AGNs) with a relativistic jet pointing close to the line (Jiang et al. 2024). In addition, there are two other
of sight, indeed have been suggested to be neutrino emit- sources 1H 0323+342 and 3HSP J095507.9+355101;
ters (e.g., Kadler et al. 2016 and references therein). the former is a radio-loud narrow-line Seyfert galaxy in
Protons possibly carried in a jet would interact with a minor γ-ray flare (Franckowiak et al. 2020) and the
low-energy photons (pγ interaction), and the produced latter had an X-ray flare (Giommi et al. 2020b).
pions decay and produce neutrinos and high-energy pho- In our examination of neutrinos of high-probability as-
tons. One notable feature of the TXS 0506+056 case trophysical origin (i.e., Gold events that have signalness
is that the blazar was undergoing a γ-ray flare; the ≥50%; Abbasi et al. 2023), we noted a ∼0.2 PeV track-
peak flux reached ≃ 5 × 10−7 ph cm−2 s−1 in the en- like event IC-190619A. For its either originally reported
2

positional uncertainty region (IceCube Collaboration

2019) or that updated in the IceCube Event Catalog
of Alert Tracks (IceCat-1; Abbasi et al. 2023), we could

14.00
4FGL J2253.7+1405
include the blazar PKS 2254+074 in it. This BL-Lac–
4FGL J2256.7+1307
type blazar has redshift z = 0.19 (Stickel et al. 1988;

13.00
4FGL J2252.6+1245
Peña-Herazo et al. 2021) and is a γ-ray source detected
with the Large Area Telescope (LAT) onboard the Fermi

12.00
Gamma-ray Space Telescope (Fermi) (e.g., Ballet et al. 4FGL J2308.9+1111

11.00
2023). We also noted that this source had been in qui-
escence at γ-rays in most of the Fermi-LAT observa- IceCube-190619A

DEC

10.00
4FGL J2306.6+0940
tion, but had a flare over the arrival time of IC-190619A.
Thus, both the positional and temporal coincidence sup-

9.00
port that PKS 2254+074 is likely another neutrino-
source case. We conducted analysis of the archival 4FGL J2257.5+0748

8.00
data for this blazar, and report the results in this pa-
per. In this work, the cosmological parameters, H0 =

7.00
67.7 km s−1 Mpc−1 , Ωm = 0.31, and ΩΛ =0.69, from the
Planck mission (Planck Collaboration et al. 2020) were 347.00 346.00 345.00 344.00 343.00 342.00 341.00 340.00
RA
used.

2. DATA ANALYSIS AND RESULTS 0 15 29 44 58 73 87 102 117 131 146

2.1. Positional coincidence analysis

Figure 1. TS map of a size of 8◦ × 8◦ centered at
On 2019 June 19 13:14:18.04 UT (MJD 58653.55), IC-190619A in the energy of 0.1–500 GeV. The dashed
IceCube detected a ∼0.2 PeV track-like event IC- cyan ellipse marks the uncertainty region with the system-
190619A with signalness ≃ 54.6%1 . The ini- atic uncertainty (by scaling the major and minor axis of
tial automated alert provided the arrival direction, the 90% error ellipse with a factor of 1.3; Giommi et al.
which was subsequently updated from the more 2020a) included, and the pink ellipse marks the uncer-
tainty region given in IceCat-1. In addition to PKS
sophisticated reconstruction algorithms to be R.A.
2254+074 (4FGL J2257.5+0748), another Fermi LAT source
= 343◦ .26+4.08
−2.63 , Decl. = +10◦ .73+1.51
−2.61 (equinox (4FGL J2306.6+0940) is also in the uncertainty region in
J2000.0; IceCube Collaboration 2019). The blazar IceCat-1.
PKS 2254+074 (4FGL J2257.5+0748) is slightly out-
√
side the uncertainty region. However, when we the detection significance of 6.9σ (≃ T S). We checked
added the systematic uncertainty often considered (e.g., its γ-ray light curve, and found it did not show any sig-
Giommi et al. 2020a; Plavin et al. 2020; Hovatta et al. nificant variations at the neutrino’s arrival time.
2021; Sahakyan et al. 2023), the source is included in
the combined uncertainty region (Figure 1). Further- 2.2. Fermi-LAT data analysis
more in IceCat-1, the neutrino’s position was updated
2.2.1. Data and source model
to be R.A. = 343◦ .52+4.13 +2.02
−3.16 , Decl. = +10 .28−2.76 , having
◦

the blazar well enclosed (Figure 1). The Fermi-LAT data used were 0.1–500 GeV photon
No γ-ray sources in the LAT 8-year Source Catalog (at events (evclass=128 and evtype=3) from the updated
the time) were within the initial 90% statistical uncer- Fermi Pass 8 database in a time range of from 2008-08-
tainty region (IceCube Collaboration 2019). However 04 15:43:36 (UTC) to 2024-04-18 00:05:53 (UTC). The
in the latest Fermi LAT Source Catalog (i.e., 4FGL- region of interest (RoI) was set to be 20◦ × 20◦ centered
DR4; Ballet et al. 2023), there is 4FGL J2306.6+0940, at PKS 2254+074. We excluded the events with zenith
classified as blazar of uncertain type, also in the uncer- angles > 90◦ to avoid the Earth-limb contamination.
tainty region in IceCat-1. This source, as well as several The expression DATA QUAL > 0 && LAT CONFIG
other sources near but outside the error region (Fig- = 1 was applied to select good time-interval events.
ure 1), was fainter than 4FGL J2257.5+0748. The test The source model was generated based on 4FGL-DR4.
statistic (TS) value for it was 48, which corresponds to In the catalog, PKS 2254+074 (4FGL J2257.5+0748)
was modeled as a point source with a power-law (PL)
spectrum, dN/dE = N 0 (E/E 0 )−Γ , where E 0 was fixed
1 https://gcn.gsfc.nasa.gov/notices amon g b/132707 54984442.amon at 1.55 GeV. We adopted this spectral model. All
 3

Figure 2. Left: 180-day binned γ-ray light curve of PKS 2254+074 in 0.1–500 GeV (upper), its optical Catalina V band and
ZTF zg and zr band light curves (middle), and MIR WISE light curves (bottom). The orange line marks the arrival time of
IC-190619A. In the top panel, the red line represents the average flux, the cyan histograms mark the blocks, and the gray region
marks the flare defined by the HOP algorithm. Right: 30-day binned γ-ray light curve during the time period marked by the
two blue dashed lines in the left panel (upper) and the arrival times of ≥10 GeV photons with probability > 68% (lower; see
Section 2.2.3). In both γ-ray light curve panels, the downward triangles are the 95% C.L. flux upper limits and red histograms
indicate the TS values of the data points.

other sources in 4FGL-DR4 within 25◦ of the target tectable for ∼10 yr. Since 2018, a γ-ray flare peaking at
were included in the source model. The spectral in- ∼MJD 58900 was observed, during which IC-190619A
dices and normalizations of the sources within 5◦ of the was detected. To define the time duration of the flare, we
target were set as free parameters and the other pa- employed the Bayesian block algorithm (Scargle et al.
rameters were fixed at the catalog values. The extra- 2013) and HOP algorithm (Meyer et al. 2019), imple-
galactic diffuse emission and the Galactic diffuse emis- mented through a python code2 (Wagner et al. 2022).
sion components, iso P8R3 SOURCE V3 v1.txt and The start/end of a segment (e.g., a flare) is determined
gll iem v07.fits respectively, were also included. Their by the flux of a block exceeding above or dropping under
normalizations were always set as free parameters in our the average flux (see details and the so-called HOP algo-
analysis. rithm in Meyer et al. 2019). The duration determined
is from MJD 58282.66 to 59182.66 (∼2.5 yr), with the
2.2.2. Light-curve analysis peak of the flare (the middle time of the highest-flux
data point) being MJD 58912.66. For the time ranges
Using the source model described above, we performed
excluding the flare, we defined them as the low state of
the standard binned likelihood analysis to the whole
the source.
data in 0.1–500 GeV for PKS 2254+074. We obtained
We also constructed the optical and mid-infrared
photon index Γ = 2.18±0.08, with a TS value of 136
(MIR) light curves of PKS 2254+074 from the data
(≃11.7σ detection signficance). The results are consis-
taken respectively from the the Catalina Real-Time
tent with those given in 4FGL-DR4.
Transient Survey (Drake et al. 2009), the Zwicky Tran-
We extracted the 0.1–500 GeV γ-ray light curve of
sient Facility (ZTF; Bellm et al. 2019), and the NEO-
the source by setting a 180-day time bin and performing
WISE Single-exposure Source Database (Mainzer et al.
the maximum likelihood analysis to the binned data. In
2014). The optical bands are Catalina V , ZTF g and
the extraction, only the normalization parameters of the
r (named as zg and zr respectively), and the MIR are
sources within 5◦ of the target were set free and the other
WISE w1 (3.4 µm) and w2 (4.6 µm). The light curves
parameters were fixed at the best-fit values obtained in
show that the optical and MIR emissions accompanied
the analysis of the whole data. For the data points with
TS<4, we computed the 95% confidence level (C.L.) up-
per limits. As revealed by the light curve (left panel 2 https://github.com/swagner-astro/lightcurves
of Figure 2), PKS 2254+074 had been nearly not de-
4
10−9
the γ-ray flare and brightened by ∼1 mag. During the Fer i-LAT low state Fermi-LAT flare
flare and afterwards, strong variations, with magnitude νμ 2.5 yr Archival data
10−10 νμ 16 yr Swift-XRT
changes as large as ∼1, are also seen.
To show more details of the γ-ray flare, we further ex-

s−1)
10−11
tracted a 30-day binned light curve from MJD 58282.66

−2
to 60262.66 (P1; the region between the two blue dashed 10−12
lines in the left panel of Figure 2); the same extraction

ν Fν (erg c
process as the above was conducted. The light curve 10−13
(right panel of Figure 2) further reveals a strong flar-
ing activity, which lasted ∼200 d from ∼MJD 58750 to 10−14
58950. During the time period, the peak flux reached
∼ 8 × 10−8 ph cm−2 s−1 . 10−15

2.2.3. Spectrum analysis 10−16 7

10 1011 1015 1019 1023 1027 1031
We performed the likelihood analysis to the data of the
ν (H()
two states (i.e., the flare and and low state) in 0.1–500 Figure 3. Broadband SED of PKS 2254+074. The gray
GeV. For the flare, the obtained best-fit Γ = 2.06±0.07 dots are the archival data from Firmamento. The Swift-
and photon flux Fγ = (1.21±0.20) × 10−8 ph cm−2 s−1 XRT X-ray spectrum is plotted as the blue dots. The γ-ray
(TS ≃ 190 or detection significance ≃ 13.8σ). For spectrum in the flare and the model fit in the low state are
the low state, Γ = 2.24±0.15 and Fγ = (2.50±0.94) plotted as red diamonds and cyan region, respectively. The
estimated neutrino fluxes for two time intervals (see Sec-
× 10−9 ph cm−2 s−1 (TS≃38 or detection significance ≃
tion 3) are shown as the red and gray lines.
6.2σ). Because of the large uncertainties in the latter,
no spectral changes between the two states could be de-
termined. we found seven high-energy photons, with probabilities
We then obtained the γ-ray spectrum of PKS ≥77%. The arrival times and energies of these photons
2254+074 in the flare by performing the binned like- are shown in the bottom right panel of Figure 2.
lihood analysis to the data in 8 evenly divided energy
2.3. Broadband spectral energy distribution
bins in logarithm from 0.1 to 500 GeV. In this anal-
ysis, the spectral normalizations of the sources in the We constructed the broadband spectral energy dis-
source model within 5◦ of the target were set free and tribution (SED) of PKS 2254+074 from radio to γ-
all other spectral parameters of the sources were fixed ray band. The archival radio to optical data were ob-
at the values obtained in the above likelihood analysis tained using the tool Firmamento3 . The X-ray spectral
to the data of the flare. For the spectrum, we only kept data points were extracted from the observations con-
the data points with TS ≥ 4. ducted with the X-ray Telescope (XRT) onboard the
For the low state, no decent spectrum can be obtained Neil Gehrels Swift Observatory (Swift). Details of the
because of the low detection significance (i.e., TS ≃ 38). X-ray data and analysis are described in Appendix Sec-
Several blazars showed spectrum hardening dur- tion A. Both the γ-ray spectrum in the flare and the
ing the flares that were temporally coincident with model fit in the low state are included in the SED (Fig-
the neutrinos’ arrival times (e.g., Garrappa et al. ure 3).
2019; Giommi et al. 2020b; Liao et al. 2022), in-
cluding TXS 0506+056 that emitted VHE photons 3. DISCUSSION
(IceCube Collaboration et al. 2018a). We thus checked Our analysis has found another blazar, PKS
the high energy, ≥10 GeV photons from PKS 2254+074. 2254+074, as a possible neutrino source. Its position
Running gtsrcprob on the whole data in an 1◦ RoI is within the uncertainty region of the IceCube Gold
with the best-fit model from the analysis of the whole event IC-190619A, and it had a flare temporally coinci-
data, we found five such photons with >68% probabili- dent with the arrival time of the neutrino. Such matches
ties. They were mostly during the flare. However, one have established the likely association of a few blazars
had 132 GeV the highest energy with 95% probability, with neutrinos. In addition, we have detected several
arrived after the flare on MJD 59987.38. The model high-energy photons during the flare, which had not
used for the target can affect the estimation of pho-
ton probability. When using the best-fit model from
3 https://firmamento.hosting.nyu.edu/data access
the analysis of the P1 data (MJD 58282.66–60262.66),
 5

been seen in the source’s long-term emission, suggest- ure 3), while the two values correspond to the neutrino-
ing the hardening of the source’s emission in the flare. emitting time intervals of 2.5 yr (flare time duration) and
This feature could be another similarity (Liao et al. 16 yr (approximate time length of the Fermi-LAT data),
2022). We note that the peak flux of this blazar’s flare respectively. The corresponding luminosities Lνµ would
reached ∼ 8 × 10−8 ph cm−2 s−1 , which is also compara- be 3.4×1045 erg s−1 and 5.4×1044 erg s−1 , and the lumi-
ble with the values of those (candidate) neutrino blazars nosity term ǫνµ Lνµ (see Giommi et al. 2020b), approxi-
(Liao et al. 2022). These similarities strongly support mated with ǫνµ Lνµ ∼ Lνµ /ln(8 PeV/80 TeV), would be
the possibility of PKS 2254+074 as another member of 7.5×1044 erg s−1 and 1.2×1044 erg s−1 , respectively.
the neutrino blazars. The neutrino flux, on the other hand, may be esti-
PKS 2254+074 is classified as a low-synchrotron mated from the observed γ-ray flux (see Giommi et al.
peaked (LSP) blazar in Ajello et al. (2022), and its peak 2020b and Jiang et al. 2024 for detailed calculations).
syn
frequency νpk ≃ 4.0+3.9 13
−2.0 × 10 Hz (given by Firma- The connection between them is given as (Murase et al.
mento using a machine learning estimator BLAST). Its 2018)
long-term average γ-ray luminosity Lγ and Γ we ob-
tained are 3.7 × 1044 erg s−1 and 2.18, and during the
(2.5 yr) flare, the values are 1.5 × 1045 erg s−1 and 2.06,
respectively. Comparing it to the previously identified
neutrino blazar candidates, it has the lowest redshift
and the lowest long-term luminosity. The other neutrino
blazars are generally in a redshift range of 0.3–1.5, but
the very recently discovered one NVSS J171822+423948
has z = 2.68 (Jiang et al. 2024). Their γ-ray luminosi-
ties are in a range of ∼ 4 × 1045 –5 × 1048 . Thus, if 6 (1 + YIC )
ǫν L ǫν ≈ ǫγ Lǫγ |ǫpπ
it is truely a neutrino source, the detection could be 5 syn

(1)
ǫγ Lǫγ |ǫpπ
 
due to its relatively close distance. The photon in- 44
≈ 8 × 10 erg s −1 syn
.
dex and luminosity values of PKS 2254+074 are typical 7 × 1044
for a BL Lac (e.g., Ackermann et al. 2015; Chen 2018).
Among the other neutrino blazars, GB6 J1040+0617
is in the same class, a BL Lac LSP. PKS B1424−418,
GB6 J2113+1121, and NVSS J171822+423948 are also
LSPs although they are flat-spectrum radio quasars (FS-
RQs). TXS 0506+056 and the other candidate 3HSP
J095507.9+355101, which are an intermediate- and a
high-synchrotron peaked blazars, respectively, have pre-
viously been noted as the outliers of the so-called blazar
Here, YIC is the Compton-Y parameter, typically ≤ 1
sequence (Giommi et al. 2020b). We also note that if we
syn (Murase et al. 2018). Basically during the photopion
take the average Lγ and νpk values of PKS 2254+074,
(pπ) process, three-eighths of the proton energy is taken
it could also be an outlier (because its Lγ < 1045 erg s−1
syn away by all-flavour neutrinos and the remaining energy
while νpk ≃ 4.0×1013 Hz; see Figure 4 in Giommi et al.
goes to the production of electrons and pionic γ-rays.
2020b). Unfortunately, we do not have information
syn Following Giommi et al. (2020b), we estimate the γ-
for its νpk in the flare. Whether its flaring varia-
ray luminosity term ǫγ Lγ ∼ Lγ / ln(500 GeV/100 MeV)
tions are similar to those of TXS 0506+056 and 3HSP
≃1.7×1044 erg s−1 (for 2.5 yr) or ≃4.4×1043 erg s−1 (for
J095507.9+355101 are not known.
16 yr). Using Eq. 1, the muon neutrino luminosi-
Following Giommi et al. (2020b) and Jiang et al.
ties (i.e., 13 ǫν Lǫν ) would be 6.7×1043 erg s−1 (2.5 yr) or
(2024), we estimated the neutrino flux given the detec-
1.7×1043 erg s−1 (16 yr). Comparing the values with the
tion of 1 muon neutrino event. Considering the effec-
above obtained by considering the detection of one muon
tive area Aeff ≃ 24 m2 (GFU Gold, see Abbasi et al.
neutrino with the IceCube, they are factors of 11 and 7,
2023), a PL neutrino energy spectrum with an in-
respectively, lower. The factors are approximately 4–14
dex of −2 from 80 TeV to 8 PeV (Oikonomou et al.
times smaller than those estimated in the cases reported
2021), the integrated muon neutrino energy flux would
by Giommi et al. (2020b) and Jiang et al. (2024), but
be 3.2×10−11 erg cm−2 s−1 and 4.9×10−12 erg cm−2 s−1
still indicate a low neutrino detection probability (i.e.,
(the corresponding differential spectra are shown in Fig-
the Poisson probability to detect one neutrino is ∼0.1).
 6

1 This work was based on observations obtained with

2 the Samuel Oschin Telescope 48-inch and the 60-inch
3 Telescope at the Palomar Observatory as part of the
4 Zwicky Transient Facility project. ZTF is supported by
5 the National Science Foundation under Grant No. AST-
6 2034437 and a collaboration including Caltech, IPAC,
7 the Weizmann Institute for Science, the Oskar Klein
8 Center at Stockholm University, the University of Mary-
9 land, Deutsches Elektronen-Synchrotron and Humboldt
10 University, the TANGO Consortium of Taiwan, the
11 University of Wisconsin at Milwaukee, Trinity College
12 Dublin, Lawrence Livermore National Laboratories, and
13 IN2P3, France. Operations are conducted by COO,
14 IPAC, and UW.
15 This work made use of data products from the Wide-
16 field Infrared Survey Explorer, which is a joint project
17 of the University of California, Los Angeles, and the
18 Jet Propulsion Laboratory/California Institute of Tech-
19 nology, funded by the National Aeronautics and Space
20 Administration.
21 We thank the referees for very detailed and help-
22 ful comments. This research is supported by the
23 Basic Research Program of Yunnan Province (No.
24 202201AS070005), the National Natural Science Foun-
25 dation of China (12273033), and the Original Inno-
26 vation Program of the Chinese Academy of Sciences
27 (E085021002). S.J. acknowledges the support of the sci-
28 ence research program for graduate students of Yunnan
29 University (KC-23234629).

REFERENCES
Aartsen, M. G., Ackermann, M., Adams, J., et al. 2017, Evans, P. A., Beardmore, A. P., Page, K. L., et al. 2009,
Journal of Instrumentation, 12, P03012, MNRAS, 397, 1177,
doi: 10.1088/1748-0221/12/03/P03012 doi: 10.1111/j.1365-2966.2009.14913.x
Abbasi, R., Ackermann, M., Adams, J., et al. 2023, ApJS, Franckowiak, A., Garrappa, S., Paliya, V., et al. 2020, ApJ,
269, 25, doi: 10.3847/1538-4365/acfa95 893, 162, doi: 10.3847/1538-4357/ab8307
Ackermann, M., Ajello, M., Atwood, W. B., et al. 2015, Garrappa, S., Buson, S., Franckowiak, A., et al. 2019, ApJ,
ApJ, 810, 14, doi: 10.1088/0004-637X/810/1/14 880, 103, doi: 10.3847/1538-4357/ab2ada
Ajello, M., Baldini, L., Ballet, J., et al. 2022, ApJS, 263, Giommi, P., Glauch, T., Padovani, P., et al. 2020a,
24, doi: 10.3847/1538-4365/ac9523 MNRAS, 497, 865, doi: 10.1093/mnras/staa2082
Ballet, J., Bruel, P., Burnett, T. H., Lott, B., & The Giommi, P., Padovani, P., Oikonomou, F., et al. 2020b,
Fermi-LAT collaboration. 2023, arXiv e-prints, A&A, 640, L4, doi: 10.1051/0004-6361/202038423
arXiv:2307.12546, doi: 10.48550/arXiv.2307.12546 HI4PI Collaboration, Ben Bekhti, N., Flöer, L., et al. 2016,
Bellm, E. C., Kulkarni, S. R., Graham, M. J., et al. 2019, A&A, 594, A116, doi: 10.1051/0004-6361/201629178
PASP, 131, 018002, doi: 10.1088/1538-3873/aaecbe Hovatta, T., Lindfors, E., Kiehlmann, S., et al. 2021, A&A,
Chen, L. 2018, ApJS, 235, 39, 650, A83, doi: 10.1051/0004-6361/202039481
doi: 10.3847/1538-4365/aab8fb IceCube Collaboration. 2013, Science, 342, 1242856,
Drake, A. J., Djorgovski, S. G., Mahabal, A., et al. 2009, doi: 10.1126/science.1242856
ApJ, 696, 870, doi: 10.1088/0004-637X/696/1/870 —. 2019, GRB Coordinates Network, 24854, 1
 7

IceCube Collaboration, Aartsen, M. G., Ackermann, M., Peña-Herazo, H. A., Massaro, F., Gu, M., et al. 2021, AJ,
et al. 2018a, Science, 361, eaat1378, 161, 196, doi: 10.3847/1538-3881/abe41d
doi: 10.1126/science.aat1378 Planck Collaboration, Aghanim, N., Akrami, Y., et al.
—. 2018b, Science, 361, 147, doi: 10.1126/science.aat2890 2020, A&A, 641, A6, doi: 10.1051/0004-6361/201833910
IceCube Collaboration, Abbasi, R., Ackermann, M., et al. Plavin, A., Kovalev, Y. Y., Kovalev, Y. A., & Troitsky, S.
2022, Science, 378, 538, doi: 10.1126/science.abg3395 2020, ApJ, 894, 101, doi: 10.3847/1538-4357/ab86bd
Icecube Collaboration, Abbasi, R., Ackermann, M., et al. Plavin, A. V., Kovalev, Y. Y., Kovalev, Y. A., & Troitsky,
2023, Science, 380, 1338, doi: 10.1126/science.adc9818 S. V. 2023, MNRAS, 523, 1799,
Jiang, X., Liao, N.-H., Wang, Y.-B., et al. 2024, ApJL, 965, doi: 10.1093/mnras/stad1467
L2, doi: 10.3847/2041-8213/ad36b9 Rodrigues, X., Paliya, V. S., Garrappa, S., et al. 2024,
Kadler, M., Krauß, F., Mannheim, K., et al. 2016, Nature A&A, 681, A119, doi: 10.1051/0004-6361/202347540
Physics, 12, 807, doi: 10.1038/nphys3715 Sahakyan, N., Giommi, P., Padovani, P., et al. 2023,
Liao, N.-H., Sheng, Z.-F., Jiang, N., et al. 2022, ApJL, 932, MNRAS, 519, 1396, doi: 10.1093/mnras/stac3607
L25, doi: 10.3847/2041-8213/ac756f Scargle, J. D., Norris, J. P., Jackson, B., & Chiang, J. 2013,
Mainzer, A., Bauer, J., Cutri, R. M., et al. 2014, ApJ, 792, arXiv e-prints, arXiv:1304.2818,
30, doi: 10.1088/0004-637X/792/1/30 doi: 10.48550/arXiv.1304.2818
Stathopoulos, S. I., Petropoulou, M., Giommi, P., et al.
Meyer, M., Scargle, J. D., & Blandford, R. D. 2019, ApJ,
2022, MNRAS, 510, 4063, doi: 10.1093/mnras/stab3404
877, 39, doi: 10.3847/1538-4357/ab1651
Stickel, M., Fried, J. W., & Kuehr, H. 1988, A&A, 191, L16
Murase, K., Oikonomou, F., & Petropoulou, M. 2018, ApJ,
Wagner, S. M., Burd, P., Dorner, D., et al. 2022, in 37th
865, 124, doi: 10.3847/1538-4357/aada00
International Cosmic Ray Conference, 868,
Oikonomou, F., Petropoulou, M., Murase, K., et al. 2021,
doi: 10.22323/1.395.0868
JCAP, 2021, 082, doi: 10.1088/1475-7516/2021/10/082
8

APPENDIX

A. X-RAY DATA ANALYSIS

The X-ray data were from the observations conducted with Swift-XRT. There are six observations between 2007
and 2012, and the information is given in Table A1. We used the online Swift-XRT data products generator tool4
(for details about the online tool, see Evans et al. 2009) to extract the 0.3–10 keV spectrum of PKS 2254+074 from
the six sets of the data. We grouped the spectrum to a minimum of 20 counts per bin using the GRPPHA task of
FTOOLS. We fitted the spectrum with an absorbed PL model in the XSPEC 12.12.1, where the Galactic hydrogen
column density NH was fixed at 4.76 × 1020 cm−2 (HI4PI Collaboration et al. 2016). The obtained PL index was
1.87±0.1 and unabsorbed flux was 1.36±0.12 × 10−12 erg cm−2 s−1 .

Table A1. Information for Swift-XRT observations

Date Obsid Exposure
(ks)
2007-04-25 00036360001 5.01
2007-04-28 00036360002 4.41
2009-01-18 00036360003 4.02
2009-01-22 00036360004 2.06
2012-04-24 00036360005 0.38
2012-07-05 00036360006 0.63

4 https://www.swift.ac.uk/user objects/