Draft version December 31, 2024

Typeset using LATEX default style in AASTeX631

Revisiting the flaring activity in early 2015 of BL Lacertae object S5 0716+714

Zhihao Ouyang ,1, 2 Hubing Xiao ,1 Marina Manganaro ,3 Shangchun Xie,1 Jingyu Wu,1 Jianzhen Chen,1
Rui Xue,4 Gege Wang,5 Shaohua Zhang,1 and Junhui Fan 2, 6, 7, 8
1 Shanghai Key Lab for Astrophysics, Shanghai Normal University, Shanghai, 200234, People’s Republic of China
2 Center for Astrophysics, Guangzhou University, Guangzhou, 510006, People’s Republic of China
arXiv:2412.20842v1 [astro-ph.HE] 30 Dec 2024

3 Department of Physics, University of Rijeka, Rijeka, 51000, Croatia

4 Department of Physics, Zhejiang Normal University, Jinhua, 321004, People’s Republic of China
5 Key Laboratory of Cosmology and Astrophysics (Liaoning) & College of Sciences, Northeastern University, Shenyang, 110819, People’s

Republic of China
6 Great Bay Brand Center of the National Astronomical Data Center, Guangzhou, 510006, People’s Republic of China
7 Key Laboratory for Astronomical Observation and Technology of Guangzhou, Guangzhou, 510006, People’s Republic of China
8 Astronomy Science and Technology Research Laboratory of Department of Education of Guangdong Province, Guangzhou, 510006,

People’s Republic of China

ABSTRACT
In this work, we analyzed multi-wavelength data of the BL Lac object S5 0716+714 to investigate
its emission mechanisms during a flaring state observed in early 2015. We examined the temporal
behavior and broadband spectral energy distributions (SEDs) during the flare. The size of the γ-ray
emission region was estimated based on the variability timescale. To explore the multi-wavelength
properties of S5 0716+714, we employed three one-zone models: the SSC model, the SSC plus EC
model, and the SSC plus pp interactions model, to reproduce the SEDs. Our findings indicate that
while the SSC model can describe the SEDs, it requires an extreme Doppler factor. In contrast, the
SSC plus EC model successfully fits the SEDs under the assumption of weak external photon fields but
requires a high Doppler factor. Additionally, the SSC plus pp interactions model also reproduces the
SEDs, with γ-ray emission originating from π 0 decay. However, this model leads to a jet power that
exceeds the Eddington luminosity, which remains plausible due to the flaring state or the presence of
a highly collimated jet.

Keywords: BL Lacertae object: S5 0716+714 — galaxies: active — high energy

1. INTRODUCTION
Blazars, a unique subclass of active galactic nuclei (AGNs), exhibit extreme observational characteristics including
strong variability, high polarization, superluminal motion, and γ-ray radiation (e.g., Urry & Padovani 1995; Fan et al.
2004; Lyutikov & Kravchenko 2017; Lister et al. 2019; Xiao et al. 2019, 2022; Abdollahi et al. 2022, and references
therein). Blazars are grouped into two subclasses based on their optical continuum: BL Lacertae objects (BL Lacs),
which show weak or absent emission lines (equivalent width, EW < 5 Å), and flat spectrum radio quasars (FSRQs),
which show prominent emission lines (EW ⩾ 5 Å; Stickel et al. 1991; Urry & Padovani 1995; Scarpa & Falomo 1997).
The spectral energy distribution (SED) exhibits a distinct two-hump structure. The low-energy hump, observed in
the infrared to X-ray range, is attributed to a synchrotron emission of relativistic electrons. The high-energy hump,
located at MeV to GeV energies, can be generated either through the inverse Compton process (IC) in a leptonic
scenario (Tavecchio et al. 1998; Ghisellini & Tavecchio 2009; Tan et al. 2020), or through a hadronic model (e.g.,
Mücke et al. 2003; Böttcher et al. 2009; Cerruti et al. 2015, 2019; Gao et al. 2019; Xue et al. 2022).

Corresponding author: Hubing Xiao, Marina Manganaro, Jianzhen Chen

hubing.xiao@shnu.edu.cn, marina.manganaro@phy.uniri.hr, jzchen@shnu.edu.cn
2

S5 0716+714 (4FGL J0721.9+7120) is classified as an intermediate-peak-frequency BL Lac (IBL) object according

to Fan et al. (2016), located at a distance of z = 0.2304 ± 0.0013 (Pichel et al. 2023). It was first discovered in
the late 1970s in the radio band (Perley et al. 1980; Kuehr et al. 1981), and early radio observations showed strong
emission and notable variability (Kraus et al. 2003). Subsequent high-resolution very long baseline interferometry
(VLBI) observations showed that S5 0716+714’s jet components exhibit fast superluminal motion, indicating it has
a strong Doppler beaming effect (e.g., Bach et al. 2005; Rani et al. 2015). Optical observations have been pivotal
in studying the variability of S5 0716+714. Previous studies demonstrated rapid variability on timescales of hours
to days (Wagner et al. 1996; Poon et al. 2009; Gupta et al. 2012; Tripathi et al. 2024), with intraday variability
suggesting a compact emission region for S5 0716+714. Long-term optical monitoring programs have reported a bluer-
when-brighter behavior (e.g., Dai et al. 2015; Xiong et al. 2020). Additionally, Ikejiri et al. (2011) performed the
photopolarimetric monitoring observations, revealing complex variability in both the degree and angle of polarization,
supporting the hypothesis that the optical emission primarily originates from synchrotron radiation. The Roentgen
Satellite (ROSAT) provided the first X-ray detection and it showed significant rapid variability and a double power-law
fitted spectrum, implying a mixture of synchrotron and inverse Compton components in the X-ray band (Cappi et al.
1994). More recent X-ray observations have offered detailed spectral and temporal analyses, they showed that the
X-ray fluxes were highly variable and the break energy between the synchrotron and inverse Compton components
shifted during different flux states (Foschini et al. 2006; Wierzcholska & Siejkowski 2015, 2016). The source was first
detected in the γ-ray band by the Energetic Gamma Ray Experiment Telescope (EGRET) on board the Compton
Gamma-Ray Observatory (CGRO) and has since been detected several times at different flux levels (Lin et al. 1995;
Hartman et al. 1999). It is one of the brightest and most variable sources in the γ-ray band, with a variability index
(VI) of 3680.86 reported by the Fermi Large Area Telescope (Fermi -LAT; Abdollahi et al. 2022). Several studies
have attempted to explore the γ-ray activity of S5 0716+714. For example, Rani et al. (2014) reported a significant
correlation between γ-ray fluxes and position angle variations in the VLBI jet, while Geng et al. (2020) found a highly
variable γ-ray flux with a spectral break between 0.93 and 6.90 GeV through long-term observations. Simultaneously
multi-wavelength observation has been employed as an effective method to study the blazar emission mechanisms.
Rani et al. (2013b) conducted a comprehensive campaign, including radio, optical, X-ray, and γ-ray observation, to
investigate a detailed temporal behavior and constructed a broadband detailed SED. Similarly, Liao et al. (2014)
performed a multi-wavelength study of S5 0716+714, finding significant variability and correlation across all bands,
and suggesting that the synchrotron self-Compton (SSC) plus external Compton (EC) model is preferred to describe
the broadband SED.
The first very-high-energy (VHE) γ-ray detection of S5 0716+714 was performed by Major Atmospheric Gamma-
ray Imaging Cherenkov telescopes (MAGIC) in 2007 (Anderhub et al. 2009). In the latter work, the observation
gave a significance of 5.8σ over 13.1 hours of observation in November 2007, and a significance of 6.9σ in April 2008.
Interestingly, the VHE observation coincided with optical high-state emission, implying a possible correlation between
VHE and optical emission. This led to the exploration of the one-zone SSC model and the structured (“spine+layer”)
jet model (Anderhub et al. 2009). In late December 2014, S5 0716+714 became brighter in the optical and infrared
bands, exhibiting an exceptionally high state in January 2015, with the highest flux recorded in these bands (Carrasco
et al. 2015; Bachev et al. 2015; Bachev & Strigachev 2015). MAGIC observations triggered by this flare revealed a
potentially variable VHE flux ranging from 4×10−11 cm−2 s−1 to 7×10−11 cm−2 s−1 above 150 GeV between January
22 and January 26, 2015 (Mirzoyan 2015; MAGIC Collaboration et al. 2018). This activity was studied in detail in
MAGIC Collaboration et al. (2018) based on the multi-wavelength observations of its flaring behavior in Jan 2015.
They stated that, due to the high level of optical flux, the broadband SED of the source can not be reproduced by
a one-zone SSC model. Instead, an interaction between a superluminal knot and a recollimation knot was found,
implying a two-zone model was preferred. However, the model underestimates the γ-ray flux in 10∼100 GeV. In
addition, the electric vector position angle (EVPA) showed a fast rotation of ∼360◦ and the high energy γ-ray flare
occurring during the γ-ray flaring activity, suggesting a shock-shock interaction in the jet. Considering the polarization
variations during the flare, Chandra et al. (2015) suggested that the magnetic reconnections were likely involved in
this flare. There is no doubt that it is necessary to further study the radiation mechanism of the S5 0716+714 flare
that occurred in Jan 2015.
In this work, we aim to further explore the high-energy emission of S5 0716+714 during the January and February
2015 flare. We will report on the multi-wavelength campaign involving Swift, NuSTAR, Fermi, and MAGIC observa-
 3

tions, and investigate the temporal behavior and spectral properties of the source. Additionally, we aim to restructure
the emission region and reproduce the broadband SED with a new hybrid model.
The paper is structured as follows. The observation introduction and the data reduction are presented in Section 2.
The result and discussion are presented in Section 3. Finally, the summary is given in Section 4.
2. OBSERVATION AND DATA REDUCTION
2.1. Swift observation
The Neil Gehrels Swift Gamma-Ray Burst Observatory (Swift; Gehrels et al. 2004) was launched in 2004 and includes
three instruments: the Ultraviolet and Optical Telescope (UVOT; Roming et al. 2005), the X-ray telescope (XRT;
Burrows et al. 2004) and the Burst Alert Telescope (BAT; Barthelmy et al. 2005). The analysis was performed using the
HEASoft package (v6.31.1) released by the NASA High Energy Astrophysics Archive Research Center (HEASARC).
2.1.1. Swift-UVOT
The Swift-UVOT observations (Roming et al. 2005) include three optical (v, b, and u) and three UV (w1, m2, and
w2) photometric bands (Poole et al. 2008; Breeveld et al. 2010). The source region was extracted from a circular
region of 5′′ centered on the source, and the background region was extracted from a circular region of 20′′ near the
source for all filters. The uvotmaghist task was used to analyze all filter data and produce the photometric light curve
data using the calibration from the release of CALDB (version 20211108). All the UVOT data were checked for the
small-scale sensitivity inhomogeneities, which occur when the source falls within the small areas of low sensitivity1 . In
addition, photometric data in which the source was saturated were excluded. Galactic extinction was corrected for the
observed magnitude with a value of E(B − V ) = 0.0268 (Schlafly & Finkbeiner 2011) using the interstellar extinction
law with RV = 3.1 (Fitzpatrick 1999). Finally, the corrected magnitudes were converted into fluxes using the standard
zero points from Breeveld et al. (2011).
2.1.2. Swift-XRT
For S5 0716+714, the Swift-XRT operated in Photon Counting (PC) and Windowed Timing (WT) readout modes
with a total exposure time of ∼ 1.55×105 s. The data were processed using the XRTDAS software package (v3.7.0) with
the release of CALDB (version 20220803). The cleaned events were produced using the xrtpipeline task, selecting
events with grades 0−12 for PC mode and grades 0−2 for WT mode. For the PC mode, the source region was extracted
from a circular region of 20 pixels (∼ 47′′ ) centered on the source. If the count rate was above 0.5 count · s−1 , pile-up
correction was applied by excluding the central region events within 3′′ −10′′ , using an annulus with an outer radius
of 20 pixels. The background region was extracted from an annulus with an inner radius of 80′′ and an outer radius
of 160′′ centered on the source. For the WT mode, the source and background regions were extracted from circular
regions of 20−30 pixels, depending on source brightness and exposure time, with the source region centered on the
source and the background region nearby. The high-level product data, including the spectra and ancillary response
files (ARF), were generated from the cleaned data using the xrtproducts task. The spectra were grouped using the
grppha (v3.1.0) tool to ensure at least one count per bin for fitting with Cash statistic (Cash 1979). The grouped
spectra were loaded into XSPEC (v12.13.0c) and fitted with an absorbed power-law model with normalization energy
E0 = 1 keV. The Galactic hydrogen column density was fixed at nH = 2.88×1020 cm−2 (HI4PI Collaboration et al.
2016). In XSPEC settings, the solar abundances used in the photoelectric absorption models were set as wilm (Wilms
et al. 2000) and the photoionization absorption cross-sections were set as vern (Verner et al. 1996). For X-ray data
fitting, the parameter errors correspond to 90% confidence errors (∆χ2 = 2.706). Finally, the unabsorbed fluxes and
photon spectral indices were obtained.
2.2. NuSTAR observation
The Nuclear Spectroscopic Telescope Array (NuSTAR), launched in 2012, operates in the hard X-ray range (3−79
keV) and features two telescopes with multilayer coatings that focus reflected X-rays onto pixillated CdZnTe focal
plane modules, FPMA and FPMB. The observation provides a spectral resolution of approximately 1 keV, and the
half-power diameter of an image of a point source is ∼1′ . Additional information can be found in Harrison et al.
(2013).

1 https://swift.gsfc.nasa.gov/analysis/uvot digest/sss check.html

4

NuSTAR observed S5 0716+714 with its two focal plane modules on 24 January 2015 (MJD 57046), with an exposure
time of ∼ 18.5 ks. The raw data were processed with the NuSTAR Data Analysis Software (NuSTARDAS, v2.1.2)
package using the calibration from the release of CALDB (version 20230307). The cleaned event files were produced
by the nupipeline task. The source region was extracted from a circular region of 45′′ centered on the centroid of
X-ray emission. The background was extracted from a position 5′ away from the centroid of the X-ray emission, using
a circular region of 1.5′ . The spectra were produced from the cleaned event files and grouped with at least one count
per bin using the nuproducts task. We focused on the energy range of 3−60 keV where the source was detected.
The XSPEC settings (including the Galactic hydrogen column density) were the same as those used in the Swift-XRT
analysis.
2.3. Fermi-LAT observation
The Fermi Gamma-ray Space Telescope was launched in 2008 and it consists of two instruments: Gamma-ray Burst
Monitor (GBM) and Large Area Telescope (LAT). The Fermi -LAT (Atwood et al. 2009) is capable of detecting γ-ray
in the energy range from 20 MeV to beyond 300 GeV. The point source sensitivity of Fermi -LAT is ∼ 2×10−13 erg
cm−2 s−1 for the north Celestial pole after 10 yrs operation (Ajello et al. 2021).
We used data events from the Fermi -LAT’s Pass 8 database in the period from MJD 57010 (2014 Dec 19) to
57075 (2015 Feb 22). The data were collected within an energy range of 0.1−100 GeV and within a 15◦ radius
region of interest (ROI) centered on S5 0716+714. A maximum zenith angle value of 90◦ was selected to avoid
background γ-rays from the Earth’s limb. We performed an unbinned likelihood analysis of the data using the latest
Fermitools (v2.2.0; Fermi Science Support Development Team 2019) and the instrument response functions (IRFs)
P8R3 SOURCE V3. The conditions “evclass=128, evtype=3” were used to filter events with a high probability of
being photons and “(DATA QUAL⩾0)&&(LAT CONFIG==1)” was used to select the good time intervals. The model file,
generated by make4FGLxml python package, included all the sources from the Fermi-LAT Fourth Source catalog (4FGL-
DR4; Abdollahi et al. 2022) within 25◦ of S5 0716+714 as well as the Galactic (gll iem v07.fits) and extragalactic
isotropic (iso P8R3 SOURCE V3 v1.txt) diffuse emission components. The spectral parameters of sources with an
average significance larger than 5σ within 5◦ of the ROI were left free, as well as sources ◦
 within 10 of the ROI with a
 −Γγ 
dN
variable index ⩾ 24.725 (Abdollahi et al. 2022). The best model between a power-law PL; dE = N0 EE0 model
"    #
 − α+β log EE
and a log-parabola LP; dN E 0 2
dE = N0 E0 model was selected by calculating TScurve = 2(log LLP −

log LPL ), where LPL /LLP represent the maximum likelihood value of power-law and log-parabola, respectively (Nolan
et al. 2012; Abdollahi et al. 2022). If TScurve ⩾ 16, corresponding to 4σ, the model was switched to the LP model.
We found that S5 0716+714 preferred the PL model rather than the LP model. We generated the light curves binned
in the one-day bin. In each time bin the normalization parameters of the sources within 5◦ of ROI and the spectral
index of the S5 0716+714 were allowed to vary freely during the spectral fitting. The rest of the parameters and other
source models were frozen. The normalization of the two diffuse emission components was also set free in the analysis.
We only included flux data points that are significantly detected with TS ⩾ 16. Meanwhile, we calculated the 95%
confidence level upper limit flux value for the case of TS < 16 using the UpperLimits3 tool. While examining the
spectral energy distribution, we fixed the spectral indices as the constant value equal to the value fitting over the whole
energy range.
2.4. MAGIC observation
We compiled the VHE γ-ray data, including the light curve and spectra, from MAGIC Collaboration et al. (2018),
who divided the observation into two periods: Phase A (MJD 57040−57050) and Phase B (MJD 57065−57070).
The spectra have been corrected by the extragalactic background light (EBL) absorption using the redshift z=0.2304
(Pichel et al. 2023) and the EBL model from Domı́nguez et al. (2011).
3. RESULT AND DISCUSSION
3.1. The γ-ray variability timescale

2 Here, the ‘log’ refers to the decimal logarithm, whereas ‘ln’ in Eq. (2) below denotes the natural logarithm.
3 https://fermi.gsfc.nasa.gov/ssc/data/analysis/scitools/upper limits.html
 5

15 (a)

(ph cm 2 s 1)
> 150 GeV
Flux × 10 11
10 95% CL Upperlimit

0
57010 57020 57030 57040 57050 57060 57070

10
(b)
1)

0.1 - 100 GeV

Flux × 10 7

95% CL Upperlimit
s
2

5
(ph cm

0
57010 57020 57030 57040 57050 57060 57070
6 (c)
(erg cm 2 s 1)

0.3 10 keV
Flux × 10 11

0
57010 57020 57030 57040 57050 57060 57070
10
(d)
(erg cm 2 s 1 Å)

u
F × 10 14

b
5 v

0
57010 57020 57030 57040 57050 57060 57070
15
(e)
(erg cm 2 s 1 Å)

uvw1
F × 10 14

10 uvw2
uvm2
5

0
57010 57020 57030 57040 57050 57060 57070
Time (MJD)

Figure 1. The multi-wavelength light curve during the period from MJD 57010 to 57075 with Swift, Fermi, and MAGIC
observations. From the top to bottom panels: (a) MAGIC VHE flux, > 150 GeV and the gray triangle is the 95% confidence
level upper limit; (b) Fermi-LAT flux in 0.1−100 GeV in one-day bin; (c) Swift-XRT flux, 0.3−10 keV; (d) Swift-UVOT, u, b
and v bands with the Galactic extinction correction; (e) Swift-UVOT, uvw1, uvm2 and uvw2 bands with the Galactic extinction
correction. The two gray time intervals are the periods MAGIC observation was taken, meaning Phase A and Phase B.

For the purpose of studying the geometry of the emission region and the property of the particle population, we
modeled the light curve to explore the time profile and variability timescale. In this work, we used the Fermi γ-ray
data to pursue this task as it is more continuously and uniformly sampled compared to the data of optical, UV, X-ray,
and VHE bands. The exponential fitting is applied to each component of the flare, the entire light curve is, thus,
expressed as a sum of exponential functions with a smoothed transition from raising to falling edge
 −1
X t0,i − t t − t0,i
F (t) = Fc + 2F0,i exp( ) + exp( ) , (1)
i
Tr,i Td,i

where Fc is the baseline or constant flux, F0 is the peak flux value at time t0 , Tr and Td are the rise and decay
time, respectively (Abdo et al. 2010). We constructed a likelihood function for fitting the exponential functions and
considered the contribution of the upper limit value to the likelihood function. The iminuit package was employed
6

12 Total Fitting
Peak1
10

Flux ( × 10 7 phcm 2 s 1)
Peak2
Peak3
8 Constant Flux
0.1 - 100 GeV
6 95% CL Upperlimit

4
2
0
57020 57030 57040 57050 57060 57070
Time (MJD)

Figure 2. The fitted light curve of the period MJD 57020−57075. The dash lines with different colors represent the different
peak components and the solid black line represents the sum-fitted time profile of each component.

Table 1. Fitting parameters of light curve in Fig. 2.

Component t0 F0 Tr Td
−7
(MJD) (× 10 ph cm−2 s−1 ) (day) (day)

(1) (2) (3) (4) (5)

1 57040.27±0.98 3.16±1.38 4.34±2.17 0.95±0.76

2 57048.33±1.99 4.75±1.76 4.75±3.89 1.48±1.02
3 57055.26±1.58 3.29±2.39 0.23±0.51 2.60±1.23
Note— The constant flux is given as Fc = (1.23 ± 0.28) × 10−7 ph cm−2 s−1 .

to perform the maximum likelihood fitting and the Bayesian Information Criterion (BIC) was utilized to ascertain the
optimal number of exponential functions required to fit the light curve:

BIC = k ln(N ) − 2 ln(L̂), (2)

where k is the number of model parameters, N is the total number of data points, and L̂ represents the maximized
value of the likelihood function for fitting the exponential functions (Edelson & Krolik 1988; Wit et al. 2012). We
tested different numbers of exponential functions and selected the model that minimized the BIC value. Finally, we
used three exponential components to fit the light curve from MJD 57020 to 57075, as shown in Fig. 2. It is important
to note that a shorter time bin light curve can reveal more detailed structures and accurate variability timescales, but
this comes with increased flux errors and reduced TS values, which can decrease the fitting quality.
Three peak profiles exhibited asymmetric shapes, reflecting the underlying particle acceleration and cooling mech-
anisms. The first two peaks showed a relatively gradual rise followed by a sharp decay, suggesting either a gradual
acceleration of particles or rapid cooling or escape of injected/accelerated electrons, resulting in a radiative cooling
timescale shorter than the acceleration timescale (Roy et al. 2019). In contrast, the third minor peak, which was also
observed in other bands (see Fig. 1) but lacked coverage in the VHE band, exhibited a relatively shorter rise timescale
compared to its decay timescale, indicating that the asymmetry may stem from changes in the bulk Lorentz factor,
the structure of the emission region’s shells (Roy et al. 2019), or particle injection processes (Wang et al. 2022). Our
findings on asymmetry differ from those of Geng et al. (2020), who reported symmetric flare profiles. This discrepancy
arises because their analysis used shorter binning to construct the light curve, increasing the associated errors and
leading to divergent results.
In addition, the rise and decay timescales can serve as a tool for constraining the geometry of the emission region,
which will be explored in the following subsection.
 7

3.2. Modeling spectral energy distributions

The emission mechanisms of blazars can be better understood by modeling their SEDs. In this context, to further
investigate the physical origin of the flaring activity of S5 0716+714, we modeled the simultaneous multi-wavelength
SEDs during two distinct periods: Phase A (MJD 57040–57050) and Phase B (MJD 57065–57070). Despite the
data we mentioned above, we collected simultaneously observed radio spectral data from MAGIC Collaboration et al.
(2018), and the X-ray spectra were corrected through a de-absorption process employing the cross-section in Morrison
& McCammon (1983) and the hydrogen column density value (HI4PI Collaboration et al. 2016). Historical archival
data were also obtained from the Space Science Data Center (SSDC)4 .
Specifically, Wierzcholska & Siejkowski (2016) performed cross-correlation analysis between the optical, UV, and
γ-ray bands during January and February 2015. Their findings indicated no evident time lags among the optical, UV,
and γ-ray emissions, suggesting they likely originate from the same region. The Swift-XRT X-ray band is also expected
to exhibit zero time lag with the γ-ray band; however, the mismatched sampling in the Swift-XRT X-ray data could
introduce artifacts affecting the correlation results (Wierzcholska & Siejkowski 2016). Therefore, the one-zone model
shall be considered during the SED modeling. The popular method for estimating the (intrinsic) size of the emission
region (Rb ) assumes that the flux variability timescale corresponds to the light travel time across the emission region.
In this context, the observed shortest doubling/halving timescale (tvar ) can be used to constrain the size of the emission
region, which is expressed as Rb ⩽ ct1+z var δ
where δ is the Doppler factor. In our estimations, tvar was calculated as
tvar = ln(2) × min {Tr , Td } (see, e.g., Rani et al. 2013a; Gasparyan et al. 2018, using the rising and decay components
in Phase A for SED modeling). Following Bach et al. (2005), we adopted δ = 30, leading to an estimated emission
region size of Rb ≃ 4 × 1016 cm during our SED modeling. In the following, we performed the SED modeling using
the public code Jets SED modeler and fitting Tool (JetSet; Tramacere et al. 2009, 2011; Tramacere 2020), and we
considered two scenarios: (i) the broad-band emission originated from the leptonic model, namely synchrotron and IC
radiation; and (ii) the emission originated from the lepto-hadronic hybrid model.
3.2.1. Leptonic scenario
The one-zone leptonic scenario assumes that the emissions originate from a spherical region (blob) of radius Rb ,
filled with a uniform magnetic field (B). This region, located within the blazar jet, moves with a bulk Lorentz
factor (Γ = √ 1 2 ∼ δ, where βΓ c is the speed of the blob) at a small viewing angle to the observer, resulting in
1−βΓ
Doppler-boosted emission characterized by a Doppler factor (δ). Blazar emission is primarily dominated by radiation
from synchrotron and IC processes. When the low-energy seed photons for the IC process originate from synchrotron
radiation, the process is referred to as synchrotron self-Compton (SSC; e.g., Finke et al. 2008). Alternatively, if the
seed photons come from external regions such as the accretion disk (AD; Dermer & Schlickeiser 1993), the broad-line
region (BLR; Sikora et al. 1994), or the dust torus (DT; Blażejowski et al. 2000), the process is termed external
Compton (EC) process.
The blob is assumed to be filled with relativistic electrons with a log-parabola with low-energy power-law branch
distribution, expressed as follows:
 −s
 γγ
 , γe, min < γ < γe, 0
e, 0
Ne (γ) = Ae ne (γ) = Ae  
 
− s+r log γ γ

. (3)
γ e, 0
, γ < γ < γ


γe, 0 e, 0 e, max

Here, Ae is defined by the actual density of relativistic electrons (Ne ) in units of cm−3 with Ae = R neN(γ)dγ
e
(see the
documentation in JetSet), s is the spectral index, r is the spectral curvature, and γe, min/0/max are the minimum,
turn-over, and maximum electron Lorentz factor, respectively.
S5 0716+714 is a BL Lac object, allowing us to model the SEDs using the one-zone SSC model. The SSC model is
described by nine parameters, where six of these parameters characterize the electron energy distribution (γe, min/0/max ,
Ne , s, and r), while the remaining three describe the properties of the emission region (Rb , B, and δ). During the
SSC modeling, the size of the emission region, Rb , was fixed at 4 × 1016 as mentioned above, and the minimum and
maximum electron Lorentz factors were fixed at 1 and 107 , respectively. The other parameters were optimized to

4 https://www.ssdc.asi.it/
8

achieve the best-fit model. The best-fitted model parameters are summarized in Tab. 2 and the resulting best-fit SSC
model SEDs are presented in Fig. 3.
However, the SSC model failed to reproduce the large separation between the low-energy peak (∼ 1014 –1015 Hz) and
the high-energy peak (∼ 1024 –1025 Hz) using a Doppler factor of δ = 30 (Bach et al. 2005). Such a large separation
required an extreme Doppler factor (δ ∼ 200, see Tab. 2); otherwise, the observed optical flux would be underestimated.
Moreover, a hard electron spectral index and a low magnetic field were also required. These difficulties could be avoided
by assuming that the HE/VHE γ-ray emissions originate from a more highly boosted substructure within the jet. For
example, the “jets-in-a-jet” model could have an extra Lorentz factor due to the plasma material outflowing from the
reconnection regions (Giannios et al. 2009). However, this scenario typically results in fast variability, as observed in
sources like Mrk 501, PKS 2155-304, 3C 279, and M87 (Albert et al. 2007; Aharonian et al. 2007; Giannios et al. 2010;
Shukla & Mannheim 2020), which contradicts the variability observed in S5 0716+714.
One can calculate the jet power and further understand the composition of the jet. The jet power (Pjet ) carried by
relativistic electrons (Pe ), cold protons (Pp, cold ), and magnetic field (PB ) is estimated via
X
Pjet = πRb2 Γ2 βΓ cUi , (4)
i

where the Ui is the energy density of the relativistic electrons (Ue ), cold protons (Up, cold ), and magnetic field (UB ),
respectively, in the co-moving frame (Ghisellini et al. 2010). These energy densities can be derived by

B2
UB = , (5)
8π
Z
2
Ue = me c γNe (γ)dγ, (6)
Z
Up, cold = mp c2 1 Ne (γ)dγ, (7)

where assuming a cold-proton-to-electron number density ratio of 1, B is the magnetic field strength obtained from the
SED modeling, me and mp are the rest mass of the electron and proton, respectively. The jet powers and the energy
densities are calculated and listed in Tab. 2. The Eddington luminosity of the supermassive black hole (SMBH) is

LEdd = 2πmp c3 RS /σT , (8)

where σT is the Thompson scattering cross-section, RS = 2GMBH /c2 is the Schwarzschild radius of the SMBH, and
MBH is the mass of SMBH. Using the MBH = 108.91 M⊙ for S5 0716+714 (Liu et al. 2019a) where M⊙ is the mass of
the Sun, the Eddington luminosity of the source is calculated as LEdd = 1.02 × 1047 erg s−1 . It is worth noting that
the jet powers in two phases under the SSC model moderately exceed the Eddington luminosity, primarily due to the
extreme Doppler factors required to account for the SSC model.
As discussed above, the HE/VHE γ-ray emissions probably require a substructure within the jet to account for the
large separation between the two peaks of the SEDs. Alternatively, these emissions could arise from an additional
radiation component, such as the EC or hadronic component, which will be explored in the upcoming content and the
next subsection, respectively.
Previous studies have conducted SED modeling for S5 0716+714; however, they also found that the SSC model failed
to reproduce the SEDs (e.g., Tagliaferri et al. 2003; Rani et al. 2013b; Liao et al. 2014). In MAGIC Collaboration et al.
(2018), they attempted the one-zone SSC model but found it underestimated the observed optical flux. Therefore,
the EC component is a plausible mechanism for accounting for the HE/VHE γ-ray emissions, assuming the scattering
of weak external emissions, despite the absence of detected thermal components or emission lines (Shaw et al. 2009;
Liao et al. 2014). Following this, we explored the SSC model combined with an EC component. Assuming a conical
jet with a half-opening angle of θopen = 5◦ , the distance of the emission region from the central engine (RH ) was
calculated to be ∼0.15 pc. We considered broad-line region as a spherical shell with an inner and outer radius of
RBLR, in = 3 × 1017 (Ldisk /1046 )1/2 cm and 1.1×RBLR, in (Kaspi et al. 2007) with a coverage factor of τBLR = 0.1, where
Ldisk is the accretion disk luminosity. The dust torus was assumed to be a radius of RDT = 2 × 1019 (Ldisk /1046 )1/2
cm (Cleary et al. 2007) with a reprocessing factor τDT = 0.1. The dust torus and accretion disk temperatures were
fixed at 1.2×103 K and 2×104 K, respectively. The actual disk luminosity (Ldisk ) is challenging to constrain due
 9
E (eV) E (eV)
10 4 10 1 102 105 108 1011 10 4 10 1 102 105 108 1011
10 8 10 8
Sync SSC Sum Phase A Sync SSC Sum Phase B

10 9 10 9
s 1)

s 1)
10 10 10 10
2

2
F (erg cm

F (erg cm
10 11 10 11

10 12 10 12

10 13 10 13

1011 1014 1017 1020 1023 1026 1011 1014 1017 1020 1023 1026
(Hz) (Hz)

Figure 3. One-zone SSC modeling. The left panel is for Phase A (MJD 57040–57050); the right one is for Phase B (MJD
57065–57070). The VHE spectra are corrected by EBL absorption adopting z=0.2304. The meanings of line styles are given in
the legend.

E (eV) E (eV)
10 4 10 1 102 105 108 1011 10 4 10 1 102 105 108 1011
10 8 10 8
Sync EC DT Phase A Sync EC DT Phase B
10 9 SSC Disk Sum 10 9 SSC Disk Sum

10 10 10 10
s 1)

s 1)

10 11 10 11
2

2
F (erg cm

F (erg cm

10 12 10 12

10 13 10 13

10 14 10 14

10 15 10 15

1011 1014 1017 1020 1023 1026 1011 1014 1017 1020 1023 1026
(Hz) (Hz)

Figure 4. One-zone SSC plus EC modeling. The left panel is for Phase A (MJD 57040–57050); the right one is for Phase B
(MJD 57065–57070). The VHE spectra are corrected by EBL absorption adopting z=0.2304. The meanings of line styles are
given in the legend, where the EC line includes the contribution from EC-BLR and EC-DT.

to the featureless optical spectra of S5 0716+714. Therefore we fixed a reasonable disk luminosity value of 2 × 1042
erg s−1 during the SED modeling in two phases. This value was estimated according to the SED modeling and is
below the upper limits reported in Ghisellini et al. (2010) and Danforth et al. (2013), making it a reasonable choice.
Furthermore, the minimum electron Lorentz factor, γe, min , was also left as a free parameter. The best-fit models are
displayed in Fig. 4, with the corresponding parameters listed in Tab. 2.
We can see that the SSC plus EC model provides a good fit to the SEDs and successfully reproduces the γ-ray
emissions. This is consistent with earlier studies, where the inclusion of EC components also successfully accounted
for the SEDs (Tagliaferri et al. 2003; Rani et al. 2013b; Liao et al. 2014). The jet powers derived from this model
using Eq. (4) are listed in Tab. 2 and remain below the Eddington luminosity in both phases. This model requires
a high Doppler factor of δ = 73.95 for Phase A and δ = 57.84 for Phase B, which are not preferred in such models
for the causality of light traveling across the emission region. While a higher Doppler factor corresponds to a larger
variability timescale in the blob frame, such variability timescale could not be associated with the light-crossing but
with the particle cooling processes or changes in external radiation fields. Moreover, such high Doppler factors derived
max
from the SED modeling remain problematic due to the maximum apparent velocity of βapp ∼ 34.4 estimated from
Lister et al. (2018). Consequently, this model is excluded from our consideration.
3.2.2. Lepto-hadronic hybrid scenario
As mentioned above, the hadronic radiation component could be the origination to produce the γ-ray emissions.
The proton-proton (pp) interactions have been used to explain the SEDs of blazars (e.g., Banik & Bhadra 2019; Banik
10

et al. 2020). In the work of Li et al. (2022), they suggested that pp interactions could be important for blazars and
have a parameter space to interpret the γ-ray spectra. Meanwhile, Xue et al. (2022) showed that pp interactions can
explain the TeV spectra. Therefore, we incorporated the pp interactions into the one-zone SSC model to reproduce the
SEDs following the parameters developed by Kelner et al. (2006). The pp interactions will produce secondary neutral
(π 0 ) and charged (π ± ) pions, which will decay into electrons/positrons (e± ), neutrinos (ν), and γ-ray emissions. The
pp interactions are comprised of the following


 p

 0
π −→ γ + γ
p + p −→ . (9)


 π + −→ µ+ + νµ −→ e+ + νe + ν̄µ + νµ
 −
π −→ µ− + ν̄µ −→ e− + ν̄e + νµ + ν̄µ

We adopted the analytical results from Li et al. (2022), which indicate that if the γ-ray emissions originate from π 0
decay and the jet power does not exceed the Eddington luminosity, the size of the emission region is constrained by
Rb σpp LEdd
⩽ , (10)
RS 12σT Lobs
TeV

where σpp ≈ 6×10−26 cm2 is the cross-section for pp interactions (Kelner et al. 2006), and Lob TeV is the observed TeV
luminosity. Applying the method from Xue et al. (2019) to calculate Lob TeV , we found that the maximum allowable Rb
13
is constrained to ∼ 10 cm. This implies a very compact emission region, smaller than the Schwarzschild radius (RS
= 2.44×1014 cm) of S5 0716+714, which is atypical for blazars. Consequently, the jet power somewhat exceeding the
Eddington luminosity appears necessary, which is plausible during a flaring state (Gao et al. 2019; Banik & Bhadra
2019; Banik et al. 2020). At the same time, Equation (11) in Li et al. (2022), which estimates the number density of
cold protons (nH ), was deemed inapplicable here.
We assumed that the relativistic protons follow a power-law with an exponential cut-off distribution:
 
γ
Np (γ) = Ap np (γ) = Ap γ −αp exp − , γp, min < γ < γp, max . (11)
γp, cut
1
Here, Ap = Np R np (γ)dγ , where Np refers to the actual density of relativistic protons in units of cm−3 . The parameter
αp is the power-law spectral index, γp, min/cut/max are the minimum, exponential cut-off, and maximum proton Lorentz
factors, respectively. The efficiency of pp interactions (fpp ) depends on the density of cold protons and can be estimated
using the expression:
fpp = Kpp σpp nH Rb , (12)
where Kpp ≈ 0.5 is the inelasticity coefficient, nH is the number density of cold protons (Kelner et al. 2006). We set
the number density of cold protons nH = 104 cm−3 and the proton Lorentz factors γp, min/cut/max = 1/102 /103 in our
modeling. The cold proton column density would be NH,cold ≃ nH Rb ≃ 4×1020 cm−2 , making the optical/UV and
X-ray emission absorbed via the photoionization absorption process. The optical depth for scattering is τsc = σsc nH Rb
where σsc is the scattering cross-section, expressed by
   
3 1 + x 2x(1 + x) 1 1 + 3x
σsc = σT − ln(1 + 2x) + ln(1 + 2x) − , (13)
4 x3 1 + 2x 2x (1 + 2x)2
with x = E/(me c2 ) and E represents the photon energy (see Liu et al. 2019b). The resulting flux should be multiplied
by a factor of (1 − exp(−τsc )) /τsc . However, the absorption effect is negligible since τsc ≪ 1 was obtained. During
the modeling, the Doppler factor was fixed at 30, the minimum electron Lorentz factor set to 1, while the remaining
parameters were treated as free. The fitting results are presented in Fig. 5 with the corresponding parameters listed
in Tab. 2.
The results indicate that the γ-ray emissions generated by π 0 decay via pp interactions successfully reproduce the
SEDs. Additionally, the synchrotron emission from secondary electron pairs contributes marginally to the radio and
X-ray fluxes compared to the emission from primary electrons. The energy densities of the cold protons and relativistic
protons in the SSC plus pp model are calculated as follows:

Up, cold = mp c2 nH , (14)

 11
E (eV) E (eV)
10 4 10 1 102 105 108 1011 10 4 10 1 102 105 108 1011
10 8 10 8
Sync ± Cascade Sum IceCube Sensitivity Sync ± Cascade Sum IceCube Sensitivity
SSC 0 Decay (pp) Phase A SSC 0 Decay (pp) Phase B
10 9 10 9
s 1)

s 1)
10 10 10 10
2

2
F (erg cm

F (erg cm
10 11 10 11

10 12 10 12

10 13 10 13

1011 1014 1017 1020 1023 1026 1011 1014 1017 1020 1023 1026
(Hz) (Hz)

Figure 5. One-zone SSC plus pp modeling. The left panel is for Phase A (MJD 57040–57050); the right one is for Phase
B (MJD 57065–57070). The VHE spectra are corrected by EBL absorption adopting z=0.2304. The meanings of line styles
are given in the legend. The dashed yellow line is the expected muon neutrino flux produced by the π ± cascade through pp
interactions. The dotted black line represents the IceCube sensitivity for declination δdel =60 using the 10 yrs dataset from
Ghiassi & Salwén (2023).
Z
Up, rel = mp c2 γNp (γ)dγ. (15)

The total jet power is estimated to be Pjet = 4.11 × 1048 erg s−1 for Phase A and Pjet = 3.09 × 1048 erg s−1 for
Phase B, which exceed the Eddington luminosity by a factor of 30-40, as detailed in Tab. 2. An enhanced activity
may result in a transient increase in jet power, potentially exceeding the Eddington luminosity, as discussed above.
Alternatively, in the case of a highly collimated jet outflow, the Eddington luminosity can be exceeded because the jet
does not interfere with the accretion flow (Gao et al. 2019). Moreover, VLBI analysis by MAGIC Collaboration et al.
(2018) suggests that a superluminal knot passing through a recollimation shock may occur during this flaring state.
Such interactions may increase the number density of electrons and protons in the emission region, further driving the
jet power to exceed the Eddington luminosity.
The dashed yellow line in Fig. 5 shows the muon neutrino (νµ ) flux produced by π ± cascade. It is necessary to
evaluate the possible neutrino emission under the hadronic model. The corresponding neutrino event rate can be
estimated via Z Eνµ , max
dNνµ
= dEνµ Aeff (Eνµ , δdel )ϕEνµ , (16)
dt Eνµ , min

where Eνµ , min and Eνµ , max are the lower and upper limits of the neutrino energy, respectively, Aeff (Eνµ , δdel ) is the
effective area in given declination (δdel ), and ϕEνµ is the muon neutrino differential energy flux. Then the expected
neutrino event rates for S5 0716+714, using the effective area from Carver (2019), are 0.69 events yr−1 for Phase A
and 0.52 events yr−1 for Phase B, respectively. However, the IceCube did not detect any neutrino events from S5
0716+715, because the expected neutrino flux was below its cumulative sensitivity, making it insufficient to detect
such events from this source, as shown in the dotted black line in Fig. 5.
In addition to the models mentioned above, the two-zone SSC model was also explored in MAGIC Collaboration
et al. (2018), which considers the interaction between a superluminal knot and a recollimation shock within the jet.
While this two-zone model provides a better fit to the SED, it cannot fully reproduce the observations in the 10-100
GeV range. Furthermore, other jet models, such as the structured jet model (MAGIC Collaboration et al. 2018)
and the helical magnetic field model (Chandra et al. 2015), have also been investigated during this flaring period.
These various models indicate the complexity of the emission mechanisms during flaring states, suggesting that further
multi-wavelength observations are needed to fully capture the underlying processes.

4. SUMMARY
In this study, we conducted a multi-wavelength analysis of the 2015 flare of S5 0716+714 to investigate its radiation
mechanisms. The data were gathered from the Swift-UVOT, Swift-XRT, NuSTAR, and Fermi -LAT databases, along
12

Table 2. Parameters for the broadband SEDs for different models

Phase A Phase B
Params Units
SSC SSC+EC SSC+pp interactions SSC SSC+EC SSC+pp interactions

(1) (2) (3) (4) (5) (6) (7) (8)

Rb † cm 4.00×1016 4.00×1016 4.00×1016 4.00×1016 4.00×1016 4.00×1016

B G 4.45×10−3 8.35×10−2 1.95×10−1 3.78×10−3 7.10×10−2 1.61×10−1
δ 204.44 73.95 30 † 188.40 57.84 30 †
γe, min 1† 77.44 1† 1† 96.69 1†
γe, max † 107 107 107 107 107 107
Ne cm−3 123.81 25.37 2071.33 104.12 23.85 1674.66
γe, 0 911.98 2540.8 1863.48 414.73 2961.56 329.75
s 1.57 2.54 1.99 1.52 2.55 1.85
r 0.64 0.78 0.87 0.47 0.59 0.48
τBLR † – 0.1 – – 0.1 –
TDT † K – 1200 – – 1200 –
τDT † – 0.1 – – 0.1 –
LDisk † erg s−1 – 2×1042 – – 2×1042 –
TDisk † K – 2×104 – – 2×104 –
θopen † deg – 5 – – 5 –
RH † cm – 4.57×1017 – – 4.57×1017
γp, min † – – 1 – – 1
γp, max † – – 1000 – – 1000
Np cm−3 – – 3185.17 – – 1318.35
γp, cut † – – 100 – – 100
αp – – 2.25 – – 2.07
nH † cm−3 – – 104 – – 104
fpp – – 1.20×10−5 – – 1.20×10−5
Ue erg cm−3 5.60×10−3 4.26×10−3 1.60×10−2 4.79×10−3 4.97×10−3 1.72×10−2
Up, cold erg cm−3 1.86×10−1 3.81×10−2 1.50×101 1.57×10−1 3.59×10−2 1.50×101
UB erg cm−3 7.89×10−7 2.77×10−4 1.51×10−3 5.68×10−7 2.01×10−4 1.03×10−3
Pe erg s−1 3.53×1046 3.51×1045 2.16×1045 2.56×1046 2.51×1045 2.33×1045
PB erg s−1 4.97×1042 2.28×1044 2.05×1044 3.04×1042 1.01×1044 1.39×1044
Pp, cold erg s−1 1.17×1048 3.14×1046 2.04×1048 8.37×1047 1.81×1046 2.04×1048
UBLR erg cm−3 – 2.70×10−10 – – 2.93×10−10 –
UDT erg cm−3 – 1.84×10−4 – – 1.13×10−4 –
Up, rel erg cm−3 – – 1.52×101 – – 7.70×100
Pp, rel erg s−1 – – 2.07×1048 – – 1.04×1048
Ptot erg s−1 1.21×1048 3.52×1046 4.11×1048 8.63×1047 2.07×1046 3.09×1048
Note— Parameters with the symbol “†” represent that they keep frozen in the SED fitting. The redshift is adopted as
0.2304. The symbol “–” represents a null value. The Eddington luminosity is calculated as LEdd = 1.02 × 1047 erg s−1 for
the SMBH mass 108.91 M⊙ for S5 0716+714 (Liu et al. 2019a).
 13

with the MAGIC VHE data from the literature. These observations allowed us to estimate the size of the emission
region, while the modeling of the broad-band SEDs for Phase A and Phase B provided valuable insights into the
physical processes occurring during the flaring periods. The main results are as follows. The size of the γ-ray emission
region was estimated using the variability timescale, determined through exponential function fitting. Subsequently,
one-zone models, including leptonic and lepto-hadronic hybrid scenarios, were employed to reproduce the SEDs for
Phase A and Phase B. However, the SSC models could not adequately describe the SEDs without invoking extreme
Doppler factors, potentially requiring an additional Lorentz factor as suggested in the “jets-in-a-jet” model. The SSC
plus EC model provided a good fit to the SEDs but required a high Doppler factor, leading to its exclusion from our
consideration. Additionally, the SSC plus pp interactions model was explored, and the results demonstrated that this
model successfully reproduced the SEDs. Nevertheless, the total jet power in this scenario exceeded the Eddington
luminosity, a situation that is still plausible due to the flaring state or the presence of a highly collimated jet.

H.B.X. acknowledges the support from the National Natural Science Foundation of China (NSFC 12203034), the
Shanghai Science and Technology Fund (22YF1431500), and the science research grants from the China Manned Space
Project. MM acknowledges the Croatian Science Foundation (HrZZ) Project IP-2022-10-4595. R.X acknowledges the
support from the NSFC under grant No. 12203043. G.W. acknowledges the support from the China Postdoctoral
Science Foundation (grant No. 2023M730523). S.H.Z acknowledges support from the National Natural Science Foun-
dation of China (Grant No. 12173026), the National Key Research and Development Program of China (Grant No.
2022YFC2807303), the Shanghai Science and Technology Fund (Grant No. 23010503900), the Program for Professor
of Special Appointment (Eastern Scholar) at Shanghai Institutions of Higher Learning and the Shuguang Program
(23SG39) of the Shanghai Education Development Foundation and Shanghai Municipal Education Commission. J.H.F
acknowledges the support from the NSFC 12433004, NSFC U2031201, the Scientific and Technological Cooperation
Projects (2020–2023) between the People’s Republic of China and the Republic of Bulgaria, the science research grants
from the China Manned Space Project with No. CMS-CSST-2021-A06, and the support for Astrophysics Key Sub-
jects of Guangdong Province and Guangzhou City. This research was partially supported by the Bulgarian National
Science Fund of the Ministry of Education and Science under grants KP-06-H38/4 (2019), KP-06-KITAJ/2 (2020) and
KP-06-H68/4 (2022).

APPENDIX

A. SWIFT OBSERVATION DATA

The Swift-XRT best-fit parameters with an absorbed power-law model are shown in Tab. 3.

The Swift-UVOT aperture photometric data after the Galactic extinction correction are shown in Tab. 4.

REFERENCES
Abdo, A. A., Ackermann, M., Ajello, M., et al. 2010, ApJ, Anderhub, H., Antonelli, L. A., Antoranz, P., et al. 2009,
722, 520, doi: 10.1088/0004-637X/722/1/520 ApJL, 704, L129, doi: 10.1088/0004-637X/704/2/L129
Abdollahi, S., Acero, F., Baldini, L., et al. 2022, ApJS, 260, Atwood, W. B., Abdo, A. A., Ackermann, M., et al. 2009,
53, doi: 10.3847/1538-4365/ac6751 ApJ, 697, 1071, doi: 10.1088/0004-637X/697/2/1071
Aharonian, F., Akhperjanian, A. G., Bazer-Bachi, A. R., Bach, U., Krichbaum, T. P., Ros, E., et al. 2005, A&A, 433,
et al. 2007, ApJL, 664, L71, doi: 10.1086/520635 815, doi: 10.1051/0004-6361:20040388
Ajello, M., Atwood, W. B., Axelsson, M., et al. 2021, ApJS, Bachev, R., Spassov, B., & Boeva, S. 2015, The
256, 12, doi: 10.3847/1538-4365/ac0ceb Astronomer’s Telegram, 6944, 1
Albert, J., Aliu, E., Anderhub, H., et al. 2007, ApJ, 669, Bachev, R., & Strigachev, A. 2015, The Astronomer’s
862, doi: 10.1086/521382 Telegram, 6957, 1
14

Table 3. Swift-XRT observations

ObsID Mode Exposure Time ΓX ΓX, err N0 N0, err Flux Fluxerr C-statistic d.o.f.

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)

00035009143 PC 921.50 57011.04 2.18 0.23 1.08E-03 1.62E-04 5.60E-12 9.57E-13 72.59 91
00035009144 PC 1066.34 57019.35 1.78 0.18 1.32E-03 1.71E-04 8.61E-12 1.63E-12 142.18 137
00035009145 PC 1051.36 57023.21 2.26 0.18 1.55E-03 1.82E-04 7.80E-12 1.00E-12 99.42 138
00035009146 PC 829.10 57029.01 1.81 0.22 1.27E-03 1.88E-04 8.08E-12 1.78E-12 119.58 105
00035009147 PC 991.42 57041.09 2.22 0.12 4.08E-03 3.01E-04 2.08E-11 1.72E-12 187.33 217
00035009148 PC 1103.80 57042.75 2.11 0.11 3.93E-03 2.92E-04 2.10E-11 1.87E-12 199.73 230
00035009149 PC 968.95 57043.41 2.05 0.15 4.01E-03 3.91E-04 2.19E-11 2.65E-12 156.73 179
00035009152 PC 1356.03 57044.02 2.29 0.09 5.85E-03 3.53E-04 2.93E-11 1.91E-12 224.78 259
00035009153 PC 6912.53 57044.29 2.19 0.06 4.98E-03 1.90E-04 2.57E-11 1.12E-12 349.31 382
00035009154 PC 998.94 57045.01 2.31 0.11 7.10E-03 4.85E-04 3.53E-11 2.57E-12 168.92 228
00035009156 PC 9574.62 57045.14 2.43 0.04 7.95E-03 2.19E-04 3.87E-11 1.09E-12 425.84 425
00035009157 PC 1688.17 57047.14 2.42 0.09 9.34E-03 5.32E-04 4.55E-11 2.65E-12 211.43 249
00035009158 PC 6557.90 57047.22 2.46 0.05 9.28E-03 2.67E-04 4.50E-11 1.31E-12 352.41 391
00035009159 PC 1490.88 57048.73 2.38 0.10 8.14E-03 5.15E-04 3.99E-11 2.62E-12 211.88 226
00035009160 PC 1490.88 57048.86 2.41 0.10 6.32E-03 3.86E-04 3.08E-11 1.93E-12 189.58 238
00035009161 PC 1490.88 57049.66 2.28 0.08 6.26E-03 3.31E-04 3.14E-11 1.79E-12 247.94 285
00035009162 WT 1007.11 57050.01 2.73 0.12 5.77E-03 4.11E-04 2.81E-11 2.05E-12 222.56 247
00035009167 PC 3161.58 57051.26 2.02 0.06 3.73E-03 1.65E-04 2.07E-11 1.15E-12 328.36 366
00035009164 WT 198.21 57051.66 2.37 0.37 4.12E-03 8.68E-04 2.02E-11 4.47E-12 60.00 87
00035009168 PC 2469.83 57051.85 2.09 0.07 3.50E-03 1.76E-04 1.88E-11 1.15E-12 291.47 329
00035009169 WT 5488.26 57052.27 2.25 0.06 4.82E-03 1.58E-04 2.43E-11 8.90E-13 472.61 515
00035009170 WT 6144.92 57052.99 2.21 0.06 4.36E-03 1.39E-04 2.23E-11 8.23E-13 548.34 553
00035009171 WT 5502.71 57054.46 2.29 0.05 5.15E-03 1.47E-04 2.57E-11 8.00E-13 471.24 518
00035009172 WT 1082.94 57055.45 2.37 0.11 5.15E-03 3.38E-04 2.53E-11 1.74E-12 242.50 266
00035009173 WT 1072.83 57056.31 2.51 0.07 1.08E-02 4.64E-04 5.20E-11 2.25E-12 301.62 329
00035009174 PC 1483.39 57057.64 2.34 0.08 7.01E-03 3.70E-04 3.47E-11 1.93E-12 262.78 269
00035009175 WT 2069.23 57058.52 2.45 0.06 7.57E-03 2.84E-04 3.68E-11 1.40E-12 348.40 384
00035009176 WT 5406.15 57058.65 2.46 0.04 8.62E-03 1.90E-04 4.18E-11 9.35E-13 494.67 549
00035009177 WT 13689.18 57059.05 2.49 0.02 8.46E-03 1.17E-04 4.10E-11 5.69E-13 739.93 696
00035009178 WT 12428.53 57060.05 2.47 0.03 7.29E-03 1.21E-04 3.53E-11 5.93E-13 636.68 682
00035009179 WT 12276.77 57061.04 2.41 0.03 6.65E-03 1.17E-04 3.25E-11 5.86E-13 602.31 683
00035009180 WT 16028.34 57062.05 2.42 0.03 5.85E-03 9.51E-05 2.85E-11 4.75E-13 605.02 701
00035009181 WT 1122.72 57063.96 2.39 0.11 5.01E-03 3.52E-04 2.45E-11 1.78E-12 206.55 252
00035009182 WT 978.13 57064.30 2.60 0.13 5.06E-03 3.92E-04 2.44E-11 1.89E-12 197.69 228
00035009184 WT 648.52 57066.03 2.27 0.14 6.74E-03 5.79E-04 3.39E-11 3.18E-12 202.84 223
00035009185 WT 4604.37 57066.09 2.46 0.04 7.65E-03 1.89E-04 3.72E-11 9.29E-13 484.02 501
00035009186 WT 4470.44 57066.10 2.47 0.05 7.80E-03 2.06E-04 3.78E-11 1.01E-12 463.70 486
00035009187 WT 998.11 57068.29 2.55 0.10 6.77E-03 3.98E-04 3.27E-11 1.92E-12 225.36 273
00035009188 WT 831.23 57068.82 2.66 0.15 4.39E-03 3.79E-04 2.12E-11 1.85E-12 186.74 212
00035009189 WT 1028.06 57069.95 2.40 0.14 3.91E-03 3.16E-04 1.91E-11 1.60E-12 209.45 228
00035009190 PC 1475.90 57067.91 2.34 0.10 7.71E-03 4.81E-04 3.81E-11 2.50E-12 218.48 243
00035009192 WT 1077.83 57070.23 2.46 0.18 3.75E-03 3.81E-04 1.82E-11 1.88E-12 200.21 207
00035009193 WT 1097.78 57070.75 2.37 0.14 3.48E-03 2.92E-04 1.71E-11 1.49E-12 202.51 228
00035009194 WT 878.14 57071.09 2.47 0.14 4.06E-03 3.47E-04 1.97E-11 1.70E-12 181.57 222
00035009195 WT 1062.83 57071.69 2.48 0.13 4.25E-03 3.23E-04 2.06E-11 1.58E-12 232.91 242
00035009196 PC 1018.89 57072.01 2.28 0.13 4.08E-03 3.37E-04 2.05E-11 1.84E-12 147.91 189
00035009197 PC 476.99 57072.75 2.21 0.17 3.95E-03 4.34E-04 2.02E-11 2.50E-12 108.49 145
00035009198 PC 1016.40 57073.01 2.31 0.12 4.28E-03 3.40E-04 2.13E-11 1.81E-12 136.03 198
00035009199 PC 1111.29 57073.61 2.18 0.10 4.15E-03 2.85E-04 2.14E-11 1.68E-12 223.43 237
00035009200 PC 1016.40 57074.14 2.19 0.12 3.53E-03 2.75E-04 1.82E-11 1.62E-12 210.63 208
00035009201 PC 1073.83 57074.61 2.17 0.14 2.83E-03 2.59E-04 1.47E-11 1.55E-12 149.23 175

Note— Col. (1): the ObsID; Col. (2): the readout mode, PC represents the Photon Counting mode and WT represents the
Windowed Timing mode; Col. (3): the net exposure time of Swift-XRT in units of second; Col. (4): the start time (MJD) of
the Swift observation. Col. (5): the Swift-XRT photon index; Col. (6): the error of Swift-XRT photon index; Col. (7): the
normalization flux in units of cm−2 s−1 keV−1 ; Col. (8): the error of normalization flux in units of cm−2 s−1 keV−1 ; Col. (9): the
unabsorbed flux in 0.3−10 keV in units of erg cm−2 s−1 ; Col. (10): the error of the unabsorbed flux in 0.3−10 keV in units of
erg cm−2 s−1 ; Col. (11): the Cash statistic; Col. (12): the degrees of freedom.
 Table 4. Swift-UVOT aperture photometric data

u band b band v band uvw1 band uvw2 band uvm2 band

ObsID MJD Fλ Fλ, err MJD Fλ Fλ, err MJD Fλ Fλ, err MJD Fλ Fλ, err MJD Fλ Fλ, err MJD Fλ Fλ, err

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19)

00035009143 57011.04 1.12E-14 4.18E-16 57011.04 1.00E-14 3.23E-16 57011.04 8.32E-15 2.87E-16 57011.04 1.30E-14 6.43E-16 57011.04 1.57E-14 6.72E-16 57011.05 1.49E-14 6.08E-16
00035009144 57019.36 1.24E-14 4.51E-16 57019.36 1.07E-14 3.36E-16 57019.36 8.82E-15 2.90E-16 57019.36 1.33E-14 6.51E-16 57019.36 1.61E-14 6.84E-16 57019.37 1.56E-14 6.19E-16
00035009145 57023.22 1.04E-14 3.84E-16 57023.22 8.71E-15 2.79E-16 57023.22 7.52E-15 2.56E-16 57023.22 1.13E-14 5.58E-16 57023.22 1.41E-14 6.05E-16 57023.23 1.37E-14 5.50E-16
00035009146 57029.01 1.30E-14 4.83E-16 57029.01 1.16E-14 3.75E-16 57029.01 9.70E-15 3.33E-16 57029.01 1.42E-14 7.03E-16 57029.01 1.79E-14 7.69E-16 57029.01 1.77E-14 7.10E-16
00035009147 57041.09 6.36E-14 2.35E-15 57041.09 5.42E-14 1.73E-15 57041.10 5.05E-14 1.40E-15 57041.09 8.53E-14 4.00E-15 57041.10 1.09E-13 4.46E-15 57041.10 1.08E-13 3.94E-15
00035009148 57042.75 5.32E-14 1.89E-15 57042.75 4.46E-14 1.35E-15 57042.76 3.93E-14 1.07E-15 57042.75 6.22E-14 2.91E-15 57042.76 7.64E-14 3.13E-15 57042.76 7.60E-14 2.77E-15
00035009149 – – – – – – – – – – – – – – – – – –
00035009152 57044.08 4.95E-14 1.77E-15 57044.08 4.05E-14 1.23E-15 57044.09 3.60E-14 9.96E-16 57044.08 5.96E-14 2.80E-15 57044.09 7.40E-14 3.04E-15 57044.09 7.46E-14 2.72E-15
00035009153 57044.29 4.55E-14 1.50E-15 – – – – – – – – – – – – – – –
57044.36 4.51E-14 1.47E-15 – – – – – – – – – – – – – – –
57044.42 4.61E-14 1.50E-15 – – – – – – – – – – – – – – –
57044.48 4.95E-14 1.63E-15 – – – – – – – – – – – – – – –
57044.55 5.18E-14 1.68E-15 – – – – – – – – – – – – – – –
57044.62 5.33E-14 1.74E-15 – – – – – – – – – – – – – – –
57044.68 5.29E-14 1.72E-15 – – – – – – – – – – – – – – –
57044.81 5.61E-14 1.84E-15 – – – – – – – – – – – – – – –
57044.95 6.02E-14 1.97E-15 – – – – – – – – – – – – – – –
00035009154 57045.01 5.76E-14 2.05E-15 57045.01 4.89E-14 1.48E-15 57045.02 4.43E-14 1.19E-15 57045.01 7.69E-14 3.59E-15 57045.02 1.00E-13 4.09E-15 57045.02 9.95E-14 3.60E-15
00035009156 – – – – – – – – – – – – 57045.15 1.03E-13 4.18E-15 57046.15 1.28E-13 4.56E-15
– – – – – – – – – – – – 57045.28 1.05E-13 4.27E-15 57046.22 1.19E-13 4.25E-15
– – – – – – – – – – – – 57045.41 1.16E-13 4.71E-15 57046.28 1.10E-13 3.90E-15
– – – – – – – – – – – – 57045.48 1.18E-13 4.76E-15 57046.40 1.19E-13 4.32E-15
– – – – – – – – – – – – 57045.68 1.24E-13 5.03E-15 57046.48 1.14E-13 4.03E-15
00035009157 57047.14 7.07E-14 2.79E-15 57047.14 5.96E-14 2.05E-15 57047.14 5.67E-14 1.65E-15 57047.14 9.90E-14 4.68E-15 57047.48 1.13E-13 4.62E-15 57047.48 1.10E-13 3.98E-15
57047.48 6.44E-14 2.40E-15 57047.48 5.44E-14 1.75E-15 57047.48 4.84E-14 1.36E-15 – – – – – – – – –
00035009158 – – – – – – – – – 57047.22 1.10E-13 5.07E-15 – – – – – –
– – – – – – – – – 57047.42 8.51E-14 3.91E-15 – – – – – –
– – – – – – – – – 57047.47 8.72E-14 4.10E-15 – – – – – –
– – – – – – – – – 57047.61 9.68E-14 4.45E-15 – – – – – –
– – – – – – – – – 57047.67 1.00E-13 4.61E-15 – – – – – –
– – – – – – – – – 57047.80 9.43E-14 4.35E-15 – – – – – –
– – – – – – – – – 57047.87 8.45E-14 3.88E-15 – – – – – –
– – – – – – – – – 57047.93 8.22E-14 3.79E-15 – – – – – –
00035009159 57048.73 6.38E-14 2.36E-15 57048.73 5.19E-14 1.64E-15 57048.73 4.70E-14 1.30E-15 57048.73 8.48E-14 3.98E-15 57048.73 1.09E-13 4.47E-15 57048.74 1.08E-13 3.84E-15
00035009160 57048.87 6.49E-14 2.11E-15 – – – – – – – – – – – – – – –
00035009161 – – – – – – – – – – – – – – – – – –
00035009162 57050.01 4.95E-14 1.92E-15 57050.01 4.22E-14 1.44E-15 57050.01 3.65E-14 1.14E-15 – – – – – – – – –
57050.14 4.41E-14 1.70E-15 57050.14 3.58E-14 1.20E-15 57050.15 3.14E-14 9.95E-16 – – – – – – – – –
00035009167 57051.26 3.56E-14 1.36E-15 57051.32 3.05E-14 9.21E-16 57051.33 2.63E-14 7.49E-16 57051.26 4.23E-14 2.00E-15 57051.33 5.34E-14 2.20E-15 57051.33 5.21E-14 1.86E-15
57051.32 3.56E-14 1.26E-15 57051.39 2.94E-14 8.88E-16 57051.40 2.47E-14 7.06E-16 57051.32 4.28E-14 2.02E-15 57051.39 5.12E-14 2.11E-15 57051.40 5.45E-14 6.72E-15
57051.39 3.52E-14 1.24E-15 57051.46 2.75E-14 8.31E-16 – – – 57051.39 4.11E-14 1.94E-15 57051.46 4.82E-14 2.01E-15 – – –
57051.46 3.31E-14 1.17E-15 – – – – – – 57051.45 3.82E-14 1.81E-15 – – – – – –
00035009164 57051.67 3.52E-14 1.32E-15 – – – – – – 57051.66 4.13E-14 1.95E-15 – – – – – –
00035009168 57051.93 3.36E-14 1.19E-15 57051.86 2.83E-14 8.54E-16 57051.86 2.40E-14 6.87E-16 57051.92 3.78E-14 1.79E-15 57051.93 4.65E-14 1.92E-15 – – –
57051.95 3.30E-14 1.17E-15 57051.93 2.77E-14 8.35E-16 57051.93 2.37E-14 6.97E-16 57051.95 3.83E-14 1.81E-15 – – – – – –
57051.99 3.06E-14 1.08E-15 57051.99 2.66E-14 8.04E-16 – – – – – – – – – – – –
00035009169 57052.27 3.04E-14 9.93E-16 57052.79 3.01E-14 8.60E-16 57052.80 2.65E-14 6.96E-16 – – – – – – – – –
57052.79 3.47E-14 1.18E-15 – – – – – – – – – – – – – – –
57052.86 3.54E-14 1.15E-15 – – – – – – – – – – – – – – –
57052.99 3.28E-14 1.08E-15 – – – – – – – – – – – – – – –
00035009170 57053.00 3.20E-14 1.04E-15 57053.73 2.22E-14 6.58E-16 57053.73 1.88E-14 5.34E-16 – – – 57053.07 4.79E-14 1.95E-15 – – –
57053.73 2.62E-14 9.14E-16 – – – – – – – – – 57053.14 4.21E-14 1.71E-15 – – –
– – – – – – – – – – – – 57053.20 4.19E-14 1.70E-15 – – –
– – – – – – – – – – – – 57053.54 4.15E-14 1.70E-15 – – –
00035009171 57054.65 2.78E-14 9.48E-16 57054.65 2.34E-14 6.72E-16 57054.66 1.94E-14 5.24E-16 57054.65 3.29E-14 1.54E-15 57054.66 4.03E-14 1.65E-15 57054.59 4.04E-14 1.44E-15
– – – – – – – – – – – – – – – 57054.66 3.92E-14 1.44E-15
15

00035009172 57055.45 3.59E-14 1.27E-15 57055.45 2.93E-14 8.80E-16 57055.46 2.44E-14 6.94E-16 57055.45 4.15E-14 1.96E-15 57055.45 5.20E-14 2.14E-15 57055.46 5.03E-14 1.85E-15
00035009173 57056.31 4.42E-14 1.58E-15 57056.31 3.54E-14 1.08E-15 57056.32 3.09E-14 8.71E-16 57056.31 5.44E-14 2.56E-15 57056.32 6.94E-14 2.85E-15 57056.32 6.80E-14 2.47E-15
00035009174 57057.64 4.21E-14 1.50E-15 57057.64 3.54E-14 1.08E-15 57057.65 2.90E-14 8.17E-16 – – – 57057.64 6.85E-14 2.81E-15 57057.65 6.69E-14 2.39E-15
00035009175 57058.52 4.43E-14 1.60E-15 57058.52 3.68E-14 1.13E-15 57058.53 3.06E-14 8.75E-16 57058.52 5.50E-14 2.59E-15 57058.52 7.11E-14 2.93E-15 57058.53 6.94E-14 2.53E-15
57058.59 4.37E-14 1.54E-15 57058.59 3.52E-14 1.06E-15 57058.59 2.98E-14 8.29E-16 57058.59 5.48E-14 2.57E-15 57058.59 7.00E-14 2.87E-15 57058.60 6.79E-14 2.47E-15
00035009176 – – – – – – – – – – – – – – – – – –
16

Banik, P., & Bhadra, A. 2019, PhRvD, 99, 103006, Edelson, R. A., & Krolik, J. H. 1988, ApJ, 333, 646,
doi: 10.1103/PhysRevD.99.103006 doi: 10.1086/166773
Banik, P., Bhadra, A., Pandey, M., & Majumdar, D. 2020, Fan, J.-H., Wang, Y.-J., Yang, J.-H., & Su, C.-Y. 2004,
PhRvD, 101, 063024, doi: 10.1103/PhysRevD.101.063024 ChJA&A, 4, 533, doi: 10.1088/1009-9271/4/6/533
Barthelmy, S. D., Barbier, L. M., Cummings, J. R., et al. Fan, J. H., Yang, J. H., Liu, Y., et al. 2016, ApJS, 226, 20,
2005, SSRv, 120, 143, doi: 10.1007/s11214-005-5096-3 doi: 10.3847/0067-0049/226/2/20
Blażejowski, M., Sikora, M., Moderski, R., & Madejski, Fermi Science Support Development Team. 2019,
G. M. 2000, ApJ, 545, 107, doi: 10.1086/317791 Fermitools: Fermi Science Tools, Astrophysics Source
Böttcher, M., Reimer, A., & Marscher, A. P. 2009, ApJ, Code Library, record ascl:1905.011
703, 1168, doi: 10.1088/0004-637X/703/1/1168 Finke, J. D., Dermer, C. D., & Böttcher, M. 2008, ApJ,
Breeveld, A. A., Landsman, W., Holland, S. T., et al. 2011, 686, 181, doi: 10.1086/590900
in American Institute of Physics Conference Series, Vol. Fitzpatrick, E. L. 1999, PASP, 111, 63, doi: 10.1086/316293
1358, Gamma Ray Bursts 2010, ed. J. E. McEnery, J. L. Foschini, L., Tagliaferri, G., Pian, E., et al. 2006, A&A,
Racusin, & N. Gehrels, 373–376, doi: 10.1063/1.3621807 455, 871, doi: 10.1051/0004-6361:20064959
Breeveld, A. A., Curran, P. A., Hoversten, E. A., et al. Gao, S., Fedynitch, A., Winter, W., & Pohl, M. 2019,
2010, MNRAS, 406, 1687, Nature Astronomy, 3, 88, doi: 10.1038/s41550-018-0610-1
doi: 10.1111/j.1365-2966.2010.16832.x Gasparyan, S., Sahakyan, N., Baghmanyan, V., &
Burrows, D. N., Hill, J. E., Nousek, J. A., et al. 2004, in Zargaryan, D. 2018, ApJ, 863, 114,
Society of Photo-Optical Instrumentation Engineers doi: 10.3847/1538-4357/aad234
(SPIE) Conference Series, Vol. 5165, X-Ray and Gehrels, N., Chincarini, G., Giommi, P., et al. 2004, ApJ,
Gamma-Ray Instrumentation for Astronomy XIII, ed. 611, 1005, doi: 10.1086/422091
K. A. Flanagan & O. H. W. Siegmund, 201–216, Geng, X., Zeng, W., Rani, B., et al. 2020, ApJ, 904, 67,
doi: 10.1117/12.504868 doi: 10.3847/1538-4357/abb603
Cappi, M., Comastri, A., Molendi, S., et al. 1994, MNRAS, Ghiassi, K., & Salwén, J. 2023, Neutrino Hotspots in the
271, 438, doi: 10.1093/mnras/271.2.438 Universe: a Sensitivity Study Using the IceCube
Carrasco, L., Porras, A., Recillas, E., et al. 2015, The Neutrino Observatory
Astronomer’s Telegram, 6902, 1 Ghisellini, G., & Tavecchio, F. 2009, MNRAS, 397, 985,
Carver, T. 2019, in International Cosmic Ray Conference, doi: 10.1111/j.1365-2966.2009.15007.x
Vol. 36, 36th International Cosmic Ray Conference Ghisellini, G., Tavecchio, F., Foschini, L., et al. 2010,
(ICRC2019), 851, doi: 10.22323/1.358.0851 MNRAS, 402, 497, doi: 10.1111/j.1365-2966.2009.15898.x
Cash, W. 1979, ApJ, 228, 939, doi: 10.1086/156922 Giannios, D., Uzdensky, D. A., & Begelman, M. C. 2009,
Cerruti, M., Zech, A., Boisson, C., et al. 2019, MNRAS, MNRAS, 395, L29, doi: 10.1111/j.1745-3933.2009.00635.x
483, L12, doi: 10.1093/mnrasl/sly210 —. 2010, MNRAS, 402, 1649,
Cerruti, M., Zech, A., Boisson, C., & Inoue, S. 2015, doi: 10.1111/j.1365-2966.2009.16045.x
MNRAS, 448, 910, doi: 10.1093/mnras/stu2691 Gupta, A. C., Krichbaum, T. P., Wiita, P. J., et al. 2012,
Chandra, S., Zhang, H., Kushwaha, P., et al. 2015, ApJ, MNRAS, 425, 1357,
809, 130, doi: 10.1088/0004-637X/809/2/130 doi: 10.1111/j.1365-2966.2012.21550.x
Cleary, K., Lawrence, C. R., Marshall, J. A., Hao, L., & Harrison, F. A., Craig, W. W., Christensen, F. E., et al.
Meier, D. 2007, ApJ, 660, 117, doi: 10.1086/511969 2013, ApJ, 770, 103, doi: 10.1088/0004-637X/770/2/103
Dai, B.-z., Zeng, W., Jiang, Z.-j., et al. 2015, ApJS, 218, 18, Hartman, R. C., Bertsch, D. L., Bloom, S. D., et al. 1999,
doi: 10.1088/0067-0049/218/2/18 ApJS, 123, 79, doi: 10.1086/313231
Danforth, C. W., Nalewajko, K., France, K., & Keeney, HI4PI Collaboration, Ben Bekhti, N., Flöer, L., et al. 2016,
B. A. 2013, ApJ, 764, 57, A&A, 594, A116, doi: 10.1051/0004-6361/201629178
doi: 10.1088/0004-637X/764/1/57 Ikejiri, Y., Uemura, M., Sasada, M., et al. 2011, PASJ, 63,
Dermer, C. D., & Schlickeiser, R. 1993, ApJ, 416, 458, 639, doi: 10.1093/pasj/63.3.327
doi: 10.1086/173251 Kaspi, S., Brandt, W. N., Maoz, D., et al. 2007, ApJ, 659,
Domı́nguez, A., Primack, J. R., Rosario, D. J., et al. 2011, 997, doi: 10.1086/512094
MNRAS, 410, 2556, Kelner, S. R., Aharonian, F. A., & Bugayov, V. V. 2006,
doi: 10.1111/j.1365-2966.2010.17631.x PhRvD, 74, 034018, doi: 10.1103/PhysRevD.74.034018
 17

Kraus, A., Krichbaum, T. P., Wegner, R., et al. 2003, A&A, Scarpa, R., & Falomo, R. 1997, A&A, 325, 109
401, 161, doi: 10.1051/0004-6361:20030118 Schlafly, E. F., & Finkbeiner, D. P. 2011, ApJ, 737, 103,
Kuehr, H., Witzel, A., Pauliny-Toth, I. I. K., & Nauber, U. doi: 10.1088/0004-637X/737/2/103
1981, A&AS, 45, 367 Shaw, M. S., Romani, R. W., Healey, S. E., et al. 2009,
Li, W.-J., Xue, R., Long, G.-B., et al. 2022, A&A, 659, ApJ, 704, 477, doi: 10.1088/0004-637X/704/1/477
A184, doi: 10.1051/0004-6361/202142051 Shukla, A., & Mannheim, K. 2020, Nature
Liao, N. H., Bai, J. M., Liu, H. T., et al. 2014, ApJ, 783, Communications, 11, 4176,
83, doi: 10.1088/0004-637X/783/2/83 doi: 10.1038/s41467-020-17912-z
Lin, Y. C., Bertsch, D. L., Dingus, B. L., et al. 1995, ApJ, Sikora, M., Begelman, M. C., & Rees, M. J. 1994, ApJ, 421,
442, 96, doi: 10.1086/175427 153, doi: 10.1086/173633
Lister, M. L., Aller, M. F., Aller, H. D., et al. 2018, ApJS, Stickel, M., Padovani, P., Urry, C. M., Fried, J. W., &
234, 12, doi: 10.3847/1538-4365/aa9c44 Kuehr, H. 1991, ApJ, 374, 431, doi: 10.1086/170133
Lister, M. L., Homan, D. C., Hovatta, T., et al. 2019, ApJ, Tagliaferri, G., Ravasio, M., Ghisellini, G., et al. 2003,
874, 43, doi: 10.3847/1538-4357/ab08ee A&A, 400, 477, doi: 10.1051/0004-6361:20021916
Liu, H. T., Feng, H. C., Xin, Y. X., et al. 2019a, ApJ, 880, Tan, C., Xue, R., Du, L.-M., et al. 2020, ApJS, 248, 27,
155, doi: 10.3847/1538-4357/ab29fc doi: 10.3847/1538-4365/ab8cc6
Liu, R.-Y., Wang, K., Xue, R., et al. 2019b, PhRvD, 99, Tavecchio, F., Maraschi, L., & Ghisellini, G. 1998, ApJ,
063008, doi: 10.1103/PhysRevD.99.063008
509, 608, doi: 10.1086/306526
Lyutikov, M., & Kravchenko, E. V. 2017, MNRAS, 467,
Tramacere, A. 2020, JetSeT: Numerical modeling and SED
3876, doi: 10.1093/mnras/stx359
fitting tool for relativistic jets, Astrophysics Source Code
MAGIC Collaboration, Ahnen, M. L., Ansoldi, S., et al.
Library, record ascl:2009.001
2018, A&A, 619, A45, doi: 10.1051/0004-6361/201832677
Tramacere, A., Giommi, P., Perri, M., Verrecchia, F., &
Mirzoyan, R. 2015, The Astronomer’s Telegram, 6999, 1
Tosti, G. 2009, A&A, 501, 879,
Morrison, R., & McCammon, D. 1983, ApJ, 270, 119,
doi: 10.1051/0004-6361/200810865
doi: 10.1086/161102
Tramacere, A., Massaro, E., & Taylor, A. M. 2011, ApJ,
Mücke, A., Protheroe, R. J., Engel, R., Rachen, J. P., &
739, 66, doi: 10.1088/0004-637X/739/2/66
Stanev, T. 2003, Astroparticle Physics, 18, 593,
Tripathi, T., Gupta, A. C., Takey, A., et al. 2024, MNRAS,
doi: 10.1016/S0927-6505(02)00185-8
527, 5220, doi: 10.1093/mnras/stad3574
Nolan, P. L., Abdo, A. A., Ackermann, M., et al. 2012,
Urry, C. M., & Padovani, P. 1995, PASP, 107, 803,
ApJS, 199, 31, doi: 10.1088/0067-0049/199/2/31
doi: 10.1086/133630
Perley, R. A., Fomalont, E. B., & Johnston, K. J. 1980, AJ,
Verner, D. A., Ferland, G. J., Korista, K. T., & Yakovlev,
85, 649, doi: 10.1086/112723
D. G. 1996, ApJ, 465, 487, doi: 10.1086/177435
Pichel, A., Donzelli, C., Muriel, H., et al. 2023, A&A, 680,
A52, doi: 10.1051/0004-6361/202245574 Wagner, S. J., Witzel, A., Heidt, J., et al. 1996, AJ, 111,
Poole, T. S., Breeveld, A. A., Page, M. J., et al. 2008, 2187, doi: 10.1086/117954
MNRAS, 383, 627, doi: 10.1111/j.1365-2966.2007.12563.x Wang, G., Fan, J., Xiao, H., & Cai, J. 2022, PASP, 134,
Poon, H., Fan, J. H., & Fu, J. N. 2009, ApJS, 185, 511, 104101, doi: 10.1088/1538-3873/ac98e0
doi: 10.1088/0067-0049/185/2/511 Wierzcholska, A., & Siejkowski, H. 2015, MNRAS, 452,
Rani, B., Krichbaum, T. P., Marscher, A. P., et al. 2015, L11, doi: 10.1093/mnrasl/slv075
A&A, 578, A123, doi: 10.1051/0004-6361/201525608 —. 2016, MNRAS, 458, 2350, doi: 10.1093/mnras/stw425
—. 2014, A&A, 571, L2, doi: 10.1051/0004-6361/201424796 Wilms, J., Allen, A., & McCray, R. 2000, ApJ, 542, 914,
Rani, B., Lott, B., Krichbaum, T. P., Fuhrmann, L., & doi: 10.1086/317016
Zensus, J. A. 2013a, A&A, 557, A71, Wit, E., Heuvel, E. v. d., & Romeijn, J.-W. 2012, Statistica
doi: 10.1051/0004-6361/201321440 Neerlandica, 66, 217,
Rani, B., Krichbaum, T. P., Fuhrmann, L., et al. 2013b, doi: https://doi.org/10.1111/j.1467-9574.2012.00530.x
A&A, 552, A11, doi: 10.1051/0004-6361/201321058 Xiao, H., Fan, J., Yang, J., et al. 2019, Science China
Roming, P. W. A., Kennedy, T. E., Mason, K. O., et al. Physics, Mechanics, and Astronomy, 62, 129811,
2005, SSRv, 120, 95, doi: 10.1007/s11214-005-5095-4 doi: 10.1007/s11433-018-9371-x
Roy, N., Chatterjee, R., Joshi, M., & Ghosh, A. 2019, Xiao, H. B., Zhu, J. T., Fan, J. H., et al. 2022, MNRAS,
MNRAS, 482, 743, doi: 10.1093/mnras/sty2748 517, 4202, doi: 10.1093/mnras/stac2959
18

Xiong, D., Bai, J., Fan, J., et al. 2020, ApJS, 247, 49, Xue, R., Wang, Z.-R., & Li, W.-J. 2022, PhRvD, 106,
doi: 10.3847/1538-4365/ab789b 103021, doi: 10.1103/PhysRevD.106.103021
Xue, R., Liu, R.-Y., Wang, X.-Y., Yan, H., & Böttcher, M.
2019, ApJ, 871, 81, doi: 10.3847/1538-4357/aaf720