betaFeedback

WorksPhysics and Astronomy

A lepto-hadronic perspective on the properties of the candidate neutrino-emitter blazar 5BZQ J0808+4950

Abstract

The role of active galactic nuclei in the production of high-energy neutrinos has been highlighted by the association of the blazar TXS 0506+056 and the Seyfert galaxy NGC 1068 with IceCube neutrino events. In particular, blazars, with their powerful jets, are potential sites of neutrino production via (photo-)hadronic interactions. The gamma-ray bright blazar 5BZQJ0808+4950 is consistent with a spatial anisotropy in the IceCube neutrino all-sky maps. An individual high-energy neutrino event, IC110929A, was also reported in the direction of this blazar. Although the correlation between the gamma-ray bright blazars and the neutrino production remains inconclusive, 5BZQ J0808+4950, with a luminous accretion disk serving as a photon field for the photo-hadronic interactions, offers the opportunity to test the relation between neutrino production and the radiative output of the blazar jet. We compile available multi-wavelength data and analyze the gamma-ray light curve of 5BZQ J0808+4950, leading us to define six contemporaneous epochs to model the source's spectral energy distribution and its evolution over time. We use a lepto-hadronic, time-dependent framework to model 5BZQ J0808+4950, taking into account external photon fields. We find that the solutions corresponding to the different states the source exhibits are explained mostly by a change in the particles' injection properties. The correlation between the predicted X-ray and the neutrino luminosity in this source is stronger than the correlation between gamma-ray and neutrino luminosity. We predict that the rate of neutrinos detectable by IceCube for this blazar ranges from ∼10−3\sim 10^{-3} to ∼10−1\sim10^{-1} neutrinos per epoch. Assuming Poisson statistics, the predicted neutrino number contemporaneous with the interval in which the high-energy neutrino event IC110929A was detected, is compatible with the detection of one neutrino.

INTRODUCTION

Active galactic nuclei (AGN), powered by accretion onto a central supermassive black hole, have since long been considered to be good candidates for the acceleration of (ultra-) high-energy cosmic rays [74]. If AGN jets are able to accelerate hadrons to sufficient energy, these particles can interact with photon fields (pγp\gamma interactions; see Mücke et al. 1999, for an application to astrophysical environments) and in turn produce neutrinos (see e.g. Mannheim & Biermann 1989; Stecker et al. 1991, 1992; Mannheim 1993; Szabo & Protheroe 1994; Mannheim 1995; Mastichiadis 1996; Protheroe 1999; Atoyan & Dermer 2001; Mücke & Protheroe 2001; Atoyan & Dermer 2003; Mücke et al. 2003; Dermer et al. 2009, 2012; Dimitrakoudis et al. 2012; Böttcher et al. 2013; Halzen 2013; Murase et al. 2014; Dermer et al. 2014; Mastichiadis 2016; Gao et al. 2017; Murase et al. 2018; Oikonomou et al. 2019). Non-jetted AGN are also considered as possible neutrino emitters, invoking hadronic interactions (pppp interactions) taking place at the core of the AGN (see e.g. Murase et al. 2016; Inoue et al. 2019, 2021; Kheirandish et al. 2021; Fiorillo et al. 2024; Karavola et al. 2025; Neronov et al. 2025). In both cases, neutrinos are the smoking-gun signatures of hadronic interactions.1

The current most sensitive detector for high-energy (≳\gtrsim TeV) astrophysical neutrino searches is the IceCube Neutrino observatory, located at the South Pole [77, 2, 3, 4 b]. As of today, only a few high-energy neutrino events detected by IceCube are spatially consistent with astrophysical sources that are thought to be capable of producing them. Currently, the strongest astrophysical neutrino signal originates from the nearby Seyfert galaxy NGC 1068 – a non-jetted, low-luminosity AGN – [80]. In the jetted AGN class, instead, the most promising sources are blazars (jetted AGN with a jet aligned with our line of sight) such as PKS 1424-41 [83], GB6 J1040+0617 [58], 3HSP J095507.9+355101 [66], PKS 1502+106 [52] and PKS 0735+178 [140].

So far, the most significant association between a high-energy neutrino event and a blazar is the one of TXS 0506+056 which was found to be spatially coincident with the 290-TeV neutrino IceCube-170922A [79], while the blazar was experiencing a gamma-ray flare [155]. Recently, the detection by the Cubic Kilometer Neutrino Telescope (KM3NeT) of an ultra-high energy (UHE) neutrino (with an estimated energy ∼220\sim220 PeV, The KM3NeT Collaboration et al. 2025a,b) challenged further our understanding of the astrophysical neutrino production. In a recent work [90], 17 blazars are found to be positionally consistent with the neutrino arrival direction, although no firm conclusion can be drawn from this study. Even if the contribution of gamma-ray bright blazars to the diffuse neutrino flux measured by IceCube is suggested to be limited to ∼27%\sim27\% [9] (Aartsen et al. 2017), several studies point out the interest of subclasses of blazars regarding high-energy neutrino correlation (Giommi et al. 2020a, Hovatta et al. 2021, Plavin et al. 2021, Zhou et al. 2021, Buson et al. 2022a, hereafter Paper I, Buson et al. 2022b, Padovani et al. 2022, Buson et al. 2023, hereafter Paper II, Karl et al. 2023, Rodrigues et al. 2024a, Moretti & Caccianiga 2025).

In addition to high-energy individual events, time-integrated studies of the IceCube data suggest statistically significant anisotropies in the spatial distribution of IceCube sky-maps events (see, e.g. Aartsen et al. 2020, Paper I, Paper II, Buson et al. 2022b). Among others, time-integrated searches performed on IceCube sky maps [22] (Paper I, Paper II, see also Buson et al. 2022b), found evidence for statistically significant correlations with the 5th edition of the Roma Blazar catalog (5BZCat; [102]). Those spatial anisotropies are referred to as neutrino hotspots. The study of the southern celestial hemisphere with 7-yrs sky maps (Paper I) and the northern celestial hemisphere with 9-yrs sky maps (Paper II) highlighted 10 and 42 spatially correlated blazars with individual neutrino hotspots, respectively.

Among the 52 sources positionally consistent with the neutrino hotspots 5BZQ J0808+4950 (RA = 122.16521∘122.16521^\circ, Dec = 49.84361∘49.84361^\circ), located at a redshift z=1.4344z = 1.4344 [146], is tentatively associated with the neutrino hotspot IC J0805+5005 (Paper II). An excess of neutrino events in the direction of 5BZQ J0808+4950 although not significant, was also reported by Smith et al. [151] using three years of IceCube data. Additionally, a high-energy neutrino event IC110929A from 2011-09-29 is reported in the first IceCube event catalog of alert tracks (IceCat-1; [9]). 5BZQ J0808+4950 is located right outside the border of the error region derived for this neutrino event. Figure 1 displays the IceCube neutrino sky map (obtained with the data presented and released in IceCube Collaboration et al. 2022) centered at the position of the hotspot IC J0805+5005. The position of the blazar 5BZQ J0808+4950 and the high-energy event IC110929A are shown with the cyan and green crosses, respectively. The pink cross is centered at the position of the hotspot IC J0805+5005 (Paper II), and we indicate the significance of the neutrino signal from this direction of the sky with the negative logarithm of the p-value.

Figure 1. Sky map of the region around the neutrino hotspot reported in PAPER II. We use the IceCube collaboration data released in IceCube Collaboration [78]. The color map indicates the significance of the neutrino signal (p-value). The position of 5BZQ J0808+4950 is indicated with the cyan square, while the location of the high-energy event reported in IceCat-1 [9] is shown with the green hexagone, and the map is centered at the position of the neutrino hotspot indicated by the pink cross. The 90% error region reported in Abbasi et al. [9] is shown with the green dashed line.

5BZQ J0808+4950 is classified as a flat-spectrum radio quasar (FSRQ, [118]), traditionally differentiated from BL Lacs objects based on the width and intensity of emission lines in their optical spectra ([164, 163]). The emission lines, expected to be strong and broader in the FSRQ spectra, are tracing the accretion mechanism, hinting for efficient accretion, and reprocessing of part of the disk photons by the broad-line region (BLR), and also by the dust torus surrounding the AGN. These reprocessed photons provide additional targets for the interactions with the blazar jet, which can be favorable in order to produce neutrinos [100, 97, 39, 30, 86, 125, 172].

In this work, we collect multi-wavelength data of 5BZQ J0808+4950 and define six time intervals, as detailed in Section 2. In order to assess the properties of the source over time, and probe whether 5BZQ J0808+4950 can be an efficient neutrino emitter, as hinted by the sky map analysis, we model the spectral energy distribution (SED) of 5BZQ J0808+4950 for each epoch. The jet model and the emission mechanisms considered for this work are described in Section 3. We present our results in Section 4, and compare our results with previous works, commenting on the neutrino production efficiency of the source in Section 5. Finally, the main conclusions and summary of this work are presented in Section 6.

Throughout the paper, we assume a flat Λ\LambdaCDM cosmology [126].

MULTIWAVELENGTH OBSERVATIONS AND DATA ACQUISITION

Fermi-LAT analysis

Over sixteen years of data provided by the Fermi-Large Area Telescope (LAT) are reduced and analyzed with the python package fermipy (v1.2)2 ([170]) and the Fermitools (v2.2.0). 5BZQ J0808+4950 is associated with the γ\gamma-ray source 4FGL J0808.5+4950 in the Fermi-LAT fourth source catalog (4FGL-DR4; Abdollahi et al. [11]; Ballet et al. [21]), for which we extract photons in a time interval between 2008-08-04 and 2024-12-31 (MJD 54682-60675) and an energy range of 100 MeV to 1 TeV for a region of interest (ROI) of 10∘×10∘10^\circ\times10^\circ centered on the source. A detailed description on the data handling, the analysis and the fitting procedure can be found in Appendix A. For the production of the light curve, we kept the flux normalization and the photon index as free parameters and used an integration time of one year per time bin. The gamma-ray light curve is displayed in the first panel of Figure 2. The variation of the photon index over time is presented

Figure 2. Multi-wavelength light curve of 5BZQ J0808+4950 from γ\gamma-rays to the radio regime, from top to bottom. The green lines separate the time intervals used for the SED modeling. The orange dashed line indicates the arrival time of the high-energy neutrino IC110929A. Top panel: γ\gamma-ray light curve with 12-month binning as observed by the Fermi-LAT. The red horizontal line shows the results of the Bayesian block analysis to identify variable states. Second panel: Same as the first panel, but for the gamma-ray photon index. Third panel: X-rays measurements from Swift-XRT and NuSTAR, between 0.3 keV and 10 keV. Fourth panel: Data from Swift-UVOT. Fifth panel: Optical observations from several instruments, corrected for galactic extinction. Sixth panel: Radio measurements with RATAN, OVRO, and VLBA.

photon index over time is presented in the second panel of the same figure. Using Bayesian blocks (Scargle et al. [143]; The Astropy Collaboration et al. [158]) with a false positive rate p0=0.05p_0 = 0.05, we determine changes in flux and photon index. In the flux light curve produced for 4FGL J0808.5+4950, we identify five distinct states, while the photon index does not show significant variations. This is consistent with a variability index of TSvar=46.01\mathrm{TS}_{\mathrm{var}} = 46.01 (16 degrees of freedom) obtained with fermipy, corresponding to a deviation from a steady state hypothesis at a ∼3.9σ\sim3.9\sigma level. For the longest state, we separated the interval into two smaller intervals due to the large errors reported in the spectral index variations, even though they are not significant enough to be reported as a change in the Bayesian blocks analysis. For each of these six epochs, labeled T1 to T6 (see Figure 2), we produce a gamma-ray spectrum.

In epoch T3, we find a significant high-energy spectral point at ∼100\sim100 GeV. This detection seems to be consistent with a similar trend observed in the SED published in 4FGL-DR4 for 4FGL J0808.5+4950. Given that the associated neutrino event IC110929A happens during the same time interval and that the source displays its highest γ\gamma-ray emission during this epoch, we performed a deeper investigation using the gtsrcprob3 method from the Fermi Science Tools. The results reveal that this excess is produced by one HE photon of ∼129\sim129 GeV detected on 2012-05-12 and is likely associated with the galactic or isotropic diffuse flux (with a probability of 41% and 58%, respectively). The probability that this photon is coming from 4FGL J0808.5+4950 is ∼1%\sim1\%. At this energy and this redshift, the optical depth of the extragalactic background light (EBL) is expected to be τγγ(E=129 GeV,z=1.4344)≈2.05−2.99\tau_{\gamma\gamma}(E = 129\,\mathrm{GeV}, z = 1.4344) \approx2.05 - 2.99 for the models of Franceschini et al. [50]; Domínguez et al. [43]; Finke et al. [46], so a photon signal at this energy would be attenuated by 82% to 95%, supporting the fact that this detection is not associated with the blazar.

X-ray observations

X-ray observations are available from the X-Ray Telescope (XRT) instrument on board the Neil Gehrels Swift Observatory (Swift-XRT) and the Nuclear Spectroscopic Telescope Array (NuSTAR). For this work, we make use of both archival data and new observations obtained by our team during Swift Cycle 20 (e.g., Proposal Number 2023182), as well as Swift Targets of Opportunity (e.g., ToO ID: 20481).

Following Sanchez Zaballa et al. [141], the Swift-XRT data (Gehrels et al. [59]) were processed using FTOOLS within the HEASoft package (v6.33)4. The Swift-XRT observations were performed in photon counting (PC) mode. Event files were calibrated and cleaned using the xrtpipeline task (v.0.13.7), applying standard filtering criteria and the latest calibration files from the Swift CALDB maintained by HEASARC. For spectral extraction, we selected events in the 0.3 - 10.0 keV energy range using xselect. After visually inspecting the images, we selected the extraction region for the source from a circular region with a radius of 20 pixels (∼47′′\sim47''), which encloses approximately 90% of the XRT point spread function, centered on the source. The background spectrum was obtained from an annular region centered on the source, with an inner radius of 40 pixels (∼1.5′\sim1.5') and an outer radius of 150 pixels (∼6.0′\sim6.0'). Exposure maps and ancillary response files (ARFs) were generated using the xrtexpomap and xrtmkarf tasks, respectively. Particularly for epochs T1 and T6 that we have multiple XRT pointings, we combined the corresponding event files and exposure maps into single datasets using the xselect and ximage tools for each epoch.

The NuSTAR observations, conducted in SCIENCE mode, were processed using the NuSTAR Data Analysis Software (NuSTARDAS) v2.1.2 (Harrison et al. [72]). Data from both Focal Plane Modules (FPM), A and B, were used in this analysis. The raw event files were calibrated with the nupipeline task, applying the latest calibration files from CALDB v20250602. Following a visual inspection of the event files, the source spectrum was extracted from a circular region centered on the source, with a radius of approximately 50′′50''. The background spectrum was derived from an annular region centered on the same position, with inner and outer radii of ∼2.0′\sim2.0' and ∼3.3′\sim3.3', respectively. Using the nuproducts script, we then generated the source and background spectra, along with the associated ancillary response files (ARFs) and redistribution matrix files (RMFs).

To enable proper spectral modeling, we applied binning to ensure a minimum of one count per spectral channel, allowing the use of C-statistics (Cash [34]). Binning was performed using grppha for the Swift and NuSTAR datasets. This approach is well-suited for low-count data, as the C-statistic is derived from the likelihood function appropriate for Poisson-distributed events. Spectral fitting was carried out using XSPEC (v12.14), with uncertainties estimated at the 1σ1\sigma confidence level (Arnaud [14]). The Galactic hydrogen column density in the direction of 5BZQ J0808+4950 was fixed at NH, gal=4.67×1020 cm−2N_{\mathrm{H,\,gal}} = 4.67 \times10^{20}\,\mathrm{cm}^{-2} (HI4PI Collaboration et al. [54]). The metal abundances were set to Solar values, adopting those from Wilms et al. [169], and the photoelectric cross-sections for all absorption components were taken from Verner et al. [166]. Additionally, we included, for all tested models, an absorption component at the redshift of the source (NH, ISMN_{\mathrm{H,\,ISM}}, fixing z=1.4344z = 1.4344) to account for interstellar medium (ISM) absorption (Abazajian et al. [8]). For the T6 epoch, which includes a simultaneous observation, we included a multiplicative constant in the spectral model to account for cross-calibration uncertainties between different instruments (Madsen et al. [96]). Specifically, the NuSTAR FPMA and FPMB spectra were analyzed in the 3.0–30.0 keV energy range, as above this energy range, the spectra were background-dominated. In the case of Swift observation, we included and fitted the data in the 0.3–10 keV range.

We modeled the spectra using the simplest spectral shape: a power-law with photon index Γ\Gamma, including Galactic absorption and an additional absorption component from the ISM, following the model phabszphabspow. For the two tested jet models, we combine all data that correspond to each state. The results of the fit are presented in Table 1. Alternative models, such as a broken power-law or a log-parabola, were also tested, but they did not provide an improvement to the fit or evidence for an additional spectral component at the 1σ1\sigma significance level. Each observation was analyzed and fitted individually. To construct the average SED model and plot it, all individually fitted datasets were rebinned to a significant detection as large as 5σ5\sigma. Additionally, for the SED modeling, a 20% fractional systematic uncertainty was added to the data above ∼10\sim10 keV, since information in this energy range is only available during the T6 state, when NuSTAR performed a single observation. This takes into account, in a conservative way, possible short and long-term hard X-ray variability, that can reach values well above 20% (see, e.g., Costamante et al. [37]; Mundo & Mushotzky [107] in blazars).

EpochObs. Date MJDInstrumentObs. IDExposure ksRate ×10−2\times10^{-2} cts/sΓ\GammaNH,ISM×1021N_{\mathrm{H,ISM}} \times10^{21} cm−2^{-2}Fabs×10−13F_{\mathrm{abs}} \times10^{-13} erg cm−2^{-2} s−1^{-1}Fint×10−13F_{\mathrm{int}} \times10^{-13} erg cm−2^{-2} s−1^{-1}Cstat/dof
T154104000361960014.12.9±0.32.9 \pm0.31.7±0.11.7 \pm0.12.6±0.92.6 \pm0.919±1.019 \pm1.022±1.022 \pm1.0368.69/402
T154389000361960024.64.3±0.34.3 \pm0.3
T154456XRT000361960033.94.3±0.34.3 \pm0.3
T154889000361960049.43.7±0.23.7 \pm0.2
T154894000361960056.44.3±0.34.3 \pm0.3
T255117XRT000361960065.11.4±0.21.4 \pm0.21.8±0.21.8 \pm0.2<2.9<2.97.3±1.37.3 \pm1.38.1±1.08.1 \pm1.064.10/62
T660645XRT000361960074.81.4±0.21.4 \pm0.2
T660660FPM6106000200238.92.1±0.12.1 \pm0.11.5±0.11.5 \pm0.1<3.7<3.76.5±0.36.5 \pm0.37.7±0.47.7 \pm0.4599.30/753
T660660XRT000975110024.41.3±0.21.3 \pm0.2
T660674XRT000361960083.91.5±0.21.5 \pm0.2

Table 1. X-rays observations of the source 5BZQ J0808+4950, and fit results for the Swift-XRT and NuSTAR observations of 5BZQ J0808+4950, including both absorbed (FabsF_{\mathrm{abs}}) and intrinsic (FintF_{\mathrm{int}}) fluxes. Uncertainties are reported at the 90% confidence level within the 0.3–10.0 keV energy range. In T6, the multiplicative constants accounting for cross-calibration between different instruments are CXRT,FPMA=0.9±0.1C_{\mathrm{XRT,FPMA}} = 0.9 \pm0.1 and CXRT,FPMB=0.8±0.1C_{\mathrm{XRT,FPMB}} = 0.8 \pm0.1 respectively.

Infrared, optical and UV observations

Spectroscopic observations in the optical are also available for 5BZQ J0808+4950 in the Seventh Data Release of the Sloan Digital Sky Survey (SDSS-DR7, analyzed in Shen et al. [146]). These provided a measure of the redshift based on the detection of a broad Mg II emission line consistent with the value z=1.4344z = 1.4344. We used this spectral profile and adopted the relation between the luminosities of the emission lines and the BLR based on the composite blazar spectrum of Francis et al. [51] to infer the LBLRL_{\mathrm{BLR}} of the source (see a similar approach in [18 hereafter Paper PI] (Sbarrato et al. 2012)). We assumed a covering factor of 0.1 for the BLR relative to a geometrically thin accretion disk and estimated the disk luminosity. Then, we estimated the virial mass of the central black hole adopting the empirical scaling laws of Shen et al. [146]; Shen & Liu [145]. The resulting values are Ldisk=(1.35±0.25)×1046L_{\mathrm{disk}} = (1.35 \pm0.25) \times10^{46} erg s−1^{-1} and MBH=(6.95±0.13)×108M⊙M_{\mathrm{BH}} = (6.95 \pm0.13) \times10^{8} M_{\odot} (Paper PI; see also [146, 121]).

There are photometric observations available for 5BZQ J0808+4950 by the Zwicky Transient Facility (ZTF; Masci et al. [101]), Gaia (Gaia Collaboration et al. [54, 55]; ); Babusiaux et al. [19]) and the Ultra-Violet Optical Telescope on board Swift (Swift-UVOT; Roming et al. [139]). While additional surveys also include this source in their catalogs, most are significantly contaminated by the nearby star 2MASS J08084034+4950387, located within <7< 7 arcsec. Swift-UVOT observations are available only in the U filter simultaneous to X-ray observations discussed in 2.2. Given the redshift of the source, Lyman-alpha absorption is expected to suppress flux below ∼2960\sim2960 Å, affecting all three UVOT UV filters. The source counts are extracted with a radius of 3–5″, while the background counts are taken from a region around the source position, avoiding visible contamination by the nearby star or other sources in the vicinity. The reddened magnitudes are then computed with the UVOTSOURCE tool provided by HEASOFT 5. Galactic extinction corrections for this and subsequent optical instruments are applied following Schlafly & Finkbeiner [144], using the source-dependent E(B−V)E(B-V) and assuming a standard reddening law with RV=3.1R_V = 3.1. Wavelength-dependent corrections follow Fitzpatrick [48]. The magnitudes are finally converted to energy flux units with the zero-points taken from Poole et al. [128] and Breeveld et al. [25]. Gaia photometry is available in the G, GBP_{\mathrm{BP}} and GRP_{\mathrm{RP}} bands. After applying the same reddening correction procedure, these are converted to energy flux units using the zero points and the filter information from Riello et al. [133]. Lastly, photometric observations performed with ZTF are available in the red (r), the green (g), and the near infrared (i) band with the filter information provided by the Spanish Virtual Observatory Filter Profile Service (SVO; Rodrigo et al. [135]; ); Rodrigo & Solano [134]). The light curves in energy flux units and uniformly corrected for galactic extinction are displayed in the third and fourth panels of Figure 2 and used for the modeling when they are contemporaneous to the time intervals T1-T6.

Radio observations

Radio light curves are available from the Academy of Sciences Radio Telescope (RATAN)6, the Very Long Baseline Array (VLBA, Lister et al. [95]) and the Owens Valley Radio Observatory 40 m telescope (OVRO, Richards et al. [132]) as published in Richards et al. [131]. They are displayed in the bottom panel of Figure 2. Data at a frequency lower than 101110^{11} Hz are used as upper limits in the modeling, since this emission likely corresponds to extended regions further away from the core, which we do not consider in our model (see, e.g. [63, 156] (Konigl 1981)).

SED MODELING

The SED of blazars is traditionally composed of two distinct components (Fossati et al. [49]). The low-energy component is commonly attributed to synchrotron radiation from relativistic electrons present in the jet, while the origin of the second hump remains debated. In purely leptonic models, the high-energy part of the SED is attributed to the inverse-Compton scattering (IC) of photons by the electrons of the jet. When considering lepto-hadronic models, in the context of (ultra) high-energy cosmic rays and/or neutrino production, the high-energy component can be explained in different ways: synchrotron radiation from the relativistic protons of the jet (synchrotron proton blazar), or a leptonic-dominated origin and a cascade component ([97, 13, 112, 36, 124]).

Jet models

We consider the case in which the emission region (hereafter "blob") is traveling over time, within a cylindrical jet (i.e., the blob is not expanding over time). We model the blob as a spherical, magnetized region of radius Rblob′R^{\prime}_{\mathrm{blob}}. We define the emission region’s bulk velocity

βj=1−Γj−2,\beta_j = \sqrt{1-\Gamma_j^{-2}},

and the jet makes an angle θ∼1/Γj\theta\sim1/\Gamma_j (typical for blazars) with the line-of-sight, defining a Doppler factor δj=[Γj(1−βjcos⁡θj)]−1\delta_j = [\Gamma_j(1-\beta_j\cos\theta_j)]^{-1}. With the small line-of-sight angle, we use: δj=Γj\delta_j = \Gamma_j.

Primary particles, namely relativistic electrons and protons, are continuously injected into the region with a constant injection rate Qi′(Ei′)=dN′/(dE′dt′dV′)Q_i^{\prime}(E_i^{\prime}) = \mathrm{d}N^{\prime}/(\mathrm{d}E^{\prime}\mathrm{d}t^{\prime}\mathrm{d}V^{\prime}) following a power-law with an super-exponential cutoff

Qi′(Ei′)∝Ei′−piexp⁡[−(Ei′Ei,max⁡′)2]for Ei′>Ei,min⁡′(1)Q_i^{\prime}(E_i^{\prime}) \propto E_i^{\prime-p_i}\exp\left[-\left(\frac{E_i^{\prime}}{E_{i,\max}^{\prime}}\right)^2\right]\quad\text{for } E_i^{\prime}>E_{i,\min}^{\prime} \tag*{(1)}

where V′∫E′Q′ dE′=Linj′V^{\prime}\int E^{\prime}Q^{\prime}\,\mathrm{d}E^{\prime}=L_{\mathrm{inj}}^{\prime} gives the normalization of the injection rate expressed as the injection luminosity. The super-exponential cutoff can be related to diffusive shock acceleration in the Bohm limit ([174]).

We use the software package7 AM3^{3} ([56, 88, 89]) to compute the lepto-hadronic interactions in the emission region. The processes taken into account are synchrotron radiation, inverse-Compton scattering, Bethe-Heitler pair production, photo-pion production, proton-proton interaction, and photon-photon annihilation. Particle escape is taken into account, with an escape time tesc=Rblob′/ct_{\mathrm{esc}} = R_{\mathrm{blob}}^{\prime}/c, with cc the velocity of light. Details about the interactions, the computational methods, and the software can be found in Klinger et al. [89].

For each quasi-simultaneous observation, we evolve the simulation over the duration of the observational periods indicated in the light curve (Figure 2, namely an observed time of 1, 1, 4.9, 4, 4, and 1.25 years respectively), corresponding to a time t′=δjt/(1+z)t^{\prime} = \delta_j t/(1+z) in the rest frame of the jet. The blob is injected at a distance RdissR_{\mathrm{diss}} from the black hole, and the emission region travels along the jet, to distances of the order of Rdiss+βjcΓjt′R_{\mathrm{diss}}+\beta_jc\Gamma_jt^{\prime} in the stationary frame of the AGN. We calculate the time-averaged SED, and transform it into the observer’s frame as Fν=(1+z)δj3Lν′/(4πdL2)F_{\nu}=(1+z)\delta_j^3L_{\nu}^{\prime}/(4\pi d_L^2), where dLd_L is the luminosity distance, and the frequencies are related through ν=δjν′/(1+z)\nu=\delta_j\nu^{\prime}/(1+z).

External photons fields

5BZQ J0808+4950 belongs to the category of the high-excitation radio galaxies (HERGs, PAPER PI), so the accretion mechanisms are expected to be important (see, e.g. [62, 23] (Sbarrato et al. 2012)). We take into account photons from the accretion disk, the broad-line region, and we also consider the presence of a dusty torus in the vicinity of the supermassive black hole. The transformation of these external photon fields into the jet frame in order to carry on the radiative calculations follows Ghisellini & Madau [60] and Ghisellini & Tavecchio [61]. Depending on the position of the emission region relative to the photon fields, parametrized as rdiss=Rdiss/RBLRr_{\mathrm{diss}}=R_{\mathrm{diss}}/R_{\mathrm{BLR}}, the photons will be seen as boosted or de-boosted in the jet frame. Empirical relations are used to derive the characteristics of the external field in order to reduce the number of free parameters in the models. Specifically, we use:

Ldisk=0.1M˙c2,L_{\mathrm{disk}}=0.1\dot{M}c^2,
L_{\mathrm{BLR}}=0.1L_{\mathrm{disk}}\quad\text{(see e.g. {citet}`Ghisellini & Madau (1996)`),}
Ltorus=ftorusLdisk,L_{\mathrm{torus}}=f_{\mathrm{torus}}L_{\mathrm{disk}},

where M˙\dot{M} is the mass accretion rate of the disk, LdiskL_{\mathrm{disk}}, LBLRL_{\mathrm{BLR}} and LtorusL_{\mathrm{torus}} are the total disk, BLR and torus luminosities respectively and ftorusf_{\mathrm{torus}} is the covering factor of the torus, that we fix to ftorus=0.5f_{\mathrm{torus}}=0.5 ([31, 68]). Furthermore,

RBLR=1017(Ldisk1045 erg/s)1/2 cm(2)R_{\mathrm{BLR}}=10^{17}\left(\frac{L_{\mathrm{disk}}}{10^{45}\,\mathrm{erg/s}}\right)^{1/2}\,\mathrm{cm} \tag*{(2)}
Rtorus=1.2×1018(Ldisk1045 erg/s)1/2(Ttorus103 K)−2 cm(3)R_{\mathrm{torus}}=1.2\times10^{18}\left(\frac{L_{\mathrm{disk}}}{10^{45}\,\mathrm{erg/s}}\right)^{1/2}\left(\frac{T_{\mathrm{torus}}}{10^3\,\mathrm{K}}\right)^{-2}\,\mathrm{cm} \tag*{(3)}

We let the torus temperature be a free parameter, but enforce that it always stays below the dust sublimation temperature Ttorus≲1300KT_{\mathrm{torus}}\lesssim1300\mathrm{K}.

Effectively, the accretion disk is modeled as a multi-temperature blackbody, where we assume that the disk is perpendicular to the jet axis ([53]), and extending between the inner radius Rin=rinGMBH/c2=rinRgR_{\mathrm{in}}=r_{\mathrm{in}}GM_{\mathrm{BH}}/c^2=r_{\mathrm{in}}R_g, with GG the gravitational constant, where RgR_g is the gravitational radius of the black hole and the outer radius Rout=500RinR_{\mathrm{out}}=500R_{\mathrm{in}}

FνAD(ν)=4πhν3c2dL2∫RinRoutR dRexp⁡[hν/kBT(R)]−1,F_{\nu}^{\mathrm{AD}}(\nu)=\frac{4\pi h\nu^3}{c^2d_L^2}\int_{R_{\mathrm{in}}}^{R_{\mathrm{out}}}\frac{R\,\mathrm{d}R}{\exp[h\nu/k_BT(R)]-1},

where the temperature T(R)T(R) is given by the standard relation for thin disks ([e.g. 53])

T(R)=[3GMM˙BH8πR3σ(1−RinR)]1/4,T(R)=\left[\frac{3GM\dot{M}_{\mathrm{BH}}}{8\pi R^3\sigma}\left(1-\sqrt{\frac{R_{\mathrm{in}}}{R}}\right)\right]^{1/4},

hh is the Planck constant, kBk_B is the Boltzmann constant and σ\sigma is the Stefan-Boltzmann constant.

Additionally, the spectrum of the torus is described as a single-temperature blackbody, normalized to the luminosity of the torus

νFνDT(ν)=15Ltorus4π5dL2(hνkBTtorus)4[exp⁡(hνkBTtorus)−1]−1\nu F_{\nu}^{\mathrm{DT}}(\nu)=\frac{15L_{\mathrm{torus}}}{4\pi^5d_L^2}\left(\frac{h\nu}{k_BT_{\mathrm{torus}}}\right)^4\left[\exp\left(\frac{h\nu}{k_BT_{\mathrm{torus}}}\right)-1\right]^{-1}

Results

We model the quasi-simultaneous epochs with the procedure described above. As described in § 2.4, the SED contains archival radio data, which we treat as upper limits, as the low-frequency emission from the jet is likely originating from the extended jet. We fix the black hole mass and the redshift of the source to the values derived in the literature (PAPER PI). We scan parameters with a least-square minimization algorithm, and assess the fit quality with the chi-square value defined as

χred2=1d.o.f.∑i=1N(yi−f(xi))2σi2(4)\chi_{\mathrm{red}}^{2} = \frac{1}{\mathrm{d.o.f.}} \sum_{i=1}^{N} \frac{(y_i-f(x_i))^2}{\sigma_i^2} \tag*{(4)}

where NN is the number of data points, (xi,yi,σi=σy,i)(x_i,y_i,\sigma_i=\sigma_{y,i}) represent the data points and the uncertainty on the yiy_i value (i.e., the error bar), respectively, and f(xi)f(x_i) is the model prediction. The number of degrees of freedom, d.o.f., is determined by the number of data points minus the number of free parameters in the models. Since the frequency bins span very different scales depending on the frequency, we also compute an effective chi-square, where we replace variance σi\sigma_i by an effective one such that

σi2→σi,eff2=[∂f∂x]i2σx,i2+σy,i2\sigma_i^2 \rightarrow\sigma_{i,\mathrm{eff}}^2 = \left[\frac{\partial f}{\partial x}\right]_i^2 \sigma_{x,i}^2+\sigma_{y,i}^2

(Orear 1982, 1984), in order to take into account the uncertainty in the simulated flux induced by the finite bin width. The upper limits in the data are not used in the calculation of the degrees of freedom, but we consider them in the fit so that the predicted flux never violates them (a weight of 10510^5 is added to the residual of a certain point if it violates an upper limit). We start the fitting procedure with the epochs T1, T2, and T6, where the X-ray observations are available, and lead to a number of degrees of freedom d.o.f. ≥1\geq1. For T3, T4, and T5, the number of free parameters is larger than the number of fitted data points. To limit this degeneracy we use the solutions obtained in T1, T2, and T6 as starting points for the fitting procedure of the remaining epochs. However, the goodness of fit in the periods T3, T4, and T5 cannot be strictly compared to the epochs T1, T2, and T6, due to the large number of parameters compared to the number of data points (we report the χ2\chi^2 value, the number of data points and if applicable, the number of degrees of freedom in the last row of Table 2).

unitT1T2T3*T4T5T6explored range
Start—2008-08-042009-08-042010-08-042015-08-032019-07-232023-08-01—
End—2009-08-042010-08-042015-08-032019-07-232023-08-012024-12-31—
Rblob′R^\prime_{\mathrm{blob}}cm1.0−0.2+0.2×10161.0^{+0.2}_{-0.2} \times10^{16}3.0−0.1+0.3×10163.0^{+0.3}_{-0.1} \times10^{16}3.0×10163.0 \times10^{16}2.0×10162.0 \times10^{16}2.8×10162.8 \times10^{16}1.5−0.2+0.09×10161.5^{+0.09}_{-0.2} \times10^{16}[5×1015,1×1017][5 \times10^{15}, 1 \times10^{17}]
B′B^\primeG3.1−0.4+0.53.1^{+0.5}_{-0.4}5.0−0.4+2.85.0^{+2.8}_{-0.4}8.5×10−18.5 \times10^{-1}5.4×10−15.4 \times10^{-1}1.01.02.5−0.3+0.12.5^{+0.1}_{-0.3}[10−1,5×101][10^{-1}, 5 \times10^{1}]
δj\delta_j—5.5−0.1+0.15.5^{+0.1}_{-0.1}5.5−0.3+0.075.5^{+0.07}_{-0.3}5.35.35.35.35.35.35.1−0.07+0.035.1^{+0.03}_{-0.07}[5.0,3×101][5.0, 3 \times10^{1}]
γe,min′\gamma^\prime_{\mathrm{e,min}}—7.0−0.3+0.5×1027.0^{+0.5}_{-0.3} \times10^{2}6.6−6+23×1016.6^{+23}_{-6} \times10^{1}1.7×1021.7 \times10^{2}7.0×1037.0 \times10^{3}1.2×1021.2 \times10^{2}8.0−0.7+0.1×1028.0^{+0.1}_{-0.7} \times10^{2}[1.0,104][1.0, 10^{4}]
γe,max′\gamma^\prime_{\mathrm{e,max}}—2.5−0.3+0.3×1042.5^{+0.3}_{-0.3} \times10^{4}2.3−0.6+1.5×1042.3^{+1.5}_{-0.6} \times10^{4}1.9×1041.9 \times10^{4}1.0×1041.0 \times10^{4}4.0×1044.0 \times10^{4}1.9−0.2+0.2×1041.9^{+0.2}_{-0.2} \times10^{4}[103,5×105][10^{3}, 5 \times10^{5}]
pep_{\mathrm{e}}—1.9−0.05+0.041.9^{+0.04}_{-0.05}2.0−0.03+0.052.0^{+0.05}_{-0.03}2.12.11.91.92.22.22.0−0.03+0.092.0^{+0.09}_{-0.03}[1.5,3.0][1.5, 3.0]
Le′L^\prime_{\mathrm{e}}erg s−1^{-1}1.9−0.1+0.1×10441.9^{+0.1}_{-0.1} \times10^{44}6.1−0.6+0.2×10436.1^{+0.2}_{-0.6} \times10^{43}2.9×10442.9 \times10^{44}1.1×10441.1 \times10^{44}2.5×10442.5 \times10^{44}8.0−0.4+0.3×10438.0^{+0.3}_{-0.4} \times10^{43}[1040,1045][10^{40}, 10^{45}]
γp,min′\gamma^\prime_{\mathrm{p,min}}—1.01.01.01.01.01.01.01.01.01.01.01.0fixed
γp,max′\gamma^\prime_{\mathrm{p,max}}—6.0−4+2×1066.0^{+2}_{-4} \times10^{6}1.0−0.4+0.2×1071.0^{+0.2}_{-0.4} \times10^{7}8.4×1068.4 \times10^{6}4.0×1064.0 \times10^{6}8.3×1068.3 \times10^{6}1.0−0.6+0.3×1071.0^{+0.3}_{-0.6} \times10^{7}[105,108][10^{5}, 10^{8}]
ppp_{\mathrm{p}}—1.9−0.05+0.041.9^{+0.04}_{-0.05}2.0−0.03+0.052.0^{+0.05}_{-0.03}2.12.11.91.92.22.22.0−0.03+0.092.0^{+0.09}_{-0.03}fixed
Lp′L^\prime_{\mathrm{p}}erg s−1^{-1}1.7−0.9+0.4×10461.7^{+0.4}_{-0.9} \times10^{46}5.4−2.6+1.0×10465.4^{+1.0}_{-2.6} \times10^{46}1.3−0.5+2.6×10471.3^{+2.6}_{-0.5} \times10^{47}1.5−1.5+0.0×10461.5^{+0.0}_{-1.5} \times10^{46}4.9−4.9+9.8×10464.9^{+9.8}_{-4.9} \times10^{46}5.5−3.3+1.8×10465.5^{+1.8}_{-3.3} \times10^{46}[0,1048][0, 10^{48}]
M˙\dot{M}M⊙ yr−1M_\odot\,\mathrm{yr}^{-1}2.60−0.66+0.222.60^{+0.22}_{-0.66}2.82−0.88+0.02.82^{+0.0}_{-0.88}2.802.802.752.752.822.822.82−2.3+0.02.82^{+0.0}_{-2.3}[1.94,2.83][1.94, 2.83]
RinR_{\mathrm{in}}RgR_{\mathrm{g}}4.5−1.5+0.34.5^{+0.3}_{-1.5}5.0−2.0+0.05.0^{+0.0}_{-2.0}4.74.76.76.74.94.94.5−0.7+0.34.5^{+0.3}_{-0.7}[3.0,5×101][3.0, 5 \times10^{1}]
Rdiss/RBLRR_{\mathrm{diss}}/R_{\mathrm{BLR}}—7.7−5.7+92.3×10−17.7^{+92.3}_{-5.7} \times10^{-1}7.7−5.7+92.3×10−17.7^{+92.3}_{-5.7} \times10^{-1}7.7×10−17.7 \times10^{-1}7.7×10−17.7 \times10^{-1}7.7×10−17.7 \times10^{-1}7.7−5.7+92.3×10−17.7^{+92.3}_{-5.7} \times10^{-1}[5×10−1,101][5 \times10^{-1}, 10^{1}]
TtorusT_{\mathrm{torus}}K1.0−0.9+0.3×1031.0^{+0.3}_{-0.9} \times10^{3}1.0−0.9+0.3×1031.0^{+0.3}_{-0.9} \times10^{3}1.0×1031.0 \times10^{3}1.0×1031.0 \times10^{3}1.0×1031.0 \times10^{3}1.0−0.9+0.3×1031.0^{+0.3}_{-0.9} \times10^{3}[102,1.3×103][10^{2}, 1.3 \times10^{3}]
ue′/uB′u^\prime_{\mathrm{e}}/u^\prime_{\mathrm{B}}—1.2×1011.2 \times10^{1}1.8×10−11.8 \times10^{-1}3.0×1013.0 \times10^{1}6.7×1016.7 \times10^{1}2.1×1012.1 \times10^{1}3.83.8—
up′/uB′u^\prime_{\mathrm{p}}/u^\prime_{\mathrm{B}}—1.0×1031.0 \times10^{3}1.6×1021.6 \times10^{2}1.3×1041.3 \times10^{4}8.7×1038.7 \times10^{3}4.1×1034.1 \times10^{3}2.6×1032.6 \times10^{3}—
LdiskL_{\mathrm{disk}}erg s−1^{-1}1.5×10461.5 \times10^{46}1.6×10461.6 \times10^{46}1.6×10461.6 \times10^{46}1.6×10461.6 \times10^{46}1.6×10461.6 \times10^{46}1.6×10461.6 \times10^{46}—
η\eta—1.7×10−21.7 \times10^{-2}1.8×10−21.8 \times10^{-2}1.8×10−21.8 \times10^{-2}1.8×10−21.8 \times10^{-2}1.8×10−21.8 \times10^{-2}1.8×10−21.8 \times10^{-2}—
Pjet′P^\prime_{\mathrm{jet}}erg s−1^{-1}2.5×10472.5 \times10^{47}8.1×10478.1 \times10^{47}1.8×10481.8 \times10^{48}2.1×10472.1 \times10^{47}6.9×10476.9 \times10^{47}7.0×10477.0 \times10^{47}—
Pjet′/LEddP^\prime_{\mathrm{jet}}/L_{\mathrm{Edd}}—2.92.99.29.22.1×1012.1 \times10^{1}2.42.47.97.98.08.0—
CDCD—4.24.22.8×10−12.8 \times10^{-1}6.96.99.59.56.96.92.12.1—
N⟨νμ+νˉμ⟩N_{\langle\nu_\mu+\bar{\nu}_\mu\rangle}—0.008−0.004+0.0020.008^{+0.002}_{-0.004}0.003−0.002+0.0010.003^{+0.001}_{-0.002}0.101−0.039+0.0340.101^{+0.034}_{-0.039}0.001−0.001+0.0000.001^{+0.000}_{-0.001}0.053−0.053+0.1110.053^{+0.111}_{-0.053}0.006−0.004+0.0030.006^{+0.003}_{-0.004}—
χ2\chi^2 (NpointsfittedN_{\mathrm{points}}^{\mathrm{fitted}})—8.48.4 (17)8.58.5 (14)1.01.0 (5)1.31.3 (5)2.62.6 (6)8.18.1 (22)—

Table 2. Table 2. Best-fit parameters of the models for the six states of 5BZQ J0808+4950. The last column displays the range of parameters explored during the fit. In total, there are 13 parameters that are allowed to vary, against 5 fixed parameters (namely the minimum proton energy, the proton injection index, the BLR and torus covering factors, and the black hole mass). The second-to-last row gives the number of neutrinos N⟨νμ+νˉμ⟩N_{\langle\nu_\mu+\bar{\nu}_\mu\rangle} per epoch predicted by the best-fit models. The explored allowed range of proton luminosity is also reported in terms of number of neutrinos for each epoch. The * indicates the time interval in which the IceCube high-energy neutrino event was detected. The last row reports the chi-square value for each epoch and the corresponding number of fitted data points (i.e. the data points above 101110^{11} Hz and the points that are not upper limits).

In order to reduce the number of free parameters we make standard assumptions, following, e.g., Cerruti et al. [36], the minimum energy of the injected protons is set to γp′=1.0\gamma'_p=1.0, and the proton index to pp=pep_p=p_e. With this, we assume that the electrons and the protons are accelerated via the same mechanism while they are injected into the emission region. Setting γp′=1\gamma'_p=1 is also a conservative choice in terms of the resulting jet power because a larger value would lead to a lower proton density needed to obtain equal results.

We let the accretion rate M˙\dot{M} vary within the uncertainty of the value inferred in PAPER PI. The ranges of parameters that we explore is presented in the last column of Table 2, along with the results of the modeling. We report the best-fit parameters of the simulations, as well as some derived quantities such as the accretion regime η=LBLR/LEdd\eta=L_{\mathrm{BLR}}/L_{\mathrm{Edd}} or the jet power Pjet=2πΓj2βjRblob′2c∑iui′P_{\mathrm{jet}}=2\pi\Gamma_j^2\beta_j R_{\mathrm{blob}}^{\prime2}c\sum_i u'_i, where we sum the energy density ui′u'_i of the magnetic field uB′=B′2/(8π)u'_B=B'^2/(8\pi), the electrons ue′=Le′/(4πRblob′2c)u'_e=L'_e/(4\pi R_{\mathrm{blob}}^{\prime2}c), the protons up′=Lp′/(4πRblob′2c)u'_p=L'_p/(4\pi R_{\mathrm{blob}}^{\prime2}c), and the radiation urad′u'_{\mathrm{rad}}. We explore a broad range of parameters, either limited by physical constraints (such as the torus temperature or the accretion rate), or in order to scan different types of solutions (e.g., compact blobs with sizes Rblob′<1016R'_{\mathrm{blob}}<10^{16} cm, or larger magnetic fields strength B′>15B'>15 G. With the large number of free parameters (13, see Table 2), other solutions may also fit the SEDs, but we report here the best ones we found with a non-negligible neutrino output and a consistent evolution throughout the six epochs.

We show the six SED models, in the observer frame, in Figure 3. Each panel shows the total time-averaged SED with the solid gray line. The primary and secondary synchrotron radiation components are shown in blue, where the solid line corresponds to primary electrons’ synchrotron radiation, the dashed line to the synchrotron radiation from electron-positron pairs created through photon-photon absorption, the dash-dotted one to the synchrotron emission from the photo-pion production secondaries, and the dotted line shows the synchrotron emission from the Bethe-Heitler pair production secondaries. The corresponding inverse-Compton scattering contribution is shown in green, with the same line style for the same process. The red and orange lines indicate the accretion disk and the dusty torus emission, respectively. The shaded areas correspond to the range of proton luminosities that fit the data (both SED and light curve points) and do not violate the SED upper limits, or if this is not a constraint, we vary the parameter until Δχeff2=1\Delta\chi_{\mathrm{eff}}^2=1. For the three epochs in which we have sufficient data, i.e. T1, T2 and T6, where d.o.f. ≥1\geq1, we repeat the same procedure as for the proton luminosity for the 13 free parameters independently, and report the corresponding ranges in Table 2. We emphasize on the fact that these range are not equivalent to statistical uncertainties, but are shown here to demonstrate which parameters are the most constrained by the data.

Figure 3. The modeled SEDs of 5BZQ J0808+4950, in the observer frame. Each panel corresponds to an epoch, labeled on the top left of the panel. The black points are the contemporaneous points of the studied epoch, while the gray points are the data from the average SED, for reference. Blue lines correspond to synchrotron radiation, while green lines represent inverse-Compton scattering. The processes or particles producing each line are indicated with different line styles, namely solid for primary electrons, loosely dashed for primary protons (visible only in T2, subdominantly around 101810^{18} Hz), dashed for photon-photon pair production, dash-dotted for photo-pion production (PPP), and dotted for Bethe-Heitler pair-production (B-H). In red and orange, we show the emission from the accretion disk and the torus. The total observed photons are shown by the gray line, and the pink line indicates the muon plus anti-muon neutrino flux. The shaded areas correspond to the 1σ1\sigma uncertainty on the proton luminosity in the models. We report the icecubepy (Paper NI, Lincetto M. et al. 2026) 10-year sensitivity at the declination of the source with the dashed black line. The goodness-of-fit is reported as the effective chi-square value reduced to the number of degrees of freedom when applicable

The gray data points are the average SED of the source, over the whole time range shown in the light curve in Figure 2. The pink solid line indicates the muon neutrino flux, and we also report the IceCube 10-year sensitivity envelope, obtained with the software received by the source with the solid black line (Lincetto et al. [92], hereafter PAPER NI, see also Lincetto M. et al. 2026).

The SED models are mostly constrained by the UV data points, tracing the interplay between the accretion disk and the jet, and the upper limits in the gamma-ray spectrum, tracing the contribution of the hadronic cascade. The gamma-ray spectral shape, in turn, specifically constrains the neutrino output of a given model. This can be seen particularly in T4, where, despite the large gamma-ray flux, the resulting neutrino output is low. The protons initiate a cascade of interactions, producing secondary particles that in turn emit mainly X-ray photons (through Bethe-Heitler pair production, dotted blue line in Figure 3), but also above GeV energies (from photon annihilation, photo-pion pair production and Bethe-Heitler pair production). At the redshift of the source, the EBL photons are responsible for the attenuation of the gamma-ray flux above Ecrit≈170(1+z)−2.38≈20E_{\mathrm{crit}}\approx170(1+z)^{-2.38}\approx20 GeV (Franceschini et al. [50]; Ackermann et al. [12]). Hence, a non-detection of photons between 1 and 20 GeV provides a strong constraint for the models, as the attenuation of the flux must be intrinsic to the source. The X-ray points, when they are available, constrain the inverse-Compton component, and we are able to explain the X-ray spectral changes between epochs T1, T6 and T2 by a transition from Compton-dominated states (T1 and T6) to a synchrotron-dominated state in T2. When X-ray data are not available, we do not constrain the X-ray flux in any specific way, and we find that our solutions lie within a factor 3 from the flux observed in epoch T1 and below a factor 10 from the lowest flux, observed in T2.

These variations are driven only by the models, and are well within long-term variability observed in blazars (e.g. with Swift Giommi et al. [67]). The size of the emission regions is similar for all epochs Rblob′≈(1−3)×1016R^\prime_{\mathrm{blob}} \approx(1-3) \times10^{16} cm. The magnetic field strength varies from ∼0.5\sim0.5 to 5 G, which is the main variation in the emission region parameters, as the Doppler factor also stays constant around δj≈5\delta_j \approx5. Finke [45] found a magnetic field strength, assuming equipartition between the electrons and the magnetic field, at a distance of 1 pc from the core of B1pc′=0.9−0.8+1.2B^\prime_{\mathrm{1pc}} = 0.9^{+1.2}_{-0.8} G. Assuming magnetic flux conservation and if the magnetic field (with a toroidal topology) stays in equipartition with the particles, B′(Rdiss′)=B1pc′(Rdiss′1 pc)−1≈9.0−8.0+120B^\prime(R^\prime_{\mathrm{diss}}) = B^\prime_{\mathrm{1pc}}\left(\frac{R^\prime_{\mathrm{diss}}}{1\ \mathrm{pc}}\right)^{-1} \approx9.0^{+120}_{-8.0} G. Our values of magnetic field strengths lie between the lower bound of this estimate and the best-fit value. These are consistent with the fact that in our case, the jet departs from equipartition, with uparticles′/uB′≫1u^\prime_{\mathrm{particles}}/u^\prime_{\mathrm{B}} \gg1 with uparticles′=up′+ue′u^\prime_{\mathrm{particles}} = u^\prime_{\mathrm{p}} + u^\prime_{\mathrm{e}}. The disk is found to extend close to the black hole, Rdisk,in∼(4.5−5)RgR_{\mathrm{disk,in}} \sim(4.5-5) R_{\mathrm{g}}. The accretion disk and BLR, providing external photons from the emission region to interact with, have properties that stay overall constant throughout the epochs. This is due to the fact that external fields play a subdominant role in the emission compared to the internal photon fields, because the Doppler boosting into the jet frame is not important with the value of the Doppler factor

that we find. Hence, the impact of the related parameters (i.e., TtorusT_{\mathrm{torus}}, rdissr_{\mathrm{diss}}) is negligible on the SED, leaving these two parameters essentially unconstrained for this set of solutions, when varying them independently. The primary electron populations injected in the emission region for each epoch have a spectral index pe∼2.0p_{\mathrm{e}} \sim2.0, typically expected for diffusive acceleration (Drury [44]) at strong shocks. Somewhat steeper values pe∼2.2p_{\mathrm{e}} \sim2.2 are expected from relativistic shocks in a weakly magnetized regime (see, e.g., Sironi et al. [150]). Although constraining a shock model is beyond the scope of this work, the energy-density ratios uparticles′/uB′≫1u^\prime_{\mathrm{particles}}/u^\prime_{\mathrm{B}} \gg1 found in the solutions here hint qualitatively towards a weakly magnetized environment, supporting efficient shock acceleration (Sironi & Spitkovsky [148]; Sironi et al. [150, 147]). The primary protons are injected with a fixed index pp=pep_{\mathrm{p}} = p_{\mathrm{e}} and a fixed minimum Lorentz factor γp,min′=1.0\gamma^\prime_{\mathrm{p,min}} = 1.0, but their injected luminosity is found to vary from 1×10461 \times10^{46} to 1×10471 \times10^{47} erg s−1^{-1}, while the maximum energy at which they are injected is Ep,max′∼4−9E^\prime_{\mathrm{p,max}} \sim4-9 PeV. In terms of the contribution to the jet power, we can compare these injected physical luminosities Γj2Lp′/2\Gamma_j^2 L^\prime_{\mathrm{p}}/2 to the Eddington luminosity LEdd≈1.26×1038(MBH/M⊙) erg s−1≈8.75×1046 erg s−1L_{\mathrm{Edd}} \approx1.26 \times10^{38}(M_{\mathrm{BH}}/M_\odot)\ \mathrm{erg\,s}^{-1} \approx8.75 \times10^{46}\ \mathrm{erg\,s}^{-1}. We exceed the Eddington luminosity by a factor 2 to 20. Due to the sub-dominance of the hadronic components, the SEDs of T4 and T5 are compatible with a proton injection luminosity Lp′=0L'_{p} = 0, however, in the other epochs, the contribution to the gamma-ray luminosity added by the hadronic component is required in order to explain the light curve (Figure 2). In T4, the neutrino prediction is below the axis limit of the plot. The primary populations injected in this epoch are strongly constrained by the gamma-ray data. In order to reproduce the cutoff in the GeV in the SED, the minimum Lorentz factor of the injected electrons is pushed to γe,min=7×103\gamma_{\mathrm{e,min}} = 7 \times10^{3}, which in turns subtancially diminishes the radio and X-ray emission, compared to T1 for example, that has similar parameters but a less constraining gamma-ray spectrum.

The shaded areas in Figure 3 demonstrate the constraints that a good multi-wavelength coverage in the SED has on the proton injection luminosity. When we vary Lp′L'_{\mathrm{p}} even in T2 and T6 where the gamma-ray flux low and only comprise upper limits, the X-ray data are importantly constraining the range of fitting values, while in T1, T3, and T4, the constraint arises from the gamma-ray flux above GeV energies.

Neutrino expectations

Over the six time intervals, the total muon neutrino luminosity varies from L⟨νμ+νˉμ⟩≈4×1043 erg s−1L_{\langle\nu_\mu+\bar{\nu}_\mu\rangle} \approx4 \times10^{43}\ \mathrm{erg\ s^{-1}} to L⟨νμ+νˉμ⟩≈4×1045 erg s−1L_{\langle\nu_\mu+\bar{\nu}_\mu\rangle} \approx4 \times10^{45}\ \mathrm{erg\ s^{-1}}. The highest predicted neutrino flux is obtained in epoch T3, which is contemporaneous to when the high-energy neutrino event IC110929A was detected. In this epoch, the protons are injected with the largest luminosity, although there is no X-ray data to constrain the model further. Our models are predicting neutrinos of ∼195−820 TeV\sim195 - 820\ \mathrm{TeV} energy, which is right above the energy range where the IceCube detector is the most sensitive, at the source’s declination (see the IceCube sensitivity curve in Figure 3).

For each epoch, we compute the predicted number of neutrinos observable by the IceCube observatory, using the publicly available effective area of the detector8, at the position of the source (Aeff(E,δ)A_{\mathrm{eff}}(E,\delta)), where δ\delta is the declination of the source:

N⟨νμ+νˉμ⟩=∫0∞∫tstarttstopdE dt Aeff(E,δ) Φ⟨νμ+νˉμ⟩(E,t),N_{\langle\nu_\mu+\bar{\nu}_\mu\rangle} = \int_{0}^{\infty}\int_{t_{\mathrm{start}}}^{t_{\mathrm{stop}}} dE\,dt\, A_{\mathrm{eff}}(E,\delta)\,\Phi_{\langle\nu_\mu+\bar{\nu}_\mu\rangle}(E,t),

where Φ⟨νμ+νˉμ⟩(E,t)\Phi_{\langle\nu_\mu+\bar{\nu}_\mu\rangle}(E,t) is the sum of the muon neutrino and anti-neutrino flux in units of neutrinos per unit energy per unit of time per unit of surface, predicted by the models. The number of neutrinos predicted per epoch is reported in the last row of Table 2.

Similarly to the flux, the highest predicted neutrino rate is obtained in epoch T3. Assuming that the predicted number of neutrinos in each epoch follows a Poisson distribution, the 3σ3\sigma confidence interval is reported along with the predicted number of neutrino per epoch in Table 2, as an asymmetrical uncertainty. For epoch T3, this corresponds to [0−2][0 - 2] neutrinos, which is compatible with the detection of one neutrino event over these 4.9 years. The corresponding probability to detect at least one neutrino during this epoch, still assuming Poisson statistics, is

P(Ndet≥1)=1−exp⁡(−N⟨νμ+νˉμ⟩)≈9.6%P(N_{\mathrm{det}} \ge1) = 1 - \exp(-N_{\langle\nu_\mu+\bar{\nu}_\mu\rangle}) \approx9.6\%.

Cumulating all six epochs ∑N⟨νμ+νˉμ⟩\sum N_{\langle\nu_\mu+\bar{\nu}_\mu\rangle}, we predict a total number of 0.17 neutrinos observable by IceCube in 16.4 years for the best-fit case, corresponding to a probability of detecting at least one neutrino P(Ndet≥1)≈16%P(N_{\mathrm{det}} \ge1) \approx16\%.

Discussion

5BZQ J0808+4950 has been studied by several authors in the context of population studies ([121, 138, 45, 75]), particularly related to estimating the Doppler factor of the core, and possible core shifts. In Finke [45] the modeling reports 1.1≤Γj≤11.91.1 \le\Gamma_{\mathrm{j}} \le11.9, 1.2≤δj≤15.11.2 \le\delta_{\mathrm{j}} \le15.1 and 0.1≤B1pc≤12.90.1 \le B_{\mathrm{1pc}} \le12.9, while Hovatta et al. [75] report a Doppler factor of ∼35\sim35, based on the light curve variability timescale. The Lorentz factor value obtained through our fitting method lies within the estimates Finke [45], which is consistent with the lack of variability that we observe in the light curve. The predicted SED is also consistent with the classification of 5BZQ J0808+4950 as a low-synchrotron-peak blazar, i.e., the synchrotron radiation from the primary electrons peaks at frequency νsyn≲1014 Hz\nu_{\mathrm{syn}} \lesssim10^{14}\ \mathrm{Hz} while the high-energy hump of the SED is peaking around νIC≳1020 Hz\nu_{\mathrm{IC}} \gtrsim10^{20}\ \mathrm{Hz} when modeled by a purely leptonic model in Paliya et al. [121].

In Paliya et al. [121] and Rodrigues et al. [138], this source is found to be consistent with purely leptonic models. This is compatible with our findings in epochs T4 and T5, given the sub-dominance of the proton-initiated components, as visible in Figure 3, where we estimate the range of possible proton luminosity to be compatible with 0. However, the previous studies were focusing only on the average SED of 5BZQ J0808+4950, which we can see in Figure 3 with the gray data points. The way we model the SED is different from their approach, because we use a time-dependent assumption. As can be seen in Figure 3, when the source exhibits a lower gamma-ray flux (T2 and T6), the SED differs significantly from the average, and the relative contribution of the hadronic components is more prominent than during the gamma-ray-enhanced states.

The jet power predicted by our model is larger than the Eddington luminosity for this source, indicating that the kinetic output is larger than what can be powered by accretion. This is a common conclusion of lepto-hadronic models for neutrino-emitter-candidate blazars [e.g. 86, 56, 35, 94], mainly due to the contribution of relativistic protons in our case. The Eddington luminosity is not in itself a fundamental limit on the jet power. Another quantity one can compare the jet power to is the power of the Blandford-Znajek process [24]. The predicted jet power we find is within the estimates of Liodakis & Petropoulou [94] PBZjet=1047.0±1.38 erg/sP_{\mathrm{BZ}}^{\mathrm{jet}} = 10^{47.0\pm1.38}\ \mathrm{erg/s}. Nevertheless, even if jets carrying powers of the order of ∼M˙c2\sim\dot{M}c^2 (i.e., Pjet≈1.5×1047 erg/sP_{\mathrm{jet}} \approx1.5 \times10^{47}\ \mathrm{erg/s} in our case) are not uncommon (Tchekhovskoy et al. 2011; Zamaninasab et al. 2014; Ghisellini et al. 2014), the typical efficiency is assumed to be of the order of ∼0.1M˙c2\sim0.1\dot{M}c^2 (e.g McKinney 2005). A way to ease the tension with the jet power would be if protons are accelerated with a harder spectral index pp→1p_{\mathrm{p}} \to1 (through magnetic reconnection, studies show that the index can be as hard as 1.2 in the case of large magnetization, see e.g Guo et al. 2014; Sironi & Spitkovsky 2014; Werner et al. 2016). With the current energy-density ratios between magnetic field and protons, we find, however, that acceleration through magnetic reconnection would be inefficient (see, e.g. Guo et al. 2024, for a review), and hence, in order to reach such hard power-law injection, the particles would need to be accelerated outside the blob, in highly magnetized regions. The photo-hadronic interactions are dominated by the high-energy protons, so a harder spectrum will lead to a larger fraction of those high-energy protons compared to the low-energy ones. However, the neutrino predictions would not drastically change, given the constraints on the subdominance of the hadronic contribution to the total SED. 5BZQ J0808+4950 shares common characteristics with other FSRQs. This can be seen in Figure 4, where we show the source’s physical luminosities (Le/pphys=Γ2Le/p′/2L^{\mathrm{phys}}_{\mathrm{e/p}} = \Gamma^2 L'_{\mathrm{e/p}}/2) in Eddington units (i.e., normalized to the Eddington luminosity of each source), baryonic loading (Lp′/Le′L'_{\mathrm{p}}/L'_{\mathrm{e}}) and neutrino production efficiency Lν,all−flavor′/Lp′L'_{\nu,\mathrm{all-flavor}}/L'_{\mathrm{p}} as a function of disk and gamma-ray luminosities. The baryonic loading is used to compare the power injected into protons, which are needed to produce neutrinos, and the power injected into electrons, which are the primary population contributing the most to the SED. However, a large baryonic loading does not necessarily imply efficient neutrino production, which is why we also compare the neutrino production efficiency to the gamma-ray luminosity. We compare our results with the study of 324 blazars proposed by Rodrigues et al. [138]. The FSRQ are represented by the cyan points, while BL Lac objects are shown in purple. The red cross shows the results of the modeling of 5BZQ J0808+4950 in the same paper, while the stars show the quantities derived in this present work. The relations between quantities derived in Rodrigues et al. [138] are shown with the black solid lines.

Figure 4. Comparison with the physical quantities from the 324-blazar sample of Rodrigues et al. [138]. The cyan blue points represent the FSRQs of the sample, while the purple points show the BL Lacs. The best-fit values are represented with the filled circles, the empty circles in the upper right and lower panels represent the sources for which the modeling is compatible with a purely leptonic model (i.e., the lowest possible proton injection luminosity is zero) while the plain triangles show the sources for which the best-fit is purely leptonic and in this case we use the maximum proton luminosity to compute the physical luminosity, the baryonic loading and the neutrino production efficiency, so these points are considered upper limits. The quantities obtained by modeling 5BZQ J0808+4950 in Rodrigues et al. [138] are shown in red, while the ones we derive are shown with the stars or with a downward triangle when the proton luminosity for the epoch is compatible with 0. We also indicate the results of this previous study for three IceCube neutrino candidate sources, PKS 0735+178, PKS 1502+106, and GB6 J1040+0617. The solid lines show the relations derived in Rodrigues et al. [138], the dotted gray lines show the equality relation, and the dashed line gives the relation derived in Palladino et al. [123]. Upper panels: physical luminosity Γj2Le/p′/2\Gamma_{\mathrm{j}}^{2}L_{\mathrm{e/p}}^{\prime}/2 as a function of the disk luminosity (obtained from Paliya et al. 2017, for the blazar sample). Lower left: baryonic loading (Lp′/Le′L_{\mathrm{p}}^{\prime}/L_{\mathrm{e}}^{\prime}) as a function of the gamma-ray luminosity. For the blazar sample, we use the Fermi-LAT flux reported in Rodrigues et al. [138], and for this work, we integrate the modeled SEDs between 100 MeV and 100 GeV. Lower right: All-flavor neutrino production efficiency Lν,all′/Lp′L_{\nu,\mathrm{all}}^{\prime}/L_{\mathrm{p}}^{\prime} as a function of the gamma-ray luminosity.

Our models predict electrons and protons luminosities characteristic of the powerful accretors, as shown in the top right and left panels of Figure 4, where 5BZQ J0808+4950 is located where the largest disk luminosities are found. In the lower left panel of Figure 4, we show the baryonic loading and the curve derived by Palladino et al. [123] that shows the evolution of the baryonic loading with respect to the gamma-ray luminosity in order not to violate the IceCube stacking limit. The predicted baryonic loading in 5BZQ J0808+4950 is consistent with the fact that no significant excess in neutrino emission has been observed towards the blazar population. Although the baryonic loading is low for 5BZQ J0808+4950 the neutrino production efficiency (that is defined as the ratio of the all-flavor comoving neutrino luminosity over the comoving proton luminosity) is larger than for other FSRQs, which is the main peculiarity of 5BZQ J0808+4950 compared to the global FSRQ population. In Figure 4, lower right panel, the source lies above the trend derived in Rodrigues et al. [138]. As a comparison, we show three candidate neutrino sources, PKS 0735+178, PKS 1502+106 and GB6 J1040+0617 in yellow, pink, and lime, respectively. In comparison with PKS 0735+178 and GB6 J1040+0617, which have gamma-ray luminosities of the same order as 5BZQ J0808+4950 we predict a higher neutrino luminosity, of the order of 1 to 40 times larger. However, this prediction relies only on the total integrated neutrino luminosity and not on the neutrino energy, which is a feature to assess whether a source can be an IceCube neutrino candidate. Our prediction lies below the efficiency given by Rodrigues et al. [138], but this value is given as an upper limit on the proton and neutrino luminosities, as they found the source to be compatible with a purely leptonic model on average.

We also compare the predicted neutrino luminosity with three energy-band luminosities from our best-fit models for the different epochs in Figure 5. Namely, we investigate whether there is a correlation with either of the gamma-ray (integrated between 0.1 GeV and 100 GeV), the X-ray (integrated between 0.3 keV and 10 keV) or the MeV (integrated between 0.1 MeV and 100 MeV) luminosity. Each luminosity is calculated from the SED prediction for each epoch.

Figure 5. Comparison between the predicted neutrino luminosity and the gamma-ray (0.1 − 100 GeV), X-ray (0.3 − 10 keV), and MeV (0.1 − 100 MeV) luminosities obtained for the best-fit solutions of 2, from left to right. The linear best-fit results are indicated in each panel. The grey shaded areas represent the 1σ1\sigma-uncertainty band on the slope and the intercept of the lines. We report the Pearson correlation coefficient and the corresponding p-value on the bottom left of each panel. The best-fit correlation found in Rodrigues et al. [138] is shown with the dashed black line in the left panel.

We perform a least-square regression to find the correlation Lνμ=b(LiLi,scale)aL_{\nu_\mu} = b\left(\frac{L_i}{L_{i,\mathrm{scale}}}\right)^{a} where LiL_i is one of the three luminosities mentioned previously, and Li,scaleL_{i,\mathrm{scale}} is the scale of this given quantity over the six epochs. The results of the fits are given in each panel of Figure 5. We report the best-fit values and their uncertainty in Table 3.

Relationα\alphab/1045b/10^{45}Lscale/1046L_{\mathrm{scale}}/10^{46}
log⁡Lνμ−log⁡Lγ\log L_{\nu_\mu} - \log L_{\gamma}0.22±0.790.22 \pm0.790.92±0.730.92 \pm0.735.005.00
log⁡Lνμ−log⁡LX\log L_{\nu_\mu} - \log L_X1.12±0.191.12 \pm0.191.50±0.391.50 \pm0.393.003.00
log⁡Lνμ−log⁡LMeV\log L_{\nu_\mu} - \log L_{\mathrm{MeV}}0.20±0.850.20 \pm0.850.78±0.620.78 \pm0.6210.010.0

Table 3. Best-fit results of the correlation between the neutrino luminosities and the other luminosities, presented in Figure 5.

We find that, if a hadronic component is present, as in the best-fit solutions we present here, the predicted muon- and anti-muon-neutrino luminosity correlates best with the X-ray luminosity, visible in the second panel of Figure 5. The MeV band has a tendency to be even less correlated to the neutrino luminosity. Even if currently there are no constraints on the MeV data, and hence, depends strongly on the model prediction, this energy band is typically where the electromagnetic cascade induced by the presence of protons emerges (see, e.g Rodrigues et al. [136]; Reimer et al. [130]; Petropoulou et al. [125]), as can be seen in epoch T2 in Figure 3. For this reason we explored whether the MeV band could be an indicator of the neutrino production, even for subdominant hadronic components. We do not find a significant correlation, but the addition of MeV data with future instruments such as the Compton Spectrometer and Imager (COSI, Tomsick et al. [162]; Tomsick & COSI collaboration [161]), All-sky Medium Energy Gamma-ray Observatory eXplorer (AMEGO-X, Caputo et al. [33]), or enhanced ASTROGAM (e-ASTROGAM, De Angelis et al. [38]) would nonetheless be crucial in constraining the SED models of this source, given the strong emission we find in this band. The neutrino-to-gamma-ray ratio Yγ=Lνμ+νˉμ/LγY_{\gamma} = L_{\nu_\mu+\bar{\nu}_\mu}/L_{\gamma} for the 6 epochs varies between 8×10−48 \times10^{-4} (in the epochs where the Compton dominance9 is larger than unity) and 1.5×10−11.5 \times10^{-1} (in T2, where the Compton dominance is the lowest). In the time interval T3, in which the neutrino flux is predicted with the largest value, Yγ≈4.3×10−2Y_{\gamma} \approx4.3 \times10^{-2}. These values are compatible with the 5 – 10% ratio found in several studies (e.g Aartsen et al. 2017; Wang & Loeb [167]; Padovani et al. [119]; Capel et al. [32]) assessing the contribution of gamma-ray blazars to the diffuse neutrino flux detected by IceCube. In comparison, the trend for the total population of FSRQs and BL Lacs studied in Rodrigues et al. [138] is shown as the dashed black line. We find a trend Lνμ∝Lγ0.22±0.79L_{\nu_\mu} \propto L_{\gamma}^{0.22 \pm0.79}, which is compatible within the uncertainty to the population trend, at the level of an individual source. The correlation between the X-ray luminosity and the muon-neutrino luminosity has a smaller uncertainty than the gamma-ray correlation and a larger Pearson r-coefficient. This hints that the X-ray variations of 5BZQ J0808+4950 are a better indicator of the neutrino activity over time.

Conclusions

We have studied the neutrino-candidate flat-spectrum radio quasar 5BZQ J0808+4950 in the context of six contemporaneous, multi-wavelength epochs. We used the numerical, semi-analytical software AM3 in order to derive time-averaged lepto-hadronic SEDs for each epoch. We modeled the emission region as a sphere moving relativistically, and took into account the influence of external photons emitted by an accretion disk and reprocessed by a BLR and a dusty torus, as the emission region travels along the jet axis. We find solutions with compact emission regions and moderate bulk Lorentz factor, where the changes in the primary particles are the main variation over the epochs, with only a subdominant hadronic component and a moderate neutrino flux peaking below PeV energies. The contribution of the hadronic cascade is visible in the high-energy gamma rays, above GeV energies, while the rest of the SED is dominated by the leptonic components. The neutrino flux predicted in this work stays below the IceCube detector sensitivity, with a predicted rate ranging from 1×10−31 \times10^{-3} to 1×10−11 \times10^{-1} neutrinos per epoch. With the future upgrades of the IceCube observatory, namely IceCube-Gen2 [7], the source may be at reach of the more sensitive detector. The sensitivity is currently expected to be better by a factor of 5. The highest rate is obtained in the epoch T3, contemporaneous with the IceCube high-energy event IC110929A. Assuming Poisson statistics on the mean number of neutrinos detectable, the 1×10−11 \times10^{-1} neutrino number predicted during the epoch T3 is compatible with the detection of the high-energy neutrino IC110929A. Our model shows a stronger correlation between the predicted neutrino luminosity and the X-ray luminosity, rather than with the gamma-ray luminosity in the case of the best solutions including a hadronic component. In the context of modeling neutrino candidates, this may suggest that monitoring the X-ray variability of a source is a better proxy of neutrino emission in a given source than the gamma-ray variability.

Acknowledgements

We thank the anonymous referee for the insightful comments on the manuscript. This work was supported by the European Research Council, ERC Starting grant MessMapp, S.B. Principal Investigator, under contract no. 949555. This research has made use of data and/or software provided by the High Energy Astrophysics Science Archive Research Center (HEASARC), which is a service of the Astrophysics Science Division at NASA/GSFC. This research furthermore has made use of the NASA/IPAC Extragalactic Database, which is funded by the National Aeronautics and Space Administration and operated by the California Institute of Technology. Part of this work is based on archival data, software or online services provided by the Space Science Data Center - ASI. We thank the developers of AM3 for providing the framework used to interpret the data used for this work. The Fermi-LAT Collaboration acknowledges support for LAT development, operation, and data analysis from NASA and DOE (United States), CEA/Irfu and IN2P3/CNRS (France), ASI and INFN (Italy), MEXT, KEK, and JAXA (Japan), and the K.A. Wallenberg Foundation, the Swedish Research Council, and the National Space Board (Sweden). Science analysis support in the operations phase from INAF (Italy) and CNES (France) is also gratefully acknowledged. This work was performed in part under DOE Contract DE-AC02-76SF00515. We acknowledge the use of public data from the Neil Gehrels Swift Observatory data archive. This work presents results from the European Space Agency (ESA) space mission Gaia. Gaia data are being processed by the Gaia Data Processing and Analysis Consortium (DPAC). Funding for the DPAC is provided by national institutions, in particular the institutions participating in the Gaia MultiLateral Agreement (MLA). The Gaia mission website is https://www.cosmos.esa.int/gaia. The Gaia archive website is https://archives.esac.esa.int/gaia. This work is based on observations obtained with the Samuel Oschin Telescope 48-inch and the 60-inch Telescope at the Palomar Observatory as part of the Zwicky Transient Facility project. ZTF is supported by the National Science Foundation under Grant No. AST-2034437 and a collaboration including Caltech, IPAC, the Weizmann Institute for Science, the Oskar Klein Center at Stockholm University, the University of Maryland, Deutsches Elektronen-Synchrotron and Humboldt University, the TANGO Consortium of Taiwan, the University of Wisconsin at Milwaukee, Trinity College Dublin, Lawrence Livermore National Laboratories, and IN2P3, France. Operations are conducted by COO, IPAC, and UW. This research has made use of the RATAN-600 data, provided by the Special Astrophysical Observatory of the Russian Academy of Sciences (SAO RAS). We use radio data from the CATS database (www.sao.ru/cats/), operated at SAO RAS, Russia [165]. This research has made use of data from the OVRO 40-m monitoring program. This monitoring program has been supported by NSF grants AST-0808050 and AST-1109911, and is currently supported by NSF grants AST-2407603 and AST-2407604. It has also been supported by NASA grants NNX08AW31G, NNX11A043G, and NNX14AQ89G.

Fermi-LAT data analysis insights and results

For the analysis of 4FGL J0808.5+4950 we extracted photons of Pass 8 SOURCE and applied a zenith angle cut of 90∘90^\circ to limit the impact of high energy (HE) photons produced in Earth’s upper atmosphere. In a next step fermipy performs a binning both in the spatial and the energy distributions. For this we adopted a pixel size of 0.08 deg and 12 energy bins per decade. We furthermore took into account energy dispersion for point sources, but excluding the isotropic diffuse component to avoid modeling inconsistencies with older diffuse templates as suggested in the fermipy documentation10. Beyond the 10∘×10∘10^\circ\times10^\circ photon information, the pre-fit ROI model furthermore includes the source models of all known sources of the 4FGL-DR4 catalog (Ballet et al. [21]; Abdollahi et al. [11]) located within 15∘15^\circ from the ROI center, as well as the corresponding galactic and isotropic diffuse emission (gll_iem_v07.fits and iso_P8R3_SOURCE_V3_v1.txt11) (Atwood et al. [17]; Bruel et al. [26]). The analysis is performed in 2 steps. First we analyzed the gamma-ray properties of the source for the full-time interval performing the following steps:

  1. Using the configuration stated before, we run an initial optimization (using fermipy’s optimize() function) to possibly account for changes in the source models that are taken from the 14-year catalog while we now analyze ∼16\sim16 years of data. The function uses an iterative fitting procedure so that ideally all the source parameters in the ROI are close to their global likelihood maxima.

  2. We then search for the existence of possible new significant gamma-ray emission using fermipy’s find_sources() function, assuming a point source with its spectral shape following a power law, an excess of at least TS=5\sqrt{TS}=5 and a minimum distance to its nearest neighbor of at least rmin⁡=0.5∘r_{\min}=0.5^\circ.

  3. With fermipy’s localize() function we then scan the likelihood surface of a local patch around the source of interest to refine its position in the ROI.

  4. The final best fit model (with fermipy’s fit() function) is then computed by freeing the normalizations and spectral parameters for 4FGL J0808.5+4950 and all sources within a radius of rmin⁡=3∘r_{\min}=3^\circ, as well as the normalization for the diffuse fluxes and all sources with TS≥10TS \ge10. We allow the fitting function to retry 3 times to converge to a good fit quality.

  5. Finally we compute the gamma-ray SED and lightcurve using fermipy’s sed() and lightcurve() functions.

In step 2 we find the new point source PS J0738.8+5336. We obtain a flux of F1 GeV−1000 GeV=(1.14±0.23)×10−10 ph cm−2 s−1F_{\mathrm{1\,GeV-1000\,GeV}}=(1.14\pm0.23)\times10^{-10}\,\mathrm{ph\,cm^{-2}\,s^{-1}} and an index of Index=−2.38±0.14\mathrm{Index}=-2.38\pm0.14 which, within its errors, is consistent with the values for FL16Y J0738.7+5336 that was added to the recent 16-year source list of Fermi-LAT (FL16Y; Ballet et al. [20]). There is no known multi-wavelength counterpart associated to this source yet. The relocalization of 4FGL J0808.5+4950 resulted in the new coordinates RA =122.17∘±0.41∘=122.17^\circ\pm0.41^\circ and DEC =49.79∘±0.41∘=49.79^\circ\pm0.41^\circ, which within its bounds is consistent with the coordinates reported in FL16Y for FL16Y J0808.5+4947, associated with our source. 4FGL J0808.5+4950 is reported to follow a log-parabolic spectral shape. For this we obtain the following best fit values: F1 GeV−1000 GeV=(1.94±0.25)×10−10 ph cm−2 s−1F_{\mathrm{1\,GeV-1000\,GeV}}=(1.94\pm0.25)\times10^{-10}\,\mathrm{ph\,cm^{-2}\,s^{-1}}, α=2.82±0.09\alpha=2.82\pm0.09 and β=0.01±0.06\beta=0.01\pm0.06. The spectral index α\alpha is steeper than the value reported in FL16Y (αref=2.70±0.09\alpha_{\mathrm{ref}}=2.70\pm0.09) and the curvature value β\beta contributed less to the spectral shape of the best fit model (βref=0.12±0.07\beta_{\mathrm{ref}}=0.12\pm0.07), however, the differences are mild and overlap in their uncertainties. They might arise from differences in the analysis procedure and differences in the energy range taken into account for the catalog. For the analysis of the shorter time intervals we repeated the same analysis procedure, however we included the new point source PS J0738.8+5336 to the ROI of each analysis and fixed 4FGL J0808.5+4950 to the position we obtained from the localization, to ensure a well constrained model and consistency throughout the analysis steps. No additional point sources are found for the shorter integration times. A summary of the results is displayed in Table A.1.

IntervalTimeF1 GeV−1000 GeV×10−10F_{1\,\mathrm{GeV}-1000\,\mathrm{GeV}} \times10^{-10} ph cm−2^{-2} s−1^{-1}α\alphaβ\betaTS
T154682 – 550471.34±0.601.34 \pm0.602.91±0.372.91 \pm0.370.38±0.480.38 \pm0.4821.09
T255047 – 554124.43--3.80
T355412 – 572372.76±0.392.76 \pm0.392.93±0.092.93 \pm0.09−0.07±0.05-0.07 \pm0.05177.47
T457237 – 586870.67±0.250.67 \pm0.253.37±0.283.37 \pm0.280.03±0.180.03 \pm0.1852.26
T558687 – 601572.43±0.622.43 \pm0.622.71±0.162.71 \pm0.16−0.04±0.06-0.04 \pm0.0665.72
T660157 – 606751.71--1.75

Table A.1. Fermi-LAT analysis results for the individual time intervals analyzed for 4FGL J0808.5+4950. The analysis procedure is summarized in Appendix A and the gamma-ray SEDs created from the analysis steps are used for the modelling shown in Figure 3. If the source is not significantly detected for the time interval, the 95% flux upper limit is reported.

References

  1. [2]Aartsen, M. G., Abbasi, R., Abdou, Y., et al. 2013, Physical Review Letters, 111, 021103, doi: 10.1103/PhysRevLett.111.021103DOI
  2. [3]Aartsen, M. G., Ackermann, M., Adams, J., et al. 2014, Physical Review Letters, 113, 101101, doi: 10.1103/PhysRevLett.113.101101DOI
  3. [4]Aartsen, M. G., Abraham, K., Ackermann, M., et al. 2015a, The Astrophysical Journal, 809, 98, doi: 10.1088/0004-637X/809/1/98DOI
  4. [7]Aartsen, M. G., Abbasi, R., Ackermann, M., et al. 2021, Journal of Physics G Nuclear Physics, 48, 060501, doi: 10.1088/1361-6471/abd48
  5. [8]Abazajian, K. N., Adelman-McCarthy, J. K., Agüeros, M. A., et al. 2009, The Astrophysical Journal Supplement Series, 182, 543, doi: 10.1088/0067-0049/182/2/543
  6. [9]Abbasi, R., Ackermann, M., Adams, J., et al. 2023a, The Astrophysical Journal, 954, 75, doi: 10.3847/1538-4357/acdfcb
  7. [11]Abdollahi, S., Acero, F., Baldini, L., et al. 2022, The Astrophysical Journal Supplement Series, 260, 53, doi: 10.3847/1538-4365/ac6751
  8. [12]Ackermann, M., Ajello, M., Allafort, A., et al. 2012, Science, 338, 1190, doi: 10.1126/science.1227160
  9. [13]Aharonian, F. A. 2000, New Astronomy, 5, 377, doi: 10.1016/S1384-1076(00)00039-7
  10. [14]Arnaud, K. A. 1996, in Astronomical data analysis software and systems V, Vol. 101, 17, https://ui.adsabs.harvard.edu/abs/1996ASPC..101...17A
  11. [17]Atwood, W., Albert, A., Baldini, L., et al. 2013, Pass 8: Toward the Full Realization of the Fermi-LAT Scientific Potential, arXiv, doi: 10.48550/ARXIV.1303.3514DOI
  12. [18]Azzollini, A., Buson, S., Coleiro, A., et al. 2025, Astronomy & Astrophysics, 700, A228, doi: 10.1051/0004-6361/202451328
  13. [19]Babusiaux, C., Fabricius, C., Khanna, S., et al. 2023, Astronomy and Astrophysics, 674, A32, doi: 10.1051/0004-6361/202243790
  14. [20]Ballet, J., Bruel, P., Burnett, T. H., & Lott, B. 2026, Fermi-LAT 16-year Source List, arXiv, doi: 10.48550/ARXIV.2602.22148DOI
  15. [21]Ballet, J., Bruel, P., Burnett, T. H., Lott, B., & The Fermi-LAT collaboration. 2023, Fermi-LAT collaboration, arXiv e-prints, arXiv:2307.12546, doi: 10.48550/arXiv.2307.12546
  16. [22]Bellenghi, C., Padovani, P., Resconi, E., & Giommi, P. 2023, The Astrophysical Journal, 955, L32, doi: 10.3847/2041-8213/acf711
  17. [23]Best, P. N., & Heckman, T. M. 2012, Monthly Notices of the Royal Astronomical Society, 421, 1569, doi: 10.1111/j.1365-2966.2012.20414.x
  18. [24]Blandford, R. D., & Znajek, R. L. 1977, Monthly Notices of the Royal Astronomical Society, 179, 433, doi: 10.1093/mnras/179.3.433
  19. [25]Breeveld, A. A., Landsman, W., Holland, S. T., et al. 2011, in American institute of physics conference series, Vol. 1358, Gamma ray bursts 2010, ed. J. E. McEnery, J. L. Racusin, & N. Gehrels, 373–376, doi: 10.1063/1.3621807
  20. [26]Bruel, P., Burnett, T. H., Digel, S. W., et al. 2018, Fermi-LAT improved Pass~8 event selection, arXiv, doi: 10.48550/ARXIV.1810.11394DOI
  21. [30]Böttcher, M., Reimer, A., Sweeney, K., & Prakash, A. 2013, The Astrophysical Journal, 768, 54, doi: 10.1088/0004-637X/768/1/54
  22. [31]Calderone, G., Sbarrato, T., & Ghisellini, G. 2012, Monthly Notices of the Royal Astronomical Society: Letters, 425, L41, doi: 10.1111/j.1745-3933.2012.01296.x
  23. [32]Capel, F., Burgess, J. M., Mortlock, D. J., & Padovani, P. 2022, Astronomy & Astrophysics, 668, A190, doi: 10.1051/0004-6361/202243116
  24. [33]Caputo, R., Ajello, M., Kierans, C. A., et al. 2022, Journal of Astronomical Telescopes, Instruments, and Systems, 8, doi: 10.1117/1.JATIS.8.4.044003
  25. [34]Cash, W. 1979, The Astrophysical Journal, 228, 939, doi: 10.1086/156922
  26. [35]Cerruti, M., Zech, A., Boisson, C., et al. 2019, Monthly Notices of the Royal Astronomical Society: Letters, 483, L12, doi: 10.1093/mnrasl/sly210
  27. [36]Cerruti, M., Zech, A., Boisson, C., & Inoue, S. 2015, Monthly Notices of the Royal Astronomical Society, 448, 910, doi: 10.1093/mnras/stu2691
  28. [37]Costamante, L., Bonnoli, G., Tav ecchio, F., et al. 2018, Monthly Notices of the Royal Astronomical Society, 477, 4257, doi: 10.1093/mnras/sty857
  29. [38]De Angelis, A., Tatischeff, V., Grenier, I., et al. 2018, Journal of High Energy Astrophysics, 19, 1, doi: 10.1016/j.jheap.2018.07.001
  30. [39]Dermer, C. D., Murase, K., & Inoue, Y. 2014, Journal of High Energy Astrophysics, 3-4, 29, doi: 10.1016/j.jheap.2014.09.001
  31. [43]Domínguez, A., Primack, J. R., Rosario, D. J., et al. 2011, Monthly Notices of the Royal Astronomical Society, 410, 2556, doi: 10.1111/j.1365-2966.2010.17631.x
  32. [44]Drury, L. O. 1983, Reports on Progress in Physics, 46, 973, doi: 10.1088/0034-4885/46/8/002
  33. [45]Finke, J. D. 2019, The Astrophysical Journal, 870, 28, doi: 10.3847/1538-4357/aaf00c
  34. [46]Finke, J. D., Razz a que, S., & Dermer, C. D. 2010, The Astrophysical Journal, 712, 238, doi: 10.1088/0004-637X/712/1/238
  35. [48]Fitzpatrick, E. L. 1999, Publications of the Astronomical Society of the Pacific, 111, 63, doi: 10.1086/316293
  36. [49]Fossati, G., Maraschi, L., Celotti, A., Comastri, A., & Ghisellini, G. 1998, Monthly Notices of the Royal Astronomical Society, 299, 433, doi: 10.1046/j.1365-8711.1998.01828.x
  37. [50]Franceschini, A., Rodighiero, G., & Vaccari, M. 2008, Astronomy & Astrophysics, 487, 837, doi: 10.1051/0004-6361:200809691
  38. [51]Francis, P. J., Hewett, P. C., Foltz, C. B., et al. 1991, The Astrophysical Journal, 373, 465, doi: 10.1086/170066
  39. [52]Franckowiak, A., Garrappa, S., Paliya, V., et al. 2020, The Astrophysical Journal, 893, 162, doi: 10.3847/1538-4357/ab8307
  40. [53]Frank, J., King, A., & Raine, D. J. 2002, Accretion Power in Astrophysics: Third Edition (Cambridge University Press). https://ui.adsabs.harvard.edu/abs/2002apa..book.....F
  41. [54]Gaia Collaboration, Prusti, T., de Bruijne, J. H. J., et al. 2016, Astronomy and Astrophysics, 595, A1, doi: 10.1051/0004-6361/201629272
  42. [55]Gaia Collaboration, Vallenari, A., Brown, A. G. A., et al. 2023, Astronomy and Astrophysics, 674, A1, doi: 10.1051/0004-6361/202243940
  43. [56]Gao, S., Fedynitch, A., Winter, W., & Pohl, M. 2019, Nature Astronomy, 3, 88, doi: 10.1038/s41550-018-0610-1
  44. [58]Garrappa, S., Buson, S., Franckowiak, A., et al. 2019, The Astrophysical Journal, 880, 103, doi: 10.3847/1538-4357/ab2ada
  45. [59]Gehrels, N., Chincarini, G., Giommi, P., et al. 2004, The Astrophysical Journal, 611, 1005, doi: 10.1086/422091
  46. [60]Ghisellini, G., & Madau, P. 1996, Monthly Notices of the Royal Astronomical Society, 280, 67, doi: 10.1093/mnras/280.1.67
  47. [61]Ghisellini, G., & Tavecchio, F. 2009, Monthly Notices of the Royal Astronomical Society, 397, 985, doi: 10.1111/j.1365-2966.2009.15007.x
  48. [62]Ghisellini, G., Tavecchio, F., Foschini, L., & Ghirlanda, G. 2011, Monthly Notices of the Royal Astronomical Society, 414, 2674, doi: 10.1111/j.1365-2966.2011.18578.x
  49. [63]Ghisellini, G., Tavecchio, F., Foschini, L., et al. 2010, Monthly Notices of the Royal Astronomical Society, 402, 497, doi: 10.1111/j.1365-2966.2009.15898.x
  50. [66]Giommi, P., Padovani, P., Oikonomou, F., et al. 2020b, Astronomy and Astrophysics, 640, L4, doi: 10.1051/0004-6361/202038423
  51. [67]Giommi, P., Perri, M., Capalbi, M., et al. 2021, Monthly Notices of the Royal Astronomical Society, 507, 5690, doi: 10.1093/mnras/stab2425
  52. [68]Gu, M. 2013, The Astrophysical Journal, 773, 176, doi: 10.1088/0004-637X/773/2/176
  53. [72]Harrison, F. A., Craig, W. W., Christensen, F. E., et al. 2013, The Astrophysical Journal, 770, 103, doi: 10.1088/0004-637X/770/2/103
  54. [74]Hillas, A. M. 1984, Annual Review of Astronomy and Astrophysics, 22, 425, doi: 10.1146/annurev.aa.22.090184.002233
  55. [75]Hovatta, T., Valtaoja, E., Tornikoski, M., & Lähteenmäki, A. 2009, Astronomy and Astrophysics, 494, 527, doi: 10.1051/0004-6361:200811150
  56. [77]IceCube Collaboration. 2013, Science, 342, doi: 10.1126/science.1242856
  57. [78]IceCube Collaboration. 2022, Evidence for neutrino emission from the nearby active galaxy NGC 1068, IceCube Neutrino Observatory, doi: 10.21234/03FQ-RH11DOI
  58. [79]IceCube Collaboration, Aartsen, M., Ackermann, M., et al. 2018, Science, 361, 147, doi: 10.1126/science.aat2890
  59. [80]IceCube Collaboration, Abbasi, R., Ackermann, M., et al. 2022, Science, 378, 538, doi: 10.1126/science.abg3395
  60. [83]Kadler, M., Kraus, F., Mannheim, K., et al. 2016, Nature Physics, 12, 807
  61. [86]Keivani, A., Murase, K., Petropoulou, M., et al. 2018, The Astrophysical Journal, 864, 84, doi: 10.3847/1538-4357/aad59a
  62. [88]Klinger, M., Rudolph, A., Rodrigues, X., et al. 2023, arXiv e-prints, arXiv:2312.13371, doi: 10.48550/arXiv.2312.13371
  63. [89]Klinger, M., Rudolph, A., Rodrigues, X., et al. 2024, The Astrophysical Journal Supplement Series, 275, 4, doi: 10.3847/1538-4365/ad725c
  64. [90]KM3NeT Collaboration, MessMapp Group, LAT Collaboration, et al. 2025, Characterizing Candidate Blazar Counterparts of the Ultra-High-Energy Event KM3-230213A, arXiv, doi: 10.48550/ARXIV.2502.08484DOI
  65. [92]Lincetto, M., Bremer, A., Buson, S., et al. 2025, in Proceedings of 39th International Cosmic Ray Conference — PoS(ICRC2025) (Geneva, Switzerland: Sissa Medialab), 1099, doi: 10.22323/1.501.1099
  66. [94]Liodakis, I., & Petropoulou, M. 2020, The Astrophysical Journal Letters, 893, L20, doi: 10.3847/2041-8213/ab830a
  67. [95]Lister, M. L., Aller, H. D., Aller, M. F., et al. 2009, The Astronomical Journal, 137, 3718, doi: 10.1088/0004-6256/137/3/3718
  68. [96]Madsen, K. K., Harrison, F. A., Markwardt, C. B., et al. 2015, The Astrophysical Journal Supplement Series, 220, 8, doi: 10.1088/0067-0049/220/1/8
  69. [97]Mannheim, K. 1993, The proton blazar., arXiv, doi: 10.48550/arXiv.astro-ph/9302006
  70. [100]Mannheim, K., Stanev, T., & Biermann, P. L. 1992, Astronomy and Astrophysics, 260, L1. https://ui.adsabs.harvard.edu/abs/1992A&A...260L...1M
  71. [101]Masci, F. J., Laher, R. R., Rusholme, B., et al. 2018, Publications of the Astronomical Society of the Pacific, 131, 018003, doi: 10.1088/1538-3873/aae8ac
  72. [102]Massaro, E., Maselli, A., Leto, C., et al. 2015, Astrophysics and Space Science, 357, doi: 10.1007/s10509-015-2254-2
  73. [107]Mundo, S. A., & Mushotzky, R. 2023, Monthly Notices of the Royal Astronomical Society, 526, 4040, doi: 10.1093/mnras/stad2991
  74. [112]Mücke, A., & Protheroe, R. J. 2001, Astroparticle Physics, 15, 121, doi: 10.1016/S0927-6505(00)00141-9
  75. [118]Owen, F. N., Porcas, R. W., Mufson, S. L., & Moffett, T. J. 1978, The Astronomical Journal, 83, 685, doi: 10.1086/112253
  76. [119]Padovani, P., Petropoulou, M., Giommi, P., & Resconi, E. 2015, Monthly Notices of the Royal Astronomical Society, 452, 1877, doi: 10.1093/mnras/stv1467
  77. [121]Paliya, V. S., Domínguez, A., Ajello, M., Olmo-García, A., & Hartmann, D. 2021, The Astrophysical Journal Supplement Series, 253, 46, doi: 10.3847/1538-4365/abe135
  78. [123]Palladino, A., Rodrigues, X., Gao, S., & Winter, W. 2019, The Astrophysical Journal, 871, 41, doi: 10.3847/1538-4357/aaf507
  79. [124]Petropoulou, M., Dimitrakoudis, S., Padovani, P., Mastichiadis, A., & Resconi, E. 2015, Monthly Notices of the Royal Astronomical Society, 448, 2412, doi: 10.1093/mnras/stv179
  80. [125]Petropoulou, M., Murase, K., Santander, M., et al. 2020, The Astrophysical Journal, 891, 115, doi: 10.3847/1538-4357/ab76d0
  81. [126]Planck Collaboration, Aghanim, N., Akrami, Y., et al. 2020, Astronomy & Astrophysics, 641, A6, doi: 10.1051/0004-6361/201833910
  82. [128]Poole, T. S., Breeveld, A. A., Page, M. J., et al. 2008, Monthly Notices of the Royal Astronomical Society, 383, 627, doi: 10.1111/j.1365-2966.2007.12563.x
  83. [130]Reimer, A., Böttcher, M., & Buson, S. 2019, The Astrophysical Journal, 881, 46, doi: 10.3847/1538-4357/ab2bff
  84. [131]Richards, J. L., Hovatta, T., Max-Moerbeck, W., et al. 2014, Monthly Notices of the Royal Astronomical Society, 438, 3058, doi: 10.1093/mnras/stt2412
  85. [132]Richards, J. L., Max-Moerbeck, W., Pavlidou, V., et al. 2011, The Astrophysical Journal Supplement Series, 194, 29, doi: 10.1088/0067-0049/194/2/29
  86. [133]Riello, M., De Angeli, F., Evans, D. W., et al. 2021, Astronomy and Astrophysics, 649, A3, doi: 10.1051/0004-6361/202039587
  87. [134]Rodrigo, C., & Solano, E. 2020, in XIV.0 scientific meeting (virtual) of the spanish astronomical society, 182
  88. [135]Rodrigo, C., Solano, E., & Bayo, A. 2012, SVO Filter Profile Service Version 1.0, Tech. rep., doi: 10.5479/ADS/bib/2012ivoa.rept.1015R
  89. [136]Rodrigues, X., Gao, S., Fedynitch, A., Palladino, A., & Winter, W. 2019, The Astrophysical Journal Letters, 874, L29, doi: 10.3847/2041-8213/ab1267
  90. [138]Rodrigues, X., Paliya, V. S., Garrappa, S., et al. 2024b, Astronomy & Astrophysics, 681, A119, doi: 10.1051/0004-6361/202347540DOI
  91. [139]Roming, P. W. A., Kennedy, T. E., Mason, K. O., et al. 2005, Space Science Reviews, 120, 95, doi: 10.1007/s11214-005-5095-4DOI
  92. [140]Sahakyan, N., Giommi, P., Padovani, P., et al. 2023, Monthly Notices of the Royal Astronomical Society, 519, 1396, doi: 10.1093/mnras/stac3607DOI
  93. [141]Sanchez Zaballa, J. M., Buson, S., Marchesi, S., et al. 2025, The Astrophysical Journal, 988, 120, doi: 10.3847/1538-4357/addfd2
  94. [143]Scargle, J. D., Norris, J. P., Jackson, B., & Chiang, J. 2013, The Astrophysical Journal, 764, 167, doi: 10.1088/0004-637X/764/2/167
  95. [144]Schlafly, E. F., & Finkbeiner, D. P. 2011, The Astrophysical Journal, 737, 103, doi: 10.1088/0004-637X/737/2/103
  96. [145]Shen, Y., & Liu, X. 2012, The Astrophysical Journal, 753, 125, doi: 10.1088/0004-637X/753/2/125
  97. [146]Shen, Y., Richards, G. T., Strauss, M. A., et al. 2011, The Astrophysical Journal Supplement Series, 194, 45, doi: 10.1088/0067-0049/194/2/45
  98. [147]Sironi, L., Keshet, U., & Lemoine, M. 2015, Space Science Reviews, 191, 519, doi: 10.1007/s11214-015-0181-8DOI
  99. [148]Sironi, L., & Spitkovsky, A. 2011, The Astrophysical Journal, 726, 75, doi: 10.1088/0004-637X/726/2/75
  100. [150]Sironi, L., Spitkovsky, A., & Arons, J. 2013, The Astrophysical Journal, 771, 54, doi: 10.1088/0004-637X/771/1/54
  101. [151]Smith, D. H., Hooper, D., & Vieregg, A. 2021, Journal of Cosmology and Astroparticle Physics, 2021, 031, doi: 10.1088/1475-7516/2021/03/031
  102. [155]Tanaka, Y. T., Buson, S., & Kocevski, D. 2017, The Astronomer’s Telegram, 10791, 1. https://ui.adsabs.harvard.edu/abs/2017ATel10791....1T
  103. [156]Tavecchio, F., Maraschi, L., & Ghisellini, G. 1998, The Astrophysical Journal, 509, 608, doi: 10.1086/306526
  104. [158]The Astropy Collaboration, Robitaille, T. P., Tollerud, E. J., et al. 2013, Astronomy & Astrophysics, 558, A33, doi: 10.1051/0004-6361/201322068
  105. [161]Tomsick, J., & COSI collaboration. 2021, in Proceedings of 37th International Cosmic Ray Conference — PoS(ICRC2021) (Berlin, Germany – Online: Sissa Medialab), 652, doi: 10.22323/1.395.0652
  106. [162]Tomsick, J., Zoglauer, A., Sleator, C., et al. 2019, The Compton Spectrometer and Imager, eprint: arXiv:1908.04334: arXiv, doi: 10.48550/arXiv.1908.04334DOI
  107. [163]Urry, C. 2004, AGN Unification: An Update, eprint: arXiv:astro-ph/0312545: arXiv, doi: 10.48550/arXiv.astro-ph/0312545DOI
  108. [164]Urry, C. M., & Padovani, P. 1995, Publications of the Astronomical Society of the Pacific, 107, 803, doi: 10.1086/133630DOI
  109. [165]Verkhodanov, O. V., Trushkin, S. A., Andernach, H., & Chernenkov, V. N. 2007, doi: 10.48550/ARXIV.0705.2959DOI
  110. [166]Verner, D. A., Ferland, G. J., Korista, K. T., & Yakovlev, D. G. 1996, The Astrophysical Journal, 465, 487, doi: 10.1086/177435
  111. [167]Wang, X., & Loeb, A. 2016, Journal of Cosmology and Astroparticle Physics, 2016, 012, doi: 10.1088/1475-7516/2016/12/012
  112. [169]Wilms, J., Allen, A., & McCray, R. 2000, The Astrophysical Journal, 542, 914, doi: 10.1086/317016
  113. [170]Wood, M., Caputo, R., Charles, E., et al. 2017, in Proceedings of 35th International Cosmic Ray Conference — PoS(ICRC2017) (Bexco, Busan, Korea: Sissa Medialab), 824, doi: 10.22323/1.301.0824
  114. [172]Zhang, B. T., Petropoulou, M., Murase, K., & Oikonomou, F. 2020, The Astrophysical Journal, 889, 118, doi: 10.3847/1538-4357/ab659a
  115. [174]Zirakashvili, V. N., & Aharonian, F. 2007, Astronomy & Astrophysics, 465, 695, doi: 10.1051/0004-6361:20066494

Paper details

vlnc:W_g3ax8nrr
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents