The Bright γ-ray Flare of 3C 279 in 2015 June: AGILE Detection and Multifrequency Follow-up Observations

Blazars are a subclass of radio-loud active galactic nuclei with relativistic jets pointing toward the observer (Urry & Padovani 1995). Their emission extends from the radio band to the γ-ray band above 100 MeV up to TeV γ-rays, and it is dominated by variable nonthermal processes. They come in two main flavors, with very different optical spectra: Flat Spectrum Radio Quasars (FSRQs) that have strong, broad optical emission lines; and BL Lacertae objects (BL Lacs) with an optical spectrum that can be completely featureless or can show, at most, weak emission lines and some absorption features (e.g., see Giommi et al. 2012 for a detailed review on blazar classification). The blazar spectral energy distribution (SED) is, in general, characterized by two broad bumps: the low-energy one, spanning from the radio to the X-ray band, is attributed to synchrotron radiation, while the high-energy one, from the X-ray to the γ-ray band, is thought to be due to inverse Compton (IC) emission. In the leptonic scenario, this second component is due to relativistic energetic electrons scattering their own synchrotron photons—Synchrotron self-Compton (SSC)—or photons external to the jet—External Compton (EC). Blazars of both flavors have been found to be highly variable and particularly so in γ-rays.³⁴ The correlated variability between X-rays and γ-rays is usually well explained in the SSC or EC framework (Ghisellini et al. 1998). In fact, a new class of "orphan" γ-ray flares from FSRQ blazars is now emerging from observations, challenging the current simple one-zone leptonic models. In particular, a number of γ-ray flares from some extensively monitored FSRQs, such as 3C 279, do not correlate with optical and soft X-ray events of comparable power and timescales; see, for example, the results of a previous multiwavelength campaign on 3C 279 during flaring states in 2013–2014 (Hayashida et al. 2015).

Gamma-ray observations of flaring blazars and simultaneous multiwavelength data are thus the key to investigate possible alternative theoretical scenarios, such as a recently proposed model based on a mirror-driven process within a clumpy jet inducing localized and transient enhancements of synchrotron photon density beyond the broad-line region (BLR; Tavani et al. 2015; Vittorini et al. 2017). Other scenarios consider special structures, such as spine-sheath jet layers radiative interplay  (Tavecchio & Ghisellini 2008; Sikora et al. 2016) or "rings" of fire, i.e., synchrotron-emitting rings of electrons representing a shocked portion of the jet sheath (MacDonald et al. 2015).

3C 279 is associated with a luminous FSRQ at z =0.536 (Lynds et al. 1965), with prominent broad emission lines detected in all accessible spectral bands and revealing highly variable emission. It consistently shows strong γ-ray emission, which are already clearly detected by EGRET (Hartman et al. 1992; Kniffen et al. 1993), AGILE (Giuliani et al. 2009), Fermi- Hayashida et al. 2012, 2015), and also detected above 100 GeV by MAGIC (Albert et al. 2008). The central black hole mass estimates are in the range of $(3\mbox{--}8)\times {10}^{8}\,{M}_{\odot }$ (Gu et al. 2001; Woo & Urry 2002; Nilsson et al. 2009). The 3C 279 jet features strings of compact plasmoids, as indicated by radio observations (Hovatta et al. 2009), which may be a by-product of the magnetic reconnection process (Petropoulou et al. 2016), even though it must be taken into account that the superluminal knots observed in Very Long Baseline Interferometry images are probably much larger structures than the reconnection plasmoids formed on kinetic plasma scales, hence this connection is uncertain (Chatterjee et al. 2008).

Here we present the results of a multiband observing campaign on the blazar 3C 279 triggered by the detection of intense γ-ray emission above 100 MeV by the AGILE satellite in 2015 June (Lucarelli et al. 2015). The source is one of the γ-ray blazars monitored by the GLAST-AGILE³⁵ Support Program (GASP) of the Whole Earth Blazar Telescope (WEBT) Collaboration³⁶ (Bottcher et al. 2007; Larionov et al. 2008; Villata et al. 2008; Abdo et al. 2010).

The AGILE-Gamma-ray Imaging Detector (-GRID) γ-ray data of 3C 279 in 2015 June are compared with as yet unpublished (R-band) optical GASP-WEBT observations during the flare, including the percentage and angle of polarization, and with Fermi-Fermi-Large Area Telescope (-LAT; Paliya et al. 2015; Ackermann et al. 2016) and other multiwavelength data from the Swift-Ultraviolet/Optical Telescope (-UVOT) and the Swift-X-Ray Telescope (-XRT) Target of Opportunities (ToO). The analysis of the source multiwavelength behavior is crucial in order to study the correlation (if any) of the γ-ray radiation with the optical-ultraviolet (-UV) and X-ray emissions. The 2015 June flaring data are also compared with nonsimultaneous archival data from the NASA/IPAC Extragalactic Database and from the ASI Space Science Data Center (SSDC, previously known as ASDC).

2.1. AGILE Observations

2.2. GASP-WEBT Observations

2.3. Swift ToO Observations

2.4. Fermi-LAT Observations

3.1. Light Curves

In Figure 1, we present the simultaneous (and as yet unpublished) AGILE γ-ray and GASP-WEBT optical light curves during the 3C 279 flare in 2015 June. In order to produce the AGILE light curve, we divided the data collected in the period from 2015 June 11 to 18 (MJD: 57184–57191) in 24 and 12 hr timebins. To derive the estimated flux of the source, we ran the AGILE Multi-Source Maximum Likelihood Analysis (ALIKE) task with an analysis radius of 10°. The ALIKE was carried out by fixing the position of the source to its nominal radio position (Johnston et al. 1995), (l, b) = (305.104, 57.062) (deg) and using Galactic and isotropic diffuse emission parameters (GAL-ISO) fixed at the values estimated during the two weeks preceding the analyzed AGILE data set.

Zoom In Zoom Out Reset image size

The extended GASP-WEBT optical light curve (R-band magnitude) of 3C 279 since the end of 2014, including the γ-ray flaring period (MJD: 57010–57220), is shown in Figure 2. It includes the polarization percentage P and electric vector polarization angle (EVPA) variations. The total brightness variation in this period is ∼1.5 magnitude, from R = 16.07 at MJD = 57142.1 to R = 14.58 at MJD = 57189.6.

Zoom In Zoom Out Reset image size

The multiwavelength behavior of the source during the flare is then summarized in Figure 3, which includes γ-ray light curves, as observed by AGILE-GRID and Fermi-LAT, the prompt Swift-XRT X-ray followup, and the simultaneous GASP-WEBT de-absorbed optical flux densities and polarimetric data.

Zoom In Zoom Out Reset image size

A well-defined maximum peaking around MJD = 57189 is visible at γ-rays, which is in agreement with the optical observations. The degree of observed polarization P remains always high, ranging between about 9% and 30%. The maximum observed value occurs at MJD = 57190.2, and the daily sampling allows to identify a 1 day delay of the P maximum after the flux peak observed at optical and γ-ray frequencies. The rise and the following decrease of P and flux is accompanied by a rotation of the electric vector polarization angle of about 30° in 10 days.

As shown in Figure 3's third panel, the X-ray flux variability also appears correlated with the γ-ray and optical ones. The peak X-ray flux value occurs at MJD = 57189.14, and it is about a factor of about four higher than the one observed one day later (see Table 2).

3.2. Spectral Energy Distribution

Figure 4 shows the 3C 279 broadband SED obtained with the help of the SSDC SED Builder tool.³⁹ Simultaneous AGILE, GASP-WEBT, Swift-XRT, and Swift-UVOT data during the 2015 June flare are shown in red. Average γ-ray flux excluding the flaring period and other public nonsimultaneous archival data in other wavelengths are shown in gray.

Zoom In Zoom Out Reset image size

We have performed the AGILE spectral analysis of the peak γ-ray activity, corresponding to the period between 2015 June 14 (MJD = 57187.0) and 2015 June 17 (MJD = 57190.0) over three energy bins: 100–200, 200–400, and 400–1000 MeV. A simple power-law spectral fitting gives a photon index of Γ_γ = (2.14 ± 0.11), which is consistent within the errors with the values reported by Fermi  (Paliya et al. 2015; Ackermann et al. 2016). Moreover, we estimated the average γ-ray fluxes obtained by integrating in the whole AGILE energy band (100 MeV–50 GeV) during three time periods, defined as a pre-outburst (MJD: 57184–57187), a flare (MJD: 57187–57190), and a post-flare (MJD: 57190–57193). The corresponding AGILE integral γ-ray fluxes and spectral indices are summarized in Table 4. Historically, this is the largest γ-ray flare (≥100 MeV) of 3C 279 ever observed, including recent activity reported in Bulgarelli et al. (2017).

Table 4.  AGILE γ-ray Fluxes and Spectral Indices

Label	Tstart	Tstop	F (E ≥100 MeV)	Γγ
	MJD	MJD	(10−6 ph cm−2 s−1)
Pre-outburst	57184.0	57187.0	(1.7 ± 0.7)	(2.0 ± 0.4)
Flare	57187.0	57190.0	(13.0 ± 1.3)	(2.1 ± 0.1)
Post-flare	57190.0	57193.0	(1.0 ± 0.5)	⋯

Note. Over the considered 3 day time periods, the source flux increases of factor of about 7, then rapidly drops more than a factor of 10 in the post-flare, with insufficient statistics for the spectral analysis.

Download table as:  ASCIITypeset image

The SED during the flare (red points in Figure 4) indicates a very high "Compton dominance": the ratio of the IC peak to the synchrotron 1 is of the order of 100. Specifically, the γ-ray spectrum integrated over 1 day timebins rises by a factor of ∼3 in a few hours (as shown in Figure 3), yielding a Compton dominance of about 100, and attaining values up to ∼200 when integrating on even shorter timescales (Ackermann et al. 2016).

In this section, we estimate the parameters of a tentative simple modeling of the multiwavelength 3C 279 data acquired during the 2015 flare. The model parameter values obtained here can be used as reference input for a more detailed further theoretical analysis.

In the framework of the one-zone leptonic model for FSRQs (see e.g., Paggi et al. 2011), the optical and UV data acquired during the 2015 June flare, presented here, would constrain the luminosity of the accretion disk to L_D ≤ 10⁴⁶ erg s⁻¹. We note that this value is larger, by a factor of about 3, than the disk luminosity previously inferred for 3C 279 (Raiteri et al. 2014).

Taking into account also the simultaneous soft X-ray data and the observed variability, we can determine empirical constraints on the model parameters: the size l, the bulk boost factor Γ, the energetic content in magnetic field B, and the electron energy distribution n_e(γ) of the emitting region. We assume that the relativistic electrons have a double power-law energy-density distribution:

$$\begin{eqnarray}&&{n}_{e}(\gamma )=\displaystyle \frac{K\,{\gamma }_{b}^{-1}}{{(\gamma /{\gamma }_{b})}^{{\zeta }_{1}}+{(\gamma /{\gamma }_{b})}^{{\zeta }_{2}}}\,[{\mathrm{cm}}^{-3}],\end{eqnarray} \tag{ 1 }$$

where K is a normalization factor, γ_b is the break Lorentz factor, ζ₁ and ζ₂ are the double power-law spectral indices below and above the break, respectively.

These electrons interact via the IC process with the synchrotron photons internal to the same emitting region, with the external photons coming from the accretion disk and from the BLR. From distances R_BLR ≃ 0.1 pc, the latter reflects a fraction ξ ≃ few % of the disk radiation. In Figure 5, we show our one-zone SED model of the 2015 June flare of 3C 279 for γ-ray fluxes averaged on 1 day timescales. If we assume the emitting region located at a distance R < R_BLR from the central black hole, then seed photons coming from BLR are good candidates to be scattered into γ-rays of observed energies ≥100 MeV (see the red line in Figure 5). As shown by the blue lines in the same figure, disk photons entering the emitting region from behind are scattered mainly in the hard X-ray observed band. Instead, the internal scattering of the synchrotron photons are seen mainly in the soft X-ray band, as shown by the green lines.

Zoom In Zoom Out Reset image size

In this model, we consider the emitting region placed at a distance R = 6  × 10¹⁶ cm from the central black hole, while the accretion disk radiates the power L_D = 10⁴⁶ erg s⁻¹; a fraction ξ = 2% of this is reflected back from the BLR placed at distance ${R}_{\mathrm{BLR}}=0.15$ pc. A summary of the best-fit flare model parameters is shown in Table 5.

When the IC scattering occurs in the Thomson regime, the Compton dominance reads $q={U}_{\mathrm{ext}}^{{\prime} }/{U}_{B}^{{\prime} }$, i.e., the ratio of the comoving energy density of BLR seed photons ${U}_{\mathrm{ext}}^{{\prime} }\simeq (1\,+{\beta }_{{\rm{\Gamma }}}^{2}){{\rm{\Gamma }}}^{2}\xi \,{L}_{D}\,/(4\pi \,c\,{R}_{\mathrm{BLR}}^{2})$ to the energy density of the magnetic field ${U}_{B}^{{\prime} }={B}^{{\prime} 2}/8\pi$, thus:

$$\begin{eqnarray}&&q\,\lesssim \,0.2\,{{\rm{\Gamma }}}^{2}\displaystyle \frac{(\xi /0.02){L}_{D,46}}{{(B^{\prime} /{\rm{G}})}^{2}{({R}_{\mathrm{BLR}}/0.1\mathrm{pc})}^{2}}.\end{eqnarray} \tag{ 2 }$$

For assumed disk luminosities L_D ≤ 10⁴⁶ erg s⁻¹, this yields a value of q  ≤  80. Moreover, the one-zone assumption has two other main consequences.

• 1.  
  First, a strict correlation of optical and γ-ray fluxes: their variations must be of the same entity, so the Compton dominance should not vary.
• 2.  
  Second, to increase the upper limit for q up to values above 100, as observed, we should consider faint magnetic fields values B ≲ 0.1 G, which would in turn imply modest electron accelerations (Mignone et al. 2013). Alternatively, we could assume bulk factors Γ  >  30 (Ackermann et al. 2016), which considerably exceeds the value Γ ≃ 20 inferred from radio observations for this source (Hovatta et al. 2009), that would imply a conspicuous kinetic load in the jet.

Noticeably, the multiwavelength light curves of the flare in Figure 3 show instead that the Compton dominance rises by a factor of three or more in a half day, attaining values up to q > 200 in few minutes when considering the very fast and strong γ-ray variations reported in Ackermann et al. (2016). While the simple one-zone model presented here could account for the SED flaring data integrated on 1 day timescales (provided you assume of a very bright underlying disk), it is anyway seriously challenged by the observed strong and fast variation of the Compton dominance.

Furthermore, we notice that a single photon of energy E = 52 GeV was detected on MJD = 57189.62 (Paliya et al. 2015) in correspondence with the peak of optical emission and is consistent with the observed polarization fraction reaching its maximum. The modeling of this specific episode of high-energy emission goes beyond the scope of this paper and provides an additional argument for alternative modes of γ-ray emission.

In this paper, we present multifrequency optical and X-ray data simultaneous with the 2015 γ flaring activity of 3C 279. We use AGILE-GRID and Fermi-LAT (Paliya et al. 2015; Ackermann et al. 2016) γ-ray data together with the Swift-UVOT, the Swift-XRT, and as yet unpublished optical GASP-WEBT observations of 3C 279 in 2015 June. We find that from the multiwavelength light curve shown in Figure 3, the high-energy flare is partially correlated with the behavior in other energy bands. Specifically, the γ-ray flux rising by a factor ≃4 in half a day shows an optical counterpart rising only by a factor 2 or less on similar timescales. The γ-ray flux during this flare exceeds the largest 3C 279 flares previously detected, although Hayashida et al. (2015) reported an even more extreme multifrequency behavior for this source in the past; e.g., in 2013 December, the γ-ray flux above 100 MeV jumped by a factor ≃5 in a few hours without optical or X-ray counterparts, and the Compton dominance attained values of about 300. Ackermann et al. (2016) discuss the variability of the 2015 γ-ray flare with minute timescales.

The observed spectral characteristics and the strong and fast variations of the Compton dominance challenge one-zone models, unless we assume significant variations in the field of seed photons to be IC scattered into γ-rays. We discuss in this paper a one-zone model and provide the model parameters that can be used as a theoretical model of reference. Models alternative to the standard SSC and EC might be considered (e.g., Ackermann et al. 2016). In the moving mirror model (Tavani et al. 2015; Vittorini et al. 2017) localized enhancements of synchrotron photon density may explain the occurrence of gamma-ray flares with faint or no counterpart in other bands. These localized enhancements would persist only for short periods of time, and this would explain the fact that the majority of FSRQ γ-ray flares are not orphan in nature.

We noticed that, as shown in Figure 3, the degree of observed optical polarization P appears to correlate with the optical flux F during the flare, with P peaking about one day after F. Moreover, the polarization angle rotates by at least 30° in the period encompassing the flare. However, the behavior of the polarization degree of the jet may be very different from the observed one, due to the big blue bump dilution effect. When deriving the intrinsic jet polarization P_jet, the presence of a very luminous disc, as assumed by the one-zone model used to interpret the observed SEDs, would imply that the correction for the thermal emission contribution becomes noticeable as the flux approaches the observed minimum level. This would lead to much higher P_jet values than the observed ones, and P_jet would not maintain the general correlation with flux shown in Figure 2.

Partly based on data taken and assembled by the WEBT Collaboration and stored in the WEBT archive at the Osservatorio Astrofisico di Torino—INAF.⁴⁰ For questions about data availability, contact the WEBT President Massimo Villata (villata@oato.inaf.it).

We would like to acknowledge the financial support of ASI under contract to INAF: ASI 2014-049-R.0 dedicated to SSDC. Part of this work is based on archival data, software, or online services provided by the ASI SSDC. The research at Boston University was supported by National Science Foundation grant AST-1615796 and NASA Swift Guest Investigator grant 80NSSC17K0309. This research was partially supported by the Bulgarian National Science Fund of the Ministry of Education and Science under grant DN 08-1/2016. The Skinakas Observatory is a collaborative project of the University of Crete, the Foundation for Research and Technology—Hellas, and the Max-Planck-Institut für Extraterrestrische Physik. The St. Petersburg University team acknowledges support from Russian Science Foundation grant 17-12-01029. This article is partly based on observations made with the telescope IAC80 operated by the Instituto de Astrofisica de Canarias in the Spanish Observatorio del Teide on the island of Tenerife. The IAC team acknowledges the support from the group of support astronomers and telescope operators of the Observatorio del Teide. Based (partly) on data obtained with the STELLA robotic telescopes in Tenerife, an AIP facility jointly operated by AIP and IAC. This work is partially based upon observations carried out at the Observatorio Astronómico Nacional on the Sierra San Pedro Mártir (OAN-SPM), Baja California, Mexico. C.P., V.V. and M.T. also thank Professor A. Cavaliere for the insightful discussion.

Software: AGILE software Package (AGILE SW 5.0 SourceCode), XRTDAS (v.3.1.0), XSPEC and HEASoft package (v. 6.17).