The high-energy -ray emission of AP Librae
Key Words.:gamma rays: observations – Galaxies : active – Galaxies : jets – BL Lacertae objects: individual objects: AP Librae
Pascal Fortin - email : email@example.com
Jonathan Biteau - email : firstname.lastname@example.org
The -ray spectrum of the low-frequency-peaked BL Lac (LBL) object AP Librae is studied, following the discovery of very-high-energy (VHE; ) -ray emission up to the TeV range by the H.E.S.S. experiment. This makes AP Librae one of the few VHE emitters of the LBL type. The measured spectrum yields a flux of above 130 GeV and a spectral index of . This study also makes use of Fermi-LAT observations in the high energy (HE, E100 MeV) range, providing the longest continuous light curve (5 years) ever published on this source. The source underwent a flaring event between MJD 56306-56376 in the HE range, with a flux increase of a factor 3.5 in the 14-day bin light curve and no significant variation in spectral shape with respect to the low-flux state. While the H.E.S.S. and (low state) Fermi-LAT fluxes are in good agreement where they overlap, a spectral curvature between the steep VHE spectrum and the Fermi-LAT spectrum is observed. The maximum of the -ray emission in the spectral energy distribution is located below the GeV energy range.
The BL Lac class of blazars constitutes about of both the First (Abdo et al. 2010b; 1LAC) and Second (Ackermann et al. 2011; 2LAC) Fermi Large Area Telescope (LAT) Catalogue of Active Galactic Nuclei (AGN), and constitutes the majority of the extragalactic very-high-energy (VHE, E100 GeV) -ray sources111An up-to-date VHE -ray catalogue can be found in the TeVCat, http://tevcat.uchicago.edu. AP Librae falls into the category of “low-frequency-peaked BL Lac” (LBL), defined by an X-ray to radio flux ratio of (Padovani & Giommi 1995), and of the more recently introduced “low frequency synchrotron peaked” (LSP) class of blazars defined by a synchrotron emission peak in the spectral energy distribution (SED) at (see Abdo et al. 2010c). This is an order of magnitude lower than the values found in the bulk of VHE -ray emitting blazars, which belong to the “high-frequency-peaked BL Lac/high frequency synchrotron peaked” (HBL/HSP) class. A continuity between these classes of blazars is suggested by the “blazar sequence” (Fossati et al. 1998), where the dominance of the high-energy component and its peak emission energy are inversely proportional to the total luminosity.
AP Librae was amongst the first few objects to be classified as a member of the BL Lac class (Strittmatter et al. 1972), for which a reliable redshift could be measured (; Disney et al. 1974). The initial redshift measurement is consistent with the most recent measurement from the 6dF galaxy survey (; Jones et al. 2009). An object coincident with AP Librae was discovered in the radio band (PKS 1514–24) during a survey made with the 210-ft reflector at Parkes (Bolton et al. 1964), but it was not until 1971 that the optically variable source AP Librae and the radio source PKS 1514–24 were formally associated (Bond 1971; Biraud 1971). The host galaxy harbors a black hole at its center with a mass, estimated using stellar velocity dispersion, of (Woo et al. 2005).
In X-rays, AP Librae was first detected by the Einstein X-Ray Observatory (1E 1514.72411; Schwartz & Ku 1983). At high energies (HE, E100 MeV), the source 3EG J15172538 (Hartman et al. 1999) was tentatively associated with AP Librae. The photon index reported in the third EGRET catalog was rather soft (), resulting in a low extrapolated flux level in the VHE range covered by atmospheric Cherenkov telescopes. Observations with the University of Durham Mark 6 -ray telescope resulted in a flux upper limit of for (Armstrong et al. 1999; Chadwick et al. 1999).
An early catalog of bright -ray sources detected by the Fermi-LAT was produced using the first three months of data (Abdo et al. 2009a). One of these sources, 0FGL J1517.92423, was associated with AP Librae, but its photon index was harder (, Abdo et al. 2009b) than that reported for 3EG J15172538. The extrapolation of its spectrum to higher energies raised the possibility of a detection by Cherenkov telescopes. In 2010, the H.E.S.S. Collaboration reported the detection of VHE -rays from AP Librae (Hofmann 2010). Following this announcement, Fortin et al. (2010) showed the first radio-to-TeV SED, based on the preliminary analysis of an HE-VHE data set included in the larger one presented here, while Kaufmann et al. (2011) also pointed out the existence of an X-ray jet resolved with Chandra, making Ap Librae the only known TeV BL Lac object with an extended jet in X-rays.
2.1 H.E.S.S. observations
The High Energy Stereoscopic System (H.E.S.S.), located in the Khomas Highland in Namibia ( S, E), is an array of telescopes (four at the time of the observations studied) that detect the Cherenkov light flashes from air showers. H.E.S.S. observed AP Librae between MJD 55326 (10 May 2010) and MJD 55689 (8 May 2011) for a total of 34 observations of 28 minutes, each passing data-quality selection criteria (described in Aharonian et al. 2006). This yields an exposure of acceptance-corrected live time with a mean zenith angle of . In order to minimize the spectral gap between Fermi-LAT and H.E.S.S., cuts achieving the lowest possible energy threshold were selected. The loose cuts (Aharonian et al. 2006), which require a minimum shower image intensity of 40 photoelectrons in each camera, were applied to the data set to perform the event selection, yielding an average energy threshold of . The Model analysis method (de Naurois & Rolland 2009) was used to analyze the data within a radius disk centered on the radio core position of AP Librae (, , Johnston et al. 1995) and further extract the spectrum and light curve, using the Reflected-Region method (Berge et al. 2007) to estimate the background contamination. With 1133 on-source events, 9042 off-source events and an on-off normalization of , the significance of the 218 -rays excess is (standard deviations, Li & Ma 1983). In Figure 1, the background (black crosses) and on-source events distributions (solid histogram) are shown as a function of the squared angular distance between the source position and the -ray direction. The H.E.S.S. point-spread function (PSF) was fitted to the on-source events and matches well both the signal and the background for large angular distances.
A point-like source model, convolved with the PSF, has been fitted to the data. The position obtained through this fit is and , compatible within the statistical errors with the location of the AP Librae core away (Johnston et al. 1995). Further morphological studies confirm the absence of source extension within the H.E.S.S. PSF.
The time-averaged photon spectrum for these data is shown in Figure 2. The best fit is a power-law function, within the energy range , with a probability of , given by:
where is the decorrelation energy. The best-fit parameters are obtained using a forward folding technique (Piron et al. 2001). Spectral points are derived with a similar approach in restricted energy ranges, with a fixed (to the best fit value) power-law index and a free normalization.
This result was cross-checked with a standard Hillas analysis (Aharonian et al. 2006) with the loose cuts, based also on a different calibration chain. It was found to be entirely compatible with the Model analysis and yielding a detection significance of and a photon index of (see also the comparison of both spectra in Figure 2). The upper limit on the flux derived from observations taken with the University of Durham Mark 6 -ray telescope (Chadwick et al. 1999), corresponding to of the Crab Nebula flux at E300 GeV, is also compatible with the H.E.S.S. spectrum since it is well above the flux level measured here.
The light curve of the integral flux above , averaged over the time between two successive full moons, is shown in Figure 3. A constant function fit to the time series yields a (/ndf = 3.2/3), which indicates that the light curve does not show any significant variability within the observed statistical errors. A 99% confidence level upper limit on the fractional variance (as defined in Vaughan et al. 2003) of is derived (Feldman & Cousins 1998). No variability is found using the Hillas analysis with the different calibration.
2.2 Fermi-LAT observations
The Fermi-LAT, launched on 2008 June 11, is a pair-conversion -ray detector
sensitive to photons in the energy range from 20 MeV to more than 300 GeV
(Atwood et al. 2009). The data for this analysis were taken from 4 August
2008 to 4 August 2013 (MJD 54682-56508, 5 years) and were analyzed using the
standard Fermi analysis software (ScienceTools v9r32p4) available from the
Fermi Science Support Center (FSSC)222http://fermi.gsfc.nasa.gov/ssc/.
Events with energy between 300 MeV and 300 GeV were selected from the Pass 7 data set. Only events passing the SOURCE class filter and
located within a square region of side length centered on AP Librae
were selected. Cuts on the zenith angle () and rocking angle
() were also applied to the data. The post-launch
P7SOURCE_V6 instrument response functions (IRFs) were used in combination
with the corresponding Galactic and isotropic diffuse emission
The model of the region includes the diffuse components and all sources from the
Second Fermi-LAT Catalog (2FGL, Nolan et al. 2012) located within a
square region of side centered on AP Librae. The spectral
parameters of the sources were left free during the fitting procedure. A
power-law correction in energy with free normalization and spectral slope was
applied to the Galactic diffuse component. Events were analyzed using the binned
maximum likelihood method as implemented in gtlike.
The source underwent a flaring episode of approximately 10 weeks between MJD 56306-56376 (flaring state). We have therefore defined a quiescent state measured during the periods MJD 54682-56305 and MJD 56377-56508.
AP Librae is detected with a high test statistic of TS=2037 (, Mattox et al. 1996) in the quiescent state. The energy spectrum evaluated using this data set is well described by a power-law with a photon index , in good agreement with the 2FGL value, with no significant indication for spectral curvature. The integral flux is , and the most energetic photon within the containment radius of the Fermi-LAT PSF has an energy of 71 GeV. The systematic uncertainties were evaluated using the bracketing IRFs technique (Ackermann et al. 2012).
Replacing the power-law with a log-parabola444The log-parabola model is defined as results in only a marginal improvement in likelihood ( for 1 degree of freedom, or approximately ). With this model, the best fit differential flux is at 5.48 GeV with an index and a curvature parameter .
The Fermi-LAT spectral error contour for the power-law model of AP Librae is presented in Figure 4. Flux values for individual energy bins were calculated independently, assuming a power-law spectral shape. For each energy bin, the spectral indices of all sources modeled in the region of interest were frozen to the best-fit values obtained for the full energy range and gtlike was used to determine the flux. The superimposed vertical error bars show the statistical uncertainties and the quadratic sum of statistical and systematic uncertainties, respectively. The latter were estimated by Ackermann et al. (2012) to be 10% of the effective area at 100 MeV, 5% at 560 MeV and 10% at 10 GeV and above. 95% confidence level upper limits were calculated for energy bins with TS values below 10. For completeness, the result of the log-parabola fit is also shown in Figure 4.
The variability analysis of the LAT data showed a significant flare starting in 2013 January. During the flaring period MJD 56306-56376, the spectrum is well fitted by a power law with a total flux and a spectral index , consistent in shape with the spectrum during the quiescent period (see Figure 4). The peak flux in the two-week bin light curve is . The flaring state is discussed more extensively in Section 3.1. During this period, some observations were performed with the H.E.S.S. array, but the resulting data were too limited to be useful555Less than 1 hour of useful time was recorded during the flare. The limited duration and poor background estimation do not even give a useful limit on the flux..
3.1 The flaring state of AP Librae
The Fermi-LAT light curve of the flaring episode above 300 MeV is shown in Figure 5. The peak flux was 3.5 times greater than the averaged flux. The fastest doubling time scale (as defined in Zhang et al. 1999), corresponds to the rising part and has a value of days. The lightcurve has also been fitted with an asymmetric profile666A symmetric Gaussian profile is rejected at a level of 20 with respect to the function used in this work. where the time of the peak is and the rise and decay time are and . is a constant that is also fitted to the data. Fitting this function to the data yields a peak at , of amplitude , above a constant value of compatible with the low-state flux, for a . The rise time and decay time are found to be days and days, respectively. The rise time (or the doubling time scale) is compatible with 0 at a level, indicating a fast process but the lack of statistics prevents a more precise probe of this event by making shorter time bins. Substructures of the flare possibly present on shorter time scale might be hidden (see Saito et al. 2013, in which study complex flare structures were found when probing smaller time scale). An asymmetry in the rise and decay has already been seen in the GeV range for PKS 1502+106 (Abdo et al. 2010a) and in the TeV range during the 2006 flare of PKS 2155304 (Aharonian et al. 2007). The opposite behavior, i.e. a smaller decay timescale, has nevertheless been observed in the TeV range in the radio galaxy M 87 (Abramowski et al. 2012). However the time scale of the event reported in this work is much longer and might be of a different origin (e.g. the onset and decay of large scale structural changes in the jet or possibly a change in accretion parameters).
3.2 The LBL AP Librae
The first evidence for VHE -rays from an LBL-class blazar was the detection of BL Lacertae () at the significance level (Albert et al. 2007) corresponding to a flux that of the Crab Nebula. Its steep VHE spectrum () did not connect smoothly with the harder Fermi-LAT spectrum (Abdo et al. 2009c) () established after the measurement in VHE, but given the significant variability of the HE -ray flux of BL Lacertae (see Sokolovsky et al. 2010; Cutini 2011, 2012 and follow-up ATels), it is possible that the source was in a high VHE flux state at the time it was detected. Further evidence for VHE -ray emission from LBL-type objects was found with the detection of S5 0716+714 (Anderhub et al. 2009), a source with a steep VHE spectrum () and a harder HE spectrum (Ackermann et al. 2011) (). It appears that AP Librae has the smallest spectral change in the HE-VHE bands, with . We note however that according to the current classification of extragalactic VHE -ray emitters in the TeVCaT, (following the classification in the 2LAC), AP Librae would currently be the only VHE -ray emitter of the LBL class.
With , AP Librae has a rather soft HE spectrum among the population of 2FGL AGN also emitting in the VHE regime, for which the average photon index is (Sanchez et al. 2013). Only the BL Lac object 1RXS J101015.9-311909, BL Lacertae, W Comae and S5 0716+714 exhibit a spectral index (Nolan et al. 2012). The observed peak high energy emission in the SED of objects with is localized roughly above 10 GeV. This quantity, generally not known before the advent of Fermi, is of paramount importance for emission modeling (Tavecchio et al. 1998). Note that, Abramowski et al. (2012) reanalyzed the Fermi data of 1RXS J101015.9-311909 above 1 GeV and found a hard index of 1.71, which constrained to be around 100 GeV.
In the next subsection, the peak of the -ray emission of Ap Librae is quantified jointly using the data from H.E.S.S. and Fermi-LAT.
3.3 Broad-band gamma-ray emission of AP Librae
To further investigate the HE-VHE spectral feature, the Fermi-LAT best fit power-law spectrum was extrapolated to energies greater than and corrected for the EBL attenuation using the model of Franceschini et al. (2008). A comparison of this extrapolation with the H.E.S.S. spectrum yields a (probability ). The H.E.S.S. systematic uncertainties were included by shifting the energy by 10%777This value is slightly more conservative than the one derived by Meyer et al. (2010) using HE and VHE Crab Nebula data., which yields an uncertainty of (see Fig. 4). The same comparison based on an extrapolation of the log-parabola spectral hypothesis yields a (i.e. ), which suggests broad band curvature.
To quantify this curvature, the HE and VHE data points (not corrected for EBL) were fitted with power-law and log-parabola models, taking into account the statistical and systematic uncertainties (Fig. 6). In practice, the fit has been done in log-log space with either a first order (power-law) or a second order (log-parabola) polynomial function. The parameters obtained are given in Table 1. The fit of the data with the power-law yields a of (probability of %), while the log-parabola yields a of (probability of %). A likelihood ratio test prefers the latter model at a level of 4.3, which confirms the presence of curvature in the measured HE-VHE spectrum of AP Librae. However, the fitting method used for the broad band HE-VHE data points differs from the methods used within each energy range and has some limitations (i.e. not taking into account correlations between energy bins). A proper method to overcome such limitations would consist of a joint fit of the data, exploiting the response functions of both space-borne and ground-based -ray instruments, which is beyond the scope of this paper.
Correcting the VHE data points for EBL attenuation and repeating the same joint fit, the log-parabola model is then preferred at 2.9. In this case the power-law yields a of (probability of %) and the log-parabola a of (probability of %). Scaling up the EBL absorption by thirty percent, as in Abramowski et al. (2013), or using the model of Finke et al. (2010), does not significantly affect the latter results, due to the rather small redshift of the source.
EBL attenuation is unlikely to be the only explanation of the spectral break observed in the data. An intrinsic spectral turnover could be due to factors such as a break in the underlying electron energy distribution, the onset of the Klein-Nishina regime in the inverse-Compton emission process, or the absorption of -rays on the circumnuclear radiation fields (see the discussion on the possibly related phenomenon of GeV breaks observed in the spectra of flat-spectrum radio quasars: e.g., Finke et al. 2008; Ackermann et al. 2010; Tanaka et al. 2011; Aleksić et al. 2011). To elucidate this conundrum would require extensive multi-wavelength modeling of the SED of this complex object888See, e.g., Tavecchio et al. (2010) who noted the modeling difficulties with simple synchrotron self-Compton radiative scenarios already when VHE measurements were not yet available and with a shorter Fermi-LAT exposure than presented here., which is beyond the scope of this Research Note.
The description of the HE-VHE emission of AP Librae by a log-parabola allows of AP Librae to be estimated at MeV. This value of about 450 MeV is compatible with the low-energy boundary of the Fermi-LAT range and could then be considered as an upper limit. It can be compared to the values of determined by Abdo et al. (2010c) for the objects BL Lacertae, W Comae and S5 0716+714, using jointly Fermi and publicly available VHE spectra (30 MeV, 4100 MeV and 800 MeV, respectively). Such low-energy emission peaks are rather uncommon with respect to the bulk of extragalactic VHE emitters, which tend to have maximum emissions at or above hundreds of GeV. The broad-band emission of AP Librae is also rather peculiar, as discussed by Fortin et al. (2010) and Kaufmann et al. (2013), with an SED dominated by inverse-Compton and an X-ray spectrum that can not be explained by synchrotron emission, and that might originate from the same mechanism as the -ray emission. This is consistent with a high-energy component shifted toward lower energies and a peak location that could be below the Fermi-LAT energy range.
The LBL class of VHE emitting objects proves to be an interesting laboratory to test radiative model scenarios, and perhaps to identify parameters on which the LBL-HBL sequence could depend. At present, only a handful of LBL objects have been detected at VHE (or just this one, depending on the selection criteria), probably due to a bias toward HBL objects in observation strategies and because LSP objects are the smallest subset of all -ray selected BL Lac objects (Shaw et al. 2013). Observations with the H.E.S.S. II telescope, and the advent of the Cherenkov Telescope Array (CTA), which will open the possibility to perform an extragalactic survey (20% of the sky in 100 hours) with a sensitivity approaching one percent of the flux of the Crab Nebula (Dubus et al. 2012), should allow more LBL-type blazars to be detected, and give better insights into the physical processes at work.
The support of the Namibian authorities and of the University of Namibia in facilitating the construction and operation of H.E.S.S. is gratefully acknowledged, as is the support by the German Ministry for Education and Research (BMBF), the Max Planck Society, the French Ministry for Research, the CNRS-IN2P3 and the Astroparticle Interdisciplinary Programme of the CNRS, the U.K. Particle Physics and Astronomy Research Council (PPARC), the IPNP of the Charles University, the South African Department of Science and Technology and National Research Foundation, and by the University of Namibia. We appreciate the excellent work of the technical support staff in Berlin, Durham, Hamburg, Heidelberg, Palaiseau, Paris, Saclay, and in Namibia in the construction and operation of the equipment.
The LAT Collaboration acknowledges support from a number of agencies and institutes for both development and the operation of the LAT as well as scientific data analysis. These include NASA and DOE in the United States, CEA/Irfu and IN2P3/CNRS in France, ASI and INFN in Italy, MEXT, KEK, and JAXA in Japan, and the K. A. Wallenberg Foundation, the Swedish Research Council and the National Space Board in Sweden. Additional support from INAF in Italy and CNES in France for science analysis during the operations phase is also gratefully acknowledged.
The authors want to acknowledge the anonymous referee for his/her help that greatly improved the paper.
- Abramowski et al. (2012) H.E.S.S. Collaboration, Abramowski, A., Acero, F., et al. 2012, A&A, 542, A94
- Abdo et al. (2010a) Abdo, A. A., Ackermann, M., Ajello, M., et al. 2010a, ApJ, 710, 810
- Abdo et al. (2010b) Abdo, A. A., Ackermann, M., Ajello, M., et al. 2010b, ApJ, 715, 429
- Abdo et al. (2010c) Abdo, A. A., Ackermann, M., Agudo, I., et al. 2010c, ApJ, 716, 30
- Abdo et al. (2010c) Abdo, A. A., Ackermann, M., Ajello, M., et al. 2010c, ApJ, 722, 520
- Abdo et al. (2009a) Abdo, A. A., Ackermann, M., Ajello, M., et al. 2009a, ApJS, 183, 46
- Abdo et al. (2009b) Abdo, A. A., Ackermann, M., Ajello, M., et al. 2009b, ApJ, 700, 597
- Abdo et al. (2009c) Abdo, A. A., Ackermann, M., Ajello, M., et al. 2009c, ApJ, 707, 1310
- Abramowski et al. (2012) Abramowski, A., Acero, F., Aharonian, F., et al. 2012, ApJ, 746, 151
- Abramowski et al. (2013) Abramowski, A., Acero, F., et al. 2013, A&A, 550, A4
- Ackermann et al. (2011) Ackermann, M., Ajello, M., Allafort, A., et al. 2011, ApJ, 743, 171
- Ackermann et al. (2012) Ackermann, M., Ajello, M., Albert, A., et al. 2012, ApJS, 203, 4
- Ackermann et al. (2010) Ackermann, M., Ajello, M., Baldini, L., et al. 2010, ApJ, 721, 1383
- Aharonian et al. (2006) Aharonian, F., Akhperjanian, A. G., Bazer-Bachi, A. R., et al. 2006, A&A, 457, 899
- Aharonian et al. (2007) Aharonian, F., Akhperjanian, A. G., Bazer-Bachi, A. R., et al. 2007, ApJ, 664, L71
- Albert et al. (2007) Albert, J., Aliu, E., Anderhub, H., et al. 2007, ApJ, 666, L17
- Aleksić et al. (2011) Aleksić, J., Antonelli, L. A., Antoranz, P., et al. 2011, ApJ, 730, L8
- Anderhub et al. (2009) Anderhub, H., Antonelli, L. A., Antoranz, P., et al. 2009, ApJ, 704, L129
- Armstrong et al. (1999) Armstrong, P., Chadwick, P. M., Cottle, P. J., et al. 1999, Experimental Astronomy, 9, 51
- Atwood et al. (2009) Atwood, W. B., Abdo, A. A., Ackermann, M., et al. 2009, ApJ, 697, 1071
- Berge et al. (2007) Berge, D., Funk, S., & Hinton, J. 2007, A&A, 466, 1219
- Biraud (1971) Biraud, F. 1971, Nature, 232, 178
- Bolton et al. (1964) Bolton, J. G., Gardner, F. F., & Mackey, M. B. 1964, Australian Journal of Physics, 17, 340
- Bond (1971) Bond, H. E. 1971, ApJ, 167, L79+
- Chadwick et al. (1999) Chadwick, P. M., Lyons, K., McComb, T. J. L., et al. 1999, ApJ, 521, 547
- Cutini (2011) Cutini, S. 2011, The Astronomer’s Telegram, 3368, 1
- Cutini (2012) Cutini, S. 2012, The Astronomer’s Telegram, 4028, 1
- de Naurois & Rolland (2009) de Naurois, M. & Rolland, L. 2009, Astroparticle Physics, 32, 231
- Disney et al. (1974) Disney, M. J., Peterson, B. A., & Rodgers, A. W. 1974, ApJ, 194, L79
- Dubus et al. (2012) Dubus, G., Contreras, J. L., Funk, S., et al. 2013, Astroparticle Physics, 43, 317
- Feldman & Cousins (1998) Feldman, G. J. & Cousins, R. D. 1998, Phys. Rev. D, 57, 3873
- Finke et al. (2008) Finke, J. D., Dermer, C. D., & Böttcher, M. 2008, ApJ, 686, 181
- Finke et al. (2010) Finke, J. D., Razzaque, S., & Dermer, C. D. 2010, ApJ, 712, 238
- Fortin et al. (2010) Fortin, P., Fegan, S., Horan, D., et al. 2010, in 25th Texas Symposium on Relativistic Astrophysics, Proceedings of Science PoS(Texas2010)199
- Fossati et al. (1998) Fossati, G., Maraschi, L., Celotti, A., Comastri, A., & Ghisellini, G. 1998, MNRAS, 299, 433
- Franceschini et al. (2008) Franceschini, A., Rodighiero, G., & Vaccari, M. 2008, A&A, 487, 837
- Hartman et al. (1999) Hartman, R. C., Bertsch, D. L., Bloom, S. D., et al. 1999, ApJS, 123, 79
- Hofmann (2010) Hofmann, W. 2010, The Astronomer’s Telegram, 2743, 1
- Johnston et al. (1995) Johnston, K. J., Fey, A. L., Zacharias, N., et al. 1995, AJ, 110, 880
- Jones et al. (2009) Jones, D. H., Read, M. A., Saunders, W., et al. 2009, MNRAS, 399, 683
- Kaufmann et al. (2013) Kaufmann, S., Wagner, S. J., & Tibolla, O. 2013, ApJ, 776, 68
- Kaufmann et al. (2011) Kaufmann, S., Wagner, S., & Tibolla, O., 2011, arXiv:1111.2775
- Li & Ma (1983) Li, T.-P. & Ma, Y.-Q. 1983, ApJ, 272, 317
- Mattox et al. (1996) Mattox, J. R., Bertsch, D. L., Chiang, J., et al. 1996, ApJ, 461, 396
- Meyer et al. (2010) Meyer, M., Horns, D., & Zechlin, H.-S. 2010, A&A, 523, A2
- Nolan et al. (2012) Nolan, P. L., Abdo, A. A., Ackermann, M., et al. 2012, ApJS, 199, 31
- Padovani & Giommi (1995) Padovani, P. & Giommi, P. 1995, ApJ, 444, 567
- Piron et al. (2001) Piron, F., Djannati-Atai, A., Punch, M., et al. 2001, A&A, 374, 895
- Saito et al. (2013) Saito, S., Stawarz, Ł., Tanaka, Y. T., et al. 2013, ApJ, 766, L11
- Sanchez et al. (2013) Sanchez, D. A., Fegan, S., & Giebels, B. 2013, A&A, 554, A75
- Schwartz & Ku (1983) Schwartz, D. A. & Ku, W. H.-M. 1983, ApJ, 266, 459
- Shaw et al. (2013) Shaw, M. S., Romani, R. W., Cotter, G., et al. 2013, ApJ, 764, 135
- Sokolovsky et al. (2010) Sokolovsky, K. V., Schinzel, F. K., & Wallace, E. 2010, The Astronomer’s Telegram, 2402, 1
- Strittmatter et al. (1972) Strittmatter, P. A., Serkowski, K., Carswell, R., et al. 1972, ApJ, 175, L7+
- Tanaka et al. (2011) Tanaka, Y. T., Stawarz, Ł., Thompson, D. J., et al. 2011, ApJ, 733, 19
- Tavecchio et al. (2010) Tavecchio, F., Ghisellini, G., Ghirlanda, G., Foschini, L., & Maraschi, L. 2010, MNRAS, 401, 1570
- Tavecchio et al. (1998) Tavecchio, F., Maraschi, L., & Ghisellini, G. 1998, ApJ, 509, 608
- Vaughan et al. (2003) Vaughan, S., Edelson, R., Warwick, R. S., & Uttley, P. 2003, MNRAS, 345, 1271
- Woo et al. (2005) Woo, J.-H., Urry, C. M., van der Marel, R. P., Lira, P., & Maza, J. 2005, ApJ, 631, 762
- Zhang et al. (1999) Zhang, Y. H., Celotti, A., Treves, A., et al. 1999, ApJ, 527, 719