Multiwavelength Observations of the Previously Unidentified Blazar RX J0648.7+1516
We report on the VERITAS discovery of very-high-energy (VHE) gamma-ray emission above 200 GeV from the high-frequency-peaked BL Lac object RX J0648.7+1516 (GB J0648+1516), associated with 1FGL J0648.8+1516. The photon spectrum above 200 GeV is fit by a power law with a photon index of and a flux normalization of TeVcms with 300 GeV. No VHE variability is detected during VERITAS observations of RX J0648.7+1516 between 2010 March 4 and April 15. Following the VHE discovery, the optical identification and spectroscopic redshift were obtained using the Shane 3–m Telescope at the Lick Observatory, showing the unidentified object to be a BL Lac type with a redshift of z . Broadband multiwavelength observations contemporaneous with the VERITAS exposure period can be used to sub-classify the blazar as a high-frequency-peaked BL Lac (HBL) object, including data from the MDM observatory, Swift-UVOT and XRT, and continuous monitoring at photon energies above 1 GeV from the Fermi Large Area Telescope (LAT). We find that in the absence of undetected, high-energy rapid variability, the one-zone synchrotron self-Compton model (SSC) overproduces the high-energy gamma-ray emission measured by the Fermi-LAT over 2.3 years. The SED can be parameterized satisfactorily with an external-Compton or lepto-hadronic model, which have two and six additional free parameters, respectively, compared to the one-zone SSC model.
1FGL J0648.8+1516 was detected by Fermi-LAT in the first 11 months of operation at greater than 10 standard deviations, (Abdo et al., 2010a). This source was flagged as a very-high-energy (VHE; E100 GeV) emitting candidate by the Fermi-LAT collaboration by searching for 30 GeV photons. This information triggered the VERITAS observations reported here. 1FGL J0648.8+1516 is found to be associated with RX J0648.7+1516, which was first discovered by ROSAT (Brinkmann et al., 1997). A radio counterpart was identified in the NRAO Green Bank survey (Becker et al., 1991). Two subsequent attempts to identify an optical counterpart were unsuccessful (Motch et al., 1998; Haakonsen et al., 2009).
At 6 off the Galactic plane and without optical spectroscopy, the nature of this object remained unknown until optical spectroscopy was obtained in response to the VERITAS detection. These observations allow the active galactic nucleus (AGN) to be classified as a BL Lac, a type of AGN that has a jet co-aligned closely with the Earth’s line of sight and displays weak emission lines. These AGN are characterized by non-thermal, double-peaked broadband spectral energy distributions (SED). Based on the radio and X-ray flux, the BL Lac can further be classified as a high-frequency-peaked BL Lac (HBL) (Padovani & Giommi, 1995), or if classified by the location of its low-energy peak, a high-synchrotron-peaked BL Lac (HSP) (Abdo et al., 2010b).
2 Observations and Analysis
VERITAS comprises four imaging atmospheric Cherenkov telescopes and is sensitive to gamma-rays between 100 GeV and 30 TeV (Weekes et al., 2002; Holder et al., 2006). The VERITAS observations of RX J0648.7+1516 were completed between 2010 March 4 and April 15 (MJD 55259-55301), resulting in 19.3 hours of quality-selected live time. These observations were taken at 0.5 offset in each of four directions to enable simultaneous background estimation using the reflected-region method (Fomin et al., 1994).
The VERITAS events are parameterized by the principal moments of the elliptical shower images, allowing cosmic-ray background rejection through a set of selection criteria (cuts) which have been optimized a priori on a simulated, soft-spectrum (photon index 4.0) source with a VHE flux 6.6% of that observed from the Crab Nebula. The cuts discard images with fewer than 50 photoelectrons. Events with at least two telescope images remaining are then cosmic-ray discriminated based on the mean-scaled-width (MSW) and the mean-scaled-length (MSL) parameters. Events with MSW 1.1, MSL 1.4, a height of maximum Cherenkov emission 8 km and an angular distance to the reconstructed source position in the camera () of less than 0.14 degrees are kept as gamma-ray candidate events. The results are reproduced in two independent analysis packages (Cogan, 2008; Daniel, 2008). After background rejection, 2711 events remain in the source region, with 16722 events remaining in the background regions (larger by a factor of 6.89). The 283 excess events result in a significance of 5.2, calculated using Equation 17 from Li & Ma (1983).
A differential power law GeV is fit to the VERITAS data from 200 to 650 GeV, shown in the top panel of Figure 1. The fit ( with 3 degrees of freedom (DOF), probability of 0.83) results in a flux normalization of photons cm s TeV and an index of , corresponding to 3.3% of the Crab Nebula flux above 200 GeV.
The angular distribution of the excess events is consistent with a point source now designated VER J0648+152, located at 102.19 0.11 RA and 15.27 0.12 Dec (J2000). The systematic pointing uncertainty of VERITAS is less than 25 (7 degrees). This position is consistent with the radio position of RX J0648.7+1516 (Becker et al., 1991). A nightly-binned VHE light curve is fit with a constant and shows a null hypothesis probability of 0.39, showing no significant variability during the observation.
The Fermi-LAT is a pair-conversion telescope sensitive to photons between 20 MeV and several hundred GeV (Atwood et al., 2009; Abdo et al., 2009). The data used in this paper encompass the time interval 2008 Aug 5 through 2010 Nov 17 (MJD 54683-55517), and were analyzed with the LAT ScienceTools software package version v9r15p6, which is available from the Fermi Science Support Center (FSSC). Only events from the “diffuse” class with energy above 1 GeV within a 5 radius of RX J0648.7+1516 and with a zenith angle were used. The background was parameterized with the files gll_iem_v02.fit and isotropic_iem_v02.txt 111The files are available at http://fermi.gsfc.nasa.gov/ssc/data/access/lat/BackgroundModels.html. The normalizations of the components were allowed to vary freely during the spectral point fitting, which was performed with the unbinned likelihood method and using the instrument response function P6_V3_DIFFUSE.
The spectral fits using energies above 1 GeV are less sensitive to possible contamination from unaccounted (transient) neighboring sources, and hence have smaller systematic errors, at the expense of slightly reducing the number of source photons. Additionally, there is no significant signal from RX J0648.7+1516 below 1 GeV. The analysis of 2.3 years between 2008 Aug 5 and 2010 Nov 17 (MJD 54683–55517) of Fermi-LAT events with energy between 0.3–1 GeV (fixing the spectral index to 1.89) yields a test statistic (TS) of 9, corresponding to 222See Mattox et al. (1996) for TS definition.. In addition to the background, the emission model includes two nearby sources from the 1FGL catalog: the pulsars PSR J0659+1414 and PSR J0633+1746. The spectra from the pulsars are parameterized with power-law functions with exponential cutoffs, and the values are fixed to the values found from 18 months of data. The spectral fluxes are determined using an unbinned maximum likelihood method. The flux systematic uncertainty is estimated as at MeV and at GeV and above.333See http://fermi.gsfc.nasa.gov/ssc/data/analysis/LAT_caveats.html
The results from the Fermi-LAT spectral analysis are shown in the bottom panel of Figure 1. There is no variability detected in four time bins evenly spread over the 2.3 years of data. The dataset corresponding in time to the VERITAS observations between between 2010 March 4 and April 15 (i.e. MJD 5525955301) does not show any significant signal and thus we report 2 upper limits that were computed using the Bayesian method (Helene, 1983), where the likelihood is integrated from zero up to the flux that encompasses 95% of the posterior probability. When using the data accumulated over the expanded full 2.3 years of data, we find that 1FGL J0648.8+1516 is significantly detected above 1 GeV with a TS of 307. The spectrum is fit using a single power-law function with photon flux photons cms and hard differential photon spectral index . The analysis is also performed on five energy ranges equally spaced on a log scale with the photon index fixed to 1.89 and only fitting the normalization. The source is detected significantly (TS25) in each energy bin except for the highest energy (100-300 GeV), for which a 95% confidence level upper limit is calculated.
The Swift-XRT (Gehrels et al., 2004; Burrows et al., 2005) data are analyzed with HEASOFT 6.9 and XSPEC version 12.6.0. Observations were taken in photon counting mode with an average count rate of counts per second and did not suffer from pile-up. Six target-of-opportunity observations summing to 10.5 ks were collected on six different days between 2010 March 18 and April 18 (MJD 55273 and 55304), inclusive. These observations were combined with a response file created from summing each observation’s exposure file using ximage. The photons are grouped by energy to require a minimum of 30 counts per bin, and fit with an absorbed power law between 0.3 and 10 keV, allowing the neutral hydrogen (HI) column density to vary. A HI column density of cm is found, only slightly higher than the cm quoted in Kalberla et al. (2005). The combined X-ray energy spectrum is extracted with a fit ( for 88 DOF, null hypothesis probability of 3.2) with a photon index of and an integral flux between 0.3 and 10 keV of ergs cm s. This corresponds to a 0.3 to 10 keV rest frame luminosity of ergs s. The deabsorbed spectrum is used to constrain modeling.
The Swift-XRT observations were supplemented with UVOT exposures taken in the U, UVM2, and UVW2 bands (centered at Hz, Hz, and Hz, respectively; Poole et al. (2008)). The UVOT photometry is performed using the HEASOFT program uvotsource. The circular source region has a radius and the background regions consist of several circles with radii between of nearby empty sky. The results are reddening corrected using R(V)=3.32 and E(B-V)=0.14 (Schlegel et al., 1998). The Galactic extinction coefficients were applied according to Fitzpatrick (1999), with the largest source of error resulting from deredenning. A summary of the UVOT analysis results is given in Table 1.
2.5 Optical MDM
The region around RX J0648.7+1516 was observed in the optical B, V, and R bands with the 1.3-m McGraw-Hill Telescope of the MDM Observatory on four nights during 2010 April 1–5 (MJD 55287-55291). Exposure times ranged from 90 sec (R-band) to 120 sec (B-band). Each night, five sequences of exposures in B, V, and R were taken. The raw data were bias subtracted and flat-field corrected using standard routines in IRAF444http://www.noao.edu/credit.html. Aperture photometry is performed using the IRAF package DAOPHOT on the object as well as five comparison stars in the same field of view. Calibrated magnitudes of the comparison stars are taken from the NOMAD catalog555http://www.nofs.navy.mil/nomad.html, and the magnitudes of the object are determined using comparative photometry methods. For the construction of the SED points, the magnitudes are extinction corrected based on the Schlegel et al. (1998) dust map with values taken from NASA Extragalactic Database (NED)666http://nedwww.ipac.caltech.edu/ : , , and . These data (summarized in Table 1) are used to constrain the modeling shown in this work, although the same conclusions result with the UVOT points as model constraint.
3 Spectroscopic Redshift Measurements
Two spectra were obtained during the nights of UT 2010 March 18 and 2010 November 6 (MJD 55245 and 55506, respectively) with the KAST double spectrograph on the Shane 3-m Telescope at UCO/Lick Observatory. During the first night, the instrument was configured with a 600/5000 grating and long slit, covering Å. A single 1800 second exposure was acquired. During the night of November 6, another 1800 second exposure was acquired with a 600/4310 grism, D55 dichroic, a grating and long slit, covering the interval Å. The data were reduced with the LowRedux pipeline777http://www.ucolick.org/xavier/LowRedux/index.html and flux calibrated using a spectro-photometric star. The flux calibration is uncertain due to non-photometric conditions. Inspection of the March spectrum reveals Ca H+K absorption lines at redshift . This redshift is confirmed in the second spectrum at higher signal-to-noise (S/N) (S/N in the blue and S/N in the red) where Ca H+K, G band, Mg I and Na I absorption lines with equivalent width Å are detected (see Figure 2 and Table 2 for details). No Ca H+K break is observed. These spectral features provide evidence for an early-type nature of the blazar host galaxy and allow for BL Lac classification, following Marcha et al. (1996) and Healey et al. (2007).
4 Broadband SED Modeling
The contemporaneous multiwavelength data are matched with archival radio data from NED and are shown in Figure 3. Since the radio data are not contemporaneous they are shown only for reference. The synchrotron peak appears at a frequency greater than Hz, representing the first subclassification of RX J0648.7+1516, specifically as an HBL. These data are used to test steady-state leptonic and lepto-hadronic jet models for the broadband blazar emission. The absorption of VHE gamma rays by the extragalactic background light (EBL) is accounted for through application of the Gilmore et al. (2009) EBL model; the model of Finke et al. (2010) provides comparable results.
Leptonic models for blazar emission attribute the higher-energy peak in the SED to the inverse-Compton scattering of lower-energy photons off a population of non-thermal, relativistic electrons. These same electrons are responsible for the lower-energy synchrotron emission making up the first peak. The target photon field involved in the Compton upscattering can either be the synchrotron photons themselves, as in synchrotron self-Compton (SSC) models, or a photon field external to the jet in the case of external Compton (EC) models.
We use the equilibrium SSC model of Böttcher & Chiang (2002), as described in Acciari et al. (2009). In this model, the emission originates from a spherical blob of relativistic electrons with radius . This blob is moving down the jet with a Lorentz factor , corresponding to a jet speed of . The jet is oriented such that the angle with respect to the line of sight is , which results in a Doppler boosting with Doppler factor . In order to minimize the number of free parameters, the modeling is completed with , for which .
Within the model, electrons are injected with a power-law distribution at a rate between the low- and high-energy cut-offs, . The electron spectral index of required for the models applied in this work might be the result of acceleration in an oblique shock. While standard shock acceleration in relativistic, parallel shocks is known to produce a canonical spectral index of 2.2, oblique magnetic-field configurations reduce the acceleration efficiency and lead to much steeper spectral indices (Meli & Quenby, 2003; Sironi & Spitkovsky, 2011). The radiation mechanisms considered lead to equilibrium between the particle injection, radiative cooling and particle escape. The particle escape is characterized with an efficiency factor , such that the escape timescale , with for this work. This results in a particle distribution streaming along the jet with a power . Synchrotron emission results from the presence of a tangled magnetic field , with a Poynting flux luminosity of . The parameters and allow the calculation of the equipartition parameter .
The top panel in Figure 3 shows the SSC model for RX J0648.7+1516, with parameters summarized in Table 3. The model is marginally in agreement with the data only through use of parameters well below equipartition. The Fermi-LAT contemporaneous 95% confidence level upper limits in the energy ranges 1-3 GeV and 3-10 GeV are just above and below the one-zone SSC model predictions. Additionally, these SSC model predictions are above the 2.3 year Fermi-LAT spectrum by more than a factor of 2, although this spectrum is not contemporaneous with the other data. Variation of the model parameters within physically reasonable values does not provide better agreement between model and data. Generally, HBLs are well characterized by one-zone SSC models and hence these observations might suggest the existence of one or more additional emission mechanisms that contribute to the higher-energy peak.
An external-Compton model is also used to describe the data. The EC model is a leptonic one-zone jet model with two additional parameters beyond the SSC parameters, the thermal blackbody temperature and radiation energy density of the external photon field, which is assumed to be isotropic and stationary in the blazar rest frame. The EC model provides a better representation of the SED, as can be seen in the middle panel of Figure 3, with the parameters listed in Table 3.
A lepto-hadronic model is also applied to the data. Within this model, ultrarelativistic protons are the main source of the high-energy emission through proton synchrotron radiation and pion production. The resulting spectra of the pion decay products are evaluated with the templates of Kelner & Aharonian (2008). Additionally, a semi-analytical description is used to account for electromagnetic cascades initiated by the internal absorption of multi-TeV photons by both the decay photons and the synchrotron emission of ultrarelativistic leptons, as explained in Böttcher (2010). Similar to the particle populations in the leptonic models described above, this lepto-hadronic model assumes a power-law distribution of relativistic protons, between a low- and high-energy cut-off, . This population of relativistic protons is propagating along the blazar jet and has a total kinetic luminosity of . The lepto-hadronic modeling results are above equipartition and are shown in the bottom panel of Figure 3 with parameters (including energy partition fractions and ) summarized in Table 3.
In conclusion, multiwavelength followup of the VERITAS detection of 1FGL J0648.7+1516 has solidified its association with RX J0648.7+1516, which is identified as a BL Lac object of the HBL subclass. Other contemporaneous SEDs of VHE-detected HBLs can be well described by one-zone SSC models close to equipartition, while for RX J0648.7+1516 this model provides a poor representation with parameters below equipartition. The addition of an external photon field for Compton up-scattering in the leptonic paradigm provides a better representation of the gamma-ray (Fermi and VERITAS) data. Alternatively, a lepto-hadronic model is successful in characterizing the higher-energy peak of the SED with synchrotron emission from protons. Both of these latter models require super-equipartition conditions.
- Abdo et al. (2009) Abdo, A. et al. 2009, Astroparticle Physics, 32, 193
- Abdo et al. (2010a) Abdo, A. et al. 2010, ApJS, 188, 405
- Abdo et al. (2010b) Abdo, A. et al. 2010, ApJ, 716, 30
- Atwood et al. (2009) Atwood, W. B., et al. 2009, ApJ, 697, 1071
- Acciari et al. (2009) Acciari, V. A. et al. 2009, ApJ, 707, 612
- Becker et al. (1991) Becker, R. et al. 1991, ApJS, 75, 1
- Brinkmann et al. (1997) Brinkmann, W. et al. 1997 A&A, 323, 739
- Böttcher & Chiang (2002) Böttcher, M., & Chiang, J., 2002, ApJ, 581, 127
- Böttcher (2010) Böttcher, M., 2010, in proc. of “Fermi Meets Jansky”, eds. T. Savolainen, E. Ros, R. W. Porcas, & J. A. Zensus; p. 41
- Burrows et al. (2005) Burrows, D.N., Hill, J.E., Nousek, J.A., et al. 2005, Space Sci. Rev., 120, 165
- Cogan (2008) Cogan, P. 2008, Proc. 30th Int. Cosmic Ray Conf., Vol 3, The VERITAS Gamma-ray Analysis Suite, ed. R. Caballero, J. C. D’Olivo, G.Medina-Tanco, L. Nellen, F. A. Sánchez & J. F. Valdeé-Galicia (Mexico City, Mexico: Universidad Nacional Autónoma de México), 1385
- Daniel (2008) Daniel, M. 2008, Proc. 30th Int. Cosmic Ray Conf., Vol 3, The VERITAS Standard Data Analysis, ed. R. Caballero, J. C. D’Olivo, G.Medina-Tanco, L. Nellen, F. A. Sánchez & J. F. Valdeé-Galicia (Mexico City, Mexico: Universidad Nacional Autónoma de México), 1325
- Gilmore et al. (2009) Gilmore, R. et al. 2009, MNRAS, 399, 1694
- Helene (1983) Helene, O. 1983, Nuclear Instruments & Methods in Physics Research, 212, 319
- Fitzpatrick (1999) Fitzpatrick, E. 1999, PASP, 111, 63
- Finke et al. (2010) Finke, J. et al. 2010, ApJ, 712, 238
- Fomin et al. (1994) Fomin, V. P. et al. 1994, Astropart. Phys., 2, 137
- Gehrels et al. (2004) Gehrels, N. et al. 2004, ApJ, 611, 1005
- Healey et al. (2007) Healey, S. et al. 2007, ApJS, 171, 61
- Haakonsen et al. (2009) Haakonsen, C. B. et al. 2009, ApJS, 184, 138
- Holder et al. (2006) Holder, J. et al. 2006, Astropart. Phys., 25, 391
- Kalberla et al. (2005) Kalberla, P. et al. 2005, A&A, 440, 775
- Kelner & Aharonian (2008) Kelner, S. R. & Aharonian, F. A., 2008, Phys. Rev. D., 78, 3, 034013
- Li & Ma (1983) Li, T. & Ma, Y. 1983, ApJ, 272, 317
- Marcha et al. (1996) Marcha, M. et al. 1996, MNRAS, 281, 425
- Mattox et al. (1996) Mattox, J. et al. 1996, ApJ, 461, 396
- Meli & Quenby (2003) Meli, A., & Quenby, J., 2003, ApJ, 19, 649
- Motch et al. (1998) Motch, C. et al. 1998, A&AS, 132, 341
- Padovani & Giommi (1995) Padovani, P. & Giommi, P., 1995, ApJ, 444, 567
- Poole et al. (2008) Poole et al. 2008, MNRAS, 383, 627
- Schlegel et al. (1998) Shlegel, D. et al. 1998, ApJ, 500, 525
- Sironi & Spitkovsky (2011) Sironi, L., & Spitkovsky, A. 2011, ApJ, 726, 75
- Weekes et al. (2002) Weekes, T. C. et al. 2002, Astropart. Phys., 17, 221
|(MJD)||(Jy Hz)||(Jy Hz)|
|Ions||Rest Wavelength||CentroidaaBased on Gaussian fit||FWHM||RedshiftbbMeasured from line centroid||Observed E. W.ccError is only statistical||Notes|
|Ca II (K)||3934.79||4639.07||20.7||0.1789||2.60 0.21|
|Ca II (H)||3969.61||4678.26||16.4||0.1785||2.470.19|
Note. –  Blanded with Mg I 5168.74 Mg I 5185.04  Blanded with Na I 5891.61 and Na I 5897.57