DISCOVERY OF A NEW TeV GAMMA-RAY SOURCE: VER J0521+211

TeV astronomy is concerned with the detection of astrophysical gamma rays with energies greater than ∼0.1 TeV. The most sensitive detectors in this energy region are the current generation of ground-based imaging atmospheric Cherenkov telescopes: VERITAS (Holder et al. 2008), H.E.S.S. (Hinton 2004), and MAGIC (Albert et al. 2008a), which can detect sources with a flux less than 0.01 times the steady flux of the Crab Nebula (hereafter "Crab"; Hillas et al. 1998) with an exposure of a few tens of hours. The angular resolution of current instruments is ∼01, and the field of view is typically limited to a diameter of <5°. Observations consist of surveying regions of the sky using many overlapping, noncontemporaneous exposures (Aharonian et al. 2006a; Weinstein 2009), or targeting locations of interest based on information provided by observations at other wavelengths.

The Large Area Telescope (LAT) on board the Fermi Gamma-ray Space Telescope is the first instrument to provide a view of the entire gamma-ray sky at energies that overlap with those accessible to ground-based telescopes. Figure 1 shows the arrival directions of all diffuse-class photon events⁴⁰ recorded during the first year of Fermi-LAT observations with energies above 30 GeV. Clearly, the highest-energy Fermi-LAT observations provide an excellent guide to the TeV sky, and can be used to select additional TeV candidates for targeted observations. Numerous authors have addressed this, producing catalogs based on >100 GeV Fermi-LAT events (Neronov et al. 2010) and lists of TeV source candidates (e.g., Abdo et al. 2009). In this work, the presence of a cluster of high-energy photons (Figure 1) spatially associated with an X-ray source was used to motivate observations with the VERITAS array of a previously unobserved location, leading to the detection of a new bright TeV source: VER J0521+211. The only other cluster flagged by this method (1ES 0502+675) was already being observed by VERITAS at the time and was subsequently detected after 13 hr of exposure (Ong 2009b).

Zoom In Zoom Out Reset image size

This paper presents the discovery of gamma-ray emission from VER J0521+211, and a detailed study of its observational properties from radio frequencies to gamma-ray energies. The paper is structured as follows: Sections 2–5 present the observations and analysis results by VERITAS (TeV), Fermi-LAT (GeV), Swift (X-ray), and MDM and Steward Observatory (optical), respectively. The identification of VER J0521+211 as a new TeV blazar is discussed in Section 6, and its main observational properties are presented in Section 7. An upper limit on the redshift of VER J0521+211 is calculated in Section 8, and the overall spectral energy distribution (SED) is discussed in Section 9. Finally, Section 10 summarizes the main conclusions of the study.

The VERITAS observatory is described in detail in Holder et al. (2008) and Holder et al. (2006). The array consists of four 12 m diameter imaging atmospheric Cherenkov telescopes, with photomultiplier (PMT) cameras covering a field of view of 35. The array has a total effective area of ∼5 × 10⁴ m² between 0.2 and 10 TeV. Following the relocation of the original prototype telescope to a more favorable position in Summer 2009 (Perkins et al. 2009), VERITAS has sensitivity to detect a source with 0.01 Crab flux in under 30 hr of observations. The angular and energy resolution for reconstructed gamma-ray showers is energy dependent, reaching ∼01 and 15%, respectively, at 1 TeV.

Inspection of the >30 GeV Fermi-LAT map (Figure 1) led to the identification of a cluster of high-energy photons, which was used to trigger VERITAS observations centered around R.A. = 05^h21^m46^s, decl. = +21°12'515 (J2000), corresponding to the position of RGB J0521.8+2112, the only radio/X-ray source within 01 of the LAT excess.

Observations took place between MJD 55126 (2009 October 22) and MJD 55212 (2010 January 16) and consisted of 20 minute exposures in wobble observing mode (Fomin et al. 1994), taken at a mean zenith angle of 16°. After excluding data taken under poor weather conditions or with hardware problems, the data set comprises 14.5 hr of dead-time-corrected exposure.

VERITAS data analysis follows the procedure outlined in Acciari et al. (2008). Cherenkov light from air showers initiated by gamma rays and cosmic rays triggers the readout of PMT signals, which are then calibrated and used to reconstruct an image of the shower in the focal plane. Individual telescope images are parameterized by simple moment analysis (Hillas 1985), and geometrical reconstruction is used to calculate the arrival direction of the primary, θ, defined with respect to the position of the candidate source location on the sky. The image shape in each telescope is compared with the expected shapes for gamma-ray showers generated using Monte Carlo simulations and the differences, averaged over all telescopes, are used to derive the mean-reduced-scaled width and mean-reduced-scaled length parameters, as defined in Acciari et al. (2008). Gamma-ray-like events are selected by applying cuts on the various image parameters. The cuts used in this work are mean-reduced-scaled width/length between −1.2 and 0.5, θ < 01, and at least three telescope images with an integrated signal size per image equivalent to >94 photoelectrons. These cuts are designed to provide optimum sensitivity to a moderately strong point-like gamma-ray source (0.05 Crab) with a Crab-like differential power-law photon index of ∼2.5. Over 99.9% of the background cosmic ray events are removed by these cuts. The number of background events remaining in the signal region is estimated from 10 off-source regions within the same field of view using the reflected region technique (Aharonian et al. 2001; Berge et al. 2007).

After data selection, reduction, and signal extraction, an excess of 221 candidate gamma-ray events over a background of 119 is detected at the location of RGB J0521.8+2112, corresponding to a significance of 15.6 standard deviations according to Equation (17) in Li & Ma (1983, with N_on = 340, N_off = 1518, and α = 0.08). The angular distribution of the excess events is compatible to that of a point-like source given the instrumental point spread function of 6' for 68% containment radius. The position of the source is derived by fitting a symmetric two-dimensional Gaussian to the uncorrelated excess map, resulting in a best-fit centroid at R.A. = 05^h21^m45^s, decl. = +21°12'514, with a statistical uncertainty of 14'' and a systematic uncertainty of 25'', dominated by the telescopes' pointing accuracy. The new TeV source is cataloged as VER J0521+211, based on the name first reported in Ong (2009a).

Figure 2 shows the time-averaged energy spectrum of VER J0521+211, extending from a threshold energy of 0.2 to ∼1 TeV. It is well described by a power law dN/dE = N₀(E/0.4 TeV)^−Γ with normalization N₀ = (1.99  ±  0.18_stat  ±  0.80_syst) × 10⁻¹¹ cm⁻² s⁻¹ TeV⁻¹ and photon index Γ = 3.44  ±  0.20_stat  ±  0.30_syst. The time-averaged integral photon flux is F_>0.2 TeV = (1.93  ±  0.13_stat  ±  0.78_syst) × 10⁻¹¹ cm⁻² s⁻¹, corresponding to 0.092  ±  0.006 Crab (1 Crab = 2.1 × 10⁻¹⁰ cm⁻² s⁻¹; Hillas et al. 1998). Compatible results were obtained using two independent analysis packages.

Zoom In Zoom Out Reset image size

Flux variability was explored by producing a 1 day binned light curve during the VERITAS observations (Figure 3, top). After a first detection in 2009 October 22–24 (Ong 2009a) with a derived TeV flux of 0.09  ±  0.01 Crab, a later round of observations revealed a higher flux level, peaking at 0.33  ±  0.07 Crab on 2009 November 27 (MJD 55162, hereafter "TeV flare"). A χ² fit to the nightly flux points for constant emission gives a probability of 7 × 10⁻⁷, indicating flux variability at a confidence level of 5.8 standard deviations.

Zoom In Zoom Out Reset image size

Spectral variability was tested by deriving a "low-state" spectrum from the data taken during the first period of VERITAS observations (2009 October 22–30, MJD 55126 − 55134), and a "flare" spectrum from the night of 2009 November 27. A power-law fit to each spectrum yields Γ_low = 2.92  ±  0.34 and Γ_flare = 3.25  ±  0.72. The difference in the reconstructed photon index does not constitute significant evidence for spectral variability in the TeV band.

Fermi-LAT is a pair-conversion telescope sensitive to gamma rays in the range from 0.02 to more than 300 GeV with a field of view of ∼2.4 sr. The effective area of the LAT changes with energy and incidence angle, being ∼0.8 m² for on-axis photons with E > 10 GeV. Full details about the instrument and its performance are given in Atwood et al. (2009) and Ackermann et al. (2012).

A GeV source spatially associated with VER J0521+211 was listed in the LAT 11 month catalog (1FGL J0521.7+2114; Abdo et al. 2010a) and confirmed in the second source catalog (2FGL J0521.7+2113; Nolan et al. 2012) and in the catalog of sources detected above 10 GeV (1FHL J0521.7+2113; Ackermann et al. 2013). The region was analyzed in more detail by selecting source-class⁴¹ events with reconstructed energy 0.3 < E < 300 GeV collected during the period of VERITAS observations (MJD 55126 − 55212) in a 20° × 20° region of interest (RoI) centered at the location of VER J0521+211. Spectral parameters were extracted by fitting a model containing the Galactic and isotropic diffuse background⁴² and point sources from the 2FGL catalog. The analysis was done using binned likelihood as implemented in the Fermi-LAT Science Tools v09-26-00 with P7V6 response functions and the standard quality selection cuts described in Nolan et al. (2012). The energy spectrum (light curve) were derived following the procedure detailed in Nolan et al. (2012), dividing the data into bins of energy (time) and performing a likelihood analysis with the spectral shape parameters frozen for all sources in the model.

During the period of VERITAS observations, VER J0521+211 was detected by the LAT with a test statistic of 237 (∼15σ). The derived integral photon flux is F_{0.3–300 GeV} = (2.8  ±  0.4) × 10⁻⁸ cm⁻² s⁻¹, with a spectrum well described by a power law with photon index Γ = 1.72  ±  0.09. An identical analysis of an extended data set collected by Fermi (∼46 months) reveals that VER J0521+211 was in a harder spectral state during the VERITAS observations, compared to a time-averaged photon index of Γ = 1.97  ±  0.03, while the flux level was comparable over both periods.

Variability was tested by extracting the source flux in 29.5 day bins.⁴³ Applying a likelihood-based variability test (Nolan et al. 2012) to the 46 months of data gives a probability of a constant flux of 2 × 10⁻²⁰, indicating flux variability in the GeV band at a confidence level of 9.3 standard deviations. The 46 month light curve (not shown) exhibits flux changes by a factor of ∼4 between the lowest and highest emission levels.

An additional test for variability above 1 GeV was performed using a Bayesian Block method (Scargle 1998). For this test, source-class events with E > 1 GeV were extracted from a 1° radius RoI centered on the source coordinates. The data were divided into time blocks over which the event rate was compatible with a constant value. The optimal width of the blocks was determined by maximum likelihood analysis using the algorithm described in Jackson et al. (2005). The effective exposure associated to each event is taken into account, correcting for exposure variations caused by the motion of the spacecraft. A 1% false-positive threshold was used for detecting variability. The method identifies three periods of different constant flux with durations between 310 and 451 days, indicating variability above 1 GeV. The highest flux state, with boundaries MJD 54993–55419, includes the VERITAS observations. However, no significant evidence for shorter time variability is found in the Fermi-LAT data above 1 GeV during VERITAS observations, or in coincidence with the X-ray and TeV flare on MJD 55162.

After the discovery of TeV emission by VERITAS (Ong 2009a), Swift X-Ray Telescope (XRT; Gehrels et al. 2004) observations were triggered. Seven exposures were obtained between 2009 October 27 and November 27. The total observation time is 16.6 ks distributed in exposures of ∼2.5 ks.

Swift/XRT data were analyzed with HEAsoft 6.9 and XSPEC 12.6.0 using the most recent calibration files as described in Burrows et al. (2005). All data were taken in photon counting mode. Pile-up effects were accounted for by extracting the signal from an annular source region when rates exceeded $0.5 \mathrm{\ c {\rm oun}ts\,s^{-1}}$.

The Swift/XRT field of view of 236 × 236 completely covers the VERITAS error circle. A single X-ray source was detected at R.A. = $05^{\rm h}21^{\rm m}45\buildrel{\mathrm{s}}\over{.}98$, decl. = +21°12'529 (with 19 location uncertainty), spatially coincident with RGB J0521.8+2112, as shown in Figure 4. The measured, pile-up-corrected count rate is significantly variable on a daily timescale (Figure 3), with a flux increase up to a factor of ∼15 between low and high states. For spectral reconstruction, each observation was binned and fit with an absorbed power law, with neutral hydrogen column density taken from Kalberla et al. (2005). The three exposures from 2009 October 27–30 showed a lower source count rate (${\sim }0.1\mathrm{\ c{\rm oun}ts}\mathrm{\ s^{-1}}$) and were combined for the spectral analysis. The derived photon index continuously hardens from 2.47  ±  0.10 when the emission is lowest to 2.00  ±  0.04 on the night of the highest X-ray flux (MJD 55162), when RGB J0521.8+2112 reaches a peak flux of (3.14  ±  0.14) × 10⁻¹¹ erg cm⁻² s⁻¹ in the 2–10 keV band.

Zoom In Zoom Out Reset image size

Beginning with the radio-interferometric position from Beasley et al. (2002), the optical counterpart of RGB J0521.8+2112 was identified on the digitized sky survey, with coordinates R.A. = $05^{\rm h}21^{\rm m}45\buildrel{\mathrm{s}}\over{.}96$, decl. = +21°12'516, and magnitudes B2 = 17.32, R2 = 15.47, and I2 = 15.06 listed in the USNO B1.0 catalog (Monet et al. 2003). Optical spectroscopic observations were made on three occasions at the 2.4 m Hiltner telescope of the MDM Observatory, using three different CCD spectrographs. The optical spectra (Figure 5) are devoid of intrinsic emission or absorption lines and stellar continuum emission in the wavelength range 4000–7500 Å, classifying the source as a BL Lac-type blazar but not revealing its redshift. Features marked in Figure 5 are absorption lines from the Galactic interstellar medium at zero redshift. Recently published observations with the Low Resolution Imaging Spectrograph at the W. M. Keck Observatory (Shaw et al. 2013) show a weak emission feature identified as [N ii] λλ 6548, 6583, which would indicate a redshift of z = 0.108. Although this feature cannot be identified in the MDM spectra, the measurements are not in conflict given the lower level of continuum emission present in the spectrum by Shaw et al. (2013) and the higher sensitivity of Keck. The first spectrum of RGB J0521.8+2112 in Figure 5 was obtained on 2009 October 27, only five days after the first VERITAS observations. The calibrated flux in this spectrum corresponds to B ∼ 17.0, R ∼ 15.2, slightly brighter than the USNO B1.0 magnitudes. It appears that the optical flux faded in the subsequent months and years by ∼1 mag, although none of the spectrophotometry is precise because of the narrow spectrograph slit.

Zoom In Zoom Out Reset image size

Optical linear polarimetry was performed using the Steward Observatory 1.54 m Kuiper Telescope, located on Mt. Bigelow, AZ. RGB J0521.8+2112 was observed on 2011 November 24, 28, and December 1 with the SPOL CCD imaging/spectropolarimeter (Schmidt et al. 1992). High signal-to-noise broadband measurements were derived by binning the polarization spectra in the range of 5000–7000 Å. All measurements are summarized in Table 1 and have been corrected for statistical bias (Wardle & Kronberg 1974). Given the low Galactic latitude of the source, two field stars were also observed, suggesting a significant interstellar polarization (ISP) along the sight line to RGB J0521.8+2112 (see Table 1). The binned spectropolarimetry of RGB J0521.8+2112, corrected for an estimate of the ISP, yields a variable degree of polarization (P) between P = 3.74%  ±  0.15% with position angle θ = 268  ±  11 (November 24) and P = 7.26%  ±  0.16% at θ = 282  ±  06 (December 1). Because of the variability in P observed for RGB J0521.8+2112, the object must be intrinsically polarized regardless of the actual level of ISP in this line of sight.

The observations that led to the discovery of VER J0521+211 were triggered by the identification of a cluster of photons with E > 30 GeV in the first year of Fermi-LAT data (Figure 1). The most likely counterpart of the LAT excess was RGB J0521.8+2112, which had no classification, although sources in the ROSAT/Green Bank catalog are generally associated with active galactic nuclei (AGNs; Laurent-Muehleisen et al. 1997). RGB J0521.8+2112 is the only radio source spatially compatible with the LAT excess in a high-completeness very long baseline interferometry sample down to 0.15 Jy, indicative of a radio-loud blazar (see Kovalev 2009). No other spatially associated X-ray, UV, or compact radio source was found in archival catalogs.

Immediately after the VERITAS detection, X-ray and optical observations were triggered to confirm the extragalactic nature of VER J0521+211. Swift/XRT establishes RGB J0521.8+2112 as the only bright X-ray source inside the 39'' VERITAS error circle (Figure 4). The identification of VER J0521+211 with RGB J0521.8+2112 is further supported by the detection of a TeV flare simultaneously seen in X-rays (Figure 3). However, the statistical significance of the overall X-ray/TeV flux correlation cannot be assessed due to the limited number of flux measurements in both bands.

The flux variability observed in the TeV band further suggests that VER J0521+211 could indeed be an AGN. To date, all known variable TeV sources are AGNs, with the exception of the four detected gamma-ray binaries (LSI +61 303, HESS J0632+057, PSR B1259-63, and LS 5039) and the Crab pulsar, which is variable at a much shorter timescale (33 ms).

The optical counterpart of VER J0521+211 was subsequently identified through observations at MDM. Optical spectroscopy revealed a continuum-dominated spectrum (Figure 5), unambiguously identifying VER J0521+211 as a BL Lac-type blazar.

The main observational properties of VER J0521+211 are summarized in Table 2. Taking the 5 GHz flux from White & Becker (1992) and the B-band apparent magnitude from MDM observations, the radio-to-optical luminosity ratio for VER J0521+211 is $R_{\mathrm{rB}}=L_{5\,\mathrm{GHz}}/L_{{{B_{\rm mag}}}}= 69$, implying a radio-loud AGN (R_rB > 10; Kellermann et al. 1989) with prominent radio jet or lobe emission (Urry & Padovani 1995). The spectrum of the unresolved core emission of VER J0521+211 in the radio band is well described by a power law, with $F_\nu \propto \nu ^{-\alpha _{\mathrm{r}}}$ and spectral index α_r = 0.16  ±  0.01 (Figure 6), compatible with α_r ⩽ 0.5, characteristic of jet-dominated flat-spectrum radio sources.

Zoom In Zoom Out Reset image size

The radio jet of RGB J0521.8+2112 has been imaged on milliarcsecond scales in five epochs between 2009 October and 2012 April with the Very Long Baseline Array (VLBA) at 15 GHz as part of the MOJAVE program (Lister et al. 2009a). The radio morphology consists of a bright radio core and an apparent one-sided jet that extends for ∼20 mas to the west while curving slightly to the northwest (Figure 7). The radio core is very compact, with a brightness temperature above 10¹¹ K at all epochs according to Gaussian model fits. The model fits to individual features in the jet do not reveal significant proper motions over the 2.5 yr observation interval. During this time, however, there were significant changes in the polarization of the core and the inner 1.5 mas of the jet. The core remained weakly linearly polarized (<1%), but increased steadily in polarized flux density from 0.7 to 1.6 mJy. The downstream jet polarization was typically much higher (up to 25%) with electric polarization vectors perpendicular to the jet ridgeline, consistent with optically thin synchrotron emission associated with a relatively well-ordered, longitudinal jet magnetic field. A bright feature ∼1 mas downstream of the core briefly flared in total intensity and polarization sometime between 2010 March and 2010 October, but returned to its earlier level by 2011 July. These short-timescale changes, as well as the one-sided jet morphology, high core compactness and jet polarization are very similar to those seen in highly Doppler-boosted blazar jets in the MOJAVE sample (Lister & Homan 2005; Lister et al. 2009b).

Zoom In Zoom Out Reset image size

Infrared observations of VER J0521+211 in the WISE Source Catalog⁴⁴ show magnitudes of 10.6, 9.8, 7.6, and 5.6 in the 3.4, 4.6, 12, and 22 μm bands, respectively. Its infrared colors are similar to those of known gamma-ray blazars (Massaro et al. 2012b), providing further support for its identification as a TeV blazar (Massaro et al. 2012a).

The optical polarimetry reveals a highly polarized object that shows significant variations in both the level and position angle of polarization on timescales shorter than a week. This is strong evidence that a good fraction of the optical continuum of RGB J0521.8+2112 is produced by synchrotron radiation from regions with well-ordered magnetic fields. During the epoch of the Steward Observatory measurements, the position angle of the polarization (20°–30°) was intermediate between being parallel and perpendicular to the axis of the radio jet.

BL Lac-type blazars are usually classified as low, intermediate, or high-frequency peaked BL Lacs according to the estimated peak frequency of their synchrotron component (LBL, log (ν_sy/Hz) < 14; IBL, 14 − 15; HBL, >15; e.g., Abdo et al. 2010b). Figure 8 shows the effective spectral indices between radio, optical, and X-ray bands for all known TeV blazars. Simultaneous flux measurements during "low-state" (2009 October 22–30) suggest that VER J0521+211 has synchrotron properties similar to those of the known IBLs. However, during the X-ray and TeV flare, the brightening and hardening of the X-ray emission causes the synchrotron component of VER J0521+211 to show HBL-like properties (Figure 8). Spectral changes in the synchrotron component are not uncommon and have been reported in other TeV blazars (Massaro et al. 2008; Abdo et al. 2011c).

Zoom In Zoom Out Reset image size

VER J0521+211 also shows similar luminosity in the synchrotron and high-energy peak of the SED (Figure 6), implying a Compton dominance (R_C = L_HE/L_sy) in the range 0.8 ≲ R_C ≲ 1.8. Such values are similar to those of high-power blazars (Meyer et al. 2012), which show synchrotron peak frequencies lower than "classical" HBLs (log (ν_sy/Hz) ≲ 14.5; Meyer et al. 2011).

With a low-state TeV flux of ∼0.1 Crab measured over three months of observations, and a flare of ∼0.3 Crab, VER J0521+211 is one of the brightest TeV blazars newly discovered by the current generation of Cherenkov telescopes. Other blazars, such as 3C 279 (Albert et al. 2008b), W Comae (Acciari et al. 2009), or PKS 1222+216 (Aleksić et al. 2011), have been detected at flux levels higher than 0.1 Crab during short periods of time (⩽1 week) and then quickly fall below instrument sensitivity. Despite being a relatively bright TeV source, VER J0521+211 was never identified as a TeV-blazar candidate before the VERITAS detection. Given its low Galactic latitude of −87, VER J0521+211 was poorly characterized in the optical and X-ray bands and was not included in blazar catalogs, which are usually incomplete at low Galactic latitudes or avoid the Galactic plane altogether (e.g., Healey et al. 2008; Massaro et al. 2009), as diffuse radio emission, confusion with local radio sources, and heavy optical extinction make candidate blazars difficult to identify. Therefore, TeV-candidate catalogs (e.g., Costamante & Ghisellini 2002; Donato et al. 2001) lacked sufficient multiwavelength information on VER J0521+211 and a firm identification as a blazar, required to predict its TeV flux.

An upper limit on the redshift of VER J0521+211 can be derived from its TeV spectrum. Following the approach in Mazin & Goebel (2007; see also Aharonian et al. 2006b; Mazin & Raue 2007), the intrinsic TeV spectrum of VER J0521+211 is reconstructed from the VERITAS measured spectrum (Figure 2) by correcting for extragalactic background light absorption, assuming the density model of Domínguez et al. (2011), under varying assumptions of z. The intrinsic spectrum is then fit with a power law of the form $dN/dE \propto E^{-\Gamma ^{\star }}$. Larger values of z result in a harder reconstructed photon index Γ^⋆. Classical leptonic emission models predict Γ^⋆ > 1.5 (see discussion in Aharonian et al. 2006b). Under this assumption, a redshift upper limit of z < 0.34 is derived for VER J0521+211 at 95% confidence level. An even more conservative redshift upper limit can be obtained by allowing an intrinsic spectral index as hard as Γ^⋆ ∼ 0.7, as suggested in Katarzyński et al. (2006; see also Aharonian et al. 2008; Sitarek & Bednarek 2008). Under this less restrictive assumption of Γ^⋆ > 0.7, the redshift of VER J0521+211 is constrained to z < 0.44.

The redshift upper limits derived from the TeV spectrum of VER J0521+211 are in agreement with the recent measurement of z = 0.108 based on a single optical spectral feature (Shaw et al. 2013).

After a successful identification of VER J0521+211 as a new TeV blazar, the available multiwavelength data were combined to construct an SED, shown in Figure 6. Most commonly accepted models attribute the low-energy emission component to synchrotron radiation by relativistic electrons in the jet magnetic field, and the high-energy to inverse-Compton scattering of ambient photons off the same electron population (see, e.g., Maraschi et al. 1992).

The multiwavelength SED can be described with a one-zone synchrotron self-Compton (SSC) emission model as described in Böttcher et al. (2013). The model parameters are adjusted to describe the quasi-simultaneous spectral points obtained during the "low" emission state (2009 October 22–30). Models were tested assuming z = 0.05, 0.1, 0.15, 0.2, 0.25, although only results for z = 0.1 are discussed, being the assumption that most closely matches the tentative redshift of z = 0.108.

The model parameters describing the low-state SED of VER J0521+211 are listed in Table 3, and show a slightly low magnetic field and large emitting region compared to other TeV blazars (Aliu et al. 2013). One-zone SSC models can only describe Compton-dominated systems (R_C ≳ 1) like VER J0521+211 with very low magnetic fields, resulting in a strongly particle-dominated jet (L_B/L_e < 0.01; L_B and L_e being the magnetic and particle power in the jet, respectively). External Compton models (EC; Dermer & Schlickeiser 1993) add a second population of low-energy photons where relativistic electrons inverse-Compton scatter, increasing the level of high-energy emission, with jet energetics close to equipartition (L_B/L_e ∼ 1). However, given the scarcity of simultaneous observations, particularly in the synchrotron component, an EC model applied to the SED of VER J0521+211 would be severely underconstrained.

VERITAS detected a new TeV source: VER J0521+211, spatially associated with the radio and X-ray source RGB J0521.8+2112. Follow-up observations in the optical and X-ray bands unambiguously identify VER J0521+211 as a new blazar of the BL Lac type, displaying all the defining properties of blazars in radio, infrared, and optical wavelengths. The detected TeV emission is variable on daily timescales with an integral flux of ∼0.09–0.33 Crab measured between 0.2 and ∼1 TeV, and a time-averaged spectrum compatible with a power law with photon index Γ = 3.44  ±  0.20_stat  ±  0.30_syst. During the observing campaign that covered 2009 October to 2010 January, VER J0521+211 transitioned to a high state on 2009 November 27, when the nightly flux increased by a factor of ∼3 in the TeV range and by a factor of ∼15 in X-rays, compared to the observed baseline values. X-ray observations show a trend of spectral hardening with increasing flux, while no significant spectral variability was found at TeV energies.

Observations from radio to X-ray frequencies show a multitude of evidence for a synchrotron origin of the emission from VER J0521+211 below a few keV, as expected for a blazar: a flat spectrum in the radio band with a turnover toward infrared frequencies, polarized emission in the optical band and in the 15 GHz radio images from VLBA, and a nonthermal power-law spectrum in X-rays. The radio images also show electric polarization vectors perpendicularly aligned to the jet ridgeline, suggesting a relatively well-ordered magnetic field in the direction of the jet axis. The multiband spectral shape of the synchrotron component of VER J0521+211 is similar to that of known TeV IBLs. However, during the TeV and X-ray flare its synchrotron properties are closer to those of HBLs. Optical spectrometry with MDM could not be used to derive a redshift, although recent measurements suggest z = 0.108. The TeV spectrum of VER J0521+211 constrains its redshift to z < 0.34 under the assumption that the intrinsic TeV photon index of the source is Γ^⋆ > 1.5, which is the limit obtained for standard leptonic emission models.

The high-energy emission from VER J0521+211 peaks in the gamma-ray band, between 10 and 200 GeV, and can be described with a leptonic one-zone SSC emission model. The resulting model parameters would indicate a relatively weak magnetic field of ≲ 0.01 G and a particle-dominated jet. Similar objects have been modeled by adding an EC component to the SSC emission (Acciari et al. 2009; Abdo et al. 2011a), or considering a structured jet with a fast-moving spine and a slower outer layer (Anderhub et al. 2009). These models could presumably describe the SED of VER J0521+211, and generally reach solutions closer to equipartition. However, being more complex than SSC emission, SSC+EC and structured jet models have more free parameters, and would be underconstrained given the data available for VER J0521+211.

Because of its low Galactic latitude, VER J0521+211 observations were not triggered by its X-ray properties like most TeV-candidate blazars, but by a cluster of E > 30 GeV photons detected in the Fermi-LAT public data released after the first year of observations. Selection criteria based on Fermi-LAT data have been successful in identifying other new TeV blazars (Aleksić et al. 2011; Aliu et al. 2012; Zech et al. 2011), particularly at low Galactic latitudes where selections based on radio and X-ray data are less powerful due to Galactic extinction. The detection of VER J0521+211 adds to previous results that have demonstrated the strength of GeV band (e.g., Mukherjee et al. 2000; Vandenbroucke et al. 2010; Kara et al. 2012) and TeV-band observations (e.g., Mariotti 2010; Aliu et al. 2011; Abramowski et al. 2011) as a tool to identify blazars located behind the Galactic plane.

With a TeV flux between 0.09 and 0.33 Crab, VER J0521+211 ranks among the brightest known TeV blazars, and can be detected with current ground-based Cherenkov telescopes in less than 1 hr exposure. Assuming a redshift of 0.108, the TeV luminosity of VER J0521+211 is L_>0.2 TeV ∼ 2.4 × 10⁴⁴ erg s⁻¹, larger than that of the "classical" northern TeV blazars (Mrk 421, Mrk 501, and 1ES 1959+650), which sample the low-luminosity end of the population of TeV blazars (Şentürk et al. 2013). Given the observed variability and its bright TeV flux, future multiwavelength observations of VER J0521+211 will be able to extend the detailed time-resolved spectral modeling available for nearby HBLs (Krawczynski et al. 2004; Acciari et al. 2011; Abdo et al. 2011b) to a more luminous non-HBL blazar.

VERITAS is supported by grants from the U.S. Department of Energy Office of Science, the U.S. National Science Foundation, and the Smithsonian Institution, by NSERC in Canada, by Science Foundation Ireland (SFI 10/RFP/AST2748), and by STFC in the UK. We acknowledge the excellent work of the technical support staff at the Fred Lawrence Whipple Observatory and at the collaborating institutions in the construction and operation of the instrument.

The Fermi-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.

M.E. acknowledges support from the NASA grants NNX10AP66G and NNX12AJ30G. Y.Y.K. was supported in part by the Russian Foundation for Basic Research (projects 11-02-00368 and 12-02-33101), the basic research program "Active Processes in Galactic and Extragalactic Objects" of the Physical Sciences Division of the Russian Academy of Sciences, and the Dynasty Foundation. M.B. acknowledges support by the South African Department of Science and Technology through the National Research Foundation under NRF SARChI Chair grant No. 64789.

This research has made use of data from the MOJAVE database that is maintained by the MOJAVE team (Lister et al. 2009a). The MOJAVE project is supported under NASA-Fermi grants NNX08AV67G and 11-Fermi11-0019. The authors thank Julie Skinner for obtaining, as a target of opportunity, the first MDM spectrum of RGB J0521.8+2112 used in this paper, and Talvikki Hovatta for providing the OVRO radio data. Observations at Steward Observatory were supported by the NASA Fermi Guest Investigator Program grant NNX09AU10G. Finally, the authors thank the Swift team for accepting and carefully scheduling the target of opportunity observations of VER J0521+211 that were used in the paper and for support from the Swift Guest Investigator program, NASA grant NNX10AF89G.