DISCOVERY OF VERY HIGH ENERGY GAMMA RAYS FROM PKS 1424+240 AND MULTIWAVELENGTH CONSTRAINTS ON ITS REDSHIFT

PKS 1424+240 was detected as a radio source by Condon et al. (1977). It was classified as a blazar by Impey & Tapia (1988) from optical polarization studies. Fleming et al. (1993) verified the polarization results and also reported nonthermal X-ray radiation, further strengthening the classification.

Blazar emission is dominated by nonthermal radiation, which is thought to be related to charged particle acceleration near a massive compact object in the center of the host galaxy, or in outflowing relativistic jets. The spectral energy distribution (SED) is characterized by two peaks. The lower peak is widely accepted to be synchrotron radiation from relativistic electrons and occurs in the IR to X-ray bands. The higher energy peak is in the gamma-ray band, sometimes at energies as high as a few TeV, and can be created via either inverse-Compton scattering by relativistic electrons or hadronic interactions (for a review, see Böttcher 2007, and references therein). The position of the synchrotron peak of PKS 1424+240 has not been measured, but it can be constrained from optical and X-ray data to be between 10¹⁵ Hz and 10¹⁷ Hz. Depending on the definition used, PKS 1424+240 is either an intermediate-frequency-peaked BL Lac (IBL; Nieppola et al. 2006) or a high-frequency-peaked BL Lac (HBL; Padovani & Giommi 1996; A. A. Abdo et al. 2010, in preparation).

Gamma-ray emission from PKS 1424+240 was not detected by EGRET (Fichtel et al. 1994), but was recently observed with the Fermi Large Area Telescope (LAT) pair-conversion telescope (Abdo et al. 2009a, 2009b). The reported flux above 100 MeV of (6.2 ± 0.8) × 10⁻⁸ cm⁻² s⁻¹ and hard spectral index Γ = 1.80 ± 0.07 (dN/dE ∝ E^−Γ) triggered VERITAS observations.

The redshift of PKS 1424+240 is not known. Scarpa & Falomo (1995) have derived a lower limit on the redshift of z> 0.06 and Sbarufatti et al. (2005) a limit of z> 0.67, both assuming a minimum luminosity of the host galaxy. The latter authors also reported evidence that the ratio of the nucleus to host luminosity is much larger than 100, which is typical for BL Lac objects but complicates photometric determination of the redshift.

We report the detection of PKS 1424+240 in very high energy (VHE) gamma rays and contemporaneous observations with Fermi, Swift, and the MDM observatory. Shortly after the VHE discovery (Ong 2009), it was confirmed by the MAGIC Collaboration (Teshima 2009). This discovery marks the first Fermi-motivated VHE discovery.

The VERITAS observatory, located in southern Arizona at 1.3 km a.s.l., is described in detail in Weekes et al. (2002) and Holder et al. (2006).

PKS 1424+240 was observed with VERITAS between 2009 February 19 and June 21 at zenith angles between 7° and 30°. The observations were performed in wobble mode (Fomin et al. 1994) with a 05 offset, enabling simultaneous background estimation. About one-third of the data were taken during low levels of moonlight. About 65% of the observations were conducted with only three telescopes due to the relocation of one telescope, which began in May and was completed in 2009 August. Of the 37.3 hr of data, 28.5 hr survive standard data quality selection.

Events are reconstructed following the procedure in Acciari et al. (2008). The recorded shower images are parameterized by their principal moments, giving an efficient suppression of the far more abundant cosmic-ray background. Two separate sets of cuts are applied to reject background events, hereafter called soft and medium. These cuts are applied to the parameters mean scaled width (MSW), and mean scaled length (MSL), apparent altitude of the maximum Cherenkov emission (shower maximum), and θ², the squared direction between the position of PKS 1424+240 and the reconstructed origin of the event. Studies on independent data sets show that a shower-maximum cut significantly improves the low energy sensitivity. Soft cuts have a higher sensitivity for sources with soft photon spectra because of a lower energy threshold resulting from a minimum size cut of 50 photoelectrons. In the medium cuts, a minimum size cut of 100 photoelectrons is applied. Size is a measure of the recorded photoelectrons from a shower and a good indicator of the energy of the primary. For the soft-cuts analysis the remaining cuts are MSW <1.06, MSL <1.30, shower maximum >7 km, and θ² < (014)², and MSW <1.04, MSL <1.28, shower maximum >5 km, and θ² < (01)² for the medium cuts. The cuts have been optimized a priori to yield the highest sensitivity for a source with 5% of the Crab Nebula gamma-ray flux. The results are independently reproduced with two different analysis packages explained in Cogan (2008) and Daniel (2008).

In the soft-cuts analysis, 1907 on-source events remain out of 1.25 × 10⁷ triggered events. The background calculated with the reflected-region method (Berge et al. 2007) is 1537 events, which leaves an excess of 370 events. Figure 1 shows the corresponding θ² distribution. The statistical significance of the observed excess is 8.5 standard deviations, σ, calculated with Equation (17) of Li & Ma (1983), and including a trials factor of 2 for the two sets of cuts. In the medium-cuts analysis, the post-trials significance is 4.8 σ (329 on-source events with an estimated background of 244). The angular distribution of the excess events is consistent with a point source. The center of gravity of the excess is 14^h27^m0^s ± 7^s_stat, 23°47'40'' ± 2'_stat coinciding with the position of PKS 1424+240 in radio (Fey et al. 2004). The VERITAS source name is VER J1427+237.

Zoom In Zoom Out Reset image size

Figure 2 shows the light curve of PKS 1424+240 in different energy bands for the time period overlapping the VERITAS observations. The flux measured by VERITAS above 140 GeV is ∼5% of the Crab Nebula flux. The VERITAS data from each dark period⁸⁴ are combined into a single bin to produce a light curve, which is consistent with a constant flux, χ² = 0.3 for 3 degrees of freedom (dof). However, even a doubling in flux would have been difficult to detect. There is no evidence for strong flaring episodes on shorter timescales.

Zoom In Zoom Out Reset image size

Figure 3 shows the differential photon spectra derived with the soft-cuts and medium-cuts analyses, with one overlapping flux point at 260 GeV. The fraction of events that are used both in the last bin in the soft-cuts analysis and in the second bin in the medium-cuts analysis is about 2%, small enough to allow a combined fit of the flux points from the two analyses, with the more significant soft-cuts result at 260 GeV used in the fit. The combined spectrum is well parameterized (χ² = 2.2 for 4 degrees of freedom) by a power law dN/dE = F₀ · (E/E₀)^−Γ, where the photon index Γ is 3.8 ± 0.5_stat ± 0.3_syst and F₀ is (5.1 ± 0.9_stat ± 0.5_syst) × 10⁻¹¹ TeV⁻¹ cm⁻² s⁻¹ for E₀ = 200 GeV. The combined spectrum is consistent with the fit of the soft-cuts points alone, albeit with half the uncertainty on the photon index.

Zoom In Zoom Out Reset image size

Gamma-ray observations with Fermi-LAT (100 MeV to 300 GeV), X-ray and optical observations with Swift XRT (0.2–10 keV) and UVOT (170–650 nm), and optical observations in the R, V, and I bands at the MDM observatory were obtained simultaneously or quasi-simultaneously with the VERITAS observations.

The LAT pair-conversion telescope on board the Fermi Gamma-ray Space Telescope continuously monitors the entire sky between 100 MeV and several hundred GeV (Atwood et al. 2009). The LAT data overlapping with the VERITAS observations were analyzed by selecting "diffuse" class events that have the highest probability of being photons. Further event selection was done by only accepting events that come within a 15° radius from PKS 1424+240 and have energies between 0.1 and 300 GeV. Events with zenith angles above 105° were excluded to limit contamination by gamma rays coming from the Earth's albedo.

The analysis of the photon spectrum and light curve were performed with the standard likelihood analysis tools available from HEASARC ScienceTools v9r15p2. Accidental coincidences with charged cosmic rays in the detector were accounted for using instrument response functions P6_V3_DIFFUSE. The background model used to extract the gamma-ray signal includes a Galactic diffuse emission component and an isotropic component.⁸⁵ The isotropic component includes contributions from the extragalactic diffuse emission as well as from residual charged particle backgrounds. The spectral shape of the isotropic component was derived from residual high-latitude events after the Galactic contribution had been modeled. The background model also takes into account unresolved gamma-ray sources in the region of interest, thus avoiding a bias in the spectral reconstruction. To further reduce systematic uncertainties in the analysis, the normalization and spectral parameters in the background model were allowed to vary freely during the spectral point fitting.

The Fermi-LAT flux measurements are shown in the broadband SED in Figure 4. The flux values are unfolded by assuming an underlying power law, giving an integrated flux over the 0.1–300 GeV band (7.04 ± 0.96_stat ± 0.38_sys) × 10⁻⁸ cm⁻² s⁻¹, and a differential photon spectral index Γ_LAT = 1.73 ± 0.07_stat ± 0.05_sys. The light curve of the integral flux above 100 MeV is plotted with 10-day bins in Figure 2. A fit with a constant yields a χ² = 11.5 for 11 degrees of freedom, suggesting no variability.

Zoom In Zoom Out Reset image size

Target of opportunity observations of nearly 16 ks, distributed over 10 observing periods, were obtained with Swift (Gehrels et al. 2004) following the detection of VHE emission from PKS 1424+240. The data reduction and calibration of the XRT data were completed with the HEASoft v6.6.3 standard tools. The XRT data were taken in photon-counting mode and contained modest pile-up for nine of the observations, which was taken into account by masking a region with 3–6 pixels radius around the source. The outer radius chosen for the signal region was 20 pixels and a background region of similar size was chosen about 5 arcmin off source.

X-ray energy spectra could be extracted from all observing periods and are well described by an absorbed power law using the fixed Galactic column density of neutral hydrogen from Dickey & Lockman (1990) (N_h = 0.264 × 10²¹ cm⁻²). The fit spectral index varies between 2.1 and 2.9 (photon index between 3.1 and 3.9) with a typical statistical uncertainty of 0.1, while the normalization changes between 1.40 × 10⁻² and 0.74 × 10⁻² photons keV⁻¹ cm⁻² s⁻¹ at 1 keV with a typical uncertainty of 0.07 × 10⁻² keV⁻¹ cm⁻² s⁻¹. For the modeling of the SED we use the average spectrum shown in Figure 4. The light curve shows that the X-ray flux is variable over the 10 days of observation. A fit to a constant flux yields a χ² of 60 for 9 degrees of freedom. UVOT observations were taken in the six V, B, U, W1, M2, and W2 bands and were calibrated using standard techniques (Poole et al. 2008). The reddening has been accounted for by interpolating the absorption values from Schlegel et al. (1998) with a galactic spectral extinction model (Fitzpatrick 1999) obtaining 0.663, 0.968, 0.922 mag for the three UV bands W1, M2, and W2 and an assumed redshift of z = 0. The corresponding light curves are shown in Figure 2.

Data in the optical bands were also obtained with the 1.3 m telescope and 4 K imager of the MDM observatory located on the west ridge of the Kitt Peak near Tucson, Arizona. The CCD was operated in unbinned mode, which produces an image scale of 0.315 arcsec pixel⁻¹. 2–4 images were obtained in the V, R, and I bands during each observation. Physical magnitudes were computed from differences in the instrumental magnitudes from the three standard stars in Fiorucci & Tosti (1996), assuming that the magnitudes quoted in that paper are exact. The magnitudes were then corrected for Galactic extinction using extinction coefficients calculated following Schlegel et al. (1998), taken from NED,⁸⁶ and were then converted into νF_ν fluxes. During the 14-day span of the optical photometry, the visual brightness increased by 14% and colors became slightly bluer.

The observed gamma-ray spectrum above 100 GeV is affected by the absorption of gamma rays via pair conversion with extragalactic background light (EBL) photons (Nikishov 1962; Gould & Schréder 1967). Depending on the redshift, this effect can result in a significant softening of the spectrum. We estimate an upper limit of the redshift of PKS 1424+240 by assuming an intrinsic VHE spectrum and making use of the recent advances in EBL modeling.

We assume that the intrinsic spectrum above 140 GeV can be described by a power law. The hardest photon index that we consider is 1.7, which is the value from the simultaneous Fermi observations. The use of Fermi observations allows a model-independent estimate of the hardest possible intrinsic spectrum (see also Abdo et al. 2010). The power law with an index of 1.7 is absorbed using recent EBL models from Franceschini et al. (2008), Gilmore et al. (2009), and Finke et al. (2009). After absorption, the shape of the spectrum is fitted to the VERITAS spectrum with the normalization as a free parameter, and the best estimate of the redshift is determined by minimizing χ². For an intrinsic index of 1.7 this best-fit redshift is z = 0.5 ± 0.1_stat ± 0.1_syst with a χ² = 4 and 5 degrees of freedom. The systematic uncertainty is estimated from the differences in the EBL models.

Instead of assuming no break in the photon spectrum, a more likely scenario is that the intrinsic spectrum softens with increasing energy. In this case, an index of 1.7 is an upper limit of the true photon index and the corresponding upper limit on the redshift is z < 0.66 with a 95% confidence level.

The SED, comprising data from all of the observations, is shown in Figure 4. We model the SED using an improved version of the leptonic one-zone jet model of Böttcher & Chiang (2002). These calculations include time-dependent particle injection and evolution, and they allow for quasi-equilibrium solutions in which a slowly varying broken power-law electron distribution arises from a single power-law injection function, dn_inj/dγ ∝ γ^−q with a low- and high-energy cutoff γ₁ and γ₂, respectively. All model fits presented here are in the fast-cooling regime, with the cooling break at γ₁. We define the magnetic-field equipartition _B as _B ≡ L_B/L_e with L_B being the Poynting flux derived from the magnetic energy density and L_e the energy flux of the electrons propagating along the jet. The corresponding partition fraction for an electron–proton plasma assuming L_p = 10 × L_e of cold protons would be one order of magnitude lower. For an in-depth description of this quasi-equilibrium jet model, see Acciari et al. (2009).

There are few observational constraints on the model parameters for PKS 1424+240, and the redshift is unknown. No superluminal motion has been resolved in this object, and it has not been monitored well enough to firmly establish a minimum variability timescale to constrain the size of the emitting region. The different sizes of the emission region R_B assumed here are compatible with the X-ray variability timescale of about a day. We therefore consider a range of plausible redshifts and adopt model parameters which were typically adequate for modeling other VHE blazars. The redshifts investigated range from z = 0.05, similar to the redshift of the nearby HBLs Mrk 421 and Mrk 501, to z = 0.7. This covers the redshift range determined in the previous section and is just above the lower limit set by Sbarufatti et al. (2005), z>0.67.

The shape of the high-energy part of the electron spectrum is well constrained by the rather steep slope of the X-ray spectrum, which has an average photon index Γ_X-ray ∼ 3.7. In all fits, the relativistic electrons are injected into the emission region with a fixed q = 5.1. Lacking direct constraints on the viewing angle θ_obs, it was chosen such that the Doppler factor D = (Γ[1 − β_Γcos θ_obs])⁻¹ = Γ, where Γ is the bulk Lorentz factor of the emitting material and β_Γc is the velocity. The model parameters that were varied are shown in Table 1. Figure 4 shows the fits, after EBL absorption using the model of Gilmore et al. (2009).

Table 1. SSC Fit Parameters for PKS 1424+240 as a Function of Assumed Redshift

Parameter	z = 0.05	z = 0.10	z = 0.2	z = 0.3	z = 0.4	z = 0.5	z = 0.7
Le (1043 erg s−1)	1.60	4.12	10.7	18.9	29.2	47.1	88.8
LB (1043 erg s−1)	1.66	5.47	16.9	31.1	45.9	49.8	66.2
γ1 (104)	3.7	3.7	3.6	3.4	3.2	3.6	3.7
γ2 (105)	4.0	4.0	4.0	4.0	4.5	4.0	4.0
D	15	18	25	30	35	45	60
B (G)	0.37	0.31	0.25	0.24	0.25	0.18	0.14
B	1.04	1.33	1.59	1.65	1.57	1.06	0.75
RB (1016 cm)	1.2	2.2	3.4	4.0	4.0	4.5	5.0

Download table as:  ASCIITypeset image

The SED modeling shows that a reasonable fit can, in principle, be achieved for any redshift in the considered range. However, the inset of Figure 4 illustrates that above z ∼ 0.2, the model VHE gamma-ray spectrum becomes increasingly too steep compared with the observed VERITAS spectrum. Furthermore, for redshifts z>0.4 the models require unreasonably large Doppler factors of D>50. We note that in particular for the lowest redshift considered, z = 0.05, a good fit can be achieved with almost equipartition between magnetic-field and electron energy densities.

An attempt to improve the fit in the gamma-ray bands, by including an external Compton component, results in a steeper VHE gamma-ray spectrum. This is in conflict with the VERITAS spectrum and a worse representation of the Fermi spectrum. We therefore conclude that a leptonic fit to the SED of PKS 1424+240 during the VERITAS observation is possible with a pure synchrotron self-Compton (SSC) model very close to equipartition, in particular if the redshift of the source is z < 0.1.

We report the detection of PKS 1424+240 in VHE gamma rays. The observation with VERITAS was motivated by the release of the first Fermi source lists (Abdo et al. 2009a, 2009b) and this is the first time that Fermi observations have led to the discovery of a new source in the adjacent VHE band.

The VHE spectrum of PKS 1424+240 has a photon index of 3.8 ± 0.5_stat ± 0.3_sys, whereas the spectrum in the Fermi energy range has a photon index of 1.73 ± 0.07_stat ± 0.05_sys, indicating a break in the spectrum at several tens of GeV. The break can be explained by a one-zone SSC model assuming a wide range of redshifts or could result from EBL absorption if the redshift is about 0.5 and the intrinsic photon index is 1.7, from which a redshift upper limit of 0.66 is inferred. The modeling favors a lower redshift but cannot exclude that PKS 1424+240 is among the most distant sources detected in the VHE regime.

PKS 1424+240 is the third extragalactic source detected in the VHE regime with an unknown or uncertain redshift. It is evident that increased efforts are needed to determine the redshifts of VHE detected blazars. A redshift measurement will allow a better understanding of the source-intrinsic mechanisms and the absorption effects which go along with the gamma-ray propagation.

VERITAS is supported by grants from the US Department of Energy, the US National Science Foundation, and the Smithsonian Institution, by NSERC in Canada, by Science Foundation Ireland, and by STFC in the UK. We acknowledge the excellent work of the technical support staff at the FLWO and the collaborating institutions in the construction and operation of the instrument. N.O. acknowledges the receipt of a Feodor Lynen fellowship of the Alexander von Humboldt Foundation.

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.

This research has made use of the SIMBAD database, operated at CDS, Strasbourg, France.

Facilities: VERITAS - Very Energetic Radiation Imaging Telescope Array System, Swift - Swift Gamma-Ray Burst Mission, Fermi - Fermi Gamma-Ray Space Telescope (formerly GLAST)