BAT AGN Spectroscopic Survey: XVI. General Physical Characteristics of BAT Blazars

TL;DR

This study analyzes 146 blazars from the Swift-BAT survey using multiwavelength data, revealing their physical characteristics, gamma-ray emission properties, and jet power, and comparing them with Fermi-LAT detected blazars.

Contribution

It provides the largest multiwavelength characterization of BAT blazars to date, including gamma-ray detection analysis and physical parameter estimation, expanding understanding of their properties.

Findings

101 blazars are gamma-ray emitters based on LAT data.

BAT blazars host massive black holes (~10^9 M_sun) and luminous accretion disks.

High-redshift BAT blazars have more powerful jets and luminous disks.

Abstract

The recently released 105-month {\it Swift}-Burst Alert Telescope (BAT) all-sky hard X-ray survey catalog presents an opportunity to study astrophysical objects detected in the deepest look at the entire hard X-ray (14$-$195 keV) sky. Here we report the results of a multifrequency study of 146 blazars from this catalog, quadrupling the number compared to past studies, by utilizing recent data from the {\it Fermi}-Large Area Telescope (LAT), Swift-BAT, and archival measurements. In our $\gamma$-ray analysis of $\sim$10 years of the LAT data, 101 are found as $\gamma$-ray emitters, whereas, 45 remains LAT undetected. We model the broadband spectral energy distributions with a synchrotron-inverse Compton radiative model. On average, BAT detected sources host massive black holes ($M_{\rm bh}\sim10^9$ M$_{\odot}$) and luminous accretion disks ($L_{\rm d}\sim10^{46}$ erg s$^{-1}$). At high-redshifts ($z>2$), BAT blazars host more powerful jets with luminous accretion disks compared to those detected only with the {\it Fermi}-LAT. We find good agreement in the black hole masses derived from the single-epoch optical spectroscopic measurements and standard accretion disk modeling approaches. Other physical properties of BAT blazars are similar to those known for {\it Fermi}-LAT detected objects.

Tables

Table 1. Table 1: Swift -BAT detected sources which were classified as ‘beamed AGN’ in 105-month catalog but excluded in this work.

BAT Name	Counterpart name	Redshift	Comments
J0048.8+3155	Mrk 348	0.015	Seyfert 2 galaxy (Hernández-García et al., 2015)
J0136.5+3906	B3 0133+388	–	No redshift
J0142.0+3922	B2 0138+39B	0.08	Fermi-LAT undetected HSP BL Lac object
J0156.5$-$5303	RBS 259	–	No redshift
J0225.8+5946	NVSS J022626+592753	–	No redshift
J0241.3$-$0816	NGC 1052	0.005	LINER galaxy (Cazzoli et al., 2018)
J0245.2+1047	4C +10.08	0.07	Fermi-LAT undetected HSP BL Lac object
J0319.7+4132	NGC 1275	0.018	Seyfert 1.5 galaxy (Véron-Cetty & Véron, 2010)
J0433.0+0521	3C 120	0.033	Radio galaxy (Tombesi et al., 2017)
J0550.7$-$3215B	2XMM J055054.3$-$321616	–	No redshift, counterpart doubtful
J0608.9$-$5507	PKS 0607$-$549	–	No redshift
J0612.2$-$4645	PMN J0612$-$4647	–	No redshift
J0632.1$-$5404	1RXS J063200.7$-$540454	0.204	Giant radio quasar (Saripalli et al., 2005)
J0640.3$-$1286	PMN J0640$-$1253	–	No redshift
J0709.3$-$1527	PKS 0706$-$15	–	No redshift
J0923.2+3850	B2 0920+39	–	No redshift
J1213.2$-$6020	1RXS J121324.5$-$601458	–	No redshift
J1304.5$-$5651	IGR J13045$-$5630	0.051	Radio and Fermi-LAT undetected, Galactic plane
J1312.1$-$5631	2MASX J13103701$-$5626551	–	No redshift
J1325.4$-$4301	Cen A	0.002	Radio galaxy
J1512.2$-$1053A	NVSS J151148$-$105023	–	No redshift
J1557.8$-$7913	PKS 1549$-$79	0.150	Radio galaxy (Guillard et al., 2012)
J1655.0$-$4998	CXOU J165551.9$-$495732	0.058	Radio and Fermi-LAT undetected, Galactic plane
J1719.7+4900	ARP 102B	0.024	LINER/Radio galaxy (Popović et al., 2014)
J1734.9$-$2074	NVSS J173459$-$204533	–	No redshift
J1742.1$-$6054	PKS 1737$-$60	0.41	Fermi-LAT undetected HSP BL Lac object
J1941.3$-$6216	PKS 1936$-$623	–	No redshift
J2037.2+4151	SSTSL2 J203705.58+415005.3	–	No redshift
J2056.8+4939	RX J2056.6+4940	–	No redshift
J2117.5+5139	2MASX J21174741+5138523	–	No redshift
J2209.4$-$4711	NGC 7213	0.006	Seyfert 1 galaxy (Halpern & Filippenko, 1984)
J2303.1$-$1837	PKS 2300$-$18	0.128	Radio galaxy (Robinson et al., 1987)

Table 2. Table 2: Swift -BAT detected sources which were not classified as blazars in 105-month catalog though exhibit broadband properties similar to beamed AGNs.

BAT Name	Counterpart name	Redshift	Old classification
J0144.8$-$2754	PMN J0145$-$2733	1.155	Unknown AGN
J0131.5$-$1007	PMN J0131$-$1009	3.515	Unknown AGN
J0201.0+0329	[HB89] 0158+031	0.765	Unknown AGN
J1810.0$-$6554	PMN J1809$-$6556	0.18	Unknown AGN
J1105.4+0200	PMN J1105+0202	0.106	U3
J1254.9+1165	CGRaBS J1254+1141	0.87	U3
J1153.9+5848	RGB J1153+585	0.202	U3
J0909.0+0358	1RXS J090915.6+035453	3.2	Sy1;broad-line AGN
J0250.8$-$3626	6dF J0250552$-$361636	1.536	Sy1;broad-line AGN
J1310.9$-$5553	IGR J13109$-$5552	0.104	Sy1
J1347.1+7325	RGB J1346+733	0.29	Sy1
J0042.9+3016B	RX J0042.6+3017	0.141	Sy1
J0506.6$-$1937	6dF J0506479$-$193651	0.094	Sy1
J0547.1$-$6427	CGRaBS J0546$-$6415	0.323	Sy1
J1132.9+1019B	[HB89] 1130+106	0.539	Sy1
J1224.1+4073	NVSS J122359+404359	0.096	Sy1
J1238.4+5349	RGB J1238+534	0.347	Sy1
J1306.4$-$7603	PMN J1307$-$7602	0.183	Sy1
J2033.3$-$2253	PKS 2030$-$23	0.131	Sy1
J2033.4+2147	4C+21.55	0.173	Sy1

Table 3. Table 3: The γ 𝛾 \gamma -ray spectral parameters of Fermi -LAT detected BAT blazars.

Name	$F_{0.1-300\mathrm{GeV}}$	${\Gamma }_{0.1-300\mathrm{GeV}}/\alpha$	${\beta }_{0.1-300\mathrm{GeV}}$	TS
[1]	[2]	[3]	[4]	[5]
J0036.0+5951	38.20$\pm$1.92	1.72$\pm$0.02	–	3405
J0122.9+3420	3.38$\pm$0.80	1.54$\pm$0.11	–	164
J0144.8$-$2754	7.02$\pm$0.29	2.59$\pm$0.04	–	912
J0213.7+5147	3.59$\pm$0.51	1.91$\pm$0.09	–	162
J0218.0+7348	10.70$\pm$0.41	2.93$\pm$0.05	–	841

Table 4. Table 4: The SED parameters associated with the modeling of the broadband emission of BAT blazars.

Name	$z$	${\theta }_v$	$M_{\mathrm{bh}}$	Type	$L_d$	$R_d$	$R_{\mathrm{blr}}$	$\delta$	$\Gamma$	$B$	$p$	$q$	${\gamma }_{\min }$	${\gamma }_b$	${\gamma }_{\max }$	$U_e$	CD
[1]	[2]	[3]	[4]	[5]	[6]	[7]	[8]	[9]	[10]	[11]	[12]	[13]	[14]	[15]	[16]	[17]	[18]
J0010.5+1057	0.09	6.0	8.70	D	45.15	0.215	0.038	9.5	11	0.2	2.2	5.1	4	286	3.0e+03	-2.16	6.2
J0017.1+8134	3.37	4.0	10.04	D	48.00	0.421	1.020	12.2	8	2.2	1.9	4.5	1	41	2.0e+03	-1.68	16.2
J0036.0+5951	0.08	3.0	0.00	N	0.00	0.115	0.000	13.6	8	0.1	2.1	3.7	50	251243	1.5e+06	-2.71	0.4
J0042.9+3016B	0.14	3.0	8.90	A	45.00	0.031	0.032	16.5	11	1.2	1.5	5.3	1	20	1.3e+03	-1.69	247.2
J0057.0+6405	0.29	3.0	8.70	A	44.00	0.120	0.010	18.6	15	1.1	2.1	4.6	5	300	3.0e+03	-2.25	0.6

Table 5. Table 5: Various jet powers derived from the SED modeling.

Name	$P_e$	$P_m$	$P_r$	$P_p$	$P_j$
[1]	[2]	[3]	[4]	[5]	[6]
J0010.5+1057	44.84	44.40	45.23	46.97	46.97
J0017.1+8134	45.62	46.59	47.35	48.13	48.14
J0036.0+5951	43.46	42.68	44.26	44.23	44.31
J0042.9+3016B	43.61	44.06	44.25	46.20	46.21
J0057.0+6405	44.51	45.44	45.01	46.51	46.55

Table 6. Table 6: The list of 28 blazars which are modeled with synchrotron and SSC processes, i.e., without invoking EC mechanism.

Name	$z$
J0036.0+5951	0.08
J0122.9+3420	0.27
J0213.7+5147	0.05
J0232.8+2020	0.14
J0244.8$-$5829	0.26
J0326.0$-$5633	0.06
J0349.2$-$1159	0.18
J0353.4$-$6830	0.09
J0507.7+6732	0.31
J0550.7$-$3212A	0.07
J0710.3+5908	0.12
J0721.0+7133	0.30
J0733.9+5156	0.07
J0930.1+4987	0.19
J0934.0$-$1721	0.25
J1031.5+5051	0.36
J1103.5$-$2329	0.19
J1104.4+3812	0.03
J1136.7+6738	0.13
J1221.3+3012	0.18
J1417.7+2539	0.24
J1428.7+4234	0.13
J1943.5+2120	0.21
J1959.6+6507	0.05
J2009.6$-$4851	0.07
J2246.7$-$5208	0.19
J2251.8$-$3210	0.25
J2359.0$-$3038	0.17

Equations

$$S(\gamma )=S_0\frac{({\gamma }_{\mathrm{b}}{)}^{-p}}{(\gamma /{\gamma }_{\mathrm{b}}{)}^p+(\gamma /{\gamma }_{\mathrm{b}}{)}^q}$$

$$F_{\nu ,\mathrm{disk}}={\nu }^3\frac{4\pi h\cos {\theta }_{\mathrm{v}}}{c^2{D_l}^2}{\int }_{R_{\mathrm{d},\mathrm{in}}}^{R_{\mathrm{d},\mathrm{out}}}\frac{R\mathrm{d}R}{e^{h\nu /kT(R)}-1}$$

$$T(R)=\frac{3R_{\mathrm{Sch}}{L}_{\mathrm{d}}}{16\pi {\eta }_{\mathrm{acc}}{\sigma }_{\mathrm{SB}}{R}^3}{\left[1-{\left(\frac{3R_{\mathrm{Sch}}}{R}\right)}^{1/2}\right]}^{1/4},$$

Full text

galaxies: active — gamma-ray: galaxies— galaxies: jets— galaxies: high-redshift— quasars: general

††slugcomment: submitted to ApJ††software: fermiPy (Wood et al., 2017)

1 Introduction

Active galactic nuclei (AGN) are among the most energetic objects in the Universe and are crucial players in the evolution of galaxies (see, e.g. Kormendy & Ho, 2013, for a review). A subset of these interesting objects host relativistic jets and when the jet is closely aligned with the line of sight to the observer, the source is termed as a blazar (Blandford & Rees, 1978). Due to their peculiar orientation, the emitted radiation from blazar jets is relativistically amplified. This phenomenon makes blazars visible even at very high-redshifts ($z>2$, e.g. Romani et al., 2004; Ackermann et al., 2017). Blazars are classified as flat spectrum radio quasars (FSRQs) and/or BL Lac objects based on the rest-frame equivalent width (EW) of the optical emission lines (Stickel et al., 1991) with FSRQs exhibiting broad emission lines (EW$>$5Å), while BL Lacs show weak or no emission lines in their optical spectra. This, in turn, suggests a radiatively efficient accretion process that illuminates the broad line region (BLR) surrounding the central engine of FSRQs (cf. Ghisellini et al., 2011). The spectral energy distribution (SED) of a blazar is characterized by a double hump structure. The low energy peak is understood to be due to synchrotron process and is typically observed between the infrared and the X-ray band, whereas, the high energy peak is typically explained by the inverse Compton scattering of the low energy photons by relativistic electrons in the jet. Alternatively, hadronic models have also been invoked to explain the high energy SEDs of blazars (e.g. Böttcher et al., 2013). Based on the location of the synchrotron peak frequency ($\nu^{\rm peak}_{\rm syn}$), blazars are also classified as low- ($\nu^{\rm peak}_{\rm syn}<$1014 Hz), intermediate- (10${}^{14}<\nu^{\rm peak}_{\rm syn}<$1015 Hz) and high- ($\nu^{\rm peak}_{\rm syn}>$1015 Hz) synchrotron peaked or LSP, ISP, and HSP, respectively (Abdo et al., 2010). Typically FSRQs are LSP/ISP blazars, whereas, BL Lacs are mostly HSP ones.

The sky-surveying capabilities of the currently operating hard X-ray and $\gamma$-ray instruments, namely Burst Alert Telescope (BAT; Barthelmy et al., 2005) onboard Neil Gehrels Swift observatory (Gehrels et al., 2004) and Large Area Telescope (LAT; Atwood et al., 2009) onboard Fermi Gamma-ray Space Telescope, have allowed to carry out the population studies of blazars and their luminosity dependent evolution (e.g. Ajello et al., 2009, 2012, 2014). In particular, Ajello et al. (2009) studied 38 blazars detected in the first 36 months of the all-sky survey by Swift-BAT and used this sample to derive the 15$-$55 keV luminosity function (LF) of blazars. Interestingly, it was noted that the evolution of luminous FSRQs peaks at higher redshifts ($\sim$4) compared to other classes of BAT detected AGNs. Ghisellini et al. (2010) used the LF reported in Ajello et al. (2009) to determine the space density of billion-solar-mass black holes residing in jetted systems and introduced an exponential cut-off in the blazar luminosity function above $z=4.3$ to remain consistent with the number of detected blazars in that redshift bin and also with the space density of massive halos. However, computation of the luminosity function depends strongly on the number of sources detected in a particular redshift bin. For example, Ackermann et al. (2017) reported the first time $\gamma$-ray detection of five $z>3.1$ radio-loud quasars thus confirming their blazar nature. This allowed them to update the space density of 109 $M_{\odot}$ black holes residing in radio-loud quasars at $z\approx 4$ and to conclude that radio-loudness may be a crucial ingredient for the rapid black hole growth in the early Universe (see also Volonteri et al., 2011). Moreover, by studying high-redshift ($z>2$) Swift-BAT blazars, Ghisellini et al. (2010) found them to host more powerful jets and more luminous accretion disks compared to Fermi-LAT detected $z>2$ sources.

A new catalog of Swift-BAT detected sources covering the first 105-month of the mission has recently been released (Oh et al., 2018) along with an extensive multi-wavelength observations of BAT detected sources by the BAT AGN Spectroscopic Survey (BASS) collaboration (Koss et al., 2017; Lamperti et al., 2017; Ricci et al., 2017). Motivated by more than quadrupling of the sources compared to past studies and also the larger redshift range it covers111The farthest known blazar in 105-month catalog, SWIFT J1430.6+4211, has $z=4.71$, significantly larger than the farthest known sources in previous BAT catalogs ($z=3.67$, e.g., Ajello et al., 2009)., we have carried out a multiwavelength study of all the blazars present in the 105-month Swift-BAT catalog. Our primary objective is to explore the physical properties of Swift-BAT blazars by applying a simple leptonic radiative model and also to compare them with that known for the Fermi-LAT detected ones which are yet to be detected by BAT.

In Section 2, we describe the sample adopted in this work and data analysis procedures are explained in Section 3. The adopted leptonic model is elaborated in Section 4 and we discuss the derived SED parameters in Section 5. In Section 6, we compare the accretion-jet connection observed in BAT blazars with Fermi-LAT detected sources, and Section 7 is devoted to the luminosity dependent evolution of blazars. We summarize our findings in Section 8. Throughout this work, we use a flat cosmology with $H_{0}=67.8$ km s*-1* Mpc*-1* and $\Omega_{\rm M}=0.308$ (Planck Collaboration et al., 2016).

2 The Sample

Our blazar sample is based on the 1632 sources included in the 105-month Swift-BAT catalog. There are 158 sources classified as ‘beamed AGN’, mainly based on their presence in the BZCAT and CRATES catalogs (Massaro et al., 2015; Healey et al., 2007). We exclude those 32 objects which are reported as ‘beamed AGN’ (based on the BZCAT catalog) but, however, do not exhibit broadband properties typically observed from the blazar class of AGN. The full list of these sources is provided in Table 1. Additionally, we explore the broadband properties of the whole BAT sample to identify potential blazar candidates which might have been misclassified as Seyferts. This was done by examining their multifrequency SEDs generated using the ASDC SED builder tool222https://tools.asdc.asi.it/ and physical properties (e.g., radio detection) reported in literature (cf. Griffith & Wright, 1993; Gregory et al., 1996). This led to the inclusion of 20 sources (Table 2). Admittedly, there could be a very few unidentified BAT objects that might be blazars though it may not be possible to pinpoint them due to lack of a confirmed optical/radio counterparts and also because of paucity of the multiwavelength information. The BAT survey completeness and redshift evolution of beamed AGN will be fully examined in a future study (Marcotulli et al. in prep.).

This paper makes use of redshift and black hole mass estimates from BASS, a large effort to collect optical spectra for all the Swift-BAT AGN. The data release 1 optical spectra were obtained from a large variety of telescopes (Koss et al., 2017), and can all be viewed at the BASS website333www.bass-survey.com. We also include 15 new Data Release 2 redshifts measurements, from Palomar, Southern Astrophysical Research (SOAR) telescope, and the Very Large Telescope (VLT) that will shortly be published (Oh et al., in prep) as well as comparison black hole mass measuremenst from this catalog. Altogether, our final sample consists of 146 BAT blazars.

3 Data Reduction and Compilation

3.1 Fermi-LAT

The primary objective of this study is to explore the average physical characteristics of BAT blazars rather than any of their specific activity states. With this in mind, we perform an analysis of Fermi-LAT data covering $\sim$10 years (2008 August 5 to 2018 March 3) of Fermi operation. The all-sky surveying capability of Fermi-LAT ensures that the generated spectra represent an average activity state of the blazar. A standard data reduction procedure is adopted to carry out the analysis444http://fermi.gsfc.nasa.gov/ssc/data/analysis/documentation/. The data reduction is performed using Science Tools (v11r5p3) and publicly available analysis package fermiPy (Wood et al., 2017). We use the latest P8R3 dataset (Atwood et al., 2013; Bruel et al., 2018) in the energy range of 0.1$-$300 GeV and consider only SOURCE class events (evclass=128). We define a ROI of 15*∘* radius centered at the target blazar and apply standard cuts (zenith angle $z_{\rm max}<90^{\circ}$ and ‘DATA_QUAL$>$0 and LAT_CONFIG==1’) to select good time intervals. To derive the optimized spectral parameters from the binned-likelihood fitting, we consider all $\gamma$-ray sources present in third Fermi-LAT detected source catalog (3FGL; Acero et al., 2015) and lying within the ROI and also the isotropic and Galactic diffuse emission models (Acero et al., 2016). Note that the presence of any extended $\gamma$-ray source in the ROI is properly taken into account by adopting the publicly available extended source templates555https://github.com/fermi-lat/extendedArchive. The parameters of all of the considered $\gamma$-ray sources are allowed to vary during the likelihood fitting.

The significance of source detection is quantified using maximum likelihood test statistic TS = 2$\Delta\log\mathcal{L}$, where $\mathcal{L}$ denotes the likelihood ratio, between models with and without a point source at the position of interest. Since the time period covered is significantly larger than that used to produce the 3FGL catalog, we adopt findsource tool available in fermiPy to generate TS maps and search unmodeled $\gamma$-ray sources that are present in the data (with TS$\gtrsim$25, $\gtrsim$5$\sigma$ detection Mattox et al., 1996) but not in the model. Once found, they are characterized with a power law and included in the sky model. This procedure is repeated until the TS map stops showing any excess residual. Flux upper limits in the $\gamma$-ray spectra are derived at 95% confidence for energy bins with TS$<9$. Finally, we compute 5$\sigma$ LAT sensitivity limits by adopting a photon index of 2.4 for each $\gamma$-ray undetected BAT FSRQ (median 0.1$-$300 GeV 5$\sigma$ flux upper limit $\sim$7$\times$10*-7* MeV cm*-2* s*-1*). The generated $\gamma$-ray spectra are corrected for extragalactic background light absorption following the prescriptions of Domínguez et al. (2011).

Overall, we found 101 $\gamma$-ray emitting sources out of 146 BAT blazars and their derived $\gamma$-ray spectral parameters are provided in Table 3.

3.2 Swift-BAT, NuSTAR and Archival Measurements

We extract 105-month averaged 14$-$195 keV spectra for all sources using publicly available spectrum and response files and by fitting a simple power law model in XSPEC (Arnaud, 1996), as done by Oh et al. (2018).

The BAT spectra of 25 FSRQs exhibit a steep falling shape (photon index $>$2), in contrast to the soft X-ray ($<$10 keV) spectral shape which predicts a hard, rising (photon index $<$2) spectrum. This is likely due to low signal-to-noise spectrum in BAT where $\Gamma$ is not well constrained ($\Gamma_{Err}>0.4$). For such sources, we search for better quality NuSTAR observations, when available, and find that in all cases NuSTAR spectra matches well with the soft X-ray one. The data from the NuSTAR focal plane modules A and B are reduced following the standard procedure. We first run the tool nupipeline to clean and calibrate the event files. The source and background regions of 30*′′* and 70*′′*, respectively, are then selected from the same chip, and the nuproducts pipeline is adopted to extract the source and background spectra along with ancillary and response matrix files. The spectra are binned to have at least 20 counts per bin and fitting is performed in XSPEC.

Both Fermi-LAT and Swift-BAT are all-sky surveying instruments and hence considering their long-time averaged spectra is justified to achieve the objectives proposed in this work. On the other hand, soft X-ray and optical-UV instruments, e.g., X-ray telescope (XRT) and ultraviolet and optical telescope (UVOT) on board Swift, operate in pointed modes and observations are often triggered during elevated activity states of blazars. With this in mind, we do not separately analyze XRT and UVOT observations and rather rely on publicly available archival SED measurements from ASDC that can be considered as representatives of the average behavior of the target of interest. Note that the BASS collaboration is running a dedicated X-ray spectroscopic followup of BAT detected objects, irrespective of their activity states, and the results of a comprehensive an X-ray analysis can be found in Ricci et al. (2017).

4 Modeling the Broadband Emission

4.1 The Model

We adopt a simple one-zone leptonic emission model (e.g., Dermer et al., 2009; Ghisellini & Tavecchio, 2009) to reproduce the multifrequency SEDs of BAT blazars. We consider a conical jet with a semi-opening angle of 0.1 radian. The emission region is assumed to be spherical and cover the whole cross-section of the jet. With this assumption, the dissipation distance ($R_{\rm d}$) constrains the size of the emitting region. The observer receives the Doppler-boosted jet emission at an angle of $\theta_{\rm v}$ from the jet. Furthermore, the radiating population of relativistic electrons has a broken power-law distribution of the following type:

[TABLE]

where $S_{0}$ is the normalization constant (cm*-3*), $\gamma$ is the the Lorentz factor of electrons, and $p$, $q$ are the spectral indices below and above the break Lorentz factor $\gamma_{\rm b}$, respectively. Note that our modeling technique does not include a time-dependent evolution of the electron spectrum considering particle injection/cooling.

The leptonic population emits synchrotron radiation in presence of uniform and randomly oriented magnetic field $B$. We also consider various photon fields, both inside and outside of the jet, to calculate the inverse Compton radiation from the emitting population (e,g., Rybicki & Lightman, 1979). This includes thermal photons originated from the accretion disk, BLR, and dusty torus (the external Compton or EC process, e.g., Sikora et al., 1994) and also non-thermal synchrotron photons produced by the same electrons (synchrotron self Compton or SSC, Marscher & Gear, 1985). Moreover, we assume the BLR and torus are spherical shells with radii $R_{\rm blr}=10^{17}\sqrt{L_{\rm d,45}}$ cm and $R_{\rm torus}=2.5\times 10^{18}\sqrt{L_{\rm d,45}}$ cm, respectively, where $L_{\rm d,45}$ is the luminosity of the accretion disk in units of 1045 erg s*-1* (Ghisellini & Tavecchio, 2009). The adopted emission profiles of these components follow a blackbody distribution peaking at Hydrogen Ly-$\alpha$ frequency and the characteristic temperature of the torus ($T_{\rm torus}$), respectively. We assume a geometrically thin and optically thick accretion disk (Shakura & Sunyaev, 1973) whose radiative profile is considered to follow a multi-colored blackbody (see, e.g., Frank et al., 2002)

[TABLE]

where $h$ is the Planck constant, $D_{l}$ is the luminosity distance, $k$ is the Boltzmann constant, $c$ is the speed of light, and $R_{\rm d,~{}in}$ and $R_{\rm d,~{}out}$ are the inner and outer disk radii, assumed as 3$R_{\rm Sch}$ and 500$R_{\rm Sch}$, respectively. $R_{\rm Sch}$ is the Schwarzschild radius. In the above equation, the disk temperature has the following dependence on the radius

[TABLE]

where $\sigma_{\rm SB}$ is the Stefan-Boltzmann constant and $\eta_{\rm acc}$ is the accretion efficiency, adopted here as 10%. The thermal emission originated from the X-ray corona is assumed to follow a power law with exponential cut-off ($L_{\rm cor}(\nu)\propto\nu^{-1}\exp(-h\nu/150\,{\rm keV})$, see Ghisellini & Tavecchio, 2009).

For more than $>50$% of sources, the near-infrared-to-ultraviolet (IR-to-UV) SED exhibits a bump which we interpret as due to accretion disk emission. By modeling this bump with a standard Shakura & Sunyaev (1973) disk, we are able to constrain both the accretion disk luminosity ($L_{\rm d}$) and the mass of the central black hole ($M_{\rm bh}$) with the assumption of the accretion efficiency $\eta_{a}=0.1$ (see also, Frank et al., 2002; Ghisellini et al., 2010). In order to get independent measurements, we use results from our own ongoing optical spectroscopic campaign (Koss et al., 2017, Mejia-Restrepo et al. in prep.) and also search for the availability of $M_{\rm bh}$ and $L_{\rm d}$ derived from single epoch optical spectroscopy available in the literature (e.g., Shaw et al., 2012). We note that such commonly-used single epoch $M_{\rm bh}$ estimates carry systematic uncertainties of order $0.3-0.5$ dex (see, e.g., Shen, 2013; Peterson, 2014, for reviews of these methods), and that the usage of the broad C iv $\lambda 1549$ emission line may carry additional uncertainties related to the virialized nature of the BLR (e.g., Trakhtenbrot & Netzer, 2012; Mejía-Restrepo et al., 2018, and references therein). Furthermore, if optical spectral parameters (e.g., line and continuum luminosities) are reported (cf. Torrealba et al., 2012), we use them to constrain $M_{\rm bh}$ following empirical relations commonly used for blazars (e.g., Shen et al., 2011; Shaw et al., 2012). We compute the luminosity of the BLR from the emission line luminosities by considering the quasar emission line weights provided by Francis et al. (1991). Assuming that the BLR reprocesses 10% of the disk radiation, we determine $L_{\rm d}$.

4.2 Modeling constraints

The SED model used in this work does not perform any statistical fitting and hence the uniqueness of the derived parameters cannot be claimed. However, our modeling efforts are driven by the quality of the multiwavelength data that provide fairly good constraints to the parameters. For example, we determine the $M_{\rm bh}$ and $L_{\rm d}$ either from the disk modeling or from the optical spectroscopy and these two parameters regulate the size and radiative energy densities of various external photon fields. The shape of the X-ray spectrum constrains the relative dominance of the EC and SSC mechanisms since different parts of the electron energy distribution are involved. The SSC radiation regulates the size of the emission region and also the magnetic field. A Compton dominated SED reflects the prevalence of the external photon densities over the magnetic one and since in our model these are a function of the distance of the emission region from the central black hole, we are able to constrain $R_{\rm d}$. In Compton-dominated blazars, the low- and high-energy indices of the broken power law electron energy distribution are determined from the spectral shapes of the X-ray and $\gamma$-ray SEDs, respectively. The optical-UV spectrum further controls the high-energy index provided it is dominated by the falling synchrotron emission. On the other hand, in blazars whose synchrotron peak lies at UV-X-ray energies, i.e., HSP sources, we use Fermi-LAT spectra to determine the low-energy index. The shape of the BAT spectrum, along with the soft X-ray SED, provides a good constraint to the high-energy index in such objects.

5 Physical Characteristics

In the left panel of Figure 1, we show the 14$-$195 keV rest-frame luminosity of various types of BAT detected AGNs as a function of their redshifts. As can be seen that blazars dominate the 105-month BAT catalog above $z\approx 0.5$. This is likely a selection effect due to relativistic beaming which makes blazar jets brighter compared to other class of astrophysical objects. Considering only blazars, the photon index versus luminosity distribution is shown in the middle panel and we highlight $z>2$ blazars by showing them with pink stars. As can be seen, these are the most powerful ones with luminosity exceeding $10^{47}$ erg s*-1*. Furthermore, there is a hint of an anti-correlation (Spearmann coefficient $\rho=-0.29\pm 0.08$, probability of no correlation or PNC = 0.01) where more luminous objects, mostly FSRQs, tend to have a harder BAT spectrum. This is because the X-ray to $\gamma$-ray emission in FSRQs primarily originates via EC process which has a hard spectral shape in X-rays (see Figure 2) which becomes more prominent at high redshifts due to shift of SED peaks to lower frequencies and $K$-correction. Less luminous HSP blazars generally have a falling synchrotron emission in the BAT energy range and thus exhibit a steep spectrum with 14$-$195 keV photon index $>$2. We plot $\gamma$-ray luminosity as a function of the hard X-ray one in the right panel of Figure 1. Interestingly, the Fermi-LAT undetected objects share the same range of the 14$-$195 keV luminosity with $\gamma$-ray detected ones. Since all of them are FSRQs, this observation indicates their SED peaks to lie at lower frequencies with respect to Fermi-LAT sources.

The modeled SEDs of BAT blazars are shown in Figure 2 and we provide associated parameters in Table 4. Note that there are 28 sources whose broadband SEDs are well represented with synchrotron-SSC processes, i.e., without invoking EC mechanism. All of them are HSP blazars and their names and redshifts are provided in Table 6.

Particle energy distribution: The histograms of the spectral indices of the electron energy distribution are shown in Figure 3 (panels a and b). The low-energy index ($p$) distribution peaks at 1.90 and when fitted with a Gaussian function, the dispersion is $\sigma=0.20$. On the other hand, the high-energy index ($q$) has a rather broad range. The distribution peaks at $\langle q\rangle=4.30$ with $\sigma=0.49$ and its tail extends up to $q\sim 6$. This is primarily due to the inclusion of $\gamma$-ray undetected blazars whose falling part of the inverse Compton spectrum needs to be steep to remain below the Fermi-LAT detection threshold. Additionally, the steep index can also be due to a few HSP blazars with falling BAT spectrum. The distribution of the break energy $\gamma_{\rm b}$, which is derived from the synchrotron peak frequency location, peaks at 100 and tails to very large values, $\gtrsim$1000. Very large $\gamma_{\rm b}$ is probably due to inefficient cooling of relativistic electrons in HSP blazars and overall the distribution has a broad width ($\sigma=0.55$, Figure 3, panel c).

Magnetic field: The magnetic field and Compton dominance666Compton dominance or CD is defined as the ratio of the inverse Compton to synchrotron peak luminosities. distributions are shown in panels (d) and (e) of Figure 3. There is a hint of bi-modality that can be understood keeping in mind that a major fraction of BAT blazars studied here are FSRQs and remaining are mostly HSP blazars with synchrotron dominated SEDs. FSRQs are known to have Compton dominated (CD$>$1) SEDs with a larger magnetic field compared to HSP blazars (see, e.g., Tavecchio et al., 2010; Paliya et al., 2017). Therefore, the distributions are skewed towards larger CD and $B$ with $\langle\log{\rm CD}\rangle=0.79$ and $\langle B\rangle=1.10$ Gauss, respectively.

Bulk Lorentz factor and Doppler factor: In panels (f) and (g) of Figure 3, we show the histograms of the bulk Lorentz factor ($\Gamma$) and Doppler factor ($\delta$). Both of them have narrow distributions peaking at $\langle\Gamma\rangle=11$ and $\langle\delta\rangle=16.5$. These values are in agreement with that found from radio studies (e.g., Savolainen et al., 2010), from the SED modeling of a large sample of $\gamma$-ray emitting blazars (Paliya et al., 2017), and also from luminosity function studies (Ajello et al., 2012).

Dissipation distance: The distribution of the location of the emission region peaks at $R_{\rm d}\sim 3\times 10^{17}$ cm with a majority of sources having $R_{\rm d}>10^{17}$ cm (see Figure 3, panel h). To understand the emission region location with respect to BLR, we plot $R_{\rm d}/R_{\rm BLR}$ as a function of $L_{\rm d}$ in Eddington units in Figure 4. The $R_{\rm BLR}$ is estimated following Ghisellini & Tavecchio (2009), i.e., $R_{\rm BLR}=10^{17}\,(L_{\rm d}/10^{45}\,{\rm erg\,s}^{-1})^{1/2}$ cm. The location is outside BLR, i.e., above the horizontal line, mostly for blazars with less luminous accretion disks which is typically the case of BL Lacs. Moreover, a majority of sources have $R_{\rm d}$ comparable to or slightly larger than $R_{\rm BLR}$, thus indicating the jet environment surrounding the emission region to be transparent enough for $\gamma$-rays to avoid absorption via pair production with BLR photons (see also, Böttcher & Els, 2016; Paliya et al., 2018, for quantitative discussions). Blazars with most luminous accretion disks, i.e. with the largest BLR, tend to have emission region well within it. These are the objects with the largest Compton dominance.

Central engine: We have derived $M_{\rm bh}$ in 92 objects using the disk modeling approach which show the signature of the accretion disk emission at optical-UV energies. In 20 sources, we use the optical spectroscopic $M_{\rm bh}$ since their optical-UV SED is synchrotron dominated. The multiwavelength SEDs of 28 HSP blazars (Table 6) are well reproduced with a synchrotron-SSC model and thus no $M_{\rm bh}$ value was estimated/used. Remaining 6 sources neither exhibit the big blue bump nor have the spectroscopic $M_{\rm bh}$. In these objects, we suitably assume a $M_{\rm bh}$ value based on the available observations. The different methods adopted to compute $M_{\rm bh}$ are tabulated in Table 4.

The left and middle panels of Figure 5 represent the $M_{\rm bh}$ and $L_{\rm d}$ histograms for BAT blazars. The $M_{\rm bh}$ distribution has a rather narrow range ($\sim 10^{8}-10^{10}$ M*⊙) and peaks at $\langle\log M_{\rm bh}\rangle=9.0$ M⊙. Fitting it with a log-normal function returns a width of $\sigma=0.44$. On the other hand, $L_{\rm d}$ peaks at 1046* erg s*-1* and has a rather broad range with $\sigma=1.14$.

In our sample, there are a total of 82 blazars that have $M_{\rm bh}$ measurements from both the disk-fitting and optical spectroscopy methods. In the right panel of Figure 5, we compare $M_{\rm bh}$ derived using these two approaches. As can be seen, both methods provide reasonably similar $M_{\rm bh}$ for BAT blazars and thus confirm the findings earlier reported for Fermi-LAT detected sources (Paliya et al., 2017).

6 Comparison with Fermi Blazars

It may be of a great interest to compare the accretion and jet powers of BAT blazars with that detected by Fermi-LAT sources. The powers that the jet carries in the form of the bulk motion of electrons ($P_{\rm e}$), ‘cold’ protons ($P_{\rm p}$, assuming one proton per electron, i.e. no pairs), radiation ($P_{\rm r}$), and magnetic field ($P_{\rm m}$) are calculated following Celotti & Ghisellini (2008). The derived jet powers, assuming a two-sided jet, are reported in Table 5. For a comparison, we use the results reported for a large sample of Fermi blazars in Ghisellini et al. (2014) and Paliya et al. (2017).

$P_{\rm r}$* versus $L_{\rm d}$:* The most robust estimate of the jet power that we can have is the one that the jet produces in the form of radiation. This is because $P_{\rm r}$ is directly proportional to the observed bolometric luminosity. In the left panel of Figure 6, we show the behavior of $P_{\rm r}$ as a function of $L_{\rm d}$. The low $P_{\rm r}$-$L_{\rm d}$ end is mostly occupied by low-power HSP sources, whereas, luminous BAT FSRQs dominate at the high-end. We compute the partial Spearmann’s coefficient $\rho$ (Padovani, 1992) and PNC to quantify the strength of the correlation and derive the parameters independent of the common redshift effect. The derived coefficients are $\rho=0.40\pm 0.08$ with PNC $<$10*-10* and $\rho=0.33\pm 0.18$ with PNC $<$10*-10* for Fermi-LAT and Swift-BAT blazar populations, respectively. A positive $P_{\rm r}-L_{\rm d}$ correlation have earlier been reported for Fermi-LAT detected sources (e.g., Ghisellini et al., 2014; Paliya et al., 2017), and our findings establishes the fact that Swift-BAT blazars also follow the same positive trend.

$P_{\rm j}$* versus $L_{\rm d}$:* In the middle panel of Figure 6, we show the total jet power ($P_{\rm j}=P_{\rm p}+P_{\rm e}+P_{\rm m}$) versus $L_{\rm d}$. A positive correlation is observed which is statistically confirmed with $\rho=0.53\pm 0.08$ with PNC $<$10*-10* and $\rho=0.69\pm 0.05$ with PNC $<$10*-10* for Fermi-LAT and Swift-BAT blazars, respectively. Similar to LAT blazars, BAT detected ones too host jets with $P_{\rm j}>L_{\rm d}$. Interestingly, at the high-end of the jet power ($P_{\rm j}>10^{48}$ erg s*-1*), BAT blazars seems to have more extreme jets compared to LAT blazars. This is probably a selection effect due to higher detection threshold of Swift-BAT which leads to the identification of the most luminous sources. In a more physical scenario, the observational shift of the SED peaks to lower frequencies with increase in the bolometric luminosity (e.g., Fossati et al., 1998) can also explain this observation. BAT blazars that likely have inverse Compton peak located at lower energies with respect to LAT detected ones, host more powerful jets and luminous accretion disks.

7 Luminosity Dependent Evolution

It is instructive to determine the average SEDs of BAT blazars with the motivation to explore their luminosity dependent evolution (Ajello et al., 2012). We divide the sources in different BAT luminosity ($L_{\rm BAT}$) bins and also based on their SED appearances to identify FSRQs/BL Lacs. The latter is necessary since 105-month BAT catalog does not provide FSRQ/BL Lac classification. Moreover, we consider only those sources that have both Fermi-LAT and Swift-BAT detections. We take the arithmetic average of the broadband SEDs in each $L_{\rm BAT}$ bin and show in Figure 7. Note that the adopted choice of $L_{\rm BAT}$ bins are mainly driven to distinguish BAT blazars of various powers and does not carry any other physical meaning.

In the left panel, we show averaged SEDs for FSRQs in four $L_{\rm BAT}$ bins and note the increasing dominance of inverse Compton peak (i.e. more Compton dominated SED) with increasing $L_{\rm BAT}$ and $L_{\rm d}$. Also, there is an indication for the shift of the inverse Compton peak to lower frequencies as their power increases, especially in the highest $L_{\rm BAT}$ bin. In the right hand panel of Figure 7, we plot the SEDs for BL Lacs. There are no BL Lacs in our sample with $L_{\rm BAT}>10^{47}$ erg s*-1* and only one (J0428.6$-$3788 or PKS 0426$-$380, $z=1.11$) has $\log L_{\rm BAT}$ in the bin $46.5-47$ and hence not plotted. The lowest $L_{\rm BAT}$ bin is dominated by HSP BL Lacs and a pronounced shift of the SED peaks to lower frequencies can be observed in the higher luminosity bin. These results are aligned with our current understanding about the blazar sequence where the most powerful objects are LSP FSRQs with Compton dominated SEDs and the luminosity dependent SED evolution is more pronounced in BL Lacs (see Ghisellini et al., 2017, for latest results).

8 Summary

In this work, we study 146 blazars detected in the first 105-month all-sky survey of Swift-BAT. We analyze the Fermi-LAT data and 14$-$195 keV BAT spectra and supplement them with multiwavelength archival spectral measurements. We determine the physical properties of the jet and central engine by modeling the broadband SED with a simple leptonic emission model. Our main findings are summarized below.

We $\sim$quadruple the number of Swift-BAT detected blazars and also number of Fermi-LAT undetected ones compared to earlier studies (146 versus 38, 45 versus 12, respectively, Ajello et al., 2009). 2. 2.

In the photon index versus 14$-$195 keV luminosity plane, we find tentative evidences ($\rho=-0.29\pm 0.08$, PNC = 0.01) of inverse correlation with more luminous blazars tend to have flatter spectrum. 3. 3.

The overall physical properties of BAT blazars resemble well with that known for LAT blazar population studied in Paliya et al. (2017). 4. 4.

The mass of the central black hole derived from the single-epoch optical spectroscopy matches reasonably well with that derived by modeling the optical-UV bump with a standard accretion disk. 5. 5.

BAT blazars exhibit a positive correlation between the power that the jet produces in the form of radiation and $L_{\rm d}$. Blazars with most luminous accretion disks host the most radiatively powerful jets. 6. 6.

Comparing the averaged SEDs of BAT blazars binned in different 14$-$195 keV luminosity bins, we find that more luminous sources have more Compton dominated SED. There is also a hint of the shift of SED peaks to lower frequencies as the power of the blazars increases and it is found to be more pronounced in low-luminosity BL Lac objects.

We are grateful to the journal referee for a constructive criticism. M.K. acknowledges support from NASA through ADAP award NNH16CT03C. CR acknowledges support from the CONICYT+PAI Convocatoria Nacional subvencion a instalacion en la academia convocatoria año 2017 PAI77170080. K.O. is an International Research Fellow of the Japan Society for the Promotion of Science (JSPS, ID: P17321). We are grateful to the Fermi-LAT Collaboration internal referee Raniere Menezes for useful suggestions and Alberto Domínguez, Judith Recusin and Philippe Bruel for critical reading of the manuscript. The Fermi LAT Collaboration acknowledges generous ongoing support from a number of agencies and institutes that have supported both the development and the operation of the LAT as well as scientific data analysis. These include the National Aeronautics and Space Administration and the Department of Energy in the United States, the Commissariat à l’Energie Atomique and the Centre National de la Recherche Scientifique / Institut National de Physique Nucléaire et de Physique des Particules in France, the Agenzia Spaziale Italiana and the Istituto Nazionale di Fisica Nucleare in Italy, the Ministry of Education, Culture, Sports, Science and Technology (MEXT), High Energy Accelerator Research Organization (KEK) and Japan Aerospace Exploration Agency (JAXA) in Japan, and the K. A. Wallenberg Foundation, the Swedish Research Council and the Swedish National Space Board in Sweden. Additional support for science analysis during the operations phase is gratefully acknowledged from the Istituto Nazionale di Astrofisica in Italy and the Centre National d’Études Spatiales in France. This work performed in part under DOE Contract DE-AC02-76SF00515. We acknowledge the work of the Swift-BAT team to make this study possible. This research has made use of data obtained through the High Energy Astrophysics Science Archive Research Center Online Service, provided by the NASA/Goddard Space Flight Center. This research has made use of the NASA/IPAC Extragalactic Database (NED), which is operated by the Jet Propulsion Laboratory, California Institute of Technology, under contract with the National Aeronautics and Space Administration. Part of this work is based on archival data, software or online services provided by the ASI Data Center (ASDC). .