1: %\documentclass{aastex}
2: %\documentstyle[11pt,aastex,epsf,flushrt]{article}
3: %\documentclass[12pt,preprint,fleqn]{aastex}
4: %\documentclass[12pt,preprint,epsf]{aastex}
5: %\documentclass[preprint2]{aastex}
6: %%\documentclass{emulateapj}
7: \documentclass[onecolumn]{emulateapj}
8: %\usepackage{color}
9:
10:
11: \newcommand{\myemail}{charles.dermer@nrl.navy.mil}
12:
13: \newcommand{\G}{\Gamma}
14: %\newcommand{\eo}{\epsilon_{obs}}
15: \newcommand{\bG}{\beta_\Gamma}
16: \newcommand{\sT}{\sigma_{\rm T}}
17: \newcommand{\pr}{^\prime}
18: \newcommand{\e}{\epsilon}
19: \newcommand{\g}{\gamma}
20: \newcommand{\gp}{\gamma^{\,\prime}}
21: \newcommand{\gptwo}{\gamma^2}
22: \newcommand{\Dp}{\Delta^\prime}
23: \newcommand{\el}{\ell_{\rm S}}
24: \newcommand{\pp}{p^\prime}
25: \newcommand{\Vp}{V^\prime}
26: \newcommand{\Np}{N^\prime}
27: \newcommand{\kB} {k_{\rm B}}
28: \newcommand{\bcm}{\beta_{cm}}
29: \newcommand{\gcm}{\g_{cm}}
30: %\newcommand{\bcm}{\beta_{cm}}
31: %\newcommand{\gcm}{\g_{cm}}
32: \newcommand{\xD}{x_\Delta}
33: \newcommand{\Op}{\Omega^\prime}
34: \newcommand{\mup}{\mu^\prime}
35: \newcommand{\gpp}{\gamma_p^\prime}
36: \newcommand{\tgg}{\tau_{\g\g}}
37: \newcommand{\tp}{t^\prime}
38: \newcommand{\ep}{\epsilon^\prime}
39: \newcommand{\Dop}{\delta_{\rm D}}
40: \newcommand{\Ep}{E^\prime}
41: \newcommand{\vrm }{{\rm v}}
42: \newcommand{\dD}{\delta_{\rm D}}
43: %\newcommand{\up}(u^\prime}
44: \newcommand{\ti}{\theta_i}
45: \newcommand{\tcl}{\theta_{cl}}
46: \newcommand{\psim}{\lower.5ex\hbox{$\; \buildrel \propto \over\sim \;$}}
47: \newcommand{\lbar}{\lower.0ex\hbox{$\; \buildrel {\lower0.0ex \hbox{-}} \over\lambda \;$}}
48: \newcommand{\es}{\epsilon_{*}}
49: \newcommand{\Os}{\Omega_{*}}
50:
51: %\slugcomment{Submitted to The Astrophysical Journal 2006}
52:
53: \shorttitle{$\gamma$-ray FSRQ Blazar Analysis}
54: \shortauthors{Dermer, Finke, Krug, and B\"ottcher}
55:
56: \begin{document}
57: \title {Gamma-Ray Studies of Blazars: Synchro-Compton Analysis of
58: Flat Spectrum Radio Quasars}
59:
60:
61: \author{Charles D.\ Dermer,$^1$ Justin D. Finke,$^{1,2}$ Hannah Krug,$^{1,3}$ \& Markus B\"ottcher$^4$}
62: \affil{$^1$U.S.\ Naval Research Laboratory, Code 7653, 4555 Overlook SW, Washington, DC
63: 20375-5352\\
64: $^2$Naval Research Laboratory/National Research Council Postdoctoral Associate\\
65: $^3$Department of Astronomy, University of Maryland, College Park, MD 20742 \\
66: $^4$Astrophysical Institute, Department of Physics and Astronomy,
67: Ohio University, Athens, Ohio 45701
68: }
69: \email{charles.dermer@nrl.navy.mil}
70:
71: \begin{abstract}
72:
73: We extend a method for modeling synchrotron and synchrotron
74: self-Compton radiations
75: in blazar jets to include external Compton processes.
76: The basic model assumption is that the blazar
77: radio through soft X-ray flux is nonthermal synchrotron radiation
78: emitted by isotropically-distributed electrons in the randomly
79: directed magnetic field of outflowing relativistic blazar jet plasma.
80: Thus the electron distribution is given by the synchrotron spectrum, depending only
81: on the Doppler factor $\delta_{\rm D}$ and mean magnetic field $B$,
82: given that the comoving emission
83: region size scale $R_b^\prime \lesssim c \dD t_v/(1+z)$, where $t_v$ is
84: variability time and $z$ is
85: source redshift. Generalizing the approach of Georganopoulos, Kirk,
86: and Mastichiadis (2001) to arbitrary anisotropic target radiation
87: fields, we use the electron spectrum implied by the synchrotron
88: component to derive accurate Compton-scattered $\gamma$-ray spectra
89: throughout the Thomson and Klein-Nishina regimes for external Compton
90: scattering processes. We derive and calculate
91: accurate $\gamma$-ray spectra produced by
92: relativistic electrons that Compton-scatter (i) a point source of
93: radiation located radially behind the jet, (ii) photons from
94: a thermal Shakura-Sunyaev accretion disk
95: and (iii) target photons from the central source
96: scattered by a spherically-symmetric shell of broad
97: line region (BLR) gas. Calculations of broadband spectral
98: energy distributions from the radio through $\gamma$-ray regimes are presented,
99: which include self-consistent $\gamma\gamma$ absorption on the same
100: radiation fields that provide target photons for Compton scattering.
101: Application of this baseline flat spectrum
102: radio/$\gamma$-ray quasar model is considered in view of
103: data from $\gamma$-ray telescopes and
104: contemporaneous multi-wavelength campaigns.
105:
106: \end{abstract}
107:
108: \keywords{radiation processes: nonthermal --- galaxies: active --- supermassive black holes}
109:
110: \section{Introduction}
111:
112: A class of radio-loud blazar active galactic nuclei (AGNs)
113: that emit luminous fluxes of $\gtrsim
114: 100$ MeV -- GeV $\gamma$ rays was discovered with the Energetic
115: Gamma Ray Experiment Telescope (EGRET) on the {\em
116: Compton Observatory} \citep{har92,fic94,har99}. This result clarified
117: the nature of 3C 273, which was first identified as a $\gamma$-ray
118: emitting AGN in COS-B satellite data \citep{her77}. The $\gamma$ rays
119: from blazars are certainly nonthermal in origin and associated
120: with the radio jets formed by the supermassive black holes that power
121: these sources. The largest subclass of EGRET AGNs are moderate redshift ($z\approx
122: 1$) flat-spectrum radio quasars (FSRQs) with blazar properties,
123: including apparent superluminal motion, rapidly variable optical
124: emission, high polarization, and intense broadened optical emission
125: lines. Another subclass of the EGRET AGNs consists of $\gamma$-ray
126: emitting BL Lacertae (BL Lac) objects, which are generally at lower
127: redshifts ($z \approx 0.1$ -- $0.3$) and, by definition, have weak or absent
128: optical emission lines in their spectra. The X-ray selected BL Lac
129: (XBL) subset were discovered to be a class of TeV $\gamma$-ray sources
130: by the Whipple Observatory \citep{pun92,wee03}.
131:
132: The common superluminal nature of the first identified $\gamma$-ray
133: blazars, namely 3C 273, 3C 279, and PKS 0528+134, led \citet{dsm92} to
134: propose a Compton-scattering origin for the $\gamma$ rays. In this
135: model, jet electrons Compton-scatter accretion-disk photons that
136: intercept the jet plasma. The nonthermal jet electrons can also
137: scatter internal synchrotron photons to produce a synchrotron
138: self-Compton (SSC) component \citep{bm96}. Given the broadened
139: emission lines in the spectra of FSRQs, accretion-disk radiation
140: scattered by surrounding gas of the broad line region (BLR) will
141: provide a further source of target photons to be scattered to
142: $\gamma$-ray energies \citep{sbr94}, as will radiation from a
143: surrounding dusty torus \citep{kat99,bsmm00}. The accretion-disk and
144: scattered radiation will attenuate jet $\gamma$-rays through
145: $\gamma\gamma$ pair-production attenuation \citep{bk95,bl95}.
146:
147: Expressions for the $\gamma$-ray spectral energy distributions (SEDs)
148: of blazars produced by Compton scattering processes have been derived
149: and calculated for many specific models of the black-hole/blazar jet
150: environment. In the case of external accretion-disk photons as the
151: target photon source, where the accretion disk is described by an
152: optically thick, geometrically thin thermal \cite{ss73} accretion
153: disk, Compton-scattered $\gamma$-ray spectra were calculated in the
154: Thomson regime by \citet{ds93,ds02}. Calculations of the
155: Thomson-scattered spectra for a quasi-isotropic target radiation field
156: formed by BLR gas or hot dust were made by \citet{sbr94}, \cite{dss97}
157: and \citet{bsmm00}. Detailed numerical calculations including both
158: accretion disk and scattered radiation fields, have been made by, e.g.,
159: \citet{kt05,bb00} and \citet{br04}.
160:
161: Compton scattering in the Klein-Nishina regime is not so simple to
162: treat compared to analyses restricted to the Thomson regime, but is
163: unavoidable for blazar analysis in the era of the {\em Fermi
164: Gamma-Ray Space Telescope} (formerly known as {\em GLAST}) and the
165: ground-based $\gamma$ Cherenkov telescopes. For surrounding isotropic
166: radiation fields in the stationary frame of the blazar AGN,
167: \citet{gkm01} suggested to transform the comoving electron
168: distribution to the stationary frame and then scatter the target
169: photons to $\gamma$-ray energies, using the formula first derived by
170: Jones \citep{jon68,bg70}. This approach is generalized in this paper
171: to surrounding anisotropic radiation fields.
172:
173: The usual spectral modeling approach proceeds by injecting power-law
174: electrons and evolving these particles while they produce the output
175: synchrotron and Compton-scattered radiation \citep[e.g.,][]{ds93,bms97}.
176: For example, \citet{mod05}
177: calculate electron energy evolution and spectral formation
178: throughout the Thomson and Klein-Nishina regimes for different
179: ratios of synchrotron and isotropic radiation field energy
180: densities. They show that reduced Compton losses
181: in the Klein-Nishina regime compared to synchrotron losses
182: can lead to spectral hardening of the synchrotron component
183: in the optical/X-ray regime \citep[noted earlier by][]{da02}.
184: %, as originally proposed to explain
185: %spectral hardenings seen in X-ray knots of radio galaxies with Chandra
186: A difficulty in this
187: approach is that the electron energy-loss rate depends on the photon
188: spectrum of the comoving radiation field, not just the total radiation
189: energy density, and this field evolves with time.
190: The modeler is faced with the prospect of
191: simultaneously fitting the synchrotron and Compton components. The
192: acceleration scenario may well be over-simplified, and non-power-law
193: particle injection distributions could be more realistic than
194: power-law injection spectra, e.g., due to nonlinear effects in Fermi
195: acceleration. Moreover, a separation between the acceleration and
196: radiation zones may not be justified.
197:
198: Here we extend a method of blazar analysis recently proposed for TeV
199: blazars \citep{fdb08} that avoids these difficulties. For a standard
200: $\gamma$-ray blazar model, where isotropically distributed electrons
201: spiral in a randomly oriented magnetic field with mean magnetic field
202: strength $B$ in the fluid frame, the measured synchrotron flux
203: directly reveals the electron spectrum responsible for the synchrotron
204: radiation. The only uncertainties are the mean magnetic field $B$, the
205: comoving size scale $R_b^\prime$ of the emitting region, and the
206: Doppler factor $\dD = [\Gamma (1-\beta\cos\theta)]^{-1}$ ($\G =
207: 1/\sqrt{1-\beta^2}$ is the bulk Lorentz factor of the outflow). With
208: this electron spectrum, we then Compton-scatter target photons of the
209: surrounding radiation fields using the head-on approximation to the
210: total Compton cross section \citep{ds93}, valid when the electron
211: Lorentz factor $\gamma \gg 1$. This generalizes the approach of
212: \citet{gkm01} to surrounding anisotropic radiation fields. The
213: temporally-evolving electron spectrum in blazars can be derived in
214: this approach from simultaneous multiwavelength blazar data. Values of
215: $B$, $\dD$, and jet power can then be deduced. The related treatment
216: for XBLs applied to PKS 2155-304, including more details about the
217: derivation of the electron spectrum from the synchrotron component,
218: the derivation and calculations of the SSC component and internal
219: $\gamma$-ray opacity by the synchrotron photons, is given by
220: \citet{fdb08}.
221:
222: Analysis of blazar SEDs using this approach is presented in Section 2,
223: where formulas to calculate Compton-scattered internal and external
224: radiation and a $\delta$-function approximation for $\gamma\gamma$
225: opacity from the internal radiation field are given. Derivations of
226: the Compton-scattered spectrum for specific examples of external
227: radiation fields consisting of a monochromatic point source of
228: radiation radially behind the jet, a Shakura-Sunyaev disk model, and a
229: model BLR radiation field are derived in Section
230: 3. Discussion of the results is found in Section 4.
231: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
232: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
233:
234:
235: \section{One-Zone Synchrotron/Synchrotron Self-Compton Model with
236: $\gamma\gamma$ Opacity}
237:
238: We consider a one-zone model for blazar flares. Multiple zones could still be
239: allowed, but the product of the duty cycle and number of
240: zones would have to be small enough that interference of
241: emissions from the different regions would still permit rapid variability.
242: In this case, the emission would still predominantly arise from a
243: single zone. Distinct zones could also emit the bulk of
244: their radiation in different wavebands. In this case, the cospatiality assumption
245: often made in blazar modeling would not apply.
246: In this regard, correlated variability data is essential to
247: test the underlying assumptions made when a one-zone model
248: is employed. Slowly varying radio/IR
249: synchrotron and hard X-ray and low-energy $\gamma$-ray Compton
250: emissions could involve extended emission regions.
251:
252: A radiative event from the source emission region that varies on a
253: comoving timescale $\tp_v\gtrsim R^\prime_b/c$ is related to the
254: observed variability timescale through the relation $\tp_v = \dD
255: t_v/(1+z)$, where $z$ is redshift; thus the comoving blob radius is
256: $R^\prime_b\lesssim c\dD t_v/(1+z)$. The inequality allows us to
257: neglect light-travel time effects from different parts of the emitting
258: volume and avoid integrations over source volume. Within this zone,
259: the nonthermal electrons with isotropic pitch angle distribution
260: are described by the total comoving electron
261: number spectrum $N_e^\prime(\gp )$ in terms of comoving electron
262: Lorentz factor $\gp $. The
263: magnetic field is assumed to be randomly oriented in the comoving
264: fluid frame. The relativistic electrons that gyrate in this field
265: radiate nonthermal synchrotron radiation, observed as the low-energy
266: component in blazar SEDs.
267:
268: \subsection{Synchrotron and Self-Compton Components}
269:
270: The $\nu F_\nu$ synchrotron radiation spectrum can
271: be approximated by the expression
272: \begin{equation}
273: f_\e^{syn} \cong {\dD^4\over 6\pi d_L^2} \; c\sigma_{\rm T} U_B
274: \gamma_s^{\prime 3} N_e^\prime (\gp_s)\;,\;
275: \label{fes}
276: \end{equation}
277: where
278: \begin{equation}
279: \gp_s = \sqrt{{\e(1+z)\over \dD\epsilon_B}}= \sqrt{{\ep \over \epsilon_B}}\;,
280: \label{gps}
281: \end{equation}
282: $d_L=d_L(z)$ is the luminosity distance, $c$ is the speed of
283: light, $\sigma_T$ is the Thomson cross-section, $z$ is the source
284: redshift, and the comoving magnetic-field energy density of the
285: randomly-oriented comoving field with comoving mean intensity
286: $B$ is $U_B \equiv B^2/8\pi\;.$ We use $\epsilon$ and
287: $\epsilon^\prime$ to refer to the dimensionless photon energy in the
288: observer and comoving frame, respectively. Here and throughout this
289: paper, unprimed quantities refer to the observer's frame, and primed
290: quantities refer to the frame comoving with the AGN's jet, with the
291: exception being $B$, the comoving magnetic field. Inverting this
292: expression gives the comoving electron distribution
293: \begin{equation}
294: N_e^\prime(\gp_s ) = V_b^\prime n_e(\gp_s )
295: \cong {6\pi d_L^2 f_{\e_{syn}}^{syn}\over c\sigma_{\rm T} U_B \dD^4\gamma_s^{\prime 3}}
296: \;,\;\;
297: \label{Neprimegps}
298: \end{equation}
299: where
300: \begin{equation}
301: \e_{syn} = {\dD\e_B\g_s^{\prime 2}\over 1+z}\;,
302: \end{equation}
303: $\e_B = B/B_{cr}$ is the ratio of $B$ and the critical magnetic field
304: $B_{cr} = m_e^2 c^3/e\hbar \cong 4.41\times 10^{13}$ G \citep{ds02},
305: and $V_b^\prime=4\pi R^{\prime 3}_b/3$ is the comoving
306: volume of the blob. Note that $U_B = \e_B^2 U_{B_{cr}} =\e_B^2
307: B_{cr}^2/8\pi$. Eq. (\ref{Neprimegps}) gives a good representation to
308: the source electron distribution when the $\nu F_\nu$ spectral index
309: $a < 4/3$ (i.e., for spectra softer than $a = 4/3$, adopting the
310: convention $f_\e\propto \e^a$) and away from the high-energy
311: exponential cutoff of the spectrum \citep[see][for comparison]{fdb08}.
312:
313: %The $\delta$-function approximation for the synchrotron
314: %photon emissivity $\dot n_{syn}(\e )$ (ph cm$^{-3}$ s$^{-1}$
315: %$\e^{-1}$) is given by
316: %\begin{equation}
317: %\dot n^{\prime }_{syn}(\ep) \cong {2\over 3}\; { c \sigma_{\rm T} u_B }
318: %\; \e^{\prime -1/2} \epsilon_B^{-3/2}\; n^\prime_e\big(\gp_s\big)\;\cong\;
319: %{2c\sigma_{\rm T}U_{B_{cr}}\over 3V_b^\prime m_ec^2}\;{N^\prime_e(\gp_s )
320: %\over \gp_s}\;,
321: %\label{dotnsyne}
322: %\end{equation}
323: %where $\gp_s \equiv \sqrt{\ep/ \e_B}$ and $u_B \equiv U_B/m_ec^2$.
324: %The target synchrotron radiation field is therefore given by
325: %\begin{equation}
326: %n^\prime_{syn}(\ep ) = {u^\prime(\ep)\over m_ec^2 \ep}\cong {R_b^\prime\over c} \; \dot n_{syn}^\prime( \ep )\;\cong
327: % {3d_L^2 f_\e^{syn}\over
328: %m_ec^3R^{\prime 2}_b \dD^4 \e_B^2 \g^{\prime 4}_s}\;.
329: %\label{nprimesynep}
330: %\end{equation}
331:
332: The SSC $\nu F_\nu$ flux is given by
333: \begin{equation}
334: f_\e^{SSC} \; = \;{\dD^4\over d_L^2}\; \ep_s L^\prime_{SSC}(\ep_s,\Op_s )\;.
335: \label{feSSC}
336: \end{equation}
337: The formula of \citet{jon68} (see also \citet{bg70}) gives the
338: SSC $\nu F_\nu$ flux,
339: \begin{equation}
340: f_\e^{SSC} = {3\over 4} c \sigma_{\rm T} \e^{\prime 2}_s\;{\dD^4\over 4\pi d_L^2}\;
341: \int_0^\infty d\ep\;{u^\prime (\ep )\over \e^{\prime 2}}
342: \int_{\gp_{min}}^{\gp_{max}}d\gp\;{N_e^\prime(\gp )\over \g^{\prime 2}}\;F_{\rm C} (q^\prime,\Gamma^\prime_e)\;,
343: \label{epsjssc}
344: \end{equation}
345: where
346: \begin{equation}
347: F_{\rm C}(q^\prime,\Gamma^\prime_e
348: )= \left[ 2q^\prime \ln q^\prime +(1+2q^\prime)(1-q^\prime) +{1\over 2}
349: {(\Gamma^\prime_e q^\prime)^2\over (1+\Gamma^\prime_e
350: q^\prime)}(1-q^\prime) \right]
351: \;H\;\left( q^\prime; {1\over 4\gamma^{\prime 2}},1\right),
352: \label{fcq}
353: \end{equation}
354: \begin{equation}
355: q^\prime \equiv {\ep_s/\gp \over \Gamma^\prime_e
356: (1-\ep_s/\gp )}\;,\;{\rm and}\; \; \Gamma^\prime_e = 4\ep\gp\;.
357: \label{jesCq}
358: \end{equation}
359: The synchrotron photons provide a target radiation field with
360: spectral energy density
361: \begin{equation}
362: u^\prime (\ep ) = \epsilon^{\prime}m_ec^2 n^\prime_{syn}(\ep ) =
363: \frac{ 3 d_L^2 f_\e^{syn} }{ cR_B^{\prime 2}\dD^4\ep }\;,
364: \label{uprime}
365: \end{equation}
366: using eq.\ (\ref{gps}).
367: The scattered photon energy in the comoving frame is related to the observed photon
368: energy $h\nu = m_ec^2 \e$ by the relation
369: \begin{equation}
370: \ep_s \;=\; {(1+z)\e\over \dD }\; \equiv\; {\e_s\over \dD}\;.
371: \label{eps}
372: \end{equation}
373: From the limits on the integration over $\gp $ implied by the limits on $q^\prime$ we find
374: \begin{equation}
375: \gp_{min} = {1\over 2} \ep_s\;\left( 1+\sqrt{1+{1\over \ep\ep_s}} \;\right)
376: \label{gpmin}
377: \end{equation}
378: and
379: \begin{equation}
380: \gp_{max} = {\ep\ep_s\over \ep - \ep_s}H(\ep - \ep_s) \;+\; \gp_2H(\ep_s - \ep )\;
381: \label{gpmax}
382: \end{equation}
383: \citep[see][for a detailed derivation of synchrotron/SSC models
384: and application to blazars]{fdb08}.
385: Here the maximum lepton Lorentz factor injected
386: into the radiating fluid is $\gp_2$, and the Heaviside
387: function $H(x;a,b)$ is defined such that $H(x;a,b) = 1$ when $a\leq x \leq b$, and
388: $H(x;a,b) = 0$ otherwise; the Heaviside function
389: with one entry is defined such that $H(x) = 1$ when $x\geq 0$, and $H(x) = 0$
390: otherwise.
391: The $\nu F_\nu$ SSC spectrum is therefore given by
392: \begin{equation}
393: f_\e^{SSC} = \left( {3\over 2}\right)^3\; {d_L^2 \e_s^{\prime 2}
394: \over
395: R_b^{\prime 2} c \dD^4 U_B}\;\int_0^\infty d\ep\;{f_{\tilde \e}^{syn}\over \e^{\prime 3}}\;
396: \int_{\gp_{min}}^{\gp_{max}}d\gp\;{F_{\rm C}(q^\prime,\Gamma^\prime_e)f_{\hat \e}^{syn}\over \g^{\prime 5} }\;,
397: \label{feSSC1}
398: \end{equation}
399: where $\tilde \e \equiv {\dD \ep/ (1+z)}$ and
400: $\hat \e \equiv {\dD \e_B \g^{\prime 2}/ (1+z)}$.
401: The maximum $\nu F_\nu$ SSC flux at photon energy
402: $\e^{SSC}_{pk}$ can be approximated in the Thomson limit by the expression
403: \begin{equation}
404: \label{fssc}
405: f_{\e_{pk}^{SSC}}^{SSC} \simeq \frac{24\pi d_L^2 (1+z)^2}{(\dD^3Bt_v)^2 c^3}\
406: \left( f_{\e_{pk}^{syn}}^{syn} \right)^2\;,
407: \end{equation}
408: where
409: the peak frequencies are related by
410: \begin{equation}
411: \e_{pk}^{syn} \;\cong \; \sqrt{ {\e_{pk}^{SSC}\e_B\dD\over 1+z}}\;
412: \label{essc}
413: \end{equation}
414: \citep{tav98,fdb08}.
415: Here $f_{\e_{pk}^{syn}}^{syn}$ is the $\nu F_\nu$ peak of the synchrotron
416: component, which reaches its maximum at $\e = \e_{pk}^{syn}$.
417:
418: Second-order SSC takes place when the SSC photons are again Compton
419: scattered by electrons in the same blob, and may account for
420: superquadratic variability of the $\gamma$-ray flux with respect to
421: the synchrotron flux \citep{per08}. In principle, these photons can
422: again be Compton scattered to arbitrarily higher orders, though
423: higher-order scatterings are negligible due to Klein-Nishina effects.
424: Calculating second-order SSC can be done accurately by replacing
425:
426: $f_{\tilde \e}^{syn}$ with $f_{\tilde \e}^{SSC}$ in eq. (\ref{feSSC1}), so
427: that
428: \begin{equation}
429: f_\e^{SSC,2} = \left( {3\over 2}\right)^3\; {d_L^2 \e_s^{\prime 2}
430: \over
431: R_b^{\prime 2} c \dD^4 U_B}\;\int_0^\infty d\ep\;{f_{\tilde \e}^{SSC}\over \e^{\prime 3}}\;
432: \int_{\gp_{min}}^{\gp_{max}}d\gp\;{F_{\rm C}(q^\prime,\Gamma^\prime_e)f_{\hat \e}^{syn}\over \g^{\prime 5} }\;.
433: \label{feSSC2ndorder}
434: \end{equation}
435:
436: \subsection{ $\gamma\gamma$ Opacity}
437:
438: Gamma-ray photons are subject
439: to $\g\g$ attenuation by synchrotron photons produced in the
440: radiating plasma, by ambient photons in the environment of
441: the black hole (starred frame), and by
442: photons of the intergalactic radiation field.
443: The $\g\g$ pair-production cross section
444: \begin{equation}
445: \sigma_{\gamma\gamma} (s)={1\over 2}\pi r_e^2 (1-\bcm^2)
446: \left[(3-\bcm^4)\ln\left({1+\bcm\over
447: 1-\bcm}\right) - 2\bcm (2-\bcm^2)\right]\;
448: \label{sigmagg}
449: \end{equation}
450: \citep{jr76,nik61,gs67,bmg73},
451: where $\gcm$ is the center-of-momentum
452: frame Lorentz factor of the produced electron and positron,
453: $\bcm = (1-\gcm^{-2})^{1/2} = \sqrt{1-s^{-1}}$,
454: \begin{equation}
455: s = \gcm^2 = {\e_*\e_1(1+z)\over 2} (1-\cos\psi) \;,
456: \label{scmenergy}
457: \end{equation}
458: and $r_e =e^2/m_ec^2
459: \cong 2.8179\times 10^{-13}$ cm
460: is the classical electron radius. The interaction
461: angle $\psi$, given by the relation
462: \begin{equation}
463: \cos\psi = \mu_* \mu_s + \sqrt{1-\mu_*^2}\sqrt{1-\mu_s^2}\cos(\phi_*-\phi_s)\;,
464: \label{cospsi}
465: \end{equation}
466: is the angle between the directions of the photon detected with energy $\e_1$
467: and the target photon with
468: energy $\e_*$.
469:
470: The absorption probability per unit pathlength is
471: \begin{equation}
472: {d\tau_{\g\g}(\e_1)\over dx} =
473: \oint d\Omega_* \;(1-\cos\psi) \int_0^\infty d\e_*
474: \;n_{ph}(\e_*,\Omega_*)\;\sigma_{\g\g}(s)\;.
475: \label{dtauggdx}
476: \end{equation}
477: For absorption by synchrotron photons
478: within the radiating volume, $\e_* \rightarrow \ep$ and $\e_1\rightarrow \ep_1 =
479: (1+z)\e_1/\dD$, and the target synchrotron
480: radiation field is given by eq.\ (\ref{uprime}). In this case,
481: the optically-thin $\gamma$-ray emission spectrum is modified by
482: the factor $3u(\tgg) / \tgg$ for a spherical geometry,
483: where
484: \begin{equation}
485: u(\tgg) = {1\over 2} + {\exp(-\tgg )\over \tgg} - {1-\exp(-\tgg )\over \tgg^2} \;.
486: \label{utgg}
487: \end{equation}
488: Here $\tau_{\g\g}$ is
489: the total $\g\g$ optical depth integrated over pathlength.
490: For absorption by ambient photons in the vicinity of the AGN, $\e_*$ is
491: the photon energy in the AGN rest frame. For cosmic $\g\g$ absorption, the target photons are given by the
492: spectrum of the intergalactic background light,
493: which evolves with redshift. In the latter two cases, the intrinsic spectrum
494: is modified by the factor $\exp(-\tau_{\g\g})$.
495:
496:
497: \section{Compton-Scattered External Radiation Fields}
498:
499: In the one-zone model, the $\nu F_\nu$ spectrum of Compton-scattered
500: external radiation fields
501: is given by the Compton spectral
502: luminosity $\e_s L_{\rm C}(\e_s,\Omega_s)$ according to the relation
503: \begin{equation}
504: f^{\rm C}_{\epsilon} = {\e_s L_{\rm C}(\e_s,\Omega_s)\over d_L^2}\;,
505: \label{fecompton}
506: \end{equation}
507: where $\e_s \equiv (1+z)\e$, from eq.\ (\ref{eps}), and
508: $\Omega_s = \Omega$. The latter equality
509: means that the photon direction is not deflected
510: in transit to the observer.
511: The Compton spectral luminosity is given by
512: $$\e_s L_{\rm C}(\e_s,\Omega_s )= m_ec^3\e_s^2\oint d\Omega_* \int_0^\infty d\e_* \;n_{ph}(\e_*,\Omega_* )
513: \oint d\Omega_e \int_1^\infty d\g\; N_e(\g ,\Omega_e)\times $$
514: \begin{equation}
515: \;(1-\cos\psi )\;{d\sigma_{\rm C}(\bar\e)\over d\e_s}\;
516: \delta (\Omega_s - \Omega_e )\;,
517: \label{esJesOs}
518: \end{equation}
519: having already introduced the approximation that the scattered photon travels
520: in the same direction as the relativistic scattering electron, i.e.,
521: $\Omega_s = \Omega_e$. Because of this approximation,
522: the cosine of the angle $\psi$ is given by eq.\ (\ref{cospsi}). The invariant collision energy
523: \begin{equation}
524: \bar\e \equiv \gamma \e_* (1-\sqrt{1-1/\g^2}\cos\psi)\cong \gamma \e_* (1-\cos\psi)\;
525: \label{bare}
526: \end{equation}
527: because $\gamma \gg 1$. The relation $n_{ph}(\e_*,\Omega_*) =
528: u(\e_*,\Omega_*)/(m_ec^2\e_*)$ gives the specific spectral number
529: density of target photons with energy $\e_*$, the starred quantities
530: referring to the frame stationary with respect to the black hole.
531:
532: The Compton cross section in the head-on approximation is given by
533: \begin{equation}
534: {d\sigma_{\rm C}\over d\e_s} \;\cong\;
535: {\pi r_e^2 \over \gamma\bar\e}\;\Xi\;H\left(\e_s ;
536: {\bar\e\over 2\gamma}, {2\gamma\bar\e\over 1+2\bar\e}\right)\,
537: \label{dsigKNdedO_1}
538: \end{equation}
539: \citep{ds93,db06},
540: where
541: \begin{equation}
542: \Xi \;\equiv \;
543: y+y^{-1} - {2\e_s\over \gamma \bar\e y} +
544: ({\e_s\over \gamma \bar\e y})^2 \;,
545: \label{Xi}
546: \end{equation}
547: \begin{equation}
548: y \;\equiv\; 1 - {\e_s\over\g} \;,
549: \label{y}
550: \end{equation}
551: $\bar \e$ is given by eq.\ (\ref{bare}).
552: The Compton spectral luminosity in the head-on approximation becomes
553: \begin{equation}
554: \e_s L_{\rm C}(\e_s,\Omega_s) = {c\pi r_e^2 \e_s^2 }\oint d\Omega_*\; \int_0^{\e_{*,hi}}
555: d\e_*\; {u_*(\e_*,\Omega_* )\over \e_*^2}
556: \int_{\g_{low}}^\infty
557: d\g\;\g^{-2} N_e(\g,\Omega_s )\;
558: \Xi\;.
559: \label{esLCesomegas}
560: \end{equation}
561: The lower limit on the electron Lorentz factor $\g_{low}$ and the upper limit $\e_{*,hi}$
562: implied by the kinematic limits on $y$ are
563: \begin{equation}
564: \g_{low} = {\e_s\over 2}\;\left[ 1 + \sqrt{1 + {2\over \e_*\e_s (1-\cos\psi)}}\;\right]\;,
565: \label{glow}
566: \end{equation}
567: and
568: \begin{equation}
569: \e_{*,hi} = {2\e_s\over 1-\cos\psi}\;.
570: \label{ehi}
571: \end{equation}
572:
573: Eq.\ (\ref{esLCesomegas}) is the starting point to calculate accurate
574: Compton-scattered spectra involving relativistic electrons
575: and external photon fields with arbitrary
576: anisotropies. In contrast to the comoving electron spectrum used in the SSC calculation,
577: the calculation of Compton-scattered radiation uses
578: the electron spectrum $N_e(\g,\Omega_e)$ and the target photon spectrum
579: defined in the stationary frame \citep{gkm01}.
580: The invariant phase volume
581: $d{\cal V} = dV d^3\vec p$ for
582: relativistic particles is given by
583: \begin{equation}
584: {dN\over d{\cal V}} = {dN\over dV d^3\vec p} \;=\;{1\over (m_ec)^3}\;{1\over
585: \gamma^2}\; {dN\over d\gamma d\Omega dV}\;=\;inv\;,
586: \label{dNdV}
587: \end{equation}
588: implying that
589: \begin{equation}
590: N_e(\gamma, \Omega) =
591: {\gamma^2\over \gamma^{\prime 2}}\;{dV\over dV^\prime}\;
592: N_e^\prime(\gamma^\prime,\Omega^\prime ) = \dD^3
593: N_e^\prime(\gp,\Omega^\prime )\;,
594: \label{PdN/dV_1}
595: \end{equation}
596: noting that $dV/dV^\prime = dt^\prime/dt = \dD$, and $\gamma =
597: \dD \gp$ when $\gp, \g \gg 1$, required for the
598: head-on approximation.
599: For an isotropic comoving distribution of electrons,
600: $N_e(\g,\Omega_s) = \dD^3 N^\prime_e(\gp )/4\pi$. Hence
601: \begin{equation}
602: \e_s L_{\rm C}(\e_s,\Omega_s) = {cr_e^2 \over 4}\;\e_s^2\dD^3\;\int_0^{2\pi}d\phi_*
603: \int_{-1}^1 d\mu_*\; \int_0^{\e_{*,hi}} d\e_*\; {u_*(\e_*,\Omega_* )\over \e_*^2}\int_{\g_{low}}^\infty
604: d\g\;\g^{-2} {N^\prime_e(\g/\dD )}\;
605: \Xi\;,
606: \label{esLCesomegas_1}
607: \end{equation}
608: %and $\e_{hi} = 2\e_s/(1-\cos\psi )$.
609: or
610: \begin{equation}
611: f_\e^{\rm EC} = {\e_s L_{\rm C}(\e_s,\Omega_s)\over d_L^2}
612: = {c\pi r_e^2 \over 4\pi d_L^2}\;\e_s^2\dD^3\;\int_0^{2\pi}d\phi_*
613: \int_{-1}^1 d\mu_*\; \int_0^{\e_{*,hi}} d\e_*\; {u_*(\e_*,\Omega_* )\over \e_*^2}\int_{\g_{low}}^\infty
614: d\g\; {N^\prime_e(\g/\dD )\over \g^2}\;
615: \Xi\;.
616: \label{esLCesomegas_26}
617: \end{equation}
618:
619:
620: In terms of the measured synchrotron $\nu F_\nu$ spectrum, eq.\ (\ref{Neprimegps}),
621: the source Compton spectrum for external Compton (EC) scattering in a standard
622: one-zone model for blazars is, in general,
623: given by the four-fold integral
624: \begin{equation}
625: f_\e^{\rm EC} = \left( {3\over 4}\right)^2\;{\e_s^2 \dD^2\over U_B}\;\int_0^{2\pi}d\phi_*
626: \int_{-1}^1 d\mu_* \int_0^{\e_{hi}} d\e_*\; {u_*(\e_*,\Omega_* )\over \e_*^2}\int_{\g_{low}}^\infty
627: d\g\; {f^{syn}_{\breve\e}\over \g^5}\;
628: \Xi\;\;,
629: \label{esLCesomegas_2}
630: \end{equation}
631: with
632: \begin{equation}
633: \breve\e \equiv {\e_B\g^2\over (1+z)\dD}\;,
634: \label{brevee}
635: \end{equation}
636: using eq.\ (\ref{Neprimegps}). The number of integrations can
637: obviously be reduced by choosing symmetrical target photon geometries.
638:
639:
640: \subsection{Point Source Radially Behind Jet}
641:
642: First we consider the flux when nonthermal electrons
643: in a relativistic jet Compton-scatter photons from a
644: point source of radiation, isotropically emitting and
645: located radially behind the outflowing
646: plasma jet. For a monochromatic point source with luminosity $L_0$
647: and energy $\e_0$,
648: the spectral luminosity can be expressed as
649: \begin{equation}
650: L_*(\e_*) \; =\; L_0 \delta(\e_* - \e_0 )\;.
651: \label{Lstarestar}
652: \end{equation}
653: The spectral energy distribution of the target photon
654: source at distance $r$ from the point source is therefore given by
655: \begin{equation}
656: u(\e_*,\Omega_* ) = {L_0\over 4\pi r^2 c}\;{\delta(\mu_* - 1)\over 2\pi }\;\delta(\e_* - \e_0 )\;.
657: \label{ueomega}
658: \end{equation}
659: Substituting eq.\ (\ref{ueomega}) into eq.\ (\ref{esLCesomegas_1})
660: and solving gives
661: \begin{equation}
662: \e_s L^{pt}_{\rm C}(\e_s,\Omega_s) = {r_e^2 \e_s^2 L_0 \dD^3\over 16\pi r^2 \e_0^2}
663: \int_{\bar\g_{low}}^\infty
664: d\g\;{N^\prime_e(\g/\dD )\over \g^2}\;
665: \bar\Xi\;.
666: \label{esLCesomegas_3}
667: \end{equation}
668: Using eq.\ (\ref{fecompton}), eq.\ (\ref{esLCesomegas_3}) becomes
669: \begin{equation}
670: f^{{\rm C},pt}_\e =
671: {r_e^2 \e_s^2 L_0 \dD^3\over 16\pi r^2d_L^2 \e_0^2}
672: \int_{\bar\g_{low}}^\infty
673: d\g\;{N^\prime_e(\g/\dD )\over \g^2}\;
674: \bar\Xi\;,
675: \label{feptNegamma17}
676: \end{equation}
677: or with eq.\ (\ref{Neprimegps}),
678: \begin{equation}
679: f^{{\rm C},pt}_\e = {3^2\over 8^2\pi}\;
680: {L_0 \e_s^2 \dD^2\over c r^2 U_B \e_0^2}\;
681: \int_{\bar\g_{low}}^\infty d\g\;{f_{\breve\e}^{syn}\over \g^5}\;
682: \bar\Xi\;,
683: \label{fept}
684: \end{equation}
685: where $\breve \e$ is defined in eq.\ (\ref{brevee}), $\bar\Xi$ is defined
686: by eq.\ (\ref{Xi}) with $\bar\e$ replaced by $\bar{\bar\e} = \g \e_0(1-\mu_s)$,
687: and
688: \begin{equation}
689: \bar\g_{low} = {\e_s\over 2}\;\left[ 1 + \sqrt{1 + {2\over \e_0\e_s (1-\mu_s)}}\;
690: \right]\;.
691: \label{barglow}
692: \end{equation}
693:
694: Eqs.\ (\ref{feptNegamma17}) and (\ref{fept}) give the Compton-scattered
695: spectrum from a point source of radiation located radially behind the jet,
696: generalizing the Thomson-regime result \citep{dsm92} to include scattering
697: in the Klein-Nishina regime.
698: A scattered disk component should be found in all blazar models, with
699: its importance strongly dependent on distance $r$ of the jet from the accretion
700: disk. The point-source approximation gives the least upscattered
701: flux in the Thomson limit, and an extended disk having the same power as a point source
702: will give a more intense flux. At sufficiently large jet heights $r \gg \Gamma^4 R_g$,
703: defining the far field,
704: where $R_g = GM/c^2
705: \cong 1.5\times 10^{13} M_8$ cm is the gravitational radius,
706: the Shakura-Sunyaev disk can be described as a point source
707: radially behind the the jet. Photons from large disk radii are important
708: in the near field $r \ll \Gamma^4 R_g$ \citep{ds02}.
709:
710:
711:
712: \subsubsection{Reduction to the Thomson Regime}
713:
714: We now derive the Thomson limit for the
715: $\nu F_\nu$ spectrum, eq.\ (\ref{feptNegamma17}). Because we consider
716: relativistic electrons $\g \gg 1$, we are restricted to the condition
717: $\bar\gamma_{low} \gg 1$, which occurs according to eq.\ (\ref{barglow})
718: when
719: either $\e_s \gg 1$ or
720: $\e_s/\e_0(1-\mu_s) \gg 1$. The Thomson condition can be expressed as
721: $\e_s \ll \gamma$, which is guaranteed when $\e_s \ll \bar\g_{low}$, in which
722: case $2\e_0\e_s (1-\mu_s) \ll 1$. Another statement of the Thomson condition
723: is that $\gamma\e_0 (1-\mu_s)\ll 1$ which, with $\e_s \ll \g$, again implies that
724: $\e_0\e_s(1-\mu_s) \ll 1$. Thus
725: \begin{equation}
726: \bar\g_{low} \rightarrow \sqrt{\e_s\over 2\e_0(1-\mu_s)}\;.
727: \label{barglow1}
728: \end{equation}
729:
730: For the scattering kernel, eq.\ (\ref{Xi}), $\e_s\ll \gamma$ and $y \rightarrow 1$
731: in the Thomson regime, so
732: \begin{equation}
733: \bar \Xi \rightarrow \bar \Xi_{\rm T} \equiv
734: 2 - 2\left({\e_s\over \g\bar{\bar \e}}\right) + \left({\e_s\over \g \bar{\bar \e}}
735: \right)^{2}\;.
736: \label{XiT}
737: \end{equation}
738: Away from the endpoints of the spectrum, $\e_s\ll \g \bar{\bar \e}$ and
739: $\bar \Xi_{\rm T} \rightarrow 2$. Hence
740: \begin{equation}
741: f_\e^{pt,{\rm T}} \; = \; {3\over 4} \;{\sT L_0\over (4\pi r d_L)^2}\;
742: \left({\e_s\over \e_0}\right)^2\dD^3\;\int_{\dD\bar{\bar\g }}^\infty d\g
743: \; {N^\prime_e(\gp )\over \g^2}\;,
744: \label{feptT}
745: \end{equation}
746: defining
747: \begin{equation}
748: \dD\bar{\bar\g } \equiv \sqrt{ {\e_s \over 2\e_0(1-\mu_s)} }\;.
749: \label{dDbarbarg}
750: \end{equation}
751: For the comoving electron distribution, eq.\ (\ref{Neprimegps}), in the
752: power-law form
753: \begin{equation}
754: N_e^\prime(\gp ) = K^\prime \g^{\prime -p}H(\gp;\gp_1,\gp_2)\;,
755: \label{Neprime}
756: \end{equation}
757: eq.\ (\ref{feptT}) becomes
758: $$f_\e^{pt,{\rm T}} \; = \; {3\over 4(p+1)} \;{\sT L_0 K^\prime\over (4\pi r d_L)^2}\;
759: \left({\e_s\over \e_0}\right)^2\dD^{3+p}\;\left[\max \left( \bar{\bar\g },
760: \gp_1\right)^{-(p+1)} - \g_2^{\prime -(p+1)} \right]
761: \;$$
762: \begin{equation}
763: \rightarrow {3\over 4(p+1)} \;{\sT L_0 K^\prime\over (4\pi r d_L)^2}\;
764: \left({\e_s\over \e_0}\right)^{(3-p)/2}\dD^{3+p}\; [2(1-\mu)]^{(p+1)/2}
765: \;,
766: \label{feptT_1}
767: \end{equation}
768: where the final expression applies in the regime $\gp_1 \ll \bar{\bar\g }\ll \gp_2$.
769: This can
770: be written as
771: \begin{equation}
772: f_\e^{pt,{\rm T}}\cong \left({3\over p+1}\right)\; \dD^6 (1-\mu_s)^2\;\left({\sigma_{\rm T} \over 4\pi d_L^2}\right)
773: \;\left( {L_0\over 4\pi r^2}\right) \;\bar {\bar\g}^{3}N^{\prime}_e(\bar{\bar \g})\;,
774: \label{feptT_2}
775: \end{equation}
776: which can be compared with the Thomson-regime expression
777: \begin{equation}
778: f_\e^{pt,{\rm T}} \cong \left({1\over 2}\right)\;\dD^6 (1-\mu_s)^2\;\left({\sigma_{\rm T} \over 4\pi d_L^2}\right)
779: \;\left( {L_0\over 4\pi r^2}\right) \;\tilde \g^{\,\prime 3}N^{\prime}_e(\tilde \gp)\;
780: \label{feptT_dsm}
781: \end{equation}
782: \citep{dsm92,ds93,ds02}, where
783: \begin{equation}
784: \tilde \gp = {1\over \dD}\;\sqrt{\e(1+z)\over \e_0(1-\mu_s)} = \sqrt{2} \bar{\bar \g}\;.
785: \end{equation}
786:
787: \subsubsection{Accurate Thomson Regime Spectrum}
788:
789: Eq.\ (\ref{feptT_2}) for the Thomson-scattered
790: spectrum was derived assuming $\Xi_{\rm T}=2$,
791: away from the endpoints of the spectrum. Using the full expression for $\Xi_{\rm T}$
792: and a power-law electron distribution gives an accurate expression for the
793: spectrum of a localized jet of isotropically entrained
794: electrons Thomson scattering a point source, monochromatic
795: radiation field that enters the jet from behind. The result is
796: \begin{equation}
797: f_\e^{pt,{\rm T}}= \;
798: {\pi r_e^2 L_0 \over (4\pi rd_L)^2}\;
799: \big({\e_s\over\e_0}\big)^2 \dD^3 \;\int_{\dD\bar{\bar \g}}^\infty
800: d\g\; \big( {2\over \g^2} - {2s\over \g^4} +{s^2\over \g^6}\big)
801: N^{~\prime}_e(\gp)\;,
802: \label{feptT_3}
803: \end{equation}
804: where
805: \begin{equation}
806: s \;\equiv\; {\e_s\over \e_0(1-\mu_s)} \; = \; 2 \dD^2 {\bar{\bar \g}}^2\;.
807: \end{equation}
808:
809: For the power-law electron distribution $N_e^\prime(\gp )$, eq.\ (\ref{Neprime}), the accurate
810: analytic Thomson-regime $\nu F_\nu$ flux from an isotropic monochromatic point source
811: radiation field located behind the jet is
812: $$f_\e^{pt,{\rm T}}= {3\over 4}\; {\sigma_{\rm T} L_0 K^\prime\over (4\pi r d_L)^2}\;
813: \big({\e_s\over \e_0}\big)^2\;\dD^{3+p}\;\big\{ {1\over 1+p}\big[\g_1^{-(1+p)}-\g_2^{-(1+p)}\big]
814: $$
815: \begin{equation}
816: -\;{s\over (3+p)}\big[\g_1^{-(3+p)}-\g_2^{-(3+p)}\big]\;
817: +
818: {s^2\over 2(5+p)}\big[\g_1^{-(5+p)}-\g_2^{-(5+p)}\big]\big\},
819: \label{feptT_4}
820: \end{equation}
821: where $\g_1 = \max(\dD\gp_1,\dD{\bar{\bar \g}})$ and $\g_2 = \dD\gp_2$.
822:
823: In the asymptotic limit $\gp_1 \ll \bar{\bar \g }\ll \gp_2$,
824: eq.\ (\ref{feptT_4}) approaches
825: \begin{equation}
826: f_\e^{pt,{\rm T}}\rightarrow {\cal A}(p) \;\big({\sigma_{\rm T} \over 4\pi d_L^2}\big)
827: \;\big( {L_0\over 4\pi r^2}\big)
828: \;\dD^6 (1-\mu_s)^2\;
829: \bar {\bar\g}^{3}N^{~\prime}_e(\bar{\bar \g})\;,
830: \label{feptT_5}
831: \end{equation}
832: where
833: \begin{equation}
834: {\cal A}(p) \equiv 3\;\left({1\over 1+p} - {2\over 3+p} + {2\over 5+p}\right)\;.
835: \label{cala}
836: \end{equation}
837: The values of ${\cal A}(p) = 1.0, 0.657$, and $ 0.5$ for $p = 1, 2$, and $3$,
838: respectively.
839: For the Thomson approximation away from the
840: endpoints of the spectrum, given by eq.\ (\ref{feptT_2}), the corresponding coefficient
841: is $3/(p+1)$.
842:
843: \subsubsection{Solution with Compton Cross Section}
844:
845: From eqs.\ (\ref{esLCesomegas_3}) and (\ref{fecompton}),
846: \begin{equation}
847: f^{pt,{\rm C}}_\e = {\pi r_e^2 L_0 \over (4\pi r d_L)^2}\;\left({\e_s\over \e_0^2}\right)
848: \;\dD^3
849: \int_{\bar\g_{low}}^\infty
850: d\g\;{N^\prime_e(\g/\dD )\over \g^2}\; \left[y+y^{-1} - {2\e_s\over \gamma \bar\e y} +
851: ({\e_s\over \gamma \bar\e y})^2\right]\;\;.
852: \label{esLCesomegas_4}
853: \end{equation}
854: Introducing
855: $u = {\e_s/ \g }$ and $v = \e_s\e_0(1-\mu_s)\;$
856: and changing variables to $u$ gives, for the power-law electron distribution,
857: eq.\ (\ref{Neprime}),
858: \begin{equation}
859: f^{pt,{\rm C}}_\e = {\pi r_e^2 L_0 K^\prime\over (4\pi r d_L)^2}\;
860: \left({\e_s\over \e_0^2}\right)
861: \;\dD^{3+p}\e_s^{-(1+p)}I_{\rm C}\;,
862: \;
863: \label{esLCesomegas_5}
864: \end{equation}
865: where
866: \begin{equation}
867: I_{\rm C} = \int_{u_1}^{u_2}
868: du\;\left[ \;
869: u^p - u^{p+1} +{u^p\over 1-u} - {2u^{p+2}\over v(1-u)}
870: +{u^{p+4}\over v^2(1-u)^2}\;
871: \right]\;,
872: \label{ic1}
873: \end{equation}
874: and $u_1 = \e_s\dD/\gp_2$, $u_2 = \e_s\min\left({1\over \bar\g_{low}},{\dD\over \gp_1}\right)$.
875: The series solution of eq.\ (\ref{ic1}) is given by
876: \begin{equation}
877: I_{\rm C} = \left\{ {u^{p+1}\over p+1} - {u^{p+2}\over p+2} +
878: \sum_{i=0}^\infty \left[ {u^{i+p+1}\over i+p+1}-{2u^{i+p+3}\over v(i+p+3)}
879: +{1\over v^2}{(i+1) u^{i+p+5}\over i+p+5}
880: \right]\right\} \Biggr|_{u_1}^{u_2}\;.
881: \label{ic2}
882: \end{equation}
883: Eq.\ (\ref{ic1}) can be solved analytically for integral $p$.
884: %The analytic expressions for $I_{\rm C}(p)$ for $p = 1,$ 2, and 3 are given in the Appendix.
885:
886:
887: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
888: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
889:
890:
891: \subsection{Shakura-Sunyaev Accretion Disk Field}
892:
893:
894: For the emission spectrum of an accretion disk surrounding the
895: supermassive black hole, we consider the cool, optically-thick
896: blackbody solution of \citet{ss73}. The
897: disk emission is approximated by a surface radiating at the blackbody
898: temperature associated with the local energy dissipation rate per unit
899: surface area, which is derived from considerations of viscous
900: dissipation of the gravitational potential energy of the accreting
901: material. The accretion luminosity is defined
902: in terms of the Eddington ratio
903: \begin{equation}
904: \ell_{\rm Edd} = {\eta\dot m c^2\over L_{\rm Edd}}\;,
905: \label{ellEdd}
906: \end{equation}
907: where $\eta\sim 0.1$ is the efficiency to transform accreted matter to
908: escaping radiant energy. The Eddington luminosity $L_{\rm Edd}=
909: 1.26\times 10^{46} M_8$ ergs s$^{-1}$, where the mass of the central
910: supermassive black hole is $M =10^8M_8 M_\odot$ and the black hole is
911: accreting mass at the rate $\dot m$ (gm s$^{-1}$).
912:
913: For steady flows where the energy is derived from the viscous dissipation
914: of the gravitational potential energy of the accreting matter, the
915: radiant surface-energy flux
916: \begin{equation}
917: {d{\cal E}\over dAdt} =
918: {3GM\dot m\over 8\pi R^3}\; \varphi(R)
919: \label{PhiE}
920: \end{equation}
921: \citep{ss73}, where
922: \begin{equation}
923: \varphi(R) =[1-\beta_i(R_i/R)^{1/2}]\;,
924: \label{varphiSchwarzschild}
925: \end{equation}
926: $\beta_i \cong 1$, and $R_i = 6GM/c^2$ for the Schwarzschild
927: metric. Integrating equation (\ref{PhiE}) over a two-sided disk gives
928: $\eta = 1/12$. Assuming that the disk is an optically-thick
929: blackbody, the effective temperature of the disk can be
930: determined by equating equation (\ref{PhiE}) with the surface energy
931: flux $\sigma_{\rm SB}T^4(R)$.
932: A monochromatic approximation for the mean
933: photon energy $m_ec^2\bar \e (R) = k_{\rm B} T(R)$ at radius
934: $R$ of the accretion disk with mean temperature
935: $T(R)$ is given by
936: \begin{equation}
937: m_ec^2\langle \e (R)\rangle \;\cong \;2.70{\kB T(R)} \; \cong \;2.70 {\kB }
938: \big[ {3GM\dot m\varphi(R)\over 8\pi
939: R^3\sigma_{\rm SB}}\big ]^{1/4}\;\cong
940: 137\;\big({\ell_{\rm Edd}\over M_8\eta}\big)^{1/4}\tilde R^{-3/4}\;{\rm eV}
941: \;,
942: \label{<e>}
943: \end{equation}
944: so
945: \begin{equation}
946: \langle \e (\tilde R)\rangle \;\cong
947: \;2.7\times 10^{-4}\;\xi \tilde R^{-3/4}
948: \;,
949: \label{<e>_1}
950: \end{equation}
951: where
952: \begin{equation}
953: \xi \equiv \left({\ell_{\rm Edd}\over
954: M_8 \eta}\right)^{1/4}\;.
955: \label{xidef1}
956: \end{equation}
957: Here $\theta = \arccos \mu_*$ is the angle between the directions of the jet and the photon
958: that intercepts the jet,
959: \begin{equation}
960: \tilde R = {R\over R_g} = \tilde r\sqrt{\mu_*^{-2}-1}\;,
961: \end{equation}
962: where $\tilde r = {r/ R_g}$.
963: Lengths marked with a tilde are normalized to $R_g$. The final expression in eq.\ (\ref{<e>}) is
964: valid when $\tilde R \gg \tilde R_{min}$, though we assume it is reasonably accurate to $\tilde R\cong
965: \tilde R_{min}$.
966:
967:
968: The intensity of the Shakura-Sunyaev accretion disk model along
969: the jet axis is
970: given by
971: \begin{equation}
972: I_\e^{\rm SS}(\Omega;R) \cong \;{3GM\dot m\over 16\pi^2
973: R^3}\varphi(R)\; \delta[\e - {2.7 k_{\rm B}\over m_ec^2}\;T(R)]\;
974: \label{uoverc}
975: \end{equation}
976: \citep{st83}. Thus
977: \begin{equation}
978: I_\e^{\rm SS}(\Omega;\tilde R) =
979: {3\over 16\pi^2}\;{\ell_{\rm Edd}L_{\rm Edd}\over \eta \tilde R R^2}\;
980: \varphi(\tilde R)\delta [\e -\langle\e(\tilde R)\rangle ]\;
981: \label{IeSS}
982: \end{equation}
983: \citep{ds02}.
984: Substituting eq.\ (\ref{IeSS}) into eq.\ (\ref{esLCesomegas_1}), using the
985: relation $I_\e (\Omega) = c u(\e ,\Omega)$, gives
986: $$f_\e^{\rm SS} = {3^2\over 2^9\pi^3}\;{\sT\e_s^2\over d_L^2 R_g^2}\;
987: { \ell_{\rm Edd} L_{\rm Edd}\over\eta \tilde r^3 }\; \dD^3\;\int_0^{2\pi}d\phi_*
988: \int_{\mu_{*,min}}^{\mu_{*,max}} d\mu_*\; \;{\varphi(\tilde R )\over (\mu_*^{-2}-1)^{3/2}
989: \langle\e(\tilde R)\rangle^2}$$
990: \begin{equation}
991: \times\int_{\bar\g_{low}}^\infty
992: d\g\;\g^{-2} {N^\prime_e(\g/\dD )}\;
993: \big[y+y^{-1} - {2\e_s\over \gamma \tilde{\bar\e} y} +
994: ({\e_s\over \gamma \tilde{\bar\e} y})^2 \big]\;.
995: \label{eslcssexyz}
996: \end{equation}
997: The integration over angle in eq.\ (\ref{eslcssexyz})
998: is limited by the inner radius of the accretion
999: disk, so that
1000: $\mu_{*,max} = \big[1 + (6/\tilde r)^2\big]^{-1/2}$ for a Schwarzschild black hole,
1001: and
1002: \begin{equation}
1003: \tilde{\bar\e} = \tilde{\bar\e}(\g,\e_s,\psi) =\g \langle\e(\tilde R)\rangle(1-\cos\psi )\;,
1004: \label{tildebare}
1005: \end{equation}
1006: and
1007: \begin{equation}
1008: \bar\g_{low} = {\e_s\over 2}\;\left( 1 + \sqrt{1+{2\over \langle\e(\tilde R)\rangle\e_s(1-\cos\psi)}}\right)\;.
1009: \end{equation}
1010: The other limit on the angular integration arises because of the restriction given
1011: by eq.\ (\ref{ehi}), so that
1012: \begin{equation}
1013: \langle \e (\tilde R )\rangle \; < \; {2\e_s\over 1-\cos\psi }\;,
1014: \label{limit}
1015: \end{equation}
1016: which restricts the integral to a maximum value of $\tilde R$ and therefore
1017: $\mu_* \gtrsim \mu_{*,min}$.
1018: In the calculation of
1019: $\cos\psi$, eq.\ (\ref{cospsi}), we take $\phi_s = 0$ without loss of generality because of the
1020: assumed azimuthal symmetry of the accretion-disk emission.
1021:
1022: The result for the accretion-disk radiation field scattered
1023: by isotropic, relativistic jet electrons is a 3-fold integral---reduced
1024: from a 4-fold integral by approximating the disk blackbody spectrum by its mean
1025: thermal energy at different radii.
1026: When expressed in terms of the measured synchrotron $\nu F_\nu$
1027: spectrum using eq.\ (\ref{Neprimegps}), the result for the
1028: accretion-disk radiation field is
1029: $$f_\e^{\rm SS} = {3^3\over 2^8\pi^2}\;{\e_s^2\over c R_g^2 U_B}\;
1030: { \ell_{\rm Edd} L_{\rm Edd}\over\eta \tilde r^3 }\;\dD^2
1031: \;\int_0^{2\pi}d\phi_*
1032: \int_{\mu_{*,min}}^{\mu_{*,max}} d\mu\; \;{\varphi(\tilde R )\over (\mu_*^{-2}-1)^{3/2}
1033: \langle\e(\tilde R)\rangle^2}$$
1034: \begin{equation}
1035: \times\int_{\bar\g_{low}}^\infty
1036: d\g\;\g^{-5} {f^{syn}_{\breve\e}}\;
1037: \left[y+y^{-1} - {2\e_s\over \gamma \tilde{\bar\e} y} +
1038: ({\e_s\over \gamma \tilde{\bar\e} y})^2 \right]\;,
1039: \label{eslcssexyza}
1040: \end{equation}
1041: recalling the definitions of $\gp_s$ and $\breve \e$ from
1042: eqs.\ (\ref{gps}) and (\ref{brevee}), respectively.
1043: This can be reduced to a 2-fold
1044: integral by approximating a typical scattering as occurring at $\phi_* = \pi/2$,
1045: so that $\tilde{\bar\e} = \gamma\e (1-\mu_*\mu_s)$.
1046: Because it is feasible to perform the 3-fold integral numerically, we show
1047: results of the more accurate calculations.
1048:
1049: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1050:
1051: \subsubsection{Regimes in Compton-Scattered Accretion-Disk Spectra}
1052:
1053: \label{trends}
1054:
1055: The limiting behaviors of the Compton scattered spectra can be understood
1056: based on simple $\delta$-approximations in the Thomson regime.
1057: External Compton scattering of the Shakura-Sunyaev
1058: disk with radiant luminosity $L_d = \ell_{\rm Edd} L_{\rm Edd}$
1059: in the
1060: near field, i.e., $\tilde r\ll \Gamma^4$, can be approximated as
1061: \begin{equation}
1062: \label{fecnf}
1063: f_\e^{ECNF} \cong \frac{\dD^2L_{d}}{2B^2 R_g^2 \tilde r^3c}\
1064: f^{syn}_{\bar{\e}_{syn}}
1065: \end{equation}
1066: where
1067: \begin{equation}
1068: \label{eecnf}
1069: \bar{\e}_{syn} = \bar{\e}_{syn}(\tilde r) = {2\e\e_B\over \dD \bar\e(\sqrt{3 r})}
1070: \cong 3 \times 10^{-10}\ \frac{\e B({\rm G})}{\dD}\
1071: \left(\frac{M_8\eta \tilde r^3}{l_{Edd}}\right)^{1/4}\ .
1072: \end{equation}
1073: In the far field ($\tilde r\gg\Gamma^4$),
1074: \begin{equation}
1075: \label{fecff}
1076: f_\e^{ECFF} \cong \frac{3}{8}\ \frac{L_{d}}{r^2cB^2\dD^2}\
1077: f_{\bar{\bar{\e}}_{syn}}\;,
1078: \end{equation}
1079: where
1080: \begin{equation}
1081: \label{eecnf1}
1082: \bar{\bar{\e}}_{syn} = 3.2\times10^{14}\ \dD^4\e\e_BM_8\
1083: \left(\frac{\eta}{l_{Edd}}\right)^{1/4}
1084: \end{equation}
1085: \citep{ds02}.
1086:
1087:
1088: \subsubsection{$\g\g$ Opacity from Accretion Disk Photons}
1089:
1090: Photons from the accretion disk will interact with high energy
1091: $\g$-rays to produce electron-positron pairs, modifying the very-high energy
1092: (VHE; multi-GeV -- TeV)
1093: $\g$-ray spectrum by a factor of $e^{-\tau_{\g\g}}$.
1094: The absorption opacity $\tau_{\g\g}$ can be calculated
1095: by inserting the
1096: photon density, $n_{ph}$ for an accretion disk
1097: into eq. (\ref{dtauggdx}) and integrating $x$ from $r$ to $\infty$.
1098: For a Shakura-Sunyaev accretion disk, the photon density is given by
1099: \begin{equation}
1100: n_{ph}^{SS}(\e_*,\Omega_*) = \frac{I_{\e_*}^{SS}(\Omega_*; \tilde{R})}{\e_* m_ec^3}
1101: \end{equation}
1102: where $I_\e^{SS}(\Omega; \tilde{R})$ is given by eq. (\ref{IeSS}).
1103: For an azimuthally symmetric accretion disk,
1104: the optical depth to $\gamma\gamma$ pair production attenuation
1105: for a photon with observed energy $\e_1$ traveling outward along the jet axis
1106: starting at height $\tilde r$
1107: is given by
1108: \begin{equation}
1109: \tau_{\g\g}^{SS}(\e_{1},\tilde r) \cong 3 \times 10^{6}\
1110: \frac{l_{Edd}^{3/4}M_8^{1/4}}{\eta^{3/4}}\
1111: \int^{\infty}_{\tilde{r}} \frac{d\tilde{x}}{\tilde{x}^2}\
1112: \int^{\infty}_6 {d\tilde{R}\over
1113: \tilde{R}^{5/4}}\;
1114: {[\phi(\tilde R )]^{1/4}H\left(\tilde s - 1\right)\over (1+\tilde R^2/\tilde x^2)^{3/2}}\;
1115: \left[{\sigma_{\g\g}(\tilde s)\over \pi r_e^2}\right]\; (1-\mu_*) \; ,
1116: \label{tauggSStilder}
1117: \end{equation}
1118: where $\tilde s \equiv \langle\e(\tilde{R})\rangle {\e_1(1+z)(1-\mu_*)/2}$ and
1119: $
1120: \mu_* = 1/\sqrt{1+\tilde{R}^2/\tilde{x}^2}.
1121: $
1122: For the Shakura-Sunyaev disk extending to the
1123: innermost stable orbit of a Schwarzschild black hole,
1124: one sees from eq.\ (\ref{tauggSStilder})
1125: and the definition of $\langle\e(\tilde{R})\rangle$, eq.\ (\ref{<e>_1}), and $\xi$,
1126: eq.\ (\ref{xidef1}),
1127: that
1128: \begin{equation}
1129: {\tau_{\g\g}^{SS}(\e_{1},\tilde r)\over M_8} \propto {\rm ~function~of~}\xi\;{\rm and~}\tilde r \; .
1130: \label{xidef}
1131: \end{equation}
1132: A first-order correction to $\gamma\gamma$ opacity for a photon
1133: traveling at a small angle angle $\theta_s =\arccos \mu_s \ll 1$
1134: along the jet axis is obtained by replacing $\mu_*$ with $\mu_*\mu_s$,
1135: which implicitly assumes that typical interactions take place at azimuth
1136: $\phi_* = \pi/2$.
1137: A detailed calculation of the $\g\g$ opacity from accretion disk photons is given by \citet{bk95}.
1138:
1139:
1140: \subsubsection{Numerical Results}
1141:
1142:
1143: \begin{deluxetable}{lll}
1144: %\rotate
1145: \tabletypesize{\scriptsize}
1146: \tablecaption{
1147: Parameters of Baseline Model
1148: }
1149: \tablewidth{0pt}
1150: \startdata
1151: \hline
1152: Redshift & $z$ & 1 \\
1153: Bulk Lorentz Factor & $\Gamma$ & 25 \\
1154: Doppler Factor & $\dD$ & 25 \\
1155: Magnetic Field & $B$ & 1 G \\
1156: Variability Timescale & $t_v$ & $10^4$ s \\
1157: Black Hole Mass ($10^8 M_\odot$) & $M_8$ & 1 \\
1158: Eddington Ratio & $l_{Edd}$ & 1 \\
1159: Jet Height (units of $R_g$) & $\tilde{r}$ & $10^3$ \\
1160: Low-Energy Electron Spectral Index & $p_1$ & 2 \\
1161: High-Energy Electron Spectral Index & $p_2$ & 4 \\
1162: Minimum Electron Lorentz Factor & $\gp_{min}$ & $10^2$ \\
1163: Maximum Electron Lorentz Factor & $\gp_{max}$ & $10^7$ \\
1164: Accretion Efficiency & $\eta$ & 1/12 \\
1165: Ratio of $B$ to Equipartition Field & $B/B_{eq}$ & 0.3 \\
1166: Jet Power in Magnetic Field & $P_{j,B}$ & $7\times10^{43}$ erg s$^{-1}$ \\
1167: Jet Power in Particles & $P_{j,par}$ & $10^{46}$ erg s$^{-1}$ \\
1168: \enddata
1169: \end{deluxetable}
1170:
1171: The electron distribution is assumed to be well-represented by
1172: the \citet{band93}-type function
1173: $$N^\prime_e(\gp) = K^\prime_e\ H(\gp;\gp_{min},\gp_{max})\;\;\big\{
1174: \g^{\prime -p_1} \exp\left(-\gp/\gp_0\right)\;H[(p_2-p_1)\gp_0 - \gp] \;
1175: +$$
1176: \begin{equation}
1177: [(p_2-p_1)\gp_0]^{p_2-p_1} \g^{\prime -p_2}\ \exp(p_2-p_1) H[\gp -(p_2-p_1) \gp_0]
1178: \big\}.
1179: \label{bandfn}
1180: \end{equation}
1181:
1182: This distribution is essentially two smoothly joined power laws with
1183: power law number indices $p_1$ and $p_2$, and low- and high-energy
1184: cutoffs, $\gp_{min}$ and $\gp_{max}$, respectively, in the electron
1185: spectrum. For illustrative purposes, we take $p_1 = 2$ and $p_2 = 4$,
1186: or a break by 2 units in the electron spectrum. This can be compared
1187: to a break by one unit expected for synchrotron and Thomson
1188: losses. Our approach is, however, to use the flaring synchrotron
1189: spectrum to imply the underlying flaring electron distribution without
1190: regard to specific acceleration and radiation processes. When the
1191: nonthermal electron distribution is obtained from analysis of blazar
1192: data, then the underlying jet physics that gives rise to the inferred
1193: electron spectrum can be examined.
1194:
1195:
1196: The total jet power in the stationary frame of the host galaxy
1197: is given by
1198: \begin{equation}
1199: \label{jetpower}
1200: P_j = 2 \pi R_b^{\prime 2}\beta\G^2c u^{\prime}_{tot}
1201: = P_{j,par}+P_{j,B}
1202: \end{equation}
1203: \citep{celotti93,celotti07,fdb08}, where $u^{\prime}_{tot}$ is the
1204: total energy density in the jet, $P_{j,par}$ is the jet power from
1205: particles, and $P_{j,B}$ is the jet power from the magnetic field.
1206: Here the factor of $2$ takes into account that the jet is two-sided.
1207: For synchrotron--only emission, it is expected that the jet will be in
1208: equipartition and $P_{j,par} \approx P_{j,B}$, which will minimize the
1209: jet power. In order to explain the Compton-scattered component,
1210: however, the energy density in electrons can be different than the
1211: magnetic field energy density.
1212:
1213:
1214: We performed a parameter study by varying model parameters with
1215: respect to a baseline model, with baseline parameters given in Table
1216: 1. We consider a $10^8 M_\odot$ supermassive black-hole jet source
1217: located at redshift $z = 1$. The jet is radiating at $10^3 R_g$ from
1218: the black hole and has Doppler factor $\dD = 25$ and bulk Lorentz
1219: factor $\Gamma = 25$, so that the angle of the jet direction to the
1220: line of sight is $\theta \cong 1/\Gamma$. The mean magnetic field is
1221: 1 G, and the variability time $t_v = 10^4$ s, corresponding to
1222: $\approx 10\times$ the light crossing time for the Schwarzschild
1223: radius of a $10^8 M_\odot$ black hole. The jet opening angle $\theta_j
1224: \approx R_b^\prime /r \lesssim c\dD t_v/(1+z) r \cong 0.25 \approx
1225: 15^\circ$ for standard parameters. While varying the parameters, the
1226: synchrotron spectrum was kept relatively constant by varying
1227: $K^\prime_e$ and $\gp_0$ following the relations given in \S
1228: \ref{trends} (except when changing angle). This was done with a
1229: $\chi^2$ fitting technique to the baseline synchrotron spectrum
1230: \citep[see][]{fdb08}. The transition between the near field and far
1231: field takes place at $\tilde r \approx \Gamma^4$, so that the baseline
1232: height of the jet is in the near field.
1233:
1234: %\clearpage
1235:
1236: \begin{figure}[t]
1237: \center
1238: {\includegraphics[scale=0.5]{f1_color}}
1239: \caption[]{Spectral energy distribution (SED) of
1240: accretion-disk/relativistic jet model using standard parameters for
1241: disk-jet system given by Table 1. Magnetic field is varied with the
1242: synchrotron flux remaining essentially constant, though with the peak
1243: synchrotron frequency varying between $\approx 2\times 10^{11}$ and
1244: $10^{13}$ eV. Separate spectral components (thin short dashed
1245: curves) are, from left to right, the synchrotron, accretion-disk,
1246: SSC, Compton-scattered accretion disk radiation, and second-order
1247: SSC components, for the high-magnetic field case, $B = 4$ G.
1248: }
1249: \label{f1}
1250: \end{figure}
1251:
1252: %\clearpage
1253:
1254: The Compton-scattered accretion disk spectra are calculated from eq.\
1255: (\ref{eslcssexyz}). Fig.\ \ref{f1} shows the effects of changing the
1256: magnetic field. With increasing $B$, fewer electrons are required to
1257: make the same synchrotron flux, so that both the first and
1258: second-order SSC flux, and the flux of the Compton-scattered accretion-disk
1259: component decrease with increasing $B$. In all of these models, the
1260: second-order SSC emission is overwelmed by the external Compton
1261: component and is not visible. The overall levels of all
1262: Compton-scattered components are $\propto B^{-2}$.
1263:
1264: %\clearpage
1265:
1266: \begin{figure}[t]
1267: \center
1268: {\includegraphics[scale=0.5]{f2_color}}
1269: \caption[]{ Same as Fig.\ \ref{f1}, except that the luminosity of the
1270: disk is changed. Separate spectral components are shown for the
1271: $l_{Edd}=10^{-4}$ case.}
1272: \label{f2}
1273: \end{figure}
1274:
1275: %\clearpage
1276:
1277: \begin{figure}[t]
1278: \center
1279: {\includegraphics[scale=0.5]{f3_color}}
1280: \caption[]{Same as Fig.\ \ref{f1}, except that the jet height is varied.
1281: Separate spectral components are shown for the
1282: $r=3.2\times10^{4}\ R_g$ case.}
1283: \label{f3}
1284: \end{figure}
1285:
1286: %\clearpage
1287:
1288: \begin{figure}[t]
1289: \center
1290: {\includegraphics[scale=0.5]{f4_color}}
1291: \caption[]{
1292: Same as Fig.\ \ref{f1}, except that the minimum electron Lorentz factor $\gp_{min}$ is changed.
1293: Separate spectral components are shown for the
1294: $\g_{min}=600$ case.}
1295: \label{f4}
1296: \end{figure}
1297:
1298: %\clearpage
1299:
1300: With increasing disk power $\ell_{\rm Edd}$, the accretion-disk
1301: radiation and also the Compton-scattered accretion-disk
1302: component becomes progressively larger, as shown in Fig.\ \ref{f2}.
1303: Note that the temperature of the disk photons at a given radius
1304: increases $\propto \ell_{\rm Edd}^{1/4}$. Fig.\ \ref{f3} displays the
1305: dependence of the blazar SED on height $\tilde r$ of the jet. As the
1306: blob's distance from the disk increases, the level of the Comptonized
1307: disk radiation decreases. In both the case of changing the
1308: $\ell_{\rm Edd}$ and $\tilde r$ the SSC components are unaffected;
1309: eventually, the Compton-scattered disk flux falls below the SSC fluxes.
1310: The second-order SSC is visible when the external Compton flux
1311: decreases enough. Fig.\ \ref{f4} shows how variations in $\g_{min}$
1312: affect the low energy part of the synchrotron, SSC, and external
1313: Compton components.
1314:
1315:
1316: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1317:
1318: %\clearpage
1319:
1320: \begin{figure}[t]
1321: \center
1322: {\includegraphics[scale=0.5]{f5_color}}
1323: \caption[]{
1324: Same as Fig.\ \ref{f1}, except that the Doppler factor is varied with the synchrotron
1325: component essentially held constant.
1326: Separate spectral components are shown for the
1327: $\dD=15$ case.}
1328: \label{f5}
1329: \end{figure}
1330:
1331: %\clearpage
1332:
1333: In Fig.\ \ref{f5}, the Doppler factor is varied from the baseline
1334: value of 25, with the same approximate synchrotron flux. As $\dD$
1335: increases, the SSC component decreases, while the Compton-scattered
1336: accretion disk component increases. This behavior follows from
1337: eqs. (\ref{fssc}) and (\ref{fecnf}), for the following reasons: For
1338: larger Doppler factors and fixed variability times, the radius becomes
1339: larger and fewer electrons are required to make the synchrotron flux
1340: at the same flux level, so the electron and internal photon energy
1341: densities decrease. Consequently the SSC flux decreases, but the
1342: larger value of $\dD$ means that the external target photon field
1343: becomes more intense. Thus the Compton-scattered accretion disk
1344: radiation increases with increasing $\dD$. Also, as the Doppler
1345: factor increases, the SSC component is shifted to lower energies while
1346: the Compton-disk component is shifted to higher energies, in
1347: accordance with eqs. (\ref{essc}) and (\ref{eecnf}), respectively.
1348:
1349: %\clearpage
1350:
1351: \begin{figure}[t]
1352: \center
1353: {\includegraphics[scale=0.5]{f6_color}}
1354: \caption[]{Same as Fig.\ \ref{f1}, except that the viewing angle is varied. The peak flux of
1355: the synchrotron, SSC, and near-field scattered disk component goes roughly
1356: as $\dD^4$, $\dD^2$, and $\dD^6$,
1357: respectively. The different observing angles $\theta = 1/\Gamma, 2/\Gamma,3/\Gamma, 4/\Gamma$
1358: correspond to Doppler factors $\dD = 25, 10, 5,$ and 2.94, respectively.
1359: Separate spectral components are shown for the
1360: $\theta = 4/\Gamma$ case.}
1361: \label{f6}
1362: \end{figure}
1363:
1364: %\clearpage
1365:
1366: Fig.\ \ref{f6} illustrates how the blazar SED is affected by changes
1367: in viewing angle. Contrary to Figs.\ (\ref{f1}) -- (\ref{f5}), we let
1368: the synchrotron flux vary. In this calculation, $\Gamma = 25$, and the
1369: observation angle and therefore $\dD$ changes. It is interesting to
1370: note that the SSC flux varies least with changes in viewing angle,
1371: while the Comptonized disk component changes most rapidly. The ratios
1372: of the various components are explained by noting that $\dD/\Gamma
1373: \cong 1, 2/5, 1/5,$ and 2/17 for $\theta \Gamma = 1, 2, 3,$ and 4,
1374: respectively. The beaming factor for synchrotron radiation is
1375: $\propto \dD^{3+\alpha} \propto \dD^4$ in the flat portion of the $\nu
1376: F_\nu$ SED, with the convention that flux density $F_\nu \propto
1377: \nu^{-\alpha}$. The relative magnitudes of the SSC components are
1378: explained from eq.\ (\ref{fssc}), noting that
1379: $f_{\e_{pk}^{syn}}^{syn}\propto \dD^4$, so that
1380: $f_{\e_{pk}^{SSC}}^{SSC} \propto \dD^2$. The peak flux of the
1381: scattered disk component in the near-field regime, $f_\e^{ECNF}
1382: \propto \dD^6$, from eq.\ (\ref{fecnf}).
1383:
1384: A larger viewing angle or smaller Doppler factor implies a smaller
1385: size scale of the emitting region for a given variability time $t_v$.
1386: The characteristic Thomson scattering depth is $n^\prime_e \sigma_{\rm
1387: T} R^\prime_b\propto R^{\prime~-2}_b$ for a constant comoving electron
1388: spectrum $N^\prime_e(\gp )$ (described in our calculations by eq.\
1389: [\ref{bandfn}]) in a region of size $R^\prime_b$. Consequently, the
1390: suppression of the SSC component due to Doppler boosting is offset by
1391: the increased scattering depth for larger observing angles.
1392:
1393: %\clearpage
1394:
1395: \begin{figure}[t]
1396: \center
1397: {\includegraphics[scale=0.5]{f7_color}}
1398: \caption[]{Opacity of a photon emitted at
1399: various jet heights $\tilde r$ in units of
1400: $R_g$, subject to $\gamma\gamma$ pair production
1401: attenuation with photons from a Shakura-Sunyaev
1402: accretion disk with parameters given in Table 1. The parameter
1403: $\xi$ is defined in eq.\ (\ref{xidef}).}
1404: \label{tau_disk}
1405: \end{figure}
1406:
1407: %\clearpage
1408:
1409: Fig.\ \ref{tau_disk} shows a calculation of the accretion-disk opacity
1410: $\tau_{\g\g}(\e_1,\tilde r)$ for photons traveling along the jet axis,
1411: using eq.\ (\ref{tauggSStilder}) and the parameters given in Table 1.
1412: Here we assume that the Shakura-Sunyaev disk reaches to the innermost
1413: stable orbit ($\tilde R_{i}=6$) of a Schwarzschild black hole. At
1414: larger jet heights, the accretion-disk opacity declines and only
1415: increasingly higher energy photons are subject to attenuation due to
1416: the threshold condition. The effect of increasing $\xi$ is to
1417: increase the mean accretion-disk photon energy at a given radius, from
1418: eq.\ (\ref{xidef1}), and therefore lower the energy at which $\gamma$
1419: rays can be attenuated. At $\tilde r = 10^3$, only $\gtrsim 10$ TeV
1420: photons are attenuated by the disk radiation field (with $\ell_{\rm
1421: Edd} = 1$). The $\tau_{\g\g}$ corrections have not been included in
1422: the SED calculations, but are only important at $\nu \gtrsim 10^{25}$
1423: Hz.
1424:
1425: In all of our models the jet is particle-dominated ($P_{j,par}\gg
1426: P_{j,B}$) and the magnetic field is below its equipartition value. The
1427: total jet power, also divided into magnetic field and
1428: particle powers, is shown in Table \ref{jetpowertable}. This is typical of
1429: results from modeling the observed X-ray and $\g$-ray emission of
1430: TeV blazars with the SSC component \citep{fdb08}.
1431:
1432:
1433: \begin{deluxetable}{llllll}
1434: \tabletypesize{\scriptsize}
1435: \tablecaption{Jet Powers for Models in Figs.\ \ref{f1} -- \ref{f6}.\tablenotemark{a}}
1436: \tablewidth{0pt}
1437: \tablehead{
1438: \colhead{ $B$ [G] } &
1439: \colhead{ $\delta_D$ } &
1440: \colhead{ $\gamma_{min} $ } &
1441: \colhead{ $L_B$ [$10^{43}$ erg s$^{-1}$] } &
1442: \colhead{ $L_{par}$ [$10^{45}$ erg s$^{-1}$] } &
1443: \colhead{ $L_{tot}$ [$10^{45}$ erg s$^{-1}$] }
1444: }
1445: \startdata
1446: 1 & 25 & 100 & 6.6 & 11 & 11 \\
1447: 1 & 15 & 100 & .85 & 36 & 36 \\
1448: 1 & 20 & 100 & 2.7 & 19 & 19 \\
1449: 1 & 30 & 100 & 14 & 7.5 & 7.6 \\
1450: 1 & 35 & 100 & 25 & 5.3 & 5.5 \\
1451: 0.5 & 25 & 100 & 1.6 & 27 & 27 \\
1452: 2 & 25 & 100 & 26 & 4.8 & 5.0 \\
1453: 4 & 25 & 100 & 110 & 2.0 & 3.1 \\
1454: 1 & 25 & 30 & 6.6 & 18 & 18 \\
1455: 1 & 25 & 300 & 6.6 & 5.7 & 5.7 \\
1456: 1 & 25 & 600 & 6.6 & 3.0 & 3.0 \\
1457: \enddata
1458: \label{jetpowertable}
1459: \tablenotetext{a}{First model is baseline model.}
1460: \end{deluxetable}
1461:
1462:
1463:
1464:
1465:
1466: \subsection{External Isotropic Radiation Field}
1467:
1468: This is the case treated by \citet{gkm01}, where jet electrons
1469: scatter a surrounding external isotropic radiation field
1470: $u_*(\e_*,\Omega_* ) = u_*(\e_* )/4\pi$. By integrating
1471: eq.\ (\ref{esLCesomegas}) over the angle variables, recognizing that
1472: $d\mu_* d\phi* = 2 \pi d\cos\psi$ for this geometry, one obtains
1473: \begin{equation}
1474: f^{{\rm C},iso}_{\e} = {3\over 4}\;{c\sigma_{\rm T} \e_s^2\over 4\pi d_L^2}\;\dD^3
1475: \;
1476: \int_0^\infty d\e_*\;{u_*(\e_* )\over\e_*^2}
1477: \int_{\gamma_{min}}^{\gamma_{max}} d\gamma\;
1478: {N^\prime_e(\g/\dD )\over \gamma^2 } \; F_{\rm C}(q,\Gamma_e)\;,
1479: \label{jesCpowerlaw}
1480: \end{equation}
1481: where $F_{\rm C}(q,\Gamma_e)$ is given by Jones's formula \citep{jon68},
1482: eq.\ (\ref{fcq}).
1483: Because the scattering is taking place in the stationary frame,
1484: \begin{equation}
1485: q \equiv {\e_s/\g \over \Gamma_e
1486: (1-\e_s/\g )}\;,
1487: \label{qnormal}
1488: \end{equation}
1489: with $\Gamma_e = 4\e_*\g\;$ and, as before, $ \e_s = (1+z) \e$.
1490: The limits $\gamma_{min}$
1491: and $\g_{max}$ are given by
1492: \begin{equation}
1493: \g_{min} = {1\over 2} \e_s\;\left( 1+\sqrt{1+{1\over \e_*\e_s}} \;\right)
1494: \label{gpmin1}
1495: \end{equation}
1496: and
1497: \begin{equation}
1498: \g_{max} = {\e_*\e_s\over \e_* - \e_s}H(\e_* - \e_s) \;+\; \g_2H(\e_s - \e_* )\;.
1499: \label{gpmax1}
1500: \end{equation}
1501:
1502:
1503: For the
1504: case of the cosmic microwave background radiation (CMBR),
1505: \begin{equation}
1506: {u_*(\e_*)\over 4\pi }\;= \; u_{bb}(\e_*,\Omega_* ) = {2m_e c^3\over \lambda_{\rm C}^3}\;
1507: {\e_*^3\over \exp(\e_*/\Theta ) - 1 }\;,
1508: \label{blackbody}
1509: \end{equation}
1510: where $\Theta = \kB T/m_ec^2$ is the dimensionless temperature
1511: of the blackbody radiation field, and $T = 2.72(1+z)$ K.
1512: Substituting eq.\ (\ref{blackbody})
1513: in eqs.\ (\ref{fecompton}) and (\ref{esLCesomegas_1}) gives
1514: \begin{equation}
1515: f_\e^{{\rm C},CMBR} = {3 m_ec^3\sigma_{\rm T}\e_s^2 \dD^3\over 2 d_L^2 \lambda_{\rm C}^3}
1516: \int_0^\infty d\e_*\; {\e_*\over \exp(\e_*/\Theta ) - 1 }\int_{\g_{min}}^{\g_{max}} d\gamma
1517: \; \gamma^{-2} N^\prime_e(\g/\dD ) \;F_{\rm C} (q,\Gamma_e)\;.
1518: \label{eslcsses}
1519: \end{equation}
1520: Eq.\ (\ref{eslcsses}) gives an accurate
1521: spectrum of radiation made when jet electrons with number spectrum
1522: $N^\prime_e(\gp )$ Compton-scatter blackbody
1523: photons \citep[see, e.g.][]{tav00,da02,bdf08}.
1524:
1525: \subsection{Scattered BLR Radiation Field}
1526:
1527: The BLR is thought to consist of dense clouds with a specified covering factor
1528: that can be determined from AGN studies. These clouds
1529: intercept central-source radiation
1530: to produce the broad emission lines in broad-line AGN
1531: \citep{kn99}. The diffuse gas will also
1532: Thomson scatter the central source radiation. The scattered radiation provides an important
1533: source of target photons that jet electrons scatter to $\gamma$-ray energies \citep{sbr94,sik97}.
1534: This radiation field also attenuates $\gamma$ rays produced within
1535: the BLR \citep{bl95,lb95,bd95,dp03,lb06,rei07,sb08}.
1536:
1537: %\clearpage
1538:
1539: \begin{figure}[t]
1540: \center
1541: {\includegraphics[scale=0.5]{f8_color}}
1542: \caption[]{
1543: Idealized geometry of the scattering region, approximated as a spherically-symmetric shell of
1544: gas with radial Thomson depth $\tau_{\rm T}$ between inner and outer radii
1545: $R_i$ and $R_o$, respectively, and a density gradient defined by index $\zeta$, where
1546: $n_e(R) \propto R^{\zeta}$. }
1547: \label{f8}
1548: \end{figure}
1549:
1550: %\clearpage
1551:
1552: Here we calculate the angular distribution of the Thomson-scattered radiation
1553: from a shell of gas with density
1554: \begin{equation}
1555: n_e(R)\; = n_0 \left( {R\over R_i} \right)^{\zeta}\;H(R;R_i,R_o)
1556: \label{ner}
1557: \end{equation}
1558: extending from inner radius $R_i$ to outer radius $R_o$ (see Fig.\ \ref{f8};
1559: the calculation of fluorescence
1560: atomic-line radiation differs by considering dense clouds with a volume filling
1561: factor).
1562: The shell is assumed to be spherically symmetric with a power-law radial density
1563: distribution and radial Thomson depth $\tau_{\rm T} = \sT\int_{R_i}^{R_o} dR\; n_e(R)$.
1564: The Thomson-scattered spectral photon density can be estimated by noting
1565: that a fraction $\approx rn_e(r)\sT $ of the central source radiation with ambient
1566: photon density $n_{ph}(\e_*; r) = \dot N_{ph}(\e_* )/4\pi r^2 c$ is scattered, giving
1567: a target scattered radiation field
1568: \begin{equation}
1569: n_{sc}(\e_*;r)\approx {n_e(r)\sT \dot N_{ph}(\e_*)\over 4\pi r c}\;
1570: \label{nscer}
1571: \end{equation}
1572: when $R_i \lesssim r \lesssim R_0$.
1573: Here $\dot N_{ph}(\e_* ) = L(\e_* )/(m_ec^2 \e_* )$ is the central source
1574: photon-production rate, assumed to radiate isotropically,
1575: and $L(\e_* )$ is its spectral luminosity.
1576:
1577: A more accurate calculation of the Thomson-scattered photon density is obtained by integrating
1578: the expression
1579: \begin{equation}
1580: n_{ph}(\e_* ; R)\; = \; \int dV \; {\dot n (\e_*; \vec R ) \over 4\pi x^2 c}\;
1581: \label{nph}
1582: \end{equation}
1583: over volume, where $dV = R^2 d\phi d\mu dR$ and $x^2 = R^2 + r^2 -2rR\cos\theta$ (Fig.\ \ref{f8}).
1584: Assuming that the photons
1585: are isotropically Thomson-scattered by an electron
1586: without change in energy, so that $ \dot n (\e_*; \vec R )
1587: = \dot N_{ph}(\e_*)\sT n_e(R)/(4\pi R^2 c)$, then
1588: \begin{equation}
1589: n_{ph}(\e_* ; r)\;
1590: = \;
1591: {L(\e_* )\sT \over 8\pi m_ec^3 \e_*} \int_{-1}^1 d\mu\; \int_{0}^\infty dR \; {n_e(R)\over x^2}
1592: \;= \;
1593: {L(\e_* )\sT \over 8\pi m_ec^3 \e_* r} \int_{0}^\infty dR \; {n_e(R)\over R}\;
1594: \ln\left|{R+r\over R-r}\right|
1595: \label{nph7}
1596: \end{equation}
1597: from a
1598: spherically-symmetric electron density distribution
1599: \citep[cf.][for a time-dependent treatment]{gou79,bd95}. In the case of an isotropic, uniform surrounding medium,
1600: $n_e(R) = n_0$, $\zeta = 0$, $R_i = 0$ and
1601: $R_o \rightarrow \infty$ in eq.\ (\ref{ner}), and eq.\ (\ref{nph7}) gives
1602: $n_{ph}(\e_*; r) = 3.324\dot N_{ph}(\e_*)\sT n_0/ (8\pi r c)$, noting that the integral $\int_0^\infty du\;
1603: \ln|(1+u)/(1-u)|/u \cong 3.324$.
1604: Thus the approximation given by eq.\ (\ref{nscer}) is accurate to
1605: within a factor of $\approx 2$ for this case.
1606:
1607: The angle-dependent scattered photon distribution $n_{ph}(\e_*, \mu_*; r)$ can
1608: be derived by imposing a $\delta$-function constraint on the angle
1609: $\theta_* = \arccos \mu_*$ in eq.\ (\ref{nph}) \citep[see also][]{dp03}, so that
1610: \begin{equation}
1611: n_{ph}(\e_*, \mu_*; r)\;=\;{\sT \dot N_{ph}(\e_* )\over 8\pi c r}\;\int _{-\mu_*}^1 d\mu\;\int_0^\infty dg\;
1612: {n_e(gr)\delta [\mu_* - \bar\mu_*(\mu,g)]\over g^2 + 1 - 2g\mu }\;,
1613: \label{nphemuz}
1614: \end{equation}
1615: after changing variables to $g = R/r$. From Fig.\ \ref{f8}, $\theta_{max} = \pi - \theta_*$,
1616: so that $\mu_{min} = \cos\theta_{max} = \pi - \theta_*$ and $-\mu_* \leq \mu \leq 1$.
1617: The law of sines gives $R^2 (1-\mu^2) = x^2 (1-\mu_*^2)$, with the result
1618: \begin{equation}
1619: \bar\mu_*(\mu,g) \;=\; {\pm (1-g\mu)\over \sqrt{1+g^2 - 2 g \mu}}\;.
1620: \label{barmustarmuy}
1621: \end{equation}
1622: Transforming the $\delta$-function in $\mu_*$ to a $\delta$-function in $g$
1623: gives, after solving,
1624: \begin{equation}
1625: n_{ph}(\e_*,\mu_*;r) \; = \; {\sT \dot N_{ph}(\e_* )\over 8\pi c r }\;
1626: {\cal N}(\mu_*,r)\;,
1627: \label{nphemuz1}
1628: \end{equation}
1629: where
1630: \begin{equation}
1631: {\cal N}(\mu_*,r) \;=\; {\cal N}[\mu_*,n(r)]\;\equiv \;
1632: \int_{-\mu_*}^1 d\mu\; n_e(\bar gr) \;{\sqrt{1+\bar g^2 - 2\bar g \mu} \over \bar g ( 1- \mu^2) }\;
1633: \label{nphemuz111}
1634: \end{equation}
1635: (units of ${\cal N}$ are $1/L^3$), and
1636: \begin{equation}
1637: \bar g \;=\; \bar g(\mu,\mu_*) \;\equiv\; {-\mu (1- \mu_*^2) + \mu_*\sqrt{(1-\mu^2 )(1-\mu_*^2)}\over
1638: \mu_*^2 -\mu^2 }\;.
1639: \label{gbar}
1640: \end{equation}
1641:
1642: %\clearpage
1643:
1644: \begin{figure}[t]
1645: \center
1646: {\includegraphics[scale=0.6]{f9_color}}
1647: \caption[]{
1648: Angle-dependent density of Thomson-scattered radiation at height $r$ along
1649: the jet axis. The radial Thomson depth of the spherically-symmetric scattering
1650: medium is $\tau_{\rm T} = 0.01$, and the scattering
1651: shell is assumed to extend from $10^2 $ to $10^5$ Schwarzschild
1652: radii for a $10^8 M_\odot$ black hole. The solid and dashed curves
1653: show results with $\zeta = 0$ and $\zeta = -2$, respectively. The values of $r$ for
1654: are labeled on the dashed curves. The ratio of the luminosity of the photon
1655: source to the dimensionless photon energy is $L_*/\e_* =10^{44}$ ergs s$^{-1}$, and
1656: the calculation assumes that $\e_*\ll 1$,
1657: so that scattering is in the Thomson regime (see eq.\ [\ref{dotNphestar}]).}
1658: \label{fignsc}
1659: \end{figure}
1660:
1661: %\clearpage
1662:
1663: Fig.\ \ref{fignsc} shows the angle dependence of the scattered radiation field
1664: in the stationary frame for this idealized geometry for the parameters
1665: given in the figure caption. As can be seen, the scattered radiation field
1666: is nearly isotropic when $r < R_i$,
1667: and starts to display increasing asymmetry peaked in the outward direction at
1668: increasingly greater heights. When $r > R_o$, all scattered radiation is outwardly
1669: directed. Most of the scattering material is near the inner
1670: edge when $\zeta = -2$, so that the intensity of the scattered radiation field
1671: is largest at $r\lesssim R_i$. By contrast, the intensity of the scattered radiation
1672: field is not so enhanced towards the inner regions when $\zeta = 0$.
1673:
1674: The Compton-scattered radiation spectrum is given, in general, by eq.\ (\ref{esLCesomegas_26}).
1675: Substituting eq.\ (\ref{nphemuz1}) for the angular distribution of the
1676: target photon source gives, for a monochromatic photon source
1677: \begin{equation}
1678: \dot N_{ph}(\e_* ) = {L_0\delta(\e_* - \e_{*0})\over m_ec^2 \e_*}\;,
1679: \label{dotNphestar}
1680: \end{equation}
1681: the $\nu F_\nu$ flux
1682: \begin{equation}
1683: f_\e^{EC,scat}(r) \; = \;{(\pi r_e^2)^2 L_0 \dD^3\over 12\pi^2 d_L^2 r}\;
1684: \left({\e_s\over \e_*}\right)^2\;\int_{\max(-1,1-2\e_s/\e_*)}^1 d\mu_*\;
1685: {\cal N}(\mu_*,r)\;\int_{\tilde \g_{low}}^\infty d\g\;{N_e^\prime(\g/\dD )
1686: \over \g^2}\;\Xi\;.
1687: \label{feEC_scat}
1688: \end{equation}
1689: In this expression,
1690: \begin{equation}
1691: \tilde \g_{low} \;\equiv \; {\e_s \over 2}\;\left[ 1 + \sqrt{1 + {2\over \e_*\e_s(1-\mu_*)}}\;\right]\;
1692: \label{tildeglow}
1693: \end{equation}
1694: (compare eq.\ [\ref{glow}]).
1695: Substituting eq.\ (\ref{nphemuz1}) into eq.\ (\ref{dtauggdx})
1696: for a monochromatic photon source, eq.\ (\ref{dotNphestar}),
1697: gives
1698: \begin{equation}
1699: \tau_{\g\g}(\e_1,r) \; = \;
1700: {\sigma_{\rm T} L_0 \over 8\pi m_ec^3 \e_*}\;\int_r^\infty {dx\over x}\;
1701: \int_{-1}^{1 - 2/[\e_1\e_*(1+z)]} d\mu_*\;(1-\mu_*)\;{\cal N}(\mu_*,x)\sigma_{\g\g}(s)\;
1702: \label{taugge1r}
1703: \end{equation}
1704: for the opacity of a photon with measured energy $\e_1$ emitted outward
1705: along the jet axis at height $r$. The $\g\g$ opacity vanishes
1706: when $\e_1 \leq 1/[\e_*(1+z)]$ due to the $\g\g$ pair-production
1707: threshold.
1708:
1709: \subsubsection{Thomson-Scattered Isotropic Monochromatic Radiation Field}
1710:
1711: Substituting
1712: $u_*(\e_*,\Omega_*) = u_{*0}\delta(\e_* - \e_{*0})/4\pi$
1713: for an isotropic monochromatic radiation field in eq.\ (\ref{esLCesomegas_2})
1714: gives, with $\Xi \rightarrow 2$ for the Thomson regime away from the endpoints
1715: of the spectrum,
1716: \begin{equation}
1717: f_\e^{{\rm T},iso} = {c\pi r_e^2 \dD^{3+p} u_{*0}\over 4\pi d_L^2}\;
1718: \left({\e_s \over \e_*}\right)^2\;\int_{\max(-1,1-2\e_s/\e_{*})}^1 d\mu_*\;
1719: \int_{\sqrt{\e_s/2\e_*(1-\mu_*)}}^\infty d\g\; \g^{-2}N_e^\prime(\g/\dD )\;.
1720: \label{feTiso}
1721: \end{equation}
1722: We consider Compton up-scattering, i.e., $\e_s >\e_*$ for the power-law electron
1723: spectrum given by eq.\ (\ref{Neprime}).
1724: The Thomson approximation is only valid far from the endpoints.
1725: The asymptotes for the Thomson-scattered spectrum of a surrounding
1726: isotropic, monochromatic radiation field therefore becomes
1727: \begin{equation}
1728: f_\e^{{\rm T},iso} \cong {2c\pi r_e^2 u_{*0}\over 4\pi d_L^2 (p+1)}\;
1729: K^\prime\left({\e_s \over \e_*}\right)^2\;\dD^2 \g_1^{\prime~-(p+1)}\;,\;{\rm for}\;\;
1730: \e_* \lesssim \e_s \ll 4\e_s(\dD^2\gamma_1^\prime)^2\;,
1731: \label{feTisocase1}
1732: \end{equation}
1733: and
1734: \begin{equation}
1735: f_\e^{{\rm T},iso} \cong {2^{p+3}\over (p+1)(p+3)}\;{c\pi r_e^2 \dD^{3+p} u_{*0}\over 4\pi d_L^2 }\;
1736: K^\prime\left({\e_s \over \e_*}\right)^{(3-p)/2}\;,\;{\rm for}\;\;
1737: 4\e_s(\dD^2\gamma_1^\prime)^2 \lesssim \e_s \ll 4\e_s(\dD^2\gamma_2^\prime)^2\;.
1738: \label{feTisocase2}
1739: \end{equation}
1740: Note the different beaming factors in the two asymptotes.
1741: Eq.\ (\ref{feTisocase2}) agrees with the Thomson expression
1742: derived by \citet{dss97}, eq.\ (22), to within factors of order unity.
1743:
1744:
1745: \subsubsection{Numerical Calculations}
1746:
1747: We now present calculations of the SED of FSRQ blazars, including the
1748: external Compton-scattering component formed by jet electrons that
1749: scatter target photons which themselves were previously scattered by
1750: BLR material. This EC BLR component is not found in conventional
1751: synchrotron/SSC models of blazars, and so distinguishes standard model
1752: blazar BL Lacs and FSRQs. The inclusion of the $\g\g$ opacity through
1753: the scattered radiation makes the calculation self-consistent.
1754:
1755: %\clearpage
1756:
1757: \begin{figure}[t] \plottwo{f10a_color}{f10b_color}
1758: \caption{Standard model FSRQ blazar with standard parameters
1759: given in Table 1. The optical depth
1760: through the BLR is $\tau_{\rm T} = 0.1$, with no
1761: gradient in the density of the scattering material. The radiating jet is
1762: located at $10^3 R_g$. Separate synchrotron, accretion-disk, SSC, EC Disk, and EC BLR
1763: components are shown. (a) The BLR extends from $100$ to $200 R_g$.
1764: (b) The BLR extends from $100$ to $1000 R_g$.}
1765: \label{f10}
1766: \end{figure}
1767:
1768: %\clearpage
1769:
1770: For the simplified shell geometry depicted in Fig.\ \ref{f8},
1771: we calculate the external Compton scattering component using eq.\ (\ref{feEC_scat}), and calculate
1772: the $\g\g$ opacity using eq.\ (\ref{taugge1r}). To simplify
1773: the calculations, the spectrum of the radiation scattered by the BLR
1774: is assumed to be monochromatic with energy $\approx 50$ eV,
1775: corresponding to the mean energy from the accretion-disk radiation for
1776: the standard model.
1777: Results of such a calculation are shown in Fig.\ \ref{f10}, using parameters for
1778: a standard FSRQ blazar model given in Table 1. The constant-density
1779: BLR is confined
1780: between 100 and 200 $R_g$, and the Thomson depth through the BLR is $\tau_{\rm T} = 0.1$.
1781: In Fig.\ \ref{f10}(a), the emission region of the jet is far outside the BLR, at 1000 $R_g$.
1782: Therefore most of the BLR photons encounter
1783: the jet from behind, and the rate of tail-on scatterings is suppressed by
1784: the rate factor. Consequently, the flux of the EC BLR component
1785: is much less that the flux of the EC Disk component.
1786: Fig. \ref{f10}(b) presents a similar calculation, except that the BLR now
1787: extends to $10^3 R_g$, the same radius where the radiating jet is assumed to
1788: be located.
1789: %The optical depth through the BLR is now increased to $\tau_{\rm T} = 0.1$.
1790: As expected, the EC BLR component is significantly enhanced
1791: compared to Fig.\ \ref{f10}(a).
1792:
1793: %\clearpage
1794:
1795: \begin{figure}[h] \plotone{f11_color}
1796: \caption{Model blazar SED comprised of synchrotron, accretion-disk, SSC,
1797: EC disk, and EC BLR components, including $\g\g$ attenuation by the
1798: scattered BLR radiation. The jet height $r = 10^3 R_g$.
1799: The outer radius, $R_o$, of the BLR is varied,
1800: keeping its Thomson depth, $\tau_{\rm T} = 0.01$, constant.}
1801: \label{f11}
1802: \end{figure}
1803:
1804: %\clearpage
1805:
1806: There is very little $\g\g$ absorption by the accretion-disk radiation
1807: or the BLR radiation when the jet is found outside the BLR. On the other
1808: hand, when the jet is within the BLR, there can be significant $\g\g$ opacity.
1809: This is shown in Fig.\ \ref{f11}, where we use the standard parameters
1810: for the jet and a BLR with $\tau_{\rm T} = 0.01$, except that the outer
1811: radius, $R_o$, of the BLR is varied. The effects of $\g\g$ attenuation by
1812: the scattered radiation field are shown. When $R_o << r$, the jet height,
1813: then the threshold is suppressed except for the highest-energy $\gamma$ rays.
1814: When the blob lies within the BLR, $r\lesssim R_o$,
1815: the opacity from the scattered
1816: BLR radiation is significant and the $\g\g$ opacity large for $\gamma$ rays
1817: with $\e\sim 2/\e_*$, which for our monochromatic radiation
1818: field with $\e_*\cong 10^{-4}$, is at $\nu \sim 5\times 10^{24}$ Hz ($\approx
1819: 40$ GeV). When $R_o\gg r$, the $\g\g$ opacity, proportional
1820: to $r n_{sc}(\e_*;r)$, declines $\propto R_0^{-1}$ for constant Thomson
1821: depth, as can be seen from eq.\ (\ref{nscer}). Note that the EC $\gamma$ rays are not very sensitive to the changes
1822: in the BLR parameters, since most of the emission is formed by the EC disk
1823: component.
1824:
1825: %\clearpage
1826:
1827: \begin{figure}[h] \plotone{f12_color}
1828: \caption{Same as Fig.\ \ref{f11}, except that the gradient $\zeta = -2$
1829: in the density distribution. }
1830: \label{f12}
1831: \end{figure}
1832:
1833: %\clearpage
1834:
1835: The effect of changing the radial gradient of BLR scattering material
1836: is shown in Fig.\ \ref{f12}. For a steeper density gradient, $\zeta = -2$,
1837: and a constant Thomson depth, the material is concentrated near the
1838: inner edge of the BLR at $R_i$, so the changes in the $\g\g$ opacity are
1839: most dramatic when $R_o \approx r$ due to geometric effects.
1840: When $R_i \ll r \ll R_o$, the radiation field is essentially unchanged
1841: for different values of $R_o$, and so also is the $\g\g$ opacity.
1842:
1843: \section{Discussion and Summary}
1844:
1845: We have presented accurate expressions for modeling synchrotron and
1846: Compton-scattered radiation from the jets of AGN that include
1847: target radiation fields from the accretion disk and BLR. This extends our
1848: technique for modeling synchrotron and SSC emission \citep{fdb08} to
1849: include external Compton scattering in a relativistic jet of thermal
1850: radiation from the accretion disk, and accretion-disk
1851: radiation Thomson-scattered by electrons in the BLR. These formulae use the full
1852: Compton cross section and are accurate throughout the Thomson and
1853: Klein-Nishina regimes at any angle with respect to the jet axis, so
1854: can also be used to model $\gamma$-ray emission from radio galaxies, e.g.,
1855: M87 \citep{aha06}. We also derive expressions for the $\g\g$ opacity
1856: through the same scattered radiation field that serves as a target
1857: photon source for the jet electrons.
1858: The expressions, eqs.\ (\ref{tauggSStilder}) and (\ref{taugge1r}), for
1859: opacity from the accretion-disk and scattered radiation field assume, however, that the photon
1860: travels along the jet axis, though it is straightforward to derive the more
1861: general case.
1862:
1863: In the results presented here (Figs.\ \ref{f1} -- \ref{f6} and
1864: \ref{f10} -- \ref{f12}) we have chosen parameters for demonstration
1865: purposes that give an exaggerated EC component, particularly a high
1866: $\ell_{\rm Edd}$. Lowering the disk luminosity lowers the radiation
1867: considerably, as seen in Fig.\ \ref{f2}. The disk radiation field in
1868: FSRQ blazars can be seen when the nonthermal blazar radiation is in a
1869: low state, as in the cases of 3C 279 \citep{pia99}, 3C 454.3
1870: \citep{rai07} and, most clearly, 3C 273
1871: \citep[e.g.,][]{mon97}. These observations can be used to assign the
1872: accretion-disk luminosity when modeling a specific blazar, though the
1873: disk brightness could also vary during the flaring epoch.
1874:
1875: The models presented here do, however, have limitations. They do not
1876: yet include enhancements from secondary cascade radiation initiated by
1877: $e^\pm$ pairs formed by $\g$ rays interacting with lower energy
1878: radiation from the disk and the BLR. For the parameters considered
1879: here, this would not make a significant difference in the calculated
1880: SEDs because the energy flux of the attenuated radiation is a small
1881: fraction of the escaping flux. But even in this case, the reinjected
1882: pairs from the attenuated radiation will be isotropized if the
1883: reinjection occurs outside the relativistic flow, and will then make
1884: only a small contribution to the Doppler-boosted radiation.
1885:
1886: Spectral features result from $\gamma\gamma$ absorption by BLR radiation,
1887: as seen in Fig.\ \ref{f12}. By assuming hard primary $\gamma$-ray
1888: emission components,
1889: \citet{aha08} argue that hard intrinsic spectra from blazars such
1890: as 1ES 1101-232 \citep{aha07-1101} could be formed
1891: through $\gamma\gamma$ attenuation. If the primary TeV radiation
1892: originates from an underlying
1893: jetted electron distribution, then a consistent model requires that a
1894: $\gamma$-ray spectrum formed by Compton-scattering processes
1895: arises from the same radiation field responsible for $\gamma\gamma$ absorption.
1896: %(in addition to a $\gamma$-ray spectral component formed by electrons that Compton-scatter
1897: %the direct accretion-disk radiation).
1898: As our calculations show, soft Compton-scattered TeV spectra are formed due
1899: to Klein-Nishina effects in scattering, whether from the accretion disk
1900: or from photons scattered by the BLR. Thus either an extremely bright
1901: GeV component would be expected in the scenario of \citet{aha08} (cf.\ Figs.\ \ref{f11}
1902: and \ref{f12}),
1903: which would easily be detected with {\em Fermi},
1904: or a hard primary $\gamma$-ray spectrum must originate from other processes.
1905: Moreover, if this explanation was correct, then blazars
1906: with stronger broad emission lines should have harder VHE $\gamma$-ray
1907: spectra than weak-lined BL Lacs.
1908:
1909: Cascade radiation induced by ultra-relativistic hadrons could inject
1910: high-energy leptons to form a hard radiation component, though a
1911: sufficiently dense target field for efficient photohadronic losses
1912: will itself severely attenuate the TeV radiation
1913: \citep{ad03}. Depending on the underlying acceleration model, a
1914: distinctive hadronic signature could appear at $\gamma$-ray energies
1915: only, though one would expect that both electrons and protons would be
1916: accelerated synchronously.
1917:
1918:
1919: Our calculations also do not as yet include absorption by the diffuse
1920: extragalactic background light (EBL), which would be important for
1921: EGRET $\gamma$-ray FSRQs, which have a broad redshift distribution
1922: with a mean value $\langle z\rangle \sim 1$, larger than the mean
1923: value $\langle z\rangle \sim 0.3$ for BL Lacs \citep{muk97}. In cases
1924: where the Compton-scattered spectra decline steeply due to
1925: Klein-Nishina effects (e.g., Fig.\ \ref{f1}), the issue of EBL
1926: absorption is secondary. In the one FSRQ, \object{3C 279} at
1927: $z=0.536$, detected from 80 to $> 300$ GeV energies with MAGIC
1928: \citep{mag08}, EBL effects cannot be neglected. For the models
1929: studied here, the soft calculated $\gamma$-ray spectra cannot in any
1930: case account for the measured, let alone intrinsic, spectrum of 3C
1931: 279. Based on the simultaneous optical -- X-ray -- VHE $\gamma$-ray
1932: SED of 3C279 during the MAGIC detection, B\"ottcher, Marscher \&
1933: Reimer (2008, in preparation) show that one-zone leptonic models have
1934: severe problems explaining the flare, and require either extremely
1935: high Doppler factors or magnetic fields well below equipartition.
1936: Correlated X-ray, {\em Fermi}, and TeV campaigns will offer
1937: opportunities to apply the results developed in this paper.
1938:
1939: In conclusion, we have derived expressions to model FSRQ blazars that self-consistently
1940: include $\g\g$ attenuation on the same target photons that are Compton-scattered by
1941: the relativistic electrons. By assuming that the lower energy, radio -- UV emission
1942: is nonthermal synchrotron radiation, then the underlying electron distribution can
1943: be determined given the Doppler factor, magnetic field, and size scale of the
1944: emission region. The equations derived here can be used to calculate the
1945: $\gamma$-ray emission spectrum from SSC and external Compton scattering processes
1946: from this electron distribution. Separately, one can determine whether the
1947: inferred electron distribution can be derived from specific acceleration scenarios.
1948: Complementary to the technique of injecting electron spectra and cooling, this
1949: method can be applied to multiwavelength data sets to analyze high-energy
1950: processes in the jets of AGN.
1951:
1952: %----------------------------------------------------------------------------------------
1953: %----------------------------------------------------------------------------------------
1954: \acknowledgements
1955:
1956: We thank the referee for asking us to address the potential importance
1957: of higher-order SSC fluxes, and for other helpful comments. The work
1958: of J.D.F. is supported by NASA {\em Swift} Guest Investigator Grant
1959: DPR-NNG05ED411 and NASA {\em GLAST} Science Investigation
1960: DPR-S-1563-Y, which also supported summer research by H.K. at NRL, and
1961: a visit by M.B. to NRL. C.D.D. is supported by the Office of Naval
1962: Research.
1963:
1964:
1965: %!****************************************************
1966:
1967:
1968: \begin{thebibliography}{}
1969:
1970: \bibitem[Aharonian et al.(2006)]{aha06} Aharonian, F.,
1971: Akhperjanian, A.~G., Bazer-Bachi, A.~R., \& et al.\ 2006, Science, 314, 1424
1972:
1973: %\bibitem[Aharonian et
1974: %al.(2007a)]{aha07-0229} Aharonian, F., et al.\ 2007a, \aap, 475, L9
1975:
1976: \bibitem[Aharonian et
1977: al.(2007b)]{aha07-1101} Aharonian, F., et al.\ 2007b, \aap, 470, 475
1978:
1979: \bibitem[Aharonian et al.(2008)]{aha08} Aharonian, F.~A.,
1980: Khangulyan, D., \& Costamante, L.\ 2008, \mnras, 387, 1206
1981:
1982: \bibitem[Atoyan \& Dermer(2003)]{ad03} Atoyan, A.~M., \& Dermer, C.~D.\ 2003, \apj, 586, 79
1983:
1984:
1985:
1986: \bibitem[Band et al.(1993)]{band93} Band, D., et al.\ 1993,
1987: \apj, 413, 281
1988:
1989: \bibitem[Becker \& Kafatos(1995)]{bk95} Becker, P.~A., \& Kafatos, M.\ 1995, \apj, 453, 83
1990:
1991: \bibitem[Blandford \& Levinson(1995)]{bl95} Blandford,
1992: R.~D., \& Levinson, A.\ 1995, \apj, 441, 79
1993:
1994: \bibitem[B{\l}a{\.z}ejowski et al.(2000)]{bsmm00}
1995: B{\l}a{\.z}ejowski, M., Sikora, M., Moderski, R., \& Madejski, G.~M.\ 2000,
1996: \apj, 545, 107
1997:
1998: \bibitem[Bloom \& Marscher(1996)]{bm96} Bloom, S.~D., \&
1999: Marscher, A.~P.\ 1996, \apj, 461, 657
2000:
2001:
2002: \bibitem[Blumenthal \& Gould(1970)]{bg70} Blumenthal, G.~R.,
2003: \& Gould, R.~J.\ 1970, Reviews of Modern Physics, 42, 237
2004:
2005: \bibitem[B{\"o}ttcher et al.(2008)]{bdf08} B{\"o}ttcher, M.,
2006: Dermer, C.~D., \& Finke, J.~D.\ 2008, \apjl, 679, L9
2007:
2008: \bibitem[B{\"o}ttcher \& Dermer(1995)]{bd95} B{\"o}ttcher, M., \& Dermer, C.~D.\ 1995, \aap, 302, 37
2009:
2010: \bibitem[B{\"o}ttcher
2011: \& Bloom(2000)]{bb00} B{\"o}ttcher, M., \& Bloom, S.~D.\ 2000, \aj, 119, 469
2012:
2013: \bibitem[B\"ottcher et al.(1997)]{bms97} B\"ottcher, M., Mause,
2014: H., \& Schlickeiser, R.\ 1997, \aap, 324, 395
2015:
2016: \bibitem[B{\"o}ttcher \& Reimer(2004)]{br04} B{\"o}ttcher,
2017: M., \& Reimer, A.\ 2004, \apj, 609, 576
2018:
2019: \bibitem[Brown, Mikaelian, \& Gould(1973)]{bmg73} Brown, R.~W., Mikaelian, K.~O.,
2020: \& Gould, R.~J.\ 1973, Astrophys.\ Lett.\ 14, 203
2021:
2022: \bibitem[Celotti \& Fabian(1993)]{celotti93}
2023: Celotti, A., \& Fabian, A.~C.\ 1993, \mnras, 264, 228
2024:
2025: \bibitem[Celotti et al.(2007)]{celotti07} Celotti, A.,
2026: Ghisellini, G., \& Fabian, A.~C.\ 2007, \mnras, 375, 417
2027:
2028: \bibitem[Dermer \& Atoyan(2002)]{da02} Dermer, C.~D., \&
2029: Atoyan, A.~M.\ 2002, \apjl, 568, L81
2030:
2031: %\bibitem[Dermer(1995)]{der95} Dermer, C.~D.\ 1995, \apjl, 446, L63
2032: \bibitem[Dermer et al.(2007)]{drl07} Dermer, C.~D.,
2033: Ramirez-Ruiz, E., \& Le, T.\ 2007, \apjl, 664, L67
2034:
2035:
2036:
2037: \bibitem[Dermer \& B\"ottcher(2006)]{db06} Dermer, C.~D.,
2038: B\"ottcher, M.\ 2006, \apj, 643, 1081
2039:
2040: \bibitem[Dermer et al.(1992)]{dsm92} Dermer, C.~D.,
2041: Schlickeiser, R., \& Mastichiadis, A.\ 1992, \aap, 256, L27
2042:
2043: \bibitem[Dermer \& Schlickeiser(1993)]{ds93} Dermer, C.~D.,
2044: \& Schlickeiser, R.\ 1993, \apj, 416, 458
2045:
2046: \bibitem[Dermer et al.(1997)]{dss97} Dermer, C.~D., Sturner,
2047: S.~J., \& Schlickeiser, R.\ 1997, \apjs, 109, 103
2048:
2049: \bibitem[Dermer \& Schlickeiser(2002)]{ds02} Dermer, C.~D.,
2050: \& Schlickeiser, R.\ 2002, \apj, 575, 667
2051:
2052: \bibitem[Donea \& Protheroe(2003)]{dp03}
2053: Donea, A.-C., \& Protheroe, R.~J.\ 2003, Astroparticle Physics, 18, 377
2054:
2055: \bibitem[Fichtel et al.(1994)]{fic94} Fichtel, C.~E., et al.\
2056: 1994, \apjs, 94, 551
2057:
2058: \bibitem[Finke et al.(2008)]{fdb08} Finke, J.\ D., Dermer, C.\ D., \& B\"ottcher,
2059: M.\ 2008, \apj, in press (arXiv:0802.1529)
2060:
2061: \bibitem[Georganopoulos et al.(2001)]{gkm01} Georganopoulos,
2062: M., Kirk, J.~G., \& Mastichiadis, A.\ 2001, \apj, 561, 111;
2063: (e) 2004, \apj, 604, 479
2064:
2065: \bibitem[Gould(1979)]{gou79} Gould, R.~J.\ 1979, \aap, 76, 306
2066:
2067: \bibitem[Gould \& Schr{\'e}der(1967)]{gs67} Gould, R.~J., \& Schr{\'e}der, G.~P.\ 1967,
2068: Physical Review, 155, 1404
2069:
2070:
2071: \bibitem[Hartman et al.(1992)]{har92} Hartman, R.~C., et al.\
2072: 1992, \apjl, 385, L1
2073:
2074:
2075: \bibitem[Hartman et al.(1999)]{har99} Hartman, R.~C., et al.\
2076: 1999, \apjs, 123, 79
2077:
2078: \bibitem[Hermsen et al.(1977)]{her77} Hermsen, W., et al.\
2079: 1977, \nat, 269, 494
2080:
2081: \bibitem[Jauch \& Rohrlich(1976)]{jr76} Jauch, J.~M., \& Rohrlich, F.\ 1976,
2082: The Theory of Photons and Electrons
2083: (New York: Springer)
2084:
2085:
2086: \bibitem[Jones(1968)]{jon68} Jones, F.~C.\ 1968, Physical
2087: Review, 167, 1159
2088:
2089: \bibitem[Kusunose \& Takahara(2005)]{kt05} Kusunose, M.,
2090: \& Takahara, F.\ 2005, \apj, 621, 285
2091:
2092: \bibitem[Levinson \& Blandford(1995)]{lb95} Levinson, A., \& Blandford, R.\ 1995, \apj, 449, 86
2093:
2094: \bibitem[Liu \& Bai(2006)]{lb06} Liu, H.~T., \& Bai, J.~M.\ 2006, \apj, 653, 1089
2095:
2096: \bibitem[Kaspi \& Netzer(1999)]{kn99} Kaspi, S., \& Netzer, H.\ 1999, \apj, 524, 71
2097:
2098:
2099: \bibitem[Kataoka et al.(1999)]{kat99} Kataoka, J., et al.\
2100: 1999, \apj, 514, 138
2101:
2102: \bibitem[MAGIC Collaboration(2008)]{mag08} MAGIC
2103: Collaboration, et al.\ 2008, Science, 320, 1752
2104:
2105:
2106: \bibitem[Moderski et al.(2005)]{mod05} Moderski, R., Sikora,
2107: M., Coppi, P.~S., \& Aharonian, F.\ 2005, \mnras, 363, 954;
2108: (e) 2005, \mnras, 364, 1488
2109:
2110: \bibitem[von Montigny et al.(1997)]{mon97} von Montigny, C., et al.\ 1997, \apj, 483, 161
2111:
2112: \bibitem[Mukherjee et al.(1997)]{muk97} Mukherjee, R., et
2113: al.\ 1997, \apj, 490, 116
2114:
2115: \bibitem[Nikishov(1961)]{nik61} Nikishov, A.\ I.\ 1961,
2116: Zh.\ Experimen.\ i Theo.\ Fiz.\ 41, 549
2117:
2118: \bibitem[Perlman et al.(2008)]{per08} Perlman, E.~S.,
2119: Addison, B., Georganopoulos, M., Wingert, B.,
2120: \& Graff, P.\ 2008, arXiv:0807.2119
2121:
2122: \bibitem[Pian et al.(1999)]{pia99} Pian, E., et al.\ 1999,
2123: \apj, 521, 112
2124:
2125:
2126: \bibitem[Punch et al.(1992)]{pun92} Punch, M., et al.\ 1992,
2127: \nat, 358, 477
2128:
2129: \bibitem[Raiteri et al.(2007)]{rai07} Raiteri, C.~M., et al.\ 2007, \aap, 473, 819
2130: \bibitem[Reimer(2007)]{rei07}Reimer, A.\ 2007, \apj, 665, 1023
2131:
2132: \bibitem[Shakura and Sunyaev(1973)]{ss73}Shakura, N. I., and Sunyaev, R. A. 1973, A\&A, 24, 337
2133:
2134: \bibitem[Shapiro \& Teukolsky(1983)]{st83} Shapiro, S., and Teukolsky, S., 1983, Black Holes, White Dwarfs, and Neutron Stars (New York: John Wiley and Sons), chpt. 14
2135:
2136: \bibitem[Sikora et al.(1994)]{sbr94} Sikora, M., Begelman,
2137: M.~C., \& Rees, M.~J.\ 1994, \apj, 421, 153
2138:
2139: \bibitem[Sikora et al.(1997)]{sik97} Sikora, M., Madejski,
2140: G., Moderski, R., \& Poutanen, J.\ 1997, \apj, 484, 108
2141:
2142: \bibitem[Sitarek \& Bednarek(2008)]{sb08}
2143: Sitarek, J., \& Bednarek, W.\ 2008, submitted to \mnras, ArXiv e-prints, 807, arXiv:0807.4228
2144:
2145: %\bibitem[Stecker \& Scully(2008)]{stecker08}
2146: %Stecker, F.~W., \& Scully, S.~T.\ 2008, \aap, 478, L1
2147:
2148: %\bibitem[Stecker et al.(2007)]{stecker07} Stecker, F.~W., Baring,
2149: %M.~G., \& Summerlin, E.~J.\ 2007, \apjl, 667, L29
2150:
2151: \bibitem[Tavecchio et al.(1998)]{tav98} Tavecchio, F.,
2152: Maraschi, L., \& Ghisellini, G.\ 1998, \apj, 509, 608
2153:
2154: \bibitem[Tavecchio et al.(2000)]{tav00} Tavecchio, F.,
2155: Maraschi, L., Sambruna, R.~M., \& Urry, C.~M.\ 2000, \apjl, 544, L23
2156:
2157: %\bibitem[Teshima et al.(2007)]{teshima08} Teshima, M., et al.\
2158: %2007, ArXiv e-prints, 709, arXiv:0709.1475
2159:
2160: \bibitem[Weekes(2003)]{wee03} Weekes, T.~C.\ 2003, Very High
2161: Energy Gamma-ray Astronomy (Institute of Physics: Bristol, UK)
2162:
2163: \end{thebibliography}
2164:
2165: \end{document}