1: \documentclass[12pt,preprint]{aastex}
2: \usepackage{graphicx}
3: \usepackage{natbib}
4: \bibliographystyle{hapj}
5:
6: \newcommand {\ie} {{\it i.e.}}
7: \newcommand {\cf} {{\it cf.}}
8: \newcommand {\eg} {{\it e.g.}}
9: \newcommand {\ea} {{\it et~al.}}
10: \newcommand {\be} {\begin{equation}}
11: \newcommand {\ee} {\end{equation}}
12:
13: \renewcommand{\labelitemi}{$\bullet$}
14:
15: %\slugcomment{Draft: v1 MS 08/25/07; v2 GM 09/08/07; v3 RM 09/10/07;
16: %v4 MS 10/23/07; v5 RM 10/26/07}
17: \shorttitle{The 'blazar zone' in 3C454.3}
18: \shortauthors{Sikora \ea}
19:
20: \begin{document}
21:
22: \title{\object{3C454.3} reveals the structure and physics of its
23: 'blazar zone'.}
24:
25: \author{Marek~Sikora\altaffilmark{1},
26: Rafa{\l}~Moderski\altaffilmark{1},
27: Greg~M. Madejski\altaffilmark{2,3}}
28:
29: \altaffiltext{1}{Nicolaus Copernicus Astronomical Center, Bartycka 18, 00-716
30: Warsaw, Poland; \tt{sikora@camk.edu.pl}}
31: \altaffiltext{2}{Stanford Linear Accelerator Center, 2575 Sand Hill Road,
32: Menlo Park, CA 94025, USA}
33: \altaffiltext{3}{Kavli Institute for Particle Astrophysics and Cosmology,
34: Stanford University, Stanford, CA 94305}
35:
36: \begin{abstract}
37: Recent multi-wavelength observations of \object{3C454.3}, in
38: particular during its giant outburst in 2005, put severe constraints
39: on the location of the 'blazar zone', its dissipative nature, and high
40: energy radiation mechanisms. As the optical, X-ray, and millimeter
41: light-curves indicate, significant fraction of the jet energy must be
42: released in the vicinity of the millimeter-photosphere, i.e. at
43: distances where, due to the lateral expansion, the jet becomes
44: transparent at millimeter wavelengths. We conclude that this region
45: is located at $\sim 10$ parsecs, the distance coinciding with the
46: location of the hot dust region. This location is consistent with the
47: high amplitude variations observed on $\sim 10$ day time scale,
48: provided the Lorentz factor of
49: a jet is $\Gamma_j \sim 20$. We argue that dissipation is driven by
50: reconfinement shock and demonstrate that X-rays and $\gamma$-rays are
51: likely to be produced via inverse Compton scattering of near/mid IR
52: photons emitted by the hot dust. We also infer that the largest
53: gamma-to-synchrotron luminosity ratio ever recorded in this object --
54: having taken place during its lowest luminosity states -- can be
55: simply due to weaker magnetic fields carried by a less powerful jet.
56: \end{abstract}
57: %
58: \keywords{galaxies: quasars: general --- galaxies: jets --- radiation
59: mechanisms: non-thermal --- gamma rays: theory --- X-rays: general}
60:
61: \section{Introduction}
62: %
63: Multi-wavelength coverage of recent activity of quasar
64: \object{3C454.3} provided exceptional data to investigate the
65: structure and physics of its blazar zone. Prior to year 2000, this
66: object spent most of its time in the low, relatively quiescent state.
67: Starting in 2000, \object{3C454.3} entered a highly active state,
68: changing optical flux by a factor tens on time scales of a few months
69: \citep{Fuhrmann06,Villata06}. The most powerful event took place in
70: the middle of 2005. This event was monitored also in the X-ray bands
71: (Swift/XRT/BAT: \citealt{Giommi06}; INTEGRAL: \citealt{Pian06};
72: Chandra: \citealt{Villata06}), and at millimeter wavelengths
73: \citep{Krichbaum07}.
74:
75: These data allow a construction of quasi-simultaneous broadband
76: spectrum around the outburst peak. As is the case for other blazars,
77: the spectrum is composed of two humps, the lower energy one produced
78: via synchrotron mechanism and peaking in the far-infrared band, and
79: the higher energy one most likely generated by inverse-Compton process
80: and peaking in the $\gamma$-ray band. The lack of coverage of the
81: event by $\gamma$-ray observatories does not allow us to determine the
82: luminosity of the high energy component. Nevertheless, X-ray data
83: suggest that luminosity ratio of the high- to the low-energy
84: components was much smaller during the outburst than during low states
85: monitored in $\gamma$-rays by CGRO
86: \citep{Mukherjee97,Hartman99,Zhang05}.
87:
88: This difference was theoretically investigated by \citet{Pian06} and
89: by \citet{Katarzynski07}. \citet{Pian06} suggested that during the
90: low states the blazar zone is located inside the broad line region
91: (BLR) and that high energy spectra are produced by the External
92: Radiation Compton (ERC) process involving scattering of broad line
93: photons \citep[via scenario described in][]{Sikora94}, while during
94: the 2005 outburst the dissipation zone moved outside the BLR where the
95: ERC becomes inefficient. In such a model, production of the optical
96: outburst doesn't require increase of a jet power. Similarly, in the
97: scenario proposed by \citet{Katarzynski07} the jet power is constant,
98: but the drop of luminosity of the high energy component is explained
99: by decrease of the Lorentz factor.
100:
101: The idea of the constant jet power might be challenged by the most
102: recent optical outburst which in July 2007 was also detected in
103: $\gamma$-rays by AGILE \citep{Vercellone07}. Bolometric luminosity of
104: this outburst was 4-5 times larger than bolometric luminosity during
105: the low optical states and the radiative output was strongly dominated
106: by the $\gamma$-ray flux. The currently available millimeter-band light
107: curves \citep{Krichbaum07} do not indicate any significant delay of
108: the millimeter flux after the bolometric flux as inferred from the
109: infrared and optical data presented in \citet{Bach07}. All of
110: the above motivated us to investigate a different scenario, with the
111: origin of the high energy peak involving ERC with IR seed photons and
112: operating in the vicinity of the millimeter photosphere of the source.
113: Basic assumptions of the scenario are described in \S\ref{sec:assum};
114: results of modeling of the broadband spectrum of the 2005 outburst are
115: presented in \S\ref{sec:outburst}; explanation of the large
116: $\gamma$-ray dominance in the low optical states is provided in
117: \S\ref{sec:model}; and the main results are summarized in
118: \S\ref{sec:discussion}.
119:
120: \section{Model assumptions}
121: \label{sec:assum}
122: %
123: \subsection{Location of the blazar zone\label{sec:location}}
124: %
125: Optical and millimeter light-curves show that the '2005 outburst' of
126: \object{3C454.3} was actually preceded by a long term gradual increase
127: in flux which started in August 2004 and continued until the middle of
128: 2005 \citep{Villata06,Krichbaum07}. The optical flux reached maximum
129: around May 9, then dropped very rapidly, but this drop was associated
130: with several local ``wiggles''. The millimeter light-curve reached
131: maximum about 18 days later and continued at that level for $\sim 3$
132: months with fluctuations on a time scale of $\sim 10$-days. The
133: outburst ceased by the August/September 2005. The lack of a high
134: luminosity plateau in the optical light curve suggests that the
135: millimeter outburst lags the optical one by $\sim 3$ months. However
136: no such long delay is seen in the growing part of the outburst.
137: Furthermore, the optical spectrum is steep and very variable which
138: makes the optical flux a very poor tracer of the {\sl bolometric}
139: luminosity. The latter, according to data presented by
140: \citet{Bach07}, presumably reached the maximum (with the peak located
141: in the far IR) by the end of June 2005, roughly in the middle of the
142: millimeter plateau. This, coupled with large millimeter luminosities
143: which require {\sl in situ} energy dissipation rate that is comparable
144: to the rate required to account for optical emission -- and similar
145: short term variability time scales in both spectral bands -- suggest
146: that regions of the optical and millimeter emission are not spatially
147: detached.
148:
149: If the above is indeed the case, it is possible to make unambiguous
150: estimates of the location of the blazar zone (with respect to the
151: central black hole) based on the variability time scales, and this in
152: turn can be verified by using millimeter data and calculations of the
153: synchrotron-self-absorption opacity of the source. Since the spectral
154: slope measured in the millimeter band during the outburst is typically
155: within the range $0.0 < \alpha_{mm} < 0.5$, the blazar zone is
156: expected to be partially opaque at these wavelengths. The resulting
157: size of the source $R_{mm}$ and its distance from the center $r_{mm}$
158: depend on the specific model parameters and for those presented in
159: Table~\ref{tab:param} are calculated to be $R_{mm} \sim 0.5$ pc and
160: $r_{mm} \sim 9$ pc (see Appendix~\ref{app:photo}).
161:
162: \subsection{Dissipation scenario}
163: %
164: While it is relatively well-established that the endpoints of most
165: quasar jets correspond to ``hot spots'' presumably involving terminal
166: shocks, there is no consensus regarding the mechanism responsible for
167: the energy dissipation within the flow and in particular in the blazar
168: zone. Most popular, presumably because it is the easiest to treat
169: quantitatively, is the internal shock scenario. In accordance with
170: this scenario, jets are radially inhomogeneous both in density and
171: velocity and shocks are formed due to collisions between jet portions
172: propagating with different Lorentz factors \citep{Sikora94,Spada01}.
173: Internal shock scenario is attractive for blazars because predicts
174: parallel polarization (electric vector position angle, EVPA, parallel
175: to the jet) of the synchrotron radiation, in agreement with
176: observations in the optical, infrared, and millimeter bands
177: \citep{Impey91,Stevens96,Nartallo98,Jorstad07}. This prediction is
178: independent of whether magnetic field is dominated by the toroidal
179: component determined by poloidal electrical currents or by turbulent
180: magnetic fields compressed in the transverse shocks \citep{Laing81}.
181: However, internal shocks are known to dissipate energy very
182: inefficiently: modulation of a jet Lorentz factor by at least a factor
183: of $4$ is required to reach a few percent of efficiency.
184:
185: More promising dissipative scenario involves reconfinement shocks
186: \citep{Komissarov97,Sokolov04}. Such shocks keep pressure balance
187: between the jet and its environment and are formed everywhere where
188: density gradient of the external medium departs from the longitudinal
189: density gradient in a jet. On sub-parsec scales the environment is
190: too weak to affect dynamically powerful jets, but at parsec and larger
191: distance, the interaction of the jet with its environment is
192: sufficiently strong to modify the opening angle and, in the case of
193: non-axisymmetric external matter density distribution, also the
194: direction of propagation \citep[see e.g.][]{Appl96}. Reconfinement
195: shock scenario provides interesting constraints on the structure and
196: intensity of magnetic fields. In such shocks compression of chaotic
197: magnetic fields leads to the perpendicular EVPA, but if magnetic field
198: intensity is dominated by the toroidal component, the EVPA is parallel
199: to the jet, in agreement with observations.
200:
201: \subsection{Radiative mechanisms and model input parameters}
202: %
203: Basic radiative processes in relativistic jets are known to be the
204: synchrotron mechanism and the inverse-Compton process. The latter
205: involves scatterings of both 'internal' synchrotron photons (the SSC
206: process) and 'external' photons (the ERC process). The ERC is
207: expected to dominate strongly over the SSC provided radiative
208: environment is strong and jets are highly relativistic
209: \citep{Dermer92,Dermer93,Sikora94,Blandford95}. At parsec distances,
210: corresponding to the likely location of the blazar zone in
211: \object{3C454.3} (see \S\ref{sec:location}), the external diffuse
212: radiation field is dominated by near/mid infrared radiation of hot
213: dust \citep[and refs. therein]{Cleary07} and therefore such dust is
214: very likely to provide the dominant source of seed photons for the
215: inverse-Compton process \citep{Blazejowski00,Arbeiter02}. This is in
216: fact the scenario suggested for the origin of the high-energy peak in
217: MeV blazars \citep[a class of blazars also encompassing 3C454.3;
218: see][]{Sikora02} and is the scenario adopted below.
219:
220: To reproduce the broadband spectrum of radiation produced in the blazar
221: zone, we apply the numerical model BLAZAR \citep{Moderski03}, updated
222: for the treatment of the Klein-Nishina regime \citep{Moderski05}.
223: Originally, the model was designed to compute radiation spectra
224: assuming the internal shock scenario, but noting that steady-state
225: radiation can be superposed from a sequence of moving sources which
226: all radiate within the same distance range, the model can be used also
227: to approximate radiation production by the standing reconfinement
228: shock.
229:
230: Possibly the most significant simplification of our model is that we
231: do not consider real geometry and kinematics of the reconfinement
232: shock, adopting instead the uniform injection/acceleration of
233: relativistic particles within the conically diverging zone. The
234: details of the physics of reconfinement shocks and in particular of
235: particle acceleration are still not known, and it is even unclear
236: whether the dissipation process and particle acceleration involve just
237: the reconfinement shock or some sort of a hybrid model incorporating
238: internal shocks amplified in the reconfinement zone
239: \citep{Komissarov97,Sokolov04}.
240:
241: The following input parameters are used in our model:
242: %
243: \begin{itemize}
244: %
245: \item radial extension of the blazar zone, $\Delta r$, and the
246: distance of its inner edge from the center, $r_0$;
247:
248: \item the jet Lorentz factor, $\Gamma_j$, and its opening (half) angle
249: $\theta_j$;
250:
251: \item magnetic field intensity, $B=B_0 \times (r_0/r)$;
252:
253: \item the electron injection function, $Q = K \gamma^{-p}$ for
254: $\gamma_{min} < \gamma < \gamma_{max}$;
255:
256: \item energy density of the diffuse component of hot dust radiation,
257: $u_{IR} = u_{IR,in} \times [1 + (r/r_{in})^2]^{-1}$, where $r_{in}$ is
258: the inner edge of the hot dust region, $u_{IR,in} \sim \xi_{IR}
259: L_{disk} / (4 \pi r_{in}^2 c)$, $L_{disk}$ is the accretion disk
260: luminosity, and $\xi_{IR}$ is the fraction of the disk radiation
261: reprocessed by dust into infrared radiation;
262:
263: \item energy of the seed photons at thermal peak in $\nu L_{\nu}$
264: vs. $\nu$ diagram, $h\nu_{IR} \simeq 3.92 \, k T$, where $T=
265: T_{in}(r_{in}/r)^{1/2}$, and $T_{in} = (L_{disk} /(4\pi \sigma_{SB}
266: r_{in}^2 c))^{1/4}$.
267:
268: \end{itemize}
269:
270: Values of these parameters are determined by our model assumptions and
271: by relations between these parameters and observables. The latter, in
272: the form of approximate formulas, are presented in
273: Appendix~\ref{app:approx}. Analytically estimated parameters are used
274: to start an iterative procedure to fit numerically the observed
275: spectrum. Because the 2005 outburst was not observed in the
276: $\gamma$-ray band and because of uncertainties regarding distribution
277: and opacity of the hot dust, the set of input parameters cannot be
278: determined uniquely. This in particular concerns the value of the jet
279: Lorentz factor. We assumed $\Gamma_j = 20$. Such a large value
280: allows us to avoid softening of the X-ray spectrum by contribution of
281: the SSC process in the soft/mid X-ray bands. Such a large value of
282: $\Gamma_j$ is also implied when we adopt the assumption of domination
283: of the toroidal magnetic component over the turbulent one. The
284: $\Gamma_j = 20$ is larger than that deduced from the VLBI observations
285: of the superluminal expansion \citep[see][and
286: refs. therein]{Jorstad01}, but the latter can be underestimated due
287: to not taking into account effects of the divergence of a jet
288: \citep{Gopal06}.
289:
290: \section{Modeling the 2005 outburst\label{sec:outburst}}
291: %
292: Results of modeling of the spectrum observed in May 2005, when the
293: optical flux was at its maximum are shown in Fig.~\ref{fig:model1} and
294: input and output parameters are specified in Table~\ref{tab:param}.
295: As it is apparent, the entire spectrum can be reproduced using a
296: single-power-law for electron injection function, with a slope index
297: $p=2$. X-ray spectrum is produced by electrons which cool on a time
298: scale longer than the blazar-zone crossing time and therefore this
299: results in the slope $\alpha_X = (p-1)/2 \simeq 0.5$. Synchrotron
300: spectrum is produced in the fast cooling regime and results in the
301: slope $\alpha_{syn} = p/2 \simeq 1.0$, but in the optical band it
302: significantly steepens due to high energy cutoff in the injection
303: function. It hardens at the millimeter wavelengths due to synchrotron
304: self-absorption.
305:
306: Our results show that even a very moderate energy density of the dust
307: radiation is sufficient to provide strong domination of the ERC
308: luminosities over the SSC luminosities. This is due to a large value
309: of $\Gamma_j$ and strong dependence of the $L_{ERC}/L_{SSC}$ ratio on
310: $\Gamma_j$. The spectrum shown in Fig.~\ref{fig:model1} is obtained
311: for an active zone enclosed within a distance range $10^{19} - 2
312: \times 10^{19}\,$cm. Jet within this distance range is opaque at
313: millimeter wavelengths.
314:
315: In order to get spectrum with the observed slopes and fluxes in the
316: millimeter band, it is necessary to assume a larger distance of the
317: blazar zone and smaller optical luminosities. In
318: Fig.~\ref{fig:model1} we show the broadband spectrum produced within a
319: distance range $2 \times 10^{19} - 4 \times 10^{19}\,$cm. Optical
320: luminosity is smaller there by a factor $\sim 5$, but assuming that
321: magnetic energy flux is proportional to the flux associated with
322: matter flow, it was possible to accommodate this by decreasing the
323: electron injection function by only a factor of $2$ (see parameters in
324: Table~\ref{tab:param}). Optical luminosity produced within this
325: distance range corresponds with optical fluxes recorded during the
326: millimeter-plateau period. Results from Fig.~\ref{fig:model1}
327: indicate that most powerful portions of the jet start to dissipate
328: energy closer to the center than the less powerful ones, but energy
329: dissipation extends, albeit with a decreasing efficiency, up to the
330: region where the plasma becomes transparent at millimeter wavelengths.
331:
332: \section{Modeling different spectral states\label{sec:model}}
333: %
334: Important observable characterizing the double-hump spectra of blazars
335: is the luminosity ratio of the high energy component to the low energy
336: component. If production of a high energy component is dominated by
337: the ERC process, then this ratio is $L_{ERC}/ L_{syn} \sim \Gamma_j^2
338: u_{IR}/u_B'$, where $u_B'$ is energy density of the magnetic field in
339: the blazar zone of a jet. Noting that energy flux of magnetic field
340: in a jet is $L_B \simeq c u_B' \pi R^2 \Gamma_j^2$ and $u_{IR} =
341: \xi_{IR} L_{disk} /(4\pi r^2 c)$, and assuming that $L_B \propto
342: L_{jet}$ and $\theta_j = R/r \sim 1/\Gamma_j$, this ratio is
343: %
344: \be {L_{ERC} \over L_{syn}} \propto {\Gamma^2 \xi_{IR} L_{disk} \over
345: L_{jet}} \label{eq:lum}\ee
346: %
347: Hence, for a fixed disk luminosity, luminosity ratio of the two components
348: depends mainly on three parameters, $\Gamma_j$, $\xi_{IR}$, and
349: $L_{jet}$. All of them can be a function of a distance in a jet, and
350: $\Gamma_j$ and $L_{jet}$ can additionally vary with time. With our
351: basic assumption that the blazar zone is related to the location of
352: the reconfinement shock and that this location is not changing
353: significantly with time, changes of the luminosity ratio from the
354: epoch to the epoch can be just a function of $L_{jet}$ and $\Gamma_j$.
355: We demonstrate in Fig.~\ref{fig:model3} and \ref{fig:model4} that
356: spectra of \object{3C454.3} taken at two epochs, during the outburst
357: and during the quiescent phase, can be reproduced just by assuming
358: changes in $L_{jet}$ and some modifications in the shape of the
359: injection function. From inspection of these spectra (including
360: Fig.~\ref{fig:model1}), it is apparent that differences between
361: synchrotron luminosities at different states are much larger than
362: differences between bolometric luminosities. This results from the
363: fact that for $L_{ERC} > L_{syn}$, $L_{ERC} \sim L_{bol} \propto
364: L_{jet}$, and when this is combined with the Eq.~(\ref{eq:lum}),
365: it gives $L_{syn} \propto L_{jet}^2$.
366:
367: \section{Discussion and conclusions\label{sec:discussion}}
368: %
369: We demonstrated in this paper that broadband spectra of
370: \object{3C454.3} can be reconstructed assuming that they are produced
371: at distances $r \sim 3-9$ parsecs. By the end of this distance range
372: the jet becomes transparent at millimeter wavelengths. Blazar
373: activity historically has been defined via observations in the
374: IR/optical bands, while ``blazar-zone'' is often considered to be
375: located deeply within the millimeter photosphere. However, the
376: optical and millimeter light-curves seem to indicate a significant
377: overlap of the blazar-zone with a region where the jet becomes
378: transparent at millimeter wavelengths (see \S\ref{sec:location}).
379: This is further supported by very large millimeter luminosities which
380: require high, {\it in situ}, dissipation rate of energy, and is
381: consistent with time scales of the fastest high amplitude variations,
382: of the order of $10$ days in both spectral bands. Furthermore, at such
383: distances the co-spatial model self-consistently incorporates
384: production of X- and $\gamma$-rays, via scatterings of near/mid IR
385: photons emitted by hot dust.
386:
387: It should be emphasized here that the input-parameter set for ERC
388: models is not unique and that high energy spectra can be reproduced
389: also by scattering of broad emission photons if taking place in the
390: sub-parsec region. However, then the high energy non-thermal
391: radiation should be accompanied by bulk-Compton features
392: \citep{Sikora00,Moderski04,Celotti07}, which so far have not been
393: observationally confirmed. Their lack or weakness can be explained by
394: assuming that in the sub-parsec region jet is still in acceleration
395: phase and the blazar zone is located at larger distances
396: \citep{Kataoka07}.
397:
398: We identify the ``blazar zone'' with a reconfinement shock. That,
399: together with optical polarization data imply domination of the
400: toroidal magnetic field over chaotic/turbulent magnetic fields.
401: However it should be noted that domination of the toroidal component
402: doesn't necessary indicate the domination of the Poynting flux over
403: the matter energy flux. It is very likely that the conversion of the
404: Poynting flux dominated jet into matter dominated jet -- and hence the
405: jet acceleration process -- are accomplished on sub-parsec scales
406: \citep{Sikora05,Komissarov07}. Similar conclusions are reached by
407: \citet{Jorstad07}, following multi-waveband polarimetric observations
408: of 15 AGN.
409:
410: During its 2005 outburst, \object{3C454.3} was the most luminous
411: object ever recorded in the optical band. To explain such an
412: outburst, the jet power larger than $7 \times 10^{47}\,$erg s$^{-1}$
413: is required (see Table~\ref{tab:param}). Is it feasible? Noting that
414: the estimates of the black hole mass in this object give $\sim 4
415: \times 10^9 M_{\odot}$ \citep{Gu01}, we infer that the jet power is on
416: the order of the Eddington luminosity. This, however, is at least by
417: a factor of few larger than the accretion luminosity, which in turn,
418: as determined from the optical luminosity of the thermal component
419: detected during the low state \citep{Smith88}, and after application
420: of the bolometric correction, is likely to be of the order
421: $10^{47}\,$erg s$^{-1}$. \object{3C454.3} is in this respect not
422: exceptional among most powerful radio-loud quasars: powers of jets
423: larger than $10^{47}\,$erg s$^{-1}$ have been inferred for several
424: other quasars from analysis of the lobe energetics \citep{Rawlings91},
425: as well as from Chandra and HST observations of gamma-ray blazars
426: \citep{Tavecchio07}.
427:
428: \acknowledgments This project was partially supported by Polish KBN
429: grant 5 P03D 00221 and NASA observing grant NNX07AB05G. This work was
430: also supported, in part, by the Department of Energy contract to SLAC
431: no.\ DE-AC3-76SF00515. This research has made use of the NASA/IPAC
432: Extragalactic Database (NED) which is operated by the Jet Propulsion
433: Laboratory, California Institute of Technology, under contract with
434: the National Aeronautics and Space Administration.
435:
436: \appendix
437:
438: \section{Analytical approximations of the model parameters\label{app:approx}}
439: %
440: \subsection{Injection function}
441: %
442: Normalization factor $K_e$ of the electron injection function $Q$ can
443: be derived using approximate formulas for production of the X-ray
444: spectrum via the ERC process in the slow cooling regime
445: \citep[see][]{Moderski03}:
446: %
447: \be \nu_{x}L_{\nu_x} = {1 \over 2} [\gamma N_{\gamma}] |\dot
448: \gamma|_{ERC}(\theta_{obs}) m_e c^2 {\cal D}^4 \ee
449: %
450: where
451: %
452: \be |\dot \gamma|_{ERC}(\theta_{obs}) = {c\sigma_T \over m_ec^2}
453: u_{IR}' \gamma^2 \left({{\cal D} \over \Gamma_j}\right)^2 \ee
454: %
455: \be N_{\gamma} = Q {\Delta r \over c \Gamma_j} \ee
456: %
457: \be u_{ext}' = {4\over 3} \Gamma_j^2 u_{ext} \ee
458: %
459: and
460: %
461: \be {\cal D} = {1 \over \Gamma_j(1-\beta \cos{\theta_{obs}})} \ee
462: %
463: In the slow cooling regime the slope $p$ of the electron injection
464: function is $p=2 \alpha_x +1$ and for $\Delta r = r$ above equations
465: give
466: %
467: \be K_e = {3\over 2} {\nu_{x}L_{\nu_x} \over \sigma_T u_{IR} r}
468: {\Gamma_j \over {\cal D}^{4+2\alpha_x}} \left({\nu_{ext} \over
469: \nu_x}\right)^{1 - \alpha_x} \ee
470: %
471:
472: In one of our models, the break in the injection functions is
473: introduced in order to get a better fit of the observed spectrum:
474: %
475: \be Q = K_e {1\over \gamma^p + \gamma_{br}^{p-q}\gamma^q }\ee
476: %
477: where $\gamma_{br}$ is the break energy and $q$ is the spectral index
478: of the injection function at high energy limit.
479:
480: \subsection{Magnetic field intensity}
481: %
482: The ERC to synchrotron peak luminosity ratio
483: %
484: \be {L_{ERC} \over L_{syn}} = {u_{ext}' ({\cal D}/\Gamma_i)^2 \over
485: u_B'} \ee
486: %
487: gives us magnetic field intensity
488: %
489: \be B' = {\cal D} \sqrt{{32 \over 3} u_{ext} {L_{ERC} \over L_{syn}}}
490: \ee
491: %
492: and magnetic energy flux
493: %
494: \be L_B = c u_B' \pi R^2 \Gamma_j^2 = \pi c u_B' r^2 (\theta_j
495: \Gamma_j)^2 \ee
496: %
497: where $u_B' = B^{\prime 2} /( 8 \pi)$ is magnetic energy density.
498: %
499: With known $B'$ we can estimate the maximum energy of injected electrons
500: %
501: \be \gamma_{max} \simeq 5.2 \times 10^{-4}
502: \sqrt{{\nu_{syn,max,obs}(1+z) \over B' {\cal D}}} \ee
503: %
504:
505: \subsection{Electron energy density}
506: %
507: Due to light travel effects, sources moving with relativistic speeds
508: are seen on the sky as stretched by a factor ${\cal D} \Gamma_j$,
509: which means that only a fraction $1/({\cal D} \Gamma_j)$ of particles
510: is seen at a given instance to be enclosed within the distance range
511: $\Delta r$. Hence the volume of the jet segment into which electrons
512: are injected at the 'observed' rate $Q$ is $\pi R^2 \lambda$, where
513: $\lambda = \Delta r /({\cal D}\Gamma_j)$. Amount of energy injected
514: into the segment during its propagation through the $\Delta r$ zone is
515: %
516: \be E_{e,inj}' = {\Delta r \over c \Gamma_j} \int{Q \gamma m_ec^2 \,
517: d\gamma} \ee
518: %
519: and energy density of injected electrons is
520: %
521: \be u_{e,inj}'(r_0 + \Delta r= 2 r_0) = { E_{e,inj}' \over \pi R^2
522: \lambda'} = {{\cal D} \over \Gamma_j} {\int{Q \gamma m_ec^2 \,
523: d\gamma} \over \pi c R^2} = {m_e c {\cal D} \Gamma_j \int{Q \gamma \,
524: d\gamma} \over 4\pi r_0^2 (\theta_{obs}\Gamma_j)^2} \ee
525: %
526: where $\lambda'=\lambda \Gamma_j$.
527:
528: \subsection{Energy dissipation efficiency}
529: %
530: In the proton inertia dominated jets acceleration of electrons is
531: powered by protons and we have
532: %
533: \be u_{e,inj}' = \eta_e u_p' (\bar \gamma_p -1) \ee
534: %
535: where $(\bar \gamma_p -1) \ll 1$ is the fraction of proton bulk
536: kinetic energy converted to the 'thermal' proton energy called
537: hereafter the efficiency of energy dissipation, and $\eta_e$ is the
538: fraction of proton 'thermal' energy tapped by electrons. Condition of
539: having matter dominated jet implies $u_p' > u_B'$, and combining this
540: with previous equation gives
541: %
542: \be (\bar \gamma_p -1) < {u_{e,inj}' \over u_B' \eta_e}
543: \label{eq:appgp} \ee
544: %
545:
546: \subsection{Pair content}
547: %
548: Using definition of particle energy densities ($u=nmc^2\bar\gamma$)
549: and noting that $\bar \gamma \gg 1$ (throughout our paper, $\gamma
550: \equiv \gamma_e$) and $\bar \gamma_p -1 \ll 1$ we obtain the pair
551: content
552: %
553: \be {n_e' \over n_p'} = {m_p \over m_e} {\bar \gamma_p -1 \over \bar
554: \gamma} < {m_p \over \eta_e \bar \gamma m_e} {u_{e,inj}' \over u_B'}
555: \ee
556: %
557: where inequality (\ref{eq:appgp}) was used and $\bar \gamma \equiv
558: \int{Q\gamma \, d\gamma}/\int{Q \, d\gamma}$.
559:
560: \subsection{Toroidal vs. turbulent magnetic field}
561: %
562: We assumed in the paper that magnetic field is dominated by the
563: toroidal component. This assumption can be verified as follows. For
564: $u_{B,tor}' >> u_{B,turb}'$, $u_{B,tor}' \simeq u_{B,tot}' \equiv
565: u_B'$ and
566: %
567: \be u_{B,tor}' \simeq {u_B' \over u_{e,inj}'} u_{e,inj}' = \eta_e
568: {u_B' \over u_{e,inj}'} u_p'(\bar \gamma_p -1) \ee
569: %
570: For $u_{B,turb}' \simeq \eta_B u_p'(\bar \gamma_p - 1)$ this gives
571: %
572: \be {u_{B,tor}' \over u_{B,turb}'} = {\eta_e \over \eta_B} {u_B' \over
573: u_{e,inj}'} \ee
574: %
575: \bigskip
576:
577: Note that all formulas which involve a Doppler factor apply for
578: 'mono-Doppler' sources only. In the case of conically diverging
579: jets, the observed radiation is contributed by jet portions moving
580: relative to the line of sight at different angles and then analytical
581: estimations differ significantly from numerical results. This in
582: particular concerns the quantity $K_{\rm e}$ because of its strong
583: dependence on ${\cal D}$. However, for $\theta_{obs} \sim \theta_j
584: \sim 1/\Gamma_j$, still reasonable analytical estimates are achievable
585: if using ${\cal D} = 1.5 \Gamma_j$, instead of ${\cal D}= \Gamma_j$.
586:
587: \section{The millimeter photosphere\label{app:photo}}
588: %
589: Optically thin synchrotron spectrum in \object{3C454.3} and other
590: quasar hosted blazars is produced by electrons in the fast cooling
591: regime. In this regime an electron distribution is steepened due to
592: radiative losses, and for a single-power-law injection function, $Q
593: \sim \gamma^{-p}$, the electrons reach a distribution with the index
594: $s= p+1$. Below we provide estimation of the millimeter photosphere
595: distance, assuming $p=2$. For such a source the synchrotron-self
596: absorption opacity $\tau(\nu_{abs}')$ is at $\nu_{\rm a}'$ equal to 1
597: for
598: %
599: \be R_{mm} = 2.7 \times 10^{-15} {\nu_{\rm a}^{\prime 7/2} \over c_n
600: B^{\prime 5/2}} \, {\rm [cm]}\ee
601: %
602: where $n_{\gamma} = c_n \gamma^{-3}$ is the electron density energy
603: distribution.
604:
605: Noting that
606: %
607: \be c_n = {C_{N} \over V'} = {C_N \Gamma_j^2 {\cal D} \over \pi r^3
608: (\Delta r/r) (\Gamma_j \theta_j)^2} \ee
609: %
610: where $C_N: N_{\gamma} = C_N \gamma^{-3}$, and that
611: %
612: \be N_{\gamma} = {\int_{\gamma} Q \, d\gamma \over |\dot\gamma|_{tot}}
613: \ee
614: %
615: where for $L_{ERC} > L_{syn}$
616: %
617: \be |\dot\gamma|_{tot} \simeq {16 c \sigma_T \gamma^2 \Gamma_j^2
618: u_{ext} \over 9 m_e c^2} \ee
619: %
620: we obtain, for $\theta_{obs} \Gamma_j=1$ and $\Delta r=r$,
621: %
622: \be R_{mm} \simeq 1.9 \times 10^7 {{\cal D}^{9/5} \over
623: \Gamma_j^{7/5}} {B_0' r_0 \over (u_{ext,in}r_{in}^2)^{2/5}} {K_e^{2/5}
624: \over [\nu_{\rm a,obs}(1+z)]^{7/5}} \, {\rm [cm]} \ee
625: %
626: and $r_{mm} = R_{mm}/\Gamma_j$. For $\nu_{\rm a,obs} = 3 \times
627: 10^{11}$ Hz ($\lambda_{\rm a,obs} = 1$mm) and parameters of the Model
628: 1 (see Table~\ref{tab:param}), this gives $R_{mm} \simeq 1.4 \times
629: 10^{18}\,$cm and $r_{mm} \simeq 2.8 \times 10^{19}\,$cm.
630:
631: \bibliography{ms}
632:
633: \clearpage
634:
635: \begin{figure}
636: \centering
637: \includegraphics[width=0.95\textwidth,bb=50 50 555 495]{fig1.eps}
638: \caption{Data points show the broadband spectrum of 3C454.3 at the
639: epoch of the optical peak during the 2005 outburst. Infrared data
640: points at $1\,$mm and $3\,$mm and upper optical data points are from
641: IRAM telescope and WEBT campaign, respectively, and were reported
642: together with Chandra data in \citet{Villata06}. Lower optical data
643: points from REM telescope and Swift data are taken from
644: \citet{Giommi06}. Integral data are from \citet{Pian06}. Continuous
645: lines show our preferred models obtained using the $BLAZAR$ code
646: \citep{Moderski03}. Thick, solid line shows the model accounting
647: for the broad-band data during the optical peak of the outburst
648: (Model 1 in Table~\ref{tab:param}); the thin, solid lines indicate
649: various components of the spectrum and illustrate that the SSC
650: component is relatively weak. Dashed lines show the model spectrum
651: produced at a distance twice as large as the thick solid line, and
652: are intended to illustrate the emission at the millimeter
653: photosphere (Model 2 in Table~\ref{tab:param}). Model parameters
654: are given in Table~\ref{tab:param}.}
655: \label{fig:model1}
656: \end{figure}
657:
658: \clearpage
659:
660: \begin{figure}
661: \centering
662: \includegraphics[width=0.95\textwidth,bb=50 50 555 495]{fig2.eps}
663: \caption{Broadband spectral observations of 3C454.3 during the 2007
664: outburst. Tuorla Observatory optical data point and Swift UV and
665: X-ray data are taken from \citet{Ghisellini07}. Agile point comes
666: from \citet{Vercellone07}. Model illustrated as a solid line has
667: parameters given in Table~\ref{tab:param} as Model~3.}
668: \label{fig:model3}
669: \end{figure}
670:
671: \clearpage
672:
673: \begin{figure}
674: \centering
675: \includegraphics[width=0.95\textwidth,bb=50 50 555 495]{fig3.eps}
676: \caption{Broadband spectrum during the low state in the 'CGRO epoch'.
677: All data points below $10^{16}\,$Hz are from NASA Extragalactic
678: Database. BeppoSAX data come from \citet{Tavecchio02}, while CGRO
679: OSSE, Comptel and EGRET data are from \citet{McNaron95},
680: \citet{Zhang05}, and \citet{Hartman99}, respectively. Model
681: accounting for those data, with parameters given in
682: Table~\ref{tab:param} (Model~4) is plotted as a solid line.}
683: \label{fig:model4}
684: \end{figure}
685:
686: \newpage
687:
688: \begin{table}
689: \caption{The model parameters.}\label{tab:param}
690: \begin{center}
691: \begin{tabular}{lcccc}
692: \tableline\tableline
693: Parameter & Model 1 & Model 2 & Model 3 & Model 4 \\
694: \tableline\tableline
695: $\gamma_{\rm min}$ & $1$ & $1$ & $1$ & $1$ \\
696: $\gamma_{\rm br}$ & --- & --- & --- & $80$ \\
697: $\gamma_{\rm max}$ & $4 \times 10^3$ & $4 \times 10^3$ & $4 \times 10^3$ & $9 \times 10^3$ \\
698: $p$ & $2.0$ & $2.0$ & $2.0$ & $1.7$ \\
699: $q$ & $2.0$ & $2.0$ & $2.0$ & $2.5$ \\
700: $K_{\rm e}\, [\rm s^{-1}]$ & $3.0 \times 10^{49}$ & $1.5 \times 10^{49}$ & $2.3 \times 10^{49}$ & $3.0\times 10^{48}$ \\
701: $\Gamma_{\rm j}$ & $20$ & $20$ & $20$ & $20$ \\
702: $\theta_{\rm j}$ \, [rad] & $0.05$ & $0.05$ & $0.05$ & $0.05$ \\
703: $\theta_{\rm obs} \, [{\rm rad}]$ & $0.05$ & $0.05$ & $0.05$ & $0.05$ \\
704: $r_0= \Delta r_0 \, [{\rm cm}]$ & $10^{19}$ & $2 \times 10^{19}$ & $2 \times 10^{19}$ & $2 \times 10^{19}$ \\
705: $B_0 \, [{\rm G}]$ & $1.4$ & $0.50$ & $0.63$ & $0.27$ \\
706: $r_{\rm in} \, [{\rm cm}]$ & $10^{19}$ & $10^{19}$ & $10^{19}$ & $10^{19}$ \\
707: $u_{\rm IR}(r_{\rm in}) \, [{\rm erg \, cm^{-3} \, s^{-1}}]$ & $1.24 \times 10^{-4}$ & $1.24 \times 10^{-4}$ & $1.24 \times 10^{-4}$ & $1.24 \times 10^{-4}$ \\
708: $h \nu_{\rm IR} \, [{\rm eV}]$ & $0.34$ & $0.34$ & $0.34$ & $0.34$ \\
709: \tableline
710: $u'_{\rm e,inj}(2r_0) \, [{\rm erg \, cm^{-3} \, s^{-1}}]$ & $3.25 \times 10^{-3}$ & $4.06 \times 10^{-4}$ & $6.22 \times 10^{-4}$ & $1.95 \times 10^{-4}$ \\
711: $u_B'(2r_0) \, [{\rm erg \, cm^{-3} \, s^{-1}}]$ & $1.95 \times 10^{-2}$ & $2.49 \times 10^{-3}$ & $3.95 \times 10^{-3}$ & $7.25 \times 10^{-4}$ \\
712: $L_{\rm j} > L_B \, [{\rm erg \, s^{-1}}]$ & $7.35 \times 10^{47}$ & $3.75 \times 10^{47}$ & $5.96 \times 10^{47}$ & $1.10 \times 10^{47}$ \\
713: $\bar \gamma_p - 1 <$ & $0.17/\eta_e$ & $0.16/\eta_e$ & $0.16/\eta_e$ & $0.26/\eta_e$ \\
714: $\bar \gamma$ & $8.3$ & $8.3$ & $8.3$ & $13.0$ \\
715: $n_e/n_p <$ & $37.5/\eta_e$ & $35.3/\eta_e$ & $35.3/\eta_e$ & $37.9/\eta_e$ \\
716: $u_{B_{tor}}' / u_{B_{turb}}'$ & $5.88 \, \eta_e/\eta_B$ & $6.25 \, \eta_e/\eta_B$ & $6.25 \, \eta_e/\eta_B$ & $3.72 \, \eta_e/\eta_B$ \\
717: \tableline\tableline
718: \end{tabular}
719: \end{center}
720: \end{table}
721:
722:
723: \end{document}
724: