1: \def\nfn{\nu F_{\nu}}
2: \def\nfnsy{\left( \nfn \right)_{\rm sy}}
3: \def\nfnssc{\left( \nfn \right)_{\rm SSC}}
4: \def\nfnerc{\left( \nfn \right)_{\rm ERC}}
5: \def\nfnsyn{F_{{\rm sy}, -10}}
6: \def\nfnsscn{F_{{\rm SSC}, -10}}
7: \def\nfnercn{F_{{\rm ERC}, -10}}
8: \def\esy{\epsilon_{\rm sy}}
9: \def\essc{\epsilon_{\rm SSC}}
10: \def\eerc{\epsilon_{\rm ERC}}
11: \def\estar{\epsilon_{\ast}}
12: \def\esyn{\epsilon_{{\rm sy}, -7}}
13: \def\esscn{\epsilon_{{\rm SSC}, -1}}
14: \def\eercn{\epsilon_{{\rm ERC}, 2}}
15: \def\estarn{\epsilon_{\ast, -5}}
16: \def\eb{\epsilon_B}
17: \def\gcr{\gamma_{\rm cr}}
18: \def\fsp{f_{\rm sp}}
19: \def\fsy{f_{\rm sy}}
20: \def\ferc{f_{\rm ERC}}
21: \def\fssc{f_{\rm SSC}}
22: \def\uext{u_{\rm ext}}
23: \def\Bcr{B_{\rm cr}}
24: \def\taur{\tau_{\rm repr}}
25: \def\ls{\lower4pt\hbox{${\buildrel < \over \sim}$}}
26: \def\gs{\lower4pt\hbox{${\buildrel > \over \sim}$}}
27:
28: \documentclass[12pt, preprint]{aastex}
29: %\documentclass[manuscript]{aastex}
30: %\documentclass[preprint2]{aastex}
31:
32: \newcommand{\vdag}{(v)^\dagger}
33: \newcommand{\myemail}{mboett@helios.phy.ohiou.edu}
34:
35: \slugcomment{Accepted for publication in {\it The Astrophysical Journal}}
36:
37: \shorttitle{Variability of 3C~279}
38: \shortauthors{M. B\"ottcher \& D. Principe}
39:
40: \begin{document}
41:
42: \title{The Optical Variability of the Quasar 3C~279: The Signature
43: of a Decelerating Jet?}
44:
45: \author{M. B\"ottcher and D. Principe\altaffilmark{1}}
46:
47: \altaffiltext{1}{Astrophysical Institute, Department of Physics and Astronomy, \\
48: Clippinger 339, Ohio University, Athens, OH 45701, USA}
49:
50: \begin{abstract}
51: A recent optical monitoring campaign on the prominent quasar
52: 3C~279 revealed at least one period of a remarkably clean exponential
53: decay of monochromatic (BVRI) fluxes with time, with a time constant
54: of $\tau_d = 12.8$~d, over about 14 days. This is clearly too
55: long to be associated with radiative cooling. Here we propose that
56: this may be the signature of deceleration of the synchrotron emitting
57: jet component. We develop a model analogous to the relativistic blast
58: wave model for gamma-ray bursts, including radiative energy losses and
59: radiation drag, to simulate the deceleration of a relativistically
60: moving plasmoid in the moderately dense AGN environment. Synchrotron,
61: SSC and external Compton emission are evaluated self-consistently.
62: We show that the observed optical light curve decay can be successfully
63: reproduced with this model.
64:
65: The decelerating plasmoid model predicts a delayed X-ray flare, about
66: 2 -- 3 weeks after the onset of the quasi-exponential light curve decay
67: in the optical. A robust prediction of this model, which can be tested
68: with Fermi and simultaneous optical monitoring, is that the peak in
69: the $\gamma$-ray light curve at $\sim 100$~MeV is expected to be delayed by
70: a few days with respect to the onset of the optical decay, while the VHE
71: $\gamma$-rays are expected to track the optical light curve closely
72: with a delay of at most a few hours.
73:
74: \end{abstract}
75:
76: \keywords{galaxies: active --- Quasars: individual (3C~279)
77: --- gamma-rays: theory --- radiation mechanisms: non-thermal}
78:
79: \section{Introduction}
80:
81: The quasar 3C~279 ($z = 0.536$) is one of the best-observed flat
82: spectrum radio quasars, in part because of its prominent
83: $\gamma$-ray flare shortly after the launch of the {\it Compton
84: Gamma-Ray Observatory (CGRO)} in 1991. It has been
85: persistently detected by the {\it Energetic Gamma-ray
86: Experiment Telescope (EGRET) on board {\it CGRO}}
87: each time it was observed, even in its very low quiescent
88: states, e.g., in the winter of 1992 -- 1993, and is
89: known to vary in $\gamma$-ray flux by roughly two orders of
90: magnitude \citep{maraschi94,wehrle98}. It has been monitored
91: intensively at radio, optical, and more recently also X-ray
92: frequencies, and has been the subject of intensive multiwavelength
93: campaigns \citep[e.g.,][]{maraschi94,hartman96,wehrle98}.
94: The most recent multiwavelength campaign on 3C~279 included
95: a Whole Earth Blazar Telescope (WEBT) campaign in the spring
96: of 2006 \citep{boettcher07b}. During this campaign, the source
97: was overall in a high optical state, with $R \sim 14.0$ -- $14.5$.
98: However, the light curves showed an extraordinary feature: An
99: unusually clean, quasi-exponential decay of the BVRI fluxes with
100: a time scale of $\tau_d = 12.8$~d, extended over about 2 weeks.
101: This paper aims at a theoretical interpretation of this
102: extraordinary light curve feature.
103:
104: Flat-spectrum radio quasars (FSRQs) and BL~Lac objects are
105: active galactic nuclei (AGNs) commonly unified in the class
106: of blazars. They exhibit some of the most violent high-energy
107: phenomena observed in AGNs to date. Their spectral energy
108: distributions (SEDs) are characterized by non-thermal continuum
109: spectra with a broad low-frequency component in the radio -- UV
110: or X-ray frequency range and a high-frequency component from
111: X-rays to $\gamma$-rays.
112: In the framework of relativistic jet models, the low-frequency (radio
113: -- optical/UV) emission from blazars is interpreted as synchrotron
114: emission from nonthermal electrons in a relativistic jet. The
115: high-frequency (X-ray -- $\gamma$-ray) emission could either be
116: produced via Compton upscattering of low frequency radiation by the
117: same electrons responsible for the synchrotron emission \citep[leptonic
118: jet models; for a recent review see, e.g.,][]{boettcher07a}, or
119: due to hadronic processes initiated by relativistic protons
120: co-accelerated with the electrons \citep[hadronic models, for
121: a recent discussion see, e.g.,][]{muecke01,muecke03}.
122: Several authors have modeled broadband SEDs of 3C~279 in various
123: states \citep[e.g.,][]{bednarek98,sikora01,hartman01,moderski03}.
124: A consistent picture emerges that the X-ray -- soft $\gamma$-ray
125: portion of the SED might be dominated by
126: synchrotron self-Compton (SSC)
127: emission, while the
128: {\it EGRET} emission might require an additional component, most
129: likely external Compton emission.
130:
131: Standard leptonic models of blazar emission generally assume that
132: a relativistic plasmoid containing ultrarelativistic nonthermal
133: electrons moves with constant bulk Lorentz factor $\Gamma$ along
134: a jet, directed at a small angle with respect to our line of sight.
135: However, for several blazars, in particular high-frequency peaked
136: BL Lac objects detected at $> 100$~GeV $\gamma$-rays, such models
137: sometimes require unexpectedly large bulk Lorentz factors ($\Gamma
138: \gtrsim 50$) and accordingly small viewing angles in order to explain
139: their SEDs and variability \citep{bfr08,gt08,finke08}.
140: Such large Lorentz factors and small viewing angles pose serious
141: problems for AGN unification schemes, according to which FR I radio
142: galaxies are believed to be the unbeamed equivalents of BL~Lac
143: objects. A possible solution to this dilemma might lie in the
144: deceleration of the emission region \citep{gk03a,gk03b} from
145: sub-pc scales, at which the optical -- X-ray -- $\gamma$-ray
146: emission is produced, towards pc and kpc scales, which can be
147: resolved with VLBA / VLBI techniques. At those scales, superluminal
148: speeds of individual jet components of $\beta_{\rm app} \lesssim 10$
149: are characteristically observed in most cases, providing an estimate
150: of the Lorentz factor of jet components at those scales of $\Gamma
151: \sim 10$. In fact, extreme deceleration of a radio-emitting plasmoid
152: (component C3) in the jet of 3C~279 may already habe been directly
153: observed in space VLBI monitoring observations \citep{piner00},
154: although the identification of this component over multiple
155: observing epochs with different instruments/arrays is highly
156: uncertain.
157:
158: In this paper, we propose a model analogous to the relativistic
159: blast wave model which has successfully predicted and explained
160: the smooth, self-similar light curves of X-ray and optical afterglows
161: of $\gamma$-ray bursts \citep{pr93,mr97,cd99}. We adapt this model
162: for the specific situation in blazars. In particular, we include
163: self-consistently radiative losses and radiation drag from
164: Comptonization of external radiation fields. A similar study,
165: with emphasis on the details of the isotropization of particle
166: distributions in the plasmoid and on spectral features from
167: various leptonic and hadronic processes, has been performed
168: by \cite{ps00}, who find good agreement of their results with
169: characteristic SEDs of blazars. Here, we adopt a simplified
170: description of the particle dynamics and radiation processes,
171: and focus on the expected monochromatic light curves dominated
172: by the plasmoid deceleration. We review the observational
173: motivation from 3C~279 in \S \ref{motivation}, describe the
174: model for the plasmoid dynamics in \S \ref{model}, and outline
175: our treatment of radiation processes in \S \ref{radiation}. As a
176: test of our numerical simulations, we develop an analytical
177: solution to the plasmoid dynamics and light curves in the
178: self-similar deceleration phase in \S \ref{deceleration}.
179: In \S \ref{results} we present results of our simulations
180: and fits to the observed exponential flux decay of 3C~279
181: in January 2006. We summarize in \S \ref{summary}.
182:
183: Throughout this paper, we refer to $\alpha$ as the energy
184: spectral index, $F_{\nu}$~[Jy]~$\propto \nu^{-\alpha}$. A
185: cosmology with $\Omega_m = 0.3$, $\Omega_{\Lambda} = 0.7$,
186: and $H_0 = 70$~km~s$^{-1}$~Mpc$^{-1}$ is used. In this cosmology,
187: and using the redshift of $z = 0.536$, the luminosity distance
188: of 3C~279 is $d_L = 3.08$~Gpc.
189:
190: \begin{figure}[ht]
191: \plotone{f1.eps}
192: \caption{Light curves of 3C~279 in various optical bands during the
193: spring of 2006. The dotted red line indicates our model fit, as
194: discussed in the following sections, to the R-band light curve during
195: the quasi-exponential decay around January 15, 2006}
196: \label{lightcurves}
197: \end{figure}
198:
199: \section{\label{motivation}Observational Motivation}
200:
201: 3C~279 was observed in a WEBT campaign at radio, near-IR, and optical
202: frequencies, throughout the spring of 2006. Details of the observations,
203: data analysis, and implications of the optical variability patterns
204: observed during that campaign have been published in \cite{boettcher07b}.
205: Fig. \ref{lightcurves} shows the optical light curves of 3C~279 during
206: spring 2006. The light curves exhibit an extraordinarily clean quasi-exponential
207: decay with a characteristic time scale of $\tau_d \sim 12.8$~days around
208: JD 2453743 -- JD 2453760. This light curve feature can not be interpreted
209: as the signature of radiative cooling since the synchrotron cooling time
210: scale for electrons emitting synchrotron radiation in the optical R band
211: is
212:
213: \begin{equation}
214: \tau_{\rm sy}^{\rm obs} \sim 3 \times 10^4 \, B_G^{-1/2} D_1^{-1/2} \; {\rm s}
215: \label{tausy}
216: \end{equation}
217: where $B_G$ is the magnetic field in Gauss and $D_1$ is the Doppler
218: factor in units of 10. This
219: is of the order of at most a few hours for typical values of the
220: magnetic field strength expected in quasars ($B \sim 1$~G). Setting the
221: synchrotron cooling time scale equal to the observed exponential decay time
222: scale, would require a magnetic field of $B \sim 7 \times 10^{-4} \, D_1^{-1}$~G,
223: which is about three orders of magnitude lower than usually inferred for quasar
224: jets. We therefore favor a model in which the light curve decay is associated
225: with the dynamics of the emission region rather than microscopic processes.
226: We note that similar quasi-exponential decays have also been observed
227: in 3C~279 repeatedly in the 2007 observing season \citep{larionov08}.
228:
229:
230: \section{\label{model}Model of a decelerating jet}
231:
232: Our treatment of a decelerating jet is borrowed from the blast wave model
233: of gamma-ray bursts. For details see, e.g., \cite{cd99}. We assume a plasmoid
234: moving ballistically with initial mass $M_0$ and bulk Lorentz factor
235: $\Gamma_0$ along the jet. Let $M$ be the relativistic mass of the the
236: plasmoid in the rest-frame of the plasmoid, then the momentum $P$ of the
237: plasmoid in the stationary AGN frame is $P = \beta \, \Gamma \, M \, c$,
238: where $\beta$ is the normalized velocity $v/c$ corresponding to the bulk
239: Lorentz factor $\Gamma$. If the plasmoid radiates isotropically in its rest
240: frame, the equation of motion of the plasmoid can be derived from momentum
241: conservation, $dP/dt = 0$. However, in the case of a quasar, a substantial
242: contribution to the (bolometrically dominant) $\gamma$-ray emission results
243: from Compton upscattering of external radiation fields (EC = External Compton),
244: and a significant transfer of plasmoid momentum to Compton-scattered external
245: radiation (``Compton drag'') has to be taken into account. We can therefore
246: write
247:
248: \begin{equation}
249: {dP \over dt} = \left( {dP \over dt} \right)_{\rm EC} = {c \, M
250: \dot\Gamma \over \beta} + \Gamma \beta \, {\dot M} \, c
251: \label{dPdt}
252: \end{equation}
253: For large enough $\Gamma$, most of the EC radiation will be beamed into a
254: narrow cone of solid angle $\Omega \sim 1/\Gamma^2$, and we can write
255: the momentum transfer to EC radiation as
256:
257: \begin{equation}
258: \left( {dP \over dt} \right)_{\rm EC} = - {1 \over c} \int\limits_{4\pi}
259: {dL/d\Omega} \cos\theta \, d\Omega \approx {\Gamma^2 \over 4 \pi \, c}
260: \, \dot{E'}_{\rm EC}
261: \label{dPdtEC}
262: \end{equation}
263:
264: where $\dot{E'}_{\rm EC}$ is the internal energy loss due to EC radiation
265: in the co-moving frame. For the purpose of an approximate, quantitative
266: analysis to extract the salient spectral and light curve features of this
267: model, we assume that all Compton scattering occurs in the Thomson regime
268: so that
269:
270: \begin{equation}
271: \dot{E'}_{\rm EC} = - {1 \over \Gamma} \, {4 \over 3} c \, \sigma_T \,
272: {u'}_{\rm ext} \int\limits_{1}^{\infty} N_e (\gamma) \, \gamma^2 \, d\gamma.
273: \label{EdotEC}
274: \end{equation}
275: The factor $1/\Gamma$ in Eq. (\ref{EdotEC}) stems from the fact that
276: $\dot{E'}$ constitutes a derivative with respect to time in the stationary
277: AGN frame, and $dt' = dt/\Gamma$. We can also use Eq. (\ref{dPdtEC})
278: for a rough estimate of the magnitude of the radiation drag force, assuming
279: that the observed $\gamma$-ray emission results from Compton scattering of
280: an isotropic radiation field. Then,
281:
282: \begin{equation}
283: \left({dP \over dt}\right)_{\rm drag} \approx - {(1 + z) \, d_L^2 \,
284: \nu F_{\nu}^{\rm pk, EC} \over \Gamma^2 c} \sim 4.3 \times 10^{33}
285: \, f_{13} \, \Gamma_1^{-2} \; {\rm dyne}
286: \label{radiationdrag}
287: \end{equation}
288: where $f_{13} = \nu F_{\nu}^{\rm pk, EC} / (10^{13}$~Jy~Hz) and $\Gamma_1
289: = \Gamma/10$. This may be compared to an estimate for the Compton rocket
290: effect due to radiation from the accretion disk. If we approximate the
291: accretion disk radiation as a point source (the most optimistic estimate)
292: with luminosity $L_D \equiv 10^{46} \, L_{46}$~erg~s$^{-1}$, impinging from
293: behind on an emission region of radius $R_b \equiv 10^{16} \, R_{16}$~cm,
294: located at a distance $r \equiv 0.1 \, r_{-1}$~pc from the accretion disk,
295: the force on the plasmoid due to the Compton rocket effect can be estimated
296: as
297:
298: \begin{equation}
299: \left({dP \over dt}\right)_{\rm acc} \approx {L_D \, R_b^2 \over 16 \,
300: r^2 \, c} \approx 2.3 \times 10^{31} \, L_{46} \, R_{16}^2 \, r_{-1}^{-2}
301: \; {\rm dyne}.
302: \label{comptonrocket}
303: \end{equation}
304: Thus, for standard parameters with $r \gtrsim 10^{-2}$~pc and $L_D \lesssim
305: 10^{46}$~erg~s$^{-1}$, the Compton rocket effect may safely be neglected in
306: our calculations.
307:
308:
309: The accumulation and radiative loss of relativistic mass $\dot M$ can be
310: calculated as
311:
312: \begin{equation}
313: \dot M = A(r) \rho(r) \Gamma(r) \, {dr/dt} + {1 \over c^2} \dot{E'}_{\rm rad}
314: \label{Mdot}
315: \end{equation}
316: where $dr/dt = \beta c$, $A(r) = \pi \, R_b^2 (r)$ is the cross section of the
317: jet, $\rho(r)$ is the density of external material being swept up by the plasmoid,
318: and
319:
320: \begin{equation}
321: \dot{E'}_{\rm rad} = - {1 \over \Gamma} \, {4 \over 3} c \, \sigma_T \, u'
322: \int\limits_{1}^{\infty} N_e (\gamma) \, \gamma^2 \, d\gamma
323: \label{Edotrad}
324: \end{equation}
325: Here, $u'$ is the sum of the energy densities, $u'_{\rm B} + u'_{\rm ext}
326: + u'_{\rm sy}$, and, again, we have assumed Compton scattering to be dominated
327: in the Thomson regime.
328:
329: Derivatives with respect to time can be converted to derivatives
330: with respect to distance $r$ from the central engine, yielding
331:
332: \begin{equation}
333: {d\Gamma \over dr} = - {\Gamma(r) \beta^2(r) \over M(r)} \,
334: {dM \over dr} + {\Gamma^2 \, \dot{E'}_{\rm EC} \over 4 \pi \, M \, c^3}
335: \label{dGdr}
336: \end{equation}
337: and
338: \begin{equation}
339: {dM \over dr} = A(r) \, \rho(r) \, \Gamma (r) + {\dot{E'}_{\rm rad} \over
340: \Gamma(r) \beta(r) \, c^3}
341: \label{dMdr}
342: \end{equation}
343:
344: We assume that a fraction of the swept-up electrons will be instantaneously
345: accelerated into a power-law, which can be described by an injection function
346: of the form
347:
348: \begin{equation}
349: Q(\gamma) = Q_0 \, \gamma^{-q} \, H(\gamma_{\rm min}, \gamma, \gamma_{\rm max}),
350: \label{Qg}
351: \end{equation}
352: where $H(x_0, x, x_1) = 1$ for $x_0 < x < x_1$ and 0 otherwise. The low- and
353: high-energy cutoffs of the electron injection function are given by
354:
355: \begin{equation}
356: \gamma_{\rm min} = {\epsilon_e \over \xi_e} \, \left( {q - 2 \over q - 1}
357: \right) \, {m_p \over m_e} \, \Gamma
358: \label{gmin}
359: \end{equation}
360: where $\epsilon_e$ is the fraction of swept-up power that is transferred to
361: relativistic electrons,
362: $\xi_e$ is the fraction of swept-up electrons which
363: is accelerated to ultrarelativistic energies. The maximum Lorentz
364: factor can be estimated by balancing the fastest conceivable acceleration
365: time scale (the Larmor time scale) with the synchrotron loss time scale:
366:
367: \begin{equation}
368: \gamma_{\max} \sim 4.7 \times 10^7 \, B_G^{-1/2}
369: \label{gmax}
370: \end{equation}
371:
372: where $B_G$ is the magnetic field in Gauss.
373: The magnetic field may be parameterized in terms of a fraction $e_B$
374: of the swept-up energy transferred to magnetic-field energy density:
375:
376: \begin{equation}
377: B (r) = \sqrt{ 32 \, \pi \, e_B \, \rho(r) \, c^2 \,} \, \Gamma (r)
378: \label{B}
379: \end{equation}
380:
381: where an additional factor of 4 has been introduced to account for
382: the compression of the pre-shock material by the strong shock.
383: We can find the normalization $Q_0$ of the
384: electron injection function through
385:
386: \begin{equation}
387: Q_0(r) = \xi_e \, {A(r) \, \Gamma(r) \, \beta(r) \, \rho(r) \, c \over m_p}
388: \, {1 - q \over \gamma_{\rm max}^{1 - q} - \gamma_{\rm min}^{1 - q}}.
389: \label{Q0}
390: \end{equation}
391:
392: Eq. (\ref{tausy}) indicates that the characteristic radiative cooling time
393: scales of particles emitting synchrotron radiation at optical or higher
394: frequencies are likely to be much shorter than the dynamical time scale
395: of the system. Therefore, the balance of relativistic particle acceleration,
396: cooling, and escape on a time scale
397:
398: \begin{equation}
399: \tau'_{\rm esc} = \eta \, {R_b \over c}
400: \label{tauesc}
401: \end{equation}
402: yields a critical electron energy $\gamma_c$ beyond which particles effectively
403: radiate their energy away on a time scale shorter than the escape time scale:
404:
405: \begin{equation}
406: \gamma_c = {3 \, m_e c^2 \over 4 \, \sigma_T \, u' \, \eta \, R_b}
407: \label{gammac}
408: \end{equation}
409: The resulting quasi-equilibrium electron energy distribution will be a broken
410: powerlaw with parameters depending on whether $\gamma_c > \gamma_{\rm min}$
411: (slow-cooling regime) or $\gamma_c < \gamma_{\rm min}$ (fast-cooling regime).
412: Re-writing the radiative cooling rate as $\dot\gamma \equiv - \nu_0 \, \gamma^2$
413: with $\nu_0 = (4/3) \, c \, \sigma_T \, u' / (m_e c^2)$ we find for
414: the slow-cooling regime
415:
416: \begin{equation}
417: N_{\rm sc} (\gamma, t) \approx
418: \cases{
419: {Q_0 \, t_{\rm esc} \over (q - 1)} \, \gamma^{-q}
420: & for $\gamma_{\rm min} < \gamma < \gamma_c$ \cr\cr
421: {Q_0 \over \nu_0 \, (q - 1)} \, \gamma^{-(q + 1)}
422: & for $\gamma_c < \gamma < \gamma_{\rm max}$ \cr
423: }
424: \label{Nsc}
425: \end{equation}
426: In the fast-cooling regime, we have
427:
428: \begin{equation}
429: N_{\rm fc} (\gamma, t) \approx
430: \cases{
431: {Q_0 \over \nu_0 \, (q - 1)} \, \gamma_{\rm min}^{1 - q} \, \gamma^{-2}
432: & for $\gamma_c < \gamma < \gamma_{\rm min}$ \cr\cr
433: {Q_0 \over \nu_0 \, (q - 1)} \, \gamma^{-(q + 1)}
434: & for $\gamma_{\rm min} < \gamma < \gamma_{\rm max}$ \cr
435: }
436: \label{Nfc}
437: \end{equation}
438:
439: The apparent large break in the spectral index around the synchrotron peak
440: in the SED of 3C~279 indicates that this peak can not be attributed to a
441: cooling break, but rather to a large value of $\gamma_{\rm min}$, and the
442: system is in the fast-cooling regime \citep{boettcher08}.
443: Eqs. (\ref{gmin}) and (\ref{gammac}) naturally yield values that
444: support the assumption of the fast-cooling regime for plausible parameter
445: values:
446:
447: \begin{equation}
448: \gamma_{\rm min} \sim 600 {\epsilon_{e, -1} \over \xi_e} \Gamma_1
449: \label{gminnum}
450: \end{equation}
451: where $\epsilon_{e, -1} \equiv \epsilon_e / 0.1$, while
452:
453: \begin{equation}
454: \gamma_c \lesssim 30 \, \Gamma_1^{-2} \, R_{16}^{-1}
455: \label{gcnum}
456: \end{equation}
457: where we have only taken into account Compton cooling on the external
458: radiation field for a characteristic value of $u'_{\rm ext}$ as given
459: in Eq. (\ref{ublr}). Consequently, $\gamma_{\rm min} > \gamma_{\rm c}$
460: as long as (formally) $\Gamma > 0.5 \, R_{16}^{-1} \, \xi_e / \epsilon_{e, -1}$,
461: which will be the case for all plausible parameter values, even when
462: the plasmoid becomes non-relativistic.
463:
464: \section{\label{radiation}Radiation}
465:
466: The relativistic electrons described by the distribution functions
467: (\ref{Nsc}) and (\ref{Nfc}), will emit synchrotron, synchrotron-self-Compton
468: (SSC) and EC radiation. For simplicity, we express the synchrotron
469: emissivity using a $\delta$ function approximation:
470:
471: \begin{equation}
472: \nu F_{\nu}^{\rm sy} (\nu_{\rm obs}) = {D^4 \, \epsilon' \, c \, \sigma_T \,
473: {u'}_B \over 6 \, \pi \, d_L^2 \, \epsilon_B \, (1 + z)} \left( {\epsilon'
474: \over \epsilon_B} \right)^{1/2} \, N_e \left( \sqrt{\epsilon' \over \epsilon_B}
475: \right)
476: \label{Fsy}
477: \end{equation}
478: where $\epsilon_B = B/B_{\rm crit}$ with $B_{\rm crit} = 4.414\times
479: 10^{14}$~G, and $\epsilon' = ([1 + z] / D) \, h \nu_{\rm obs} / (m_e c^2)$.
480: The corresponding photon number density is
481:
482: \begin{equation}
483: {n'}_{\rm sy} ({\epsilon'}_{\rm sy}) = {3 \, \sigma_T \, {u'}_B \over 8 \,
484: \pi \, \epsilon_B \, {\epsilon'}_{\rm sy} \, m_e c^2 \, R_b^2} \,
485: \left( {{\epsilon'}_{\rm sy} \over \epsilon_B} \right)^{1/2} \,
486: N_e \left( \sqrt{{\epsilon'}_{\rm sy}
487: \over \epsilon_B} \right).
488: \label{nsy}
489: \end{equation}
490: Writing the electron distribution (\ref{Nsc}) or (\ref{Nfc}) as
491:
492: \begin{equation}
493: N_e (\gamma) = N_0 \, \cases{
494: \left( {\gamma \over \gamma_b} \right)^{-p_1} & for $\gamma_1 < \gamma < \gamma_b$ \cr\cr
495: \left( {\gamma \over \gamma_b} \right)^{-p_2} & for $\gamma_b < \gamma < \gamma_2$ \cr
496: }
497: \label{Nnew}
498: \end{equation}
499: with $(\gamma_1, \gamma_b, \gamma_2) = (\gamma_{\rm min}, \gamma_c, \gamma_{\rm max})$
500: and $(p_1, p_2) = (q, q+1)$ for the slow-cooling regime and $(\gamma_1, \gamma_b,
501: \gamma_2) = (\gamma_c, \gamma_{\rm min}, \gamma_{\rm max})$ and $(p_1, p_2) =
502: (2, q+1)$ for the fast-cooling regime, we can express the photon energy density
503: as
504:
505: \begin{equation}
506: {u'}_{\rm syn} = {16 \over 9} \, \sigma_T \, {u'}_B { R_B \, N_0 \over V_B}
507: \left( \gamma_b^{p_1} {\gamma_b^{3 - p_1} - \gamma_1^{3 - p_1} \over 3 - p_1}
508: + \gamma_b^{p_2} {\gamma_2^{3 - p_2} - \gamma_b^{3 - p_2} \over 3 - p_2}
509: \right).
510: \label{usy}
511: \end{equation}
512:
513: The photon number density (\ref{nsy}) can be used in the \cite{jones68} formula
514: to evaluate the SSC flux:
515:
516: \begin{equation}
517: \nu F_{\nu}^{\rm SSC} (\nu_{\rm obs}) = {D^4 \, {\epsilon'}^2 \, m_e c^2 \
518: \over 4 \pi \, d_L^2 \, (1 + z)} \int\limits_1^{\infty} d\gamma \,
519: N_e (\gamma) \, \int\limits_0^{\infty} d{\epsilon'}_{\rm sy} \, {n'}_{\rm sy}
520: ({\epsilon'}_{\rm sy}) \, g(\epsilon', {\epsilon'}_{\rm sy}, \gamma)
521: \label{nFnssc}
522: \end{equation}
523: with
524: \begin{equation}
525: g(\epsilon', {\epsilon'}_{\rm sy}, \gamma) = {c \, \pi \, r_e^2 \over 2 \,
526: \gamma^4 {\epsilon'}_{\rm sy}}
527: \left( {4 \, \gamma^2 \, \epsilon' \over {\epsilon'}_{\rm sy}} - 1 \right)
528: \;\;\;\;\;\; {\rm if}
529: \;\;\;\;\;\; {{\epsilon'}_{\rm sy} \over 4 \, \gamma^2} \le \epsilon' \le
530: {\epsilon'}_{\rm sy},
531: \label{g1}
532: \end{equation}
533: and
534: $$
535: g(\epsilon', {\epsilon'}_{\rm sy}, \gamma) = {2 \, c \, \pi \, r_e^2 \over
536: \gamma^2 {\epsilon'}_{\rm sy}} \left( 2 \, q \, \ln q + (1 + 2 q) (1 - q) +
537: {(4 \, {\epsilon'}_{\rm sy} \, \gamma \, q)^2 \over (1 + 4 \, {\epsilon'}_{\rm sy}
538: \, \gamma \, q)} \, {(1 - q) \over
539: 2} \right)
540: $$
541: \begin{equation}
542: {\rm if} \;\;\;\;\;\;\;\;\;\; {\epsilon'}_{\rm sy} \le \epsilon' \le
543: {4 \, {\epsilon'}_{\rm sy} \, \gamma^2 \over 1 + 4 \, {\epsilon'}_{\rm sy}
544: \, \gamma}
545: \label{g2}
546: \end{equation}
547: where
548: \begin{equation}
549: q = {\epsilon' \over 4 \, {\epsilon'}_{\rm sy} \, \gamma^2 \left( 1 -
550: {\epsilon' \over \gamma} \right)}.
551: \label{q}
552: \end{equation}
553:
554: We evaluate the external-Compton photon spectrum with a $\delta$ function
555: approximation for the external radiation field,
556:
557: \begin{equation}
558: {n'}_{\rm ext} (\epsilon, \Omega) \approx {{u'}_{\rm ext}
559: \over {\epsilon'}_{\rm ext} \, m_e c^2} \, \delta(\epsilon - {\epsilon'}_{\rm ext})
560: \, \delta(\mu' + 1)
561: \label{next}
562: \end{equation}
563: where $\mu' = \cos{\theta'}_{\rm ext}$ refers to the angle of incidence of the
564: external photons with respect to the jet axis, and ${\epsilon'}_{\rm ext} =
565: \Gamma \epsilon_{\rm ext}$. The Compton cross section is also approximated by
566: a $\delta$ function,
567:
568: \begin{equation}
569: {d^2 \sigma_C \over d\Omega' \, d\epsilon'} \approx \sigma_T \, \delta (\Omega'
570: - {\Omega'}_e) \, \delta (\epsilon' - \gamma^2 \, {\epsilon'}_{\rm ext} \,
571: [1 - \beta \mu]).
572: \label{sigmac}
573: \end{equation}
574: With these simplifications, the EC flux can be calculated as
575:
576: \begin{equation}
577: \nu F_{\nu}^{\rm EC} (\nu_{\rm obs}) = {D^4 \, c \, \sigma_T \, {u'}_{\rm ext}
578: \over 8 \, \pi \, d_L^2 \, (1 + z)} \, \left( {{\epsilon'} \over
579: {\epsilon'}_{\rm ext}} \right)^{3/2} \, \sqrt{1 + \mu_{\rm obs}} \,
580: N_e \left( \sqrt{\epsilon' \over {\epsilon'}_{\rm ext} \, (1 + \mu_{\rm obs})}
581: \right).
582: \label{nFnec}
583: \end{equation}
584:
585:
586: \section{\label{deceleration}Asymptotic Behaviour in the Deceleration Phase}
587:
588: As in the well-known case of expanding blast waves in Gamma-Ray Bursts
589: (GRBs), the plasmoid in a decelerating jet (with constant cross section
590: $A$) is starting out in a coasting phase, in which the initial mass $M_0$
591: greatly exceeds the swept-up relativistic mass in the co-moving frame.
592: During this phase, the effect of the inertia of the swept-up mass is
593: negligible, and the Lorentz factor remains roughly constant. This phase
594: is followed by the deceleration phase. In the asymptotic limit of that
595: phase, the initial mass of the plasmoid becomes negligible.
596:
597: Although we have properly included the effect of radiative cooling and
598: radiation drag, in most cases the fraction of swept-up energy which is
599: transferred to ultrarelativistic electrons and can therefore be radiated
600: away efficiently, will be small. Therefore, we can approximate the equation
601: of motion of the plasmoid by an adiabatic solution, as
602:
603: \begin{equation}
604: {d\Gamma \over dr} \approx {- \Gamma^2 \, A \, \rho_{\rm ext} \over M}
605: \label{dGdrdecel}
606: \end{equation}
607: and
608: \begin{equation}
609: M(r) \approx A \, \rho_{\rm ext} \int\limits_{r_0}^r \Gamma (r') \, dr'
610: \label{Mdecel}
611: \end{equation}
612:
613: where we have, for simplicity, assumed a constant external density
614: $\rho(r) \equiv \rho_{\rm ext} =$~const. This system has a self-similar
615: solution of the form
616:
617: \begin{equation}
618: \Gamma(r) = \Gamma_0 \, \left( {r \over r_0} \right)^{-1/2}.
619: \label{Gammadecel}
620: \end{equation}
621: Assuming for the purpose of an analytic estimate that we are looking
622: right down the jet ($\theta_{\rm obs} = 0$), the observer's time
623: as a function of Lorentz factor can be expressed as
624:
625: \begin{equation}
626: t_{\rm obs} = t_0 + \int\limits_{r_0}^r { (1 - \beta(r') \cos\theta_{\rm obs})
627: \over \beta(r') c} \, dr' \approx {1 \over 2 \, c} \int\limits_{r_0}^r
628: {dr' \over \Gamma^2 (r')} \approx {r^2 \over 4 \, r_0 \, \Gamma_0^2 \, c}.
629: \label{tobsdecel}
630: \end{equation}
631: This yields a solution for the Lorentz factor as a function of observer's
632: time:
633:
634: \begin{equation}
635: \Gamma(t_{\rm obs}) \approx \sqrt{\Gamma_0 \over 2} \, \left( {r_0 \over
636: c \, t_{\rm obs}} \right)^{1/4} \propto t_{\rm obs}^{-1/4}.
637: \label{Gamma_tobs}
638: \end{equation}
639:
640: The observed steep spectral index of the optical synchrotron emission
641: from 3C~279 ($\alpha \sim 1.7$), indicates $p \sim 4.4$. This, in turn,
642: signifies that the system is in the fast cooling regime since otherwise
643: a cooling break would not produce a $\nu F_{\nu}$ peak at the synchrotron
644: frequency corresponding to $\gamma_b$. Furthermore,
645: electrons synchrotron radiating at optical frequencies are most likely
646: beyond the break energy, i.e., $\gamma > \gamma_{\rm min}$. For the
647: prediction of synchrotron light curves we may therefore use Eq. \ref{Fsy}
648: together with the lower branch of Eq. \ref{Nfc}. For $\gamma_{\rm max}
649: \gg \gamma_{\rm min}$, the coefficient $Q_0$ in Eq. \ref{Nfc} may be
650: approximated as
651:
652: \begin{equation}
653: Q_0(r) \approx {\xi_e \, A \, \Gamma(r) \, \rho_{\rm ext} \, c \over m_p}
654: \, (q - 1) \, \gamma_{\rm min}^{q - 1}(r) \propto \Gamma^q.
655: \label{Q0decel}
656: \end{equation}
657: The cooling coefficient $\nu_0$ is expected to be dominated by synchrotron
658: and/or external Compton cooling. With the magnetic-field scaling from
659: Eq. \ref{B}, both $u'_{\rm ext}$ and $u'_B$ carry a dependence $\propto
660: \Gamma^2$. Thus, we find for the normalization of the ultrarelativistic
661: particle population:
662:
663: \begin{equation}
664: N_e (\gamma, t_{\rm obs}) \propto \Gamma^{q - 2} \, \gamma^{-(q + 1)}
665: \propto t_{\rm obs}^{(2 - q)/4} \, \gamma^{-(q + 1)}.
666: \label{Nedecel}
667: \end{equation}
668:
669: In order to use Eq. \ref{Fsy} for a light curve estimate, we assume,
670: again, for simplicity, $\theta_{\rm obs} = 0$ and therefore $D \approx
671: 2 \, \Gamma$. Consequently, the characteristic electron energy
672: $\gamma = \sqrt{\epsilon' / \epsilon_B} \propto \Gamma^{-1}$.
673: This yields an expected light curve decay in the fast-cooling
674: synchrotron regime of
675:
676: \begin{equation}
677: \nu F_{\nu}^{\rm sy} (\nu_{\rm obs}, t_{\rm obs}) \propto
678: \nu_{\rm obs}^{(2 - q)/2} \, \Gamma^{2 \, (1 + q)} \propto
679: \nu_{\rm obs}^{(2 - q)/2} \, t_{\rm obs}^{- (1 + q)/2}.
680: \label{Fsylightcurve}
681: \end{equation}
682: In particular, for an injection index of $q = 3.4$, as inferred from the
683: optical spectral index, a light curve of $F_{\nu} \propto t_{\rm obs}^{-2.2}$
684: is expected. However, it should be pointed out that this
685: can only be considered an upper limit to the steepness of the decay.
686: Any non-zero observing angle will flatten the decay of the light curve
687: as it introduces a shallower decay of the Doppler factor $D$ with
688: decreasing $\Gamma$ and therefore with time. In any case, by the
689: time the plasmoid is in the self-similar deceleration phase, the
690: monochromatic flux has already decreased by about an order of magnitude
691: from its initial peak value, and is likely to be overwhelmed by other
692: emission components in the jet. We do therefore not expect to observe
693: the limiting deceleration case directly. We only developed this analytical
694: case to demonstrate the agreement of our numerical simulations with the
695: analytical expectation in the following section.
696:
697: \begin{figure}[ht]
698: \plotone{f2.eps}
699: \caption{Simulated snap-shot SEDs from our decelerating plasmoid model,
700: compared to various observed SEDs of 3C~279 in various observing epochs
701: and activity states. The time sequence goes from red - dotted $\to$ dashed
702: $\to$ dot-dashed $\to$ long-dashed $\to$ solid $\to$ orange - dotted $\to$
703: dashed $\to$ dot-dashed $\to$ long-dashed. This simulation provides a good
704: fit to the optical light curve decay of 3C~279 around January 15, 2006
705: (see Fig. \ref{lcfits}). }
706: \label{spsequence}
707: \end{figure}
708:
709: \section{\label{results}Numerical Results}
710:
711: In order to highlight the salient features of our model, we assume a simple
712: cylindrical jet geometry with a constant cross section of the jet, $A \equiv
713: \pi R_b^2$, as well as a homogeneous external medium with density $\rho (r)
714: \equiv \rho_{\rm ext}$.
715: We achieved good fits to the observed optical light
716: curves and overall SED shape of 3C~279 with the parameters listed in Table
717: \ref{parameters}. Our choice of the external density corresponds to
718: a number density of $n_{\rm ext} = 100$~cm$^{-3}$. This is at least about
719: an order of magnitude lower than typical particle densities in the broad
720: line regions of quasars. This is quite reasonable since we expect that the
721: jet trajectory is already partially evacuated from previous ejection events.
722:
723: Fig. \ref{spsequence} compares snap-shot SEDs of our simulation to various
724: observed broadband SEDs of 3C~279. It can be seen that the optical continuum
725: spectra in various intensity states are well represented by the model, the
726: X-ray spectral slope corresponds to the characteristically observed shape,
727: and for much of the plasmoid evolution, the simulated X-ray flux is in the
728: range of observed values. The $\gamma$-ray flux represented by our simulation
729: corresponds to a medium to low state of 3C~279. Note that this work is not
730: intended to attempt a model interpretation of the very-high-energy $\gamma$-ray
731: flux detected by MAGIC on Feb. 23, 2006 \citep{albert08}, in particular
732: since we did not take into account any effects of intrinsic or intergalactic
733: $\gamma\gamma$ absorption. For model implications of this high-energy detection
734: see \cite{boettcher08}.
735:
736:
737: The upper panel in figure \ref{complete} illustrates
738: the evolution of the plasmoid bulk Lorentz factor $\Gamma$ as a function of
739: distance from the central engine, obtained by numerically solving the coupled
740: system of Eqs. (\ref{dGdr}) and (\ref{dMdr}).
741:
742: \begin{figure}[ht]
743: \plotone{f3.eps}
744: \caption{Top: Evolution of Lorentz factor $\Gamma$ as a function of distance
745: $r$ from the central engine for the simulation illustrated in Fig.
746: \ref{spsequence}. Bottom: R-band light curve from this simulation.}
747: \label{complete}
748: \end{figure}
749:
750: As in the well-known case of the blast-wave model for gamma-ray bursts,
751: the plasmoid evolves through a coasting phase with approximately constant
752: Lorentz factor $\Gamma_0$. Around the characteristic deceleration radius
753: $r_d$, at with the swept-up relativistic mass equals the mass of the initial
754: ejecta $M_0$, the evolution makes a gradual transition into the asymptotic
755: self-similar deceleration phase with $\Gamma (r) \propto r^{-1/2}$, treated
756: analytically in \S \ref{deceleration}. For the input parameters listed in
757: Table \ref{parameters}, this occurs at
758:
759: \begin{equation}
760: r_d = {M_0 \over \Gamma_0 \, \pi \, R_b^2 \, \rho_{\rm ext}} \approx 1.1
761: \times 10^{18} \; {\rm cm}.
762: \label{rd}
763: \end{equation}
764:
765: Note, however, that substantial deceleration happens long before $r_d$ is
766: reached. The lower panel of Fig. \ref{complete} shows the resulting R-band light
767: curve over the entire evolution of the plasmoid, down to a mildly relativistic
768: speed. Again, we see the gradual turnover to the self-similar deceleration phase
769: with the expected light curve decay as $F_{\rm sy} \propto t_{\rm obs}^{-2.2}$
770: (see \S \ref{deceleration}). During the transition from the coasting to the
771: self-similar deceleration phase, the light curve is reasonably well represented
772: by an exponential decay.
773:
774: \begin{figure}[ht]
775: \plotone{f4.eps}
776: \caption{Fit of our decelerating-jet model to the V, R, and I band fluxes
777: of 3C~279. }
778: \label{lcfits}
779: \end{figure}
780:
781: The detailed fits to the V, R, and I band light curves of 3C~279 during the
782: quasi-exponential decay around January 15, 2006, are shown in Fig. \ref{lcfits}.
783: The B and U band light curves were rather poorly sampled during this time
784: period so that the (equally good) fits to those light curves does not provide
785: substantial additional information.
786:
787: Using characteristic parameters for the accretion disk luminosity, $L_D \equiv
788: 10^{45} \, L_{45}$~ergs~s$^{-1}$ and the Thomson depth $\tau_{\rm T, BLR} \equiv
789: 10^{-1} \, \tau_{-1}$ and radius $R_{\rm BLR} \equiv 0.1 \, R_{-1}$~pc of
790: the BLR of 3C~279, we estimate an external radiation energy density of
791:
792: \begin{equation}
793: u_{\rm ext} \approx {L_D \, \tau_{\rm T, BLR} \over 4 \, \pi \,
794: R_{\rm BLR}^2 \, c} \sim 3 \times 10^{-3} \, L_{45} \, \tau_{-1} \,
795: R_{-1}^{-2} \; {\rm ergs \; cm}^{-3}
796: \label{ublr}
797: \end{equation}
798: and $\epsilon_{\rm ext} \sim 10^{-5}$ \citep[see, e.g.,][]{boettcher08},
799: which we use for our calculation of the EC $\gamma$-ray emission component
800: and the associated radiation drag term. The predicted light curves of the
801: combined SSC + EC emissions at X-rays and $\gamma$-rays are compared to
802: the R-band light curves in Fig. \ref{Xgammalc}.
803:
804: \begin{figure}[ht]
805: \plotone{f5.eps}
806: \caption{Predicted X-ray and $\gamma$-ray light curves for parameters from
807: our fit to the optical light curves of 3C~279. }
808: \label{Xgammalc}
809: \end{figure}
810:
811: The most remarkable feature of the predicted light curves is a delayed
812: X-ray outburst about 2 -- 3 weeks after the onset of the optical decline.
813: This is a consequence of the shift of the SSC peak towards lower
814: frequencies as the blob decelerates, which follows from the dependence
815: $\nu_{\rm pk, SSC} \propto B^2 \, \gamma_{\rm min}^4 \propto \Gamma^6$.
816: This SSC peak frequency decrease is much more rapid than the decrease
817: of the SSC peak flux. This spectral evolution of the SSC component is
818: illustrated in Fig. \ref{spsequence}.
819: In fact, RXTE monitoring of the source \citep[see][]{boettcher08}
820: did detect a transition of 3C~279 from a quiescent state throughout the first
821: half of January 2006, to a very active, high X ray flux state with a substantial
822: X-ray outburst around Feb. 5, 2006. It is tempting to postulate that this may
823: have been the X-ray signature of the plasmoid deceleration observed during
824: the optical decay around January 15.
825:
826: A prediction of this scenario is that the $\sim 100$~MeV $\gamma$-ray light
827: curve peaks a few days after the optical before it begins its quasi-exponential
828: decay. This is a consequence of the fact that this energy range consists of
829: comparable contributions from the SSC and the EC radiation components. As the
830: SSC component shifts towards lower energies during the early deceleration, the
831: EC component just moves into the 100 MeV range, before it also decays away to
832: lower peak energies and lower flux levels. The very-high-energy ($\gtrsim 100$~GeV)
833: light curve is expected to follow closely the optical one, with only a small
834: delay, which critically depends on the initial bulk Lorentz factor and the
835: mean photon energy of the external photon field.
836:
837: We caution that any predictions concerning absolute flux levels at X-rays
838: and $\gamma$-rays are very model-parameter dependent and vary substantially
839: for different choices of initial mass, initial bulk Lorentz factor, initial
840: pladmoid radius, external matter density, external radiation field, etc.
841: Furthermore, we have not included any effects of intrinsic \citep[in particular,
842: in the radiation field of the BLR, see, e.g.,][]{dp03,reimer07,sb08}, nor
843: intergalactic $\gamma\gamma$ absorption. However, while this may significantly
844: affect the overall $\gamma$-ray flux level, the light curve features discussed
845: above are predominantly a consequence of the plasmoid dynamics, which are
846: dictated by the observed optical light curves, and are therefore robust
847: predictions of the decelerating-jet model.
848:
849: \section{\label{summary}Summary and Conclusions}
850:
851: Motivated by an extraordinarily clean quasi-exponential decay of the
852: V, R, and I band light curves of 3C~279 over a period of about 2 weeks
853: during a recent WEBT campaign, we proposed a model of a decelerating
854: plasmoid in the jet of this quasar. We take into account self-consistently
855: the inertia of swept-up mass as the plasmoid propagates through the gas
856: of the AGN environment, as well as radiation drag and radiative cooling.
857: We have demonstrated that, similar to the relativistic blast wave model
858: for GRBs, the plasmoid makes a transition from a coasting phase with
859: approximately constant Lorentz factor, to a self-similar phase. In the
860: case of a homogeneous external medium and a cylindrical jet, the self-similar
861: deceleration phase is described by $\Gamma (r) \propto r^{-1/2} \propto
862: t_{\rm obs}^{-1/4}$. The resulting optical synchrotron light curves are
863: well approximated by quasi-exponential decays during the transition from
864: the coasting to the self-similar decay phase. In the asymptotic limit
865: of the self-similar decay phase, the synchrotron light curves at frequencies
866: corresponding to electron energies above the cooling break of the electron
867: spectrum follow a behaviour of $\nu F_{\nu} (\nu_{\rm obs} \, t_{\rm obs})
868: \propto \nu_{\rm obs}^{(2 - q)/2} \, t_{\rm obs}^{-(1 + q)/2}$, where
869: $q$ is the injection spectral index of ultrarelativistic electrons in
870: the plasmoid.
871:
872: We note that the choice of a conical jet instead of a cylindrical
873: one would recover the isotropic blast wave model for gamma-ray bursts
874: \citep[e.g.,][]{cd99}: It would lead to a scaling of the plasmoid surface
875: as $A \propto r^2$. In the self-similar deceleration phase, this recovers
876: the well-known adiabatic blast wave solution with $\Gamma (r) \propto
877: r^{-3/2} \propto t_{\rm obs}^{-8/3}$. This case describes a much faster
878: deceleration of the plasmoid and therefore substantially steeper light
879: curves than the cylindrical jet geometry assumed here.
880:
881: We have demonstrated that this model can adequately reproduce the observed
882: optical light curves of 3C~279 during the $\sim 2$ week long quasi-exponential
883: decay phase in January 2006. This model predicts a delayed (SSC-dominated)
884: X-ray outburst about 2 -- 3 weeks after the onset of the optical decay.
885: We speculate that the X-ray flare around Feb. 5, 2006, detected by {\it RXTE}
886: monitoring, may have been the X-ray signature of the plasmoid deceleration
887: seen earlier in the optical bands.
888:
889: A robust prediction of the decelerating plasmoid model which can be tested
890: with Fermi and simultaneous optical monitoring is that the peak in the
891: $\gamma$-ray light curve at $\sim 100$~MeV is expected to be delayed by
892: a few days with respect to the onset of a quasi-exponential light curve
893: decay in the optical, while the VHE $\gamma$-rays are expected to track
894: the optical light curve closely.
895:
896: The quasar 3C~279 is one of the most active blazars known. The ejection of
897: $\gamma$-ray emitting plasmoids from the nucleus of 3C~279 might be a frequent
898: event. We do therefore expect to observe the quasi-exponential decay phase
899: following such ejection events only in a temporary quiescent phase, in which
900: it is not overwhelmed by subsequent ejection events. This might be the reason
901: why these quasi-exponential decays in the light curves of this blazar are not
902: more frequently observed. In addition to one or two occurrences during the
903: 2006 WEBT campaign, several more quasi-exponentially decaying light curve
904: segments can also be identified in the 2007 WEBT campaign data \citep{larionov08},
905: indicating that these are rare, but not unique events.
906:
907:
908: \acknowledgments
909: We thank the anonymous referee for a very constructive report.
910: We thank Matthew G. Baring for stimulating discussions. This work
911: was supported by NASA through XMM-Newton Guest Observer Grant
912: NNX08AD67G.
913: \begin{thebibliography}{}
914:
915: \bibitem[Albert et al.(2008)]{albert08}Albert, J., et al., 2008, Science,
916: vol. 320, no. 5884, p. 1752
917:
918: \bibitem[Bednarek(1998)]{bednarek98}Bednarek, W., 1998, A\&A, 336, 123
919:
920: \bibitem[Begelman et al.(2008)]{bfr08}Begelman, M. C., Fabian, A. C., \&
921: Rees, M. J., 2008, MNRAS, 384, L19
922:
923: \bibitem[B\"ottcher, Mause, \& Schlickeiser(1997)]{bms97}B\"ottcher, M.,
924: Mause, H., \& Schlickeiser, R., 1997, A\&A, 324, 395
925:
926: \bibitem[B\"ottcher \& Bloom(2000)]{bb00}B\"ottcher, M., \& Bloom, S. D., 2000,
927: AJ, 119, 469
928:
929: \bibitem[B\"ottcher(2007a)]{boettcher07a}B\"ottcher, M., 2007a, in proc.
930: ``The Multimessenger Approach to Gamma-Ray Sources'', ApSS, 309, 95
931:
932: \bibitem[B\"ottcher et al.(2007b)]{boettcher07b}B\"ottcher, M., et al., 2007, ApJ,
933: 670, 970
934:
935: \bibitem[B\"ottcher, Reimer, \& Marscher(2008)]{boettcher08}B\"ottcher, M.,
936: Reimer, A. F., \& Marscher, A. P., 2008, ApJ, submitted
937:
938: \bibitem[Chiang \& Dermer(1999)]{cd99}Chiang, J., \& Dermer, C. D., 1999,
939: ApJ, 512, 699
940:
941: \bibitem[Donea \& Protheroe(2003)]{dp03}Donea, A.-C., \& Protheroe, R. J., 2003,
942: Astrop. Phys., vol. 18, Issue 4, p. 337
943:
944: \bibitem[Finke et al.(2008)]{finke08}Finke, J. D., Dermer, C. D., \& B\"ottcher, M.,
945: 2008, ApJ, in press (arXiv:0802.1529)
946:
947: \bibitem[Georganopoulos \& Kazanas(2003a)]{gk03a}Georganopoulos, M., \&
948: Kazanas, D., 2003a, ApJ, 589, L5
949:
950: \bibitem[Georganopoulos \& Kazanas(2003b)]{gk03b}Georganopoulos, M., \&
951: Kazanas, D., 2003b, ApJ, 594, L27
952:
953: \bibitem[Ghisellini \& Tavecchio(2008)]{gt08}Ghisellino, G.,
954: \& Tavecchio, F., 2008, MNRAS, 386, L28
955:
956: \bibitem[Hartman et al.(1996)]{hartman96}Hartman, R. C., et al.,
957: 1996, ApJ, 461, 698
958:
959: \bibitem[Hartman et al.(2001)]{hartman01}Hartman, R. C., et al.,
960: 2001, ApJ, 553, 683
961:
962: \bibitem[Jones(1968)]{jones68}Jones, T. W., 1968, Phys. Rev., 167, 1159
963:
964: \bibitem[Larionov et al.(2008)]{larionov08}Larionov, V. M., et al., 2008,
965: A\&A, submitted
966:
967: \bibitem[Maraschi et al.(1994)]{maraschi94}Maraschi, L., et al., 1994,
968: ApJ, 435, L91
969:
970: %\bibitem[Marscher et al.(2002)]{marscher02}Marscher, A. P., et al., 2002,
971: %Nature, 417, 625
972:
973: \bibitem[M\'esz\'aros \& Rees(1997)]{mr97}M\'esz\'aros, P., \& Rees, M. J.,
974: 1997, ApJ, 476, 232
975:
976: \bibitem[Moderski et al.(2003)]{moderski03}Moderski, R., Sikora, M.,
977: Blazejowski, M., 2003, A\&A, 406, 855
978:
979: \bibitem[M\"ucke \& Protheroe(2001)]{muecke01}M\"ucke, A., \& Protheroe, R. J.,
980: 2001, Astropart. Phys., 15, 121
981:
982: \bibitem[M\"ucke et al.(2003)]{muecke03}M\"ucke, A., Protheroe, R. J.,
983: Engel, R., Rachen, J. P., \& Stanev, T., 2003, Astropart. Phys.,
984: 18, 593
985:
986: \bibitem[Paczy\'nski \& Roads(1993)]{pr93}Paczy\'nski, B., \& Roads, J., 1993,
987: ApJ, 418, L5
988:
989: %\bibitem[Pian et al.(1999)]{pian99}Pian, E., et al., 1999, ApJ, 521, 112
990:
991: \bibitem[Piner et al.(2000)]{piner00}Piner, B. G., Edwards, P. G., Wehrle, A. E.,
992: Hirabayashi, H., Lovell, J. E. J., \& Unwin, S. C., 2000, ApJ, 537, 91
993:
994: \bibitem[Pohl \& Schlickeiser(2000)]{ps00}Pohl, M., \& Schlickeiser, R.,
995: 2000, A\&A, 354, 395
996:
997: %\bibitem[Primack et al.(2005)]{primack05}Primack, J. R., Bullock, J. S.,
998: %\& Somerville, R. S., in ``High-Energy Gamma-Ray Astronomy'', AIP Conf.
999: %Series, F. Aharonian, H. Voelk, D. Horns, Eds. (AIP, Heidelberg, 2005),
1000: %vol. 745, p. 23
1001:
1002: \bibitem[Reimer(2007)]{reimer07}Reimer, A., 2007, ApJ, 665, 1023
1003:
1004: \bibitem[Sikora et al.(2001)]{sikora01}Sikora, M., Blazejowski, M.,
1005: Begelman, M. C., \& Moderski, R., 2001, ApJ, 554, 1;
1006: Erratum: ApJ, 561, 1154 (2001)
1007:
1008: \bibitem[Sitarek \& Bednarek(2008)]{sb08}Sitarek, J., \& Bednarek, W.,
1009: 2008, MNRAS, in press (arXiv:0807.4228)
1010:
1011: \bibitem[Wehrle et al.(1998)]{wehrle98}Wehrle, A. E., et al., 1998, ApJ,
1012: 497, 178
1013:
1014: \end{thebibliography}
1015:
1016:
1017: \begin{deluxetable}{ccccc}
1018: \tabletypesize{\scriptsize}
1019: \tablecaption{Parameters of our plasmoid evolution simulation providing a fit
1020: to the V, R, and I band light curves of 3C~279 around January 15, 2006.}
1021: \tablewidth{0pt}
1022: \tablehead{
1023: \colhead{Parameter} & \colhead{Symbol} & \colhead{Value} }
1024: \startdata
1025: Initial Lorentz factor & $\Gamma_0$ & $50$ \\
1026: External matter density & $n_{\rm ext} = \rho_{\rm ext} / m_p$ & $100$~cm$^{-3}$ \\
1027: Plasmoid radius & $R_b$ & $3 \times 10^{16}$~cm \\
1028: Initial mass & $M_0$ & $2.6 \times 10^{31}$~g \\
1029: Electron injection index & $q$ & $3.4$ \\
1030: B-field equipartition parameter & $e_B$ & $10^{-3}$ \\
1031: Electron acceleration efficiency & $\epsilon_e$ & $0.1$ \\
1032: Relativistic electron fraction & $\xi_e$ & $0.5$ \\
1033: Electron escape time scale parameter & $\eta$ & $10$ \\
1034: External radiation energy density & $u_{\rm ext}$ & $3 \times 10^{-3}$~erg~cm$^{-3}$ \\
1035: Observing angle & $\theta_{\rm obs}$ & $5^o$ \\
1036: \enddata
1037: \label{parameters}
1038: \end{deluxetable}
1039:
1040:
1041: \end{document}
1042:
1043: