1: \documentclass[12pt,preprint]{aastex}
2: \begin{document}
3:
4: \title{The Outburst of the Blazar AO 0235+164 in 2006 December: Shock-in-Jet Interpretation}
5: \author{V.A. Hagen-Thorn \altaffilmark{1,4}, V.M. Larionov \altaffilmark{1,2,4},
6: S.G. Jorstad \altaffilmark{1,3}, A.A. Arkharov \altaffilmark{1,2},
7: E.I. Hagen-Thorn \altaffilmark{1,2}, N.V. Efimova \altaffilmark{1,2},
8: L.V. Larionova \altaffilmark{1}, A.P. Marscher \altaffilmark{3}}
9:
10:
11: \altaffiltext{1}{Astronomical Institute of St. Petersburg State University, Universitetskiy Pr. 28, Petrodvorets, 198504, St. Petersburg, Russia; HTH-home@yandex.ru}
12: \altaffiltext{2}{Main (Pulkovo) Astronomical Observatory of RAS, Pulkovskoe Sh. 60, 196140, St. Petersburg, Russia; vlar@astro.spbu.ru}
13: \altaffiltext{3}{Institute for Astrophysical Research, Boston University,
14: 725 Commonwealth Ave., Boston, MA 02215-1401; jorstad@bu.edu}
15: \altaffiltext{4}{Isaac Newton Institute of Chile, St.-Petersburg Branch}
16: \shorttitle{Outburst of A0235+164 in 2006 December}
17: \shortauthors{Hagen-Thorn et al.}
18: \begin{abstract}
19: We present the results of polarimetric ($R$ band) and multicolor photometric
20: ($BVRIJHK$) observations of the blazar AO 0235+16 during an outburst in 2006
21: December. The data reveal a short timescale of variability (several hours),
22: which increases from optical to near-IR wavelengths; even shorter variations are
23: detected in polarization. The flux density correlates
24: with the degree of polarization, and at maximum degree of polarization
25: the electric vector tends to align with the parsec-scale jet direction.
26: We find that a variable component with
27: a steady power-law spectral energy distribution and very high optical
28: polarization (30-50\%)
29: is responsible for the variability. We interpret these properties of the blazar within
30: a model of a transverse shock propagating down the jet. In this case a small
31: change in the viewing angle of the jet, by $\lesssim 1^o$, and a decrease
32: in the shocked plasma compression by a factor of $\sim$1.5 are sufficient to
33: account for the variability.
34: \end{abstract}
35: \keywords{galaxies: active --- galaxies: BL Lacertae objects: individual
36: (AO 0235+164) --- galaxies: jet --- polarization}
37:
38: \section{Introduction}
39: Blazars are active galactic nuclei (AGNs) that
40: possess extreme properties such as violent variability, high polarization,
41: relativistic jets, and bright $\gamma$-ray emission.
42: It is commonly thought that the radio through
43: optical emission in blazars is synchrotron radiation
44: that originates in the jets. Polarimetric observations provide
45: crucial data for investigating the magnetic field structure
46: that forms the framework of such radiation. However,
47: the polarization behavior is usually too complex for straightforward
48: interpretation. For example, long-term observations of BL~Lac
49: reveal a change of degree of polarization from 0.5\%
50: to $>$20\% with all possible electric vector position angles
51: (EVPAs), although the distribution of EVPAs indicates the existence of a
52: preferential direction aligned with the parsec-scale jet
53: \citep{HT94}. Difficulties are even more severe for
54: physical models aimed to explain both the brightness and polarization
55: behavior. A connection between polarization and brightness is rarely seen
56: at optical wavelengths \citep{SMITH96}.
57: However, there are strong indications that the optical
58: polarized emission is linked to the mm-wave VLBI core
59: of blazars jets \citep{LS00,GRSS06,FRANI07,J07}.
60:
61: At radio wavelengths some progress in interpetation of both polarization and brightness variability
62: has been achieved by using a model in which a shock propagates down the jet
63: \citep[e.g., ][]{HAA89}. At optical wavelengths, where polarization and brightness
64: can vary on timescale of a few hours \citep[e.g., ][]{HT72,MN96,CARINI06},
65: a proper test of models requires very intensive monitoring of both
66: polarization and brightness. Such simultaneous photometric
67: and polarimetric measurements are difficult to
68: obtain, but they provide a vital tool to distinguish
69: among different models responsible for the variability.
70:
71: We have observed the BL~Lac object AO~0235+164 (z=0.94) during the relatively short
72: outburst in 2006 December as part of the polarimetric and photometric
73: monitoring program of blazars that has been carried out at Saint Petersburg
74: State University for many years. AO~0235+164 shows extreme blazar properties, such
75: as intraday variability \citep[e.g., ][]{R05} and very high apparent speeds in
76: the radio jet, $\beta_{\rm app}=(26\pm 7)c$~ \citep{P06}, that have drawn
77: the attention of many investigators to this object \citep{R01,R05,R06,JUNK04,HT07}.
78: In addition, AO~0235+164 is one of the most highly polarized BL~Lac objects known.
79: During the outburst in 1983 January the optical polarization of the object
80: varied from 12\% to 24\% over a 24 hour period \citep{SMITH96}.
81: Our polarimetric monitoring program has caught the source in a similar
82: state but with an even higher degree of polarization, $\gtrsim~30\% $.
83: The polarimetric observations are enhanced by multicolor photometric
84: measurements. Here we present the observations and show that the correlations
85: that we have uncovered provide strong evidence in support of the shock model.
86:
87: \section{Observational data}
88:
89: Observational data at optical wavelengths ($BVRI$ bands) were obtained at
90: the 70-cm reflector of the Crimean Astrophysical Observatory with the
91: photometer-polarimeter of the Astronomical Institute at St. Petersburg
92: State University based on an ST-7 CCD. Methods of observation and
93: data reduction of the photometric mode are described in detail in \citet{HT06}.
94: We used the comparison stars recommended by the Whole Earth
95: Blazar Telescope (WEBT) program \footnote{http://www.to.astro/blazars/webt/}.
96: The photometric errors do not exceed $0.03^m$ in $B$ band and $0.02^m$
97: in other bands when the object is brighter than $17^m$.
98:
99: Observations at near-IR wavelengths ($JHK$ bands) were carried out at the
100: 1.1-m telescope of the Main (Pulkovo) Astronomical Observatory of the Russian
101: Academy of Sciences located at Campo Imperatore (Italy) and equipped
102: with the SWIRCAM camera. Methods of reduction
103: are the same as for the optical observations. Errors in $JHK$ do not exceed
104: $0.02^m$ when the object is brighter than $16^m$ in J band.
105: In all cases the calibration of \citet{MEAD90} was used
106: to transform magnitudes to flux densities
107: (subsequently referred to as ``fluxes'').
108:
109: Polarimetric observations were performed in $R$ band using two Savart plates
110: rotated by $45^\circ$ relative to each another. By swapping the plates,
111: the observer can obtain either the relative Stokes $q$ or $u$ parameter from the
112: two split images of each source in the field.
113: Instrumental polarization was found via stars located near the object under
114: the assumption that their radiation is unpolarized. The Galactic latitude
115: of AO~0235+164 is $-39^\circ$ so that interstellar polarization (ISP) in this direction is less
116: than 0.7\%. To correct for the ISP, the mean relative Stokes parameters
117: of nearby stars were subtracted from the relative Stokes parameters of
118: the object. This accounts for the instrumental polarization as well.
119: The errors in degree of polarization are less than 2\% for
120: $R<16.5^m$, while the EVPA is determined with a precision of
121: several degrees.
122:
123: The results of photometric observations in the optical and IR bands are given in
124: Tables \ref{TAB_1} and \ref{TAB_2}, respectively. The light curves are shown in Figure
125: \ref{FDATA} for the convenience of visual inspection.
126: One can see that during the outburst the
127: brightness increases by more than $3^m$. Table \ref{TAB_3}
128: lists the results of the polarimetric observations.
129: In Figure \ref{PDATA} the upper panels illustrate the photometric $R$ and
130: polarimetric observations for three consecutive Julian dates
131: during the most dramatic period of the outburst. Intraday variations (IDV) are
132: evident within a given Julian date. The bottom panels
133: present absolute Stokes parameters during the same period.
134: The existence of systematic variations during each night
135: is a prominent feature of all the curves.
136:
137: \section{Global Properties of Variability}
138: \subsection{Timescale of Variability}
139: We have determined the timescale of variability of the flux in each band
140: using the definition proposed by
141: \citet{BJO74}: $\tau=dt/{\rm ln(F_1/F_2)}$, where $dt$ is the time interval
142: between flux measurements $F_1$ and $F_2$, with $F_1>F_2$. We have calculated
143: all possible timescales $\tau_{\rm ij}$ for any pair of
144: observations for which $|F_i-F_j|>\sigma_{F_i}+\sigma_{F_j}$ at frequency $\nu$.
145: The minimum timescale at frequency $\nu$ is determined as
146: $\tau_\nu={\rm min}\{\tau_{\rm ij,\nu}\}$, where $i$=1,...$N-1$, $j$=i+1,...$N$,
147: and $N$ is the number of observations. Table \ref{TAB_var}
148: lists the results of these calculations as well as the timescale of the polarized flux
149: variability. The uncertainties of the values of $\tau_\nu$ are determined
150: by errors of the flux measurements. However, the minimum timescale of variability
151: that we derive is mainly constrained by the sampling,
152: for example, the $B$, $V$, and $I$ light curves have
153: only two measurements during JD~2454077 (Table \ref{TAB_1}), the date at which
154: we find the minimum timescales in $R$ band and at IR wavelengths. In Table \ref{TAB_var}
155: we enclose the timescales derived from these light curves in brackets
156: to indicate that they can not be compared with the timescales derived from
157: the better sampled $R$, $J$, $H$, and $K$ light curves.
158:
159: The values of $\tau_\nu$ indicate that during the event the timescale
160: of variability was only a few hours and the shortest timescale, $\sim$0.3~hr,
161: occured in the polarized flux. Comparison of the timescales of the total
162: and polarized flux variability in $R$ band implies that the polarized emission
163: originates in a subsection ($\sim$1/5) of the optical emission region.
164: Moreover, the light curves with similar sampling during the event
165: ($R,J,H$, and $K$ data) reveal an increase of timescale with wavelength.
166:
167: \subsection{Multicolor Behavior}
168: We have calculated the discrete cross-correlation function, $DCF$,
169: between the $R$ band light curve and those at other wavelengths using the
170: approach suggested by \citet{EK98} for unequally spaced data.
171: The functions were computed for lags between light curves from
172: $-6$ to $+6$~hr with a step of $0.5$~hr and with the constraint
173: that at least 5 pairs of measurements are needed to define the coefficient
174: of correlation at a given delay. Figure \ref{cor} presents the results
175: of the calculations. The absence of points at some lags indicates that
176: the data are not capable of determining the $DCF$ for such delays
177: owing to the constraint.
178:
179: Figure \ref{cor} shows that the $DCF$s have maxima at different
180: delays between the light curves. The maxima are distinct only
181: for the optical light curves, and indicate that variations in $B$ band
182: lead variations in $R$ band by $\sim$0.5-1~hr, while variations
183: in $V$ and $I$ bands are simultaneous within 0.5~hr with the variability
184: in $R$ band. Although the peaks of the $DCF$s for the near-IR light
185: curves are broad and their positions are uncertain within 4-5~hrs,
186: the delay of variations in the near-IR region compared to those
187: in $R$ band is statistically significant. The peaks of the $DCF$ and their delays
188: for each light curve with respect to the $R$ band light curve are
189: listed in Table \ref{TAB_var}. Such delays between the light curves
190: are expected if the sizes of the emission regions at different wavelengths
191: scale with the timescales of the variability found above.
192:
193: \subsection{Polarization Properties}
194: Table 3 shows that very high degrees of polarization (up to 30\%) were observed
195: during the December 2006 outburst. Figure \ref{PTheta} demonstrates that the degree of
196: polarization increased when the object brightened. The latter is confirmed
197: by the high value
198: of the $DCF$ peak between the flux in $R$ band and the degree of polarization
199: (Table \ref{TAB_var}). The $DCF$ indicates also that the $R$ band and fractional polarization
200: variations were simultaneous within 0.5~hr (Fig. \ref{cor}).
201: Figure \ref{PTheta} shows a wide range, $\sim$90$^\circ$, of variability of
202: the polarization position angle. However, $\Theta_\circ$ aligns closely with
203: the parsec jet direction, $\sim -15^\circ$ \citep{J01,P06}, when the
204: degree of polarization exceeds 20\%.
205:
206: Visual inspection of Figure \ref{PDATA} reveals a correlation between Stokes
207: parameters - between $Q$ and $I$ and between $U$ and $I$ -- and different behavior
208: of the Stokes parameters from one epoch to another.
209: The correlation coefficients calculated separately
210: for each Julian date, given in Table \ref{TAB_Pvar},
211: confirm the tight correlation between the Stokes parameters.
212:
213: \section{Properties of the Source Responsible for Variability}
214: \subsection{Method of Analysis}
215: Our current understanding of active galactic nuclei pictures an AGN as a complex system in
216: which different components - the host galaxy, accretion disk, corona above the disk, broad line
217: regions, and jet - contribute to the total flux. In some cases multicolor photometric and polarimetric
218: observations of variability may make it possible to derive the characteristics of
219: the variable source (spectral energy distribution, SED, and polarization
220: parameters) without knowledge of the absolute contribution of the other
221: components to the total radiation.
222:
223: Let us suppose that the variability within some time interval is due to a single variable component.
224: If the variability is caused only by its flux variation but the relative Stokes parameters $q$ and $u$
225: (for polarimetry) or relative SED (for photometry) remain unchanged,
226: then in the space of absolute Stokes parameters $\{I, Q, U\}$ or in the $n$-dimensional flux space $\{F_1, ..., F_n\}$ ($n$ is the number of
227: spectral bands used in multicolor observations) the observational points must lie on straight lines.
228: The slopes of these lines are relative Stokes parameters of the variable component (for polarimetry)
229: or flux ratios for different pairs of bands as determined by its SED (for photometry).
230:
231: With small caveats, the opposite is also true. In the case of photometry,
232: a linear relation between observed fluxes in two different wavelengths during some
233: period of flux variability implies that the flux ratio does not change.
234: Such a relation for several bands indicates that the relative SED of
235: the variable component
236: remains steady and can be derived from the slopes of the lines. In the case of polarimetry,
237: a linear relation between observed Stokes parameters
238: suggests that the variable component has constant
239: degree of polarization and polarization position angle, the values of which
240: can be obtained from the relative Stokes parameters $q$ and $u$ that
241: correspond to the slope of the $Q$ vs. $I$ and $U$ vs. $I$ lines, respectively.
242: Further details of the method of analysis are provided in \citet{HM99}.
243:
244: \subsection{Polarization Parameters of the Variable Component}
245:
246: Correlation plots of the absolute Stokes parameters for three consecutive epochs
247: during the outburst (JD 2454077, 2454078, and 2454079) are given in Figure \ref{STOKES} (a,b, and c,
248: respectively). One can see that in all cases the points lie on straight lines that
249: leads us to conclude that during a given epoch the relative Stokes parameters
250: of the variable component remained unchanged. The slopes of the straight lines in Figure
251: \ref{STOKES} are determined by the orthogonal regression method. They give the relative
252: Stokes parameters $q_{var}$ and $u_{var}$ of the variable component that we use to calculate
253: its degree and position angle of polarization, $p_{var}$ and
254: $\Theta_{var}$ (Table \ref{TAB_Pvar}). Table \ref{TAB_Pvar} indicates
255: that the degree of polarization of the variable component is very high and decreases
256: from 50\% to 30\% as the object fades. This is accompaned by a systematic variation of
257: the polarization direction. At the highest degree of polarization the EVPA tends
258: to align to within 15$^\circ$ with the direction of the parsec-scale jet
259: inferred from the VLBI maps of AO 0235+164 \citep{J01,P06}.
260:
261: \subsection{Spectral Energy Distribution of the Variable Component}
262: We compare the fluxes in different bands relative to the $R$ and $J$ fluxes for
263: the optical and near-IR data, respectively (see Fig. \ref{Fcompar}), since these
264: two bands have
265: the largest number of observations. According to Figure \ref{PDATA}, the data reveal
266: intraday variability that can
267: affect the flux-flux comparison because the measurements are not completely simultaneous.
268: For example, during Julian dates with intensive monitoring, delays of $\sim10-15$
269: minutes between observations in different bands occur.
270: For such epochs the fluxes in $R$ band were interpolated to
271: the corresponding times of the $B$, $V$, and $I$ bands measurements
272: (the same for $J$ band with respect to $H$ and $K$ bands). For epochs with a few measurements,
273: the average at each band was used.
274:
275: The slopes of the lines were derived using the orthogonal regression method. They
276: are corrected for extinction based on the information provided by \citet{R05}.
277: The logarithms of the corrected values are listed in Table \ref{TAB_var}.
278: Using these values, we have constructed SEDs
279: in the optical and near-IR regions as well as the combined SED
280: (via the flux-flux diagram for $R$ and $J$) for the whole range from $B$ to $K$ bands (Fig. \ref{Spectr}).
281: Figure \ref{Spectr} shows that the spectra can be represented by a power law,
282: $F_{\nu} \sim \nu^{-\alpha}$ , where $\alpha$ is the spectral index. In the near-IR region
283: $\alpha = 1.10 \pm 0.03$, while in the optical region $\alpha = 1.39 \pm 0.12$. Although there appears to be a slight steepening of the spectrum in the optical region, this
284: falls within the 2$\sigma$
285: uncertainty, so that the combined SED can be fit by a single power law with
286: $\alpha = 1.04 \pm 0.06$. This indicates that the same component with the
287: relative SED shown in Figure \ref{Spectr} is responsible for
288: the source variability in the whole range from $B$ to $K$ bands.
289:
290: \section{Discussion}
291: The results of the polarimetric and photometric observations shows that
292: the dramatic variability
293: seen in AO~0235+164 in 2006 December might be explained by a single variable component that
294: possesses the following properties:
295: (i) steady spectral shape;
296: (ii) increase of size of the emitting region with wavelength;
297: (iii) very high degree of polarization, up to 50\%;
298: (iv) strong correlation between degree of polarization and
299: flux; and (v) tendency of the magnetic field at the highest
300: polarization state to be perpendicular to the inner radio jet direction.
301: These properties are characteristics of the model that
302: describes an outburst in a blazar as a transverse shock
303: propagating down a relativistic jet with turbulent magnetic field
304: \citep{MG85,HAA85,MAR96}.
305: The timescale of variability relates to the thickness of
306: the shock front, which is determined by the lifetime
307: of relativistic electrons accelerated at the front. This thickness decreases
308: with frequency owing to synchrotron radiation losses. The shock
309: orders the turbulent magnetic field along the front, which is
310: perpendicular to the jet direction.
311:
312: We consider a model of a square-wave transverse shock moving through a
313: turbulent plasma with a constant Lorentz factor, $\Gamma$.
314: The size and shape of the emission region at a given frequency are steady.
315: In the observer's frame the flux of the shocked region is boosted as
316: follows:
317: \begin{equation}
318: F=F_\circ\nu^{-\alpha}\delta'^{(2+\alpha)}\delta^{(3+\alpha)}, \label{eq1}
319: \end{equation}
320: where $\delta'$ is the Doppler factor of the shocked
321: plasma in the rest frame of the shock front and $\delta$ is the Doppler
322: factor of the shock front in the observer's frame.
323: Determination of $\delta'$ requires knowledge of the speed and equation of
324: state of the plasma behind the shock front \citep{CAW91}, which is difficult
325: to specify due to a lack of the necessary information. Instead, we assume
326: that the velocity of the shocked plasma in the frame of the shock front $\ll c$ and
327: $\delta'\approx$1. We can then approximate that
328: $\delta=[\Gamma(1-\beta~cos\Phi)]^{-1}$, where
329: $\beta=\sqrt{1-\Gamma^{-2}}$ is the speed of the shock and $\Phi$ is the
330: angle between the line of sight and the velocity of the centroid
331: of the emission region.
332:
333: The degree of polarization depends
334: on the viewing angle of the shock, $\Psi$, and ratio of the density in the
335: shocked region with respect to unshocked region, $\eta=n_{\rm shock}/n_{\rm unshock}$ \citep{HM91}:
336: \begin{equation}
337: p\approx \frac{\alpha+1}{\alpha+5/3}~\frac{(1-\eta^{-2})\sin^2\Psi}{2-(1-\eta^{-2})\sin^2\Psi}. \label{eq2}
338: \end{equation}
339: The viewing angle of the shock in the
340: observer's frame is subjected to relativistic
341: aberration and determined by the bulk Lorentz factor
342: and viewing angle of the jet:
343: \begin{equation}
344: \Psi=\tan^{-1}[\sin\Phi/(\Gamma(\cos\Phi-\sqrt{1-\Gamma^{-2}}))] \label{eq3}
345: \end{equation}
346:
347: We use the derived spectral index of the variable component $\alpha=1.04$
348: and assume $\Gamma=30$, consistent with the highest reliable apparent speed seen
349: in the jet of AO~0235+16 \citep{J01,P06}. We apply equation (\ref{eq1})
350: to the photometric data in the $R$ band to estimate the Doppler factor
351: as a function of time. The scaling factor $F_\circ$ is determined under the assumption
352: that $F_\circ=F_{\rm max}\nu^{\alpha}/\delta_\circ^{(3+\alpha)}$, where
353: $F_{\rm max}$ is the maximum observed flux in $R$ band and $\delta_\circ$ is
354: obtained for $\Phi_\circ$, which is calculated
355: from equations (\ref{eq2}) and (\ref{eq3}) for the maximum derived
356: degree of polarization of the variable component (Table \ref{TAB_Pvar}).
357: In this case we adopt $\eta$=2.5 that is the minimum possible compression
358: that produces polarization as high as 50\%. This yields $\delta_\circ=23.3$
359: at the maximum polarization.
360: Figure \ref{model} shows the values of the Doppler factor and viewing angle
361: of the jet derived from the observed flux variability.
362: The latter allows us to estimate the viewing angle of the shock (eq. \ref{eq3})
363: and value of the plasma compression as functions of time using the light curve
364: of fractional polarization (eq. \ref{eq2}). The values of $\Psi$ and $\eta$ are
365: shown in Figure \ref{model} as well.
366:
367: Figure \ref{model} demonstrates that the observed flux and polarization
368: behavior can be explained by a small change in the viewing angle $\lesssim 1^\circ$
369: and by variations of shocked plasma compression while the bulk Lorentz
370: factor remains constant. A small viewing angle ($\sim 2.5^\circ$) agrees
371: with the highly core-dominated
372: radio images of the blazar. A change in the viewing angle can be caused either
373: by curvature of the jet by $\lesssim 1^\circ$ or by a curved trajectory of the
374: shock-causing perturbation
375: inside the jet for an opening angle $\gtrsim 1^\circ$. In the brightest
376: flux state the viewing angle is the nearest to the line of sight, the shock
377: is seen almost edge-on ($\Psi\sim 103^\circ$), the plasma compression
378: is the highest ($\eta\sim 1.7$), and the parameters change on timescale
379: of hours. In the end of the outburst $\eta\sim 1.1$ and the parameters
380: are steady within a week. Changes in the rate of injection of
381: relativistic electrons could be related to variations in $\eta$.
382:
383: In this scenario a shock is responsible for
384: the injection of the relativistic electrons into the emitting region,
385: leading to a steady spectral shape and a short timescale of
386: the variability in the optical and IR bands.
387: Therefore, we assume that the total
388: flux and linear polarization variability is caused by changes in the
389: viewing angle and in injection of
390: relativistic electrons on timescales of hours. Radiative losses
391: determine the size of the emitting region behind the shock front,
392: which is of the order of 2-3 light hours (Table \ref{TAB_var}).
393: A steady spectral shape implies a quasi-steady state of the emission,
394: which is re-established every 2-3~hrs, between the rate of injection
395: and the radiative losses.
396: In the observer's frame the synchrotron lifetime of electrons emitting
397: at frequency $\nu$ is:
398: \begin{equation}
399: t_{\rm loss}\approx 4.75\times 10^2{\left[\frac{1+z}{\delta\nu_{GHz}B_G^3}\right]}^{1/2}~~{\rm days}, \label{eq4}
400: \end{equation}
401: where $B_G$ is the magnetic field in Gauss.
402: Since $t_{\rm loss}\approx\tau_\nu$, for $\delta=23.3$
403: (the maximum derived Doppler factor) and $\tau_\nu=1.7$~hr in the $R$ band
404: (Table \ref{TAB_var}), eq.(\ref{eq4}) yields an estimate of
405: the magnetic field, $B\approx 0.5~G$,
406: that agrees with other estimates of the
407: typical magnetic field in a blazar core
408: \citep[e.g., ][]{MG85}.
409: Equation (\ref{eq4}) implies also that the timescale of variability
410: in $K$ band should be longer than that in $R$ band
411: by a factor of 1.8, consistent with the values of $\tau_\nu$
412: derived from the observations.
413: The highest degree of polarization, $p_{\rm var}=50$\%,
414: is close to the maximum possible polarization (75\%)
415: for a synchrotron source
416: with $\alpha=1.04$ and a uniform magnetic field \citep{PAH70}.
417: This indicates an extremely well ordered magnetic field
418: behind the shock front along the plane transverse to the jet direction.
419: Therefore, we consider that the polarimetric and multicolor photometric
420: behavior of AO~0235+164 during the outbust in 2006
421: December is consistent with the properties of the
422: shock-in-jet model.
423:
424: \section{Conclusion}
425: In 2006 December the BL~Lac object AO~0235+164 underwent a spectacular
426: outburst with an increase in optical flux by a factor of 4 and
427: with degree of polarization reaching 30\%. Analysis of the
428: multicolor photometric data and polarization measurements in $R$
429: band indicate that:
430:
431: (i) The flux increase has an achromatic character.
432:
433: (ii) The spectral energy distribution of the variable component
434: is a power law with $\alpha$=1.04$\pm$0.06.
435:
436: (iii) Variations on timescales of hours are present at all wavelengths.
437: The minimum timescale of variability increases with wavelength.
438:
439: (iv) There is a delay between the variations at different wavelengths,
440: with fluctuations occurring first in $B$ band and last at the near-IR
441: wavelengths.
442:
443: (v) The fractional polarization correlates with the flux.
444:
445: (vi) The direction of polarization at the highest polarization states tends
446: to align with the inner radio jet direction.
447:
448: These properties of the variable component
449: responsible for the outburst match the characteristics of a transverse
450: shock propagating down the jet. We find that the observed photometric
451: and polarimetric variability can be reproduced in detail
452: by the shock propagating down a twisted jet such that the viewing angle
453: changes by $\lesssim 1^\circ$ and the plasma compression decreases
454: by a factor of $\sim$1.5 while the bulk Lorentz factor, $\Gamma=30$, remains constant.
455:
456: \acknowledgments
457: This work is supported by the Russian Fund of Basic Research, grant 05-02-17562.
458: S. Jorstad and A. Marscher acknowledge support from the
459: National Science Foundation under grant no. AST-0406865.
460:
461: \begin{thebibliography}{99}
462: \bibitem[Burbidge, Jones, \& O'Dell(1974)]{BJO74} Burbidge, G. R., Jones, T. W., \& O'Dell, S. L. 1974, \apj, 193, 43
463: \bibitem[Carini (2006)]{CARINI06} Carini, M. 2006, in Blazar Variability Workshop II: Entering the GLAST Era, ed. H.R. Miller, K. Marshall, J.R. Webb, \& M.F. Aller, ASP Conference Series, 350, 55
464: \bibitem[Cawthorne (1991)]{CAW91} Cawthorne, T. V. 1991, in Beams and Jets in Astrophysics, ed. P. A. Hughes,
465: Cambridge Astrophysics Series, (Cambridge Univ. Press), 187
466: \bibitem[D'Arcangelo et al. (2007)]{FRANI07} D'Arcangelo, F. D. et al. 2007, \apj, 659, L107
467: \bibitem[Edelson \& Krolik (1988)]{EK98} Edelson, R. A., \& Krolik, J. H. 1988, \apj, 333, 646
468: \bibitem[Gabuzda et al. (2006)]{GRSS06} Gabuzda, D. C., Rastorgueva, E. A., Smith, P. S., \& O'Sullivan, S. P. 2006, \mnras, 369, 1596
469: \bibitem[Hagen-Thorn (1972)]{HT72} Hagen-Thorn, V. A. 1972, Astron. Circ., 714, 5 (in Russian)
470: \bibitem[Hagen-Thorn \& Marchenko (1999)]{HM99} Hagen-Thorn, V. A. \& Marchenko, S. G. 1999, Baltic Astron., 8, 575
471: \bibitem[Hagen-Thorn et al. (1994)]{HT94} Hagen-Thorn, V. A., Marchenko, S. G., Yurchenko, A. V., Yakovleva, V. A. 1994, Astron. Reports, 38, 635
472: \bibitem[Hagen-Thorn et al. (2006)]{HT06} Hagen-Thorn, V. A., et al. 2006, Astron. Reports, 50, 458
473: \bibitem[Hagen-Thorn et al. (2007)]{HT07} Hagen-Thorn, V. A., et al. 2007, Astron. Reports, in press
474: \bibitem[Hughes, Aller, \& Aller (1985)]{HAA85} Hughes, P. A., Aller, H. D., \& Aller, M. F. 1985, \apj, 298, 301
475: \bibitem[Hughes, Aller, \& Aller (1989)]{HAA89} Hughes, P. A., Aller, H. D., \& Aller, M. F. 1989, \apj, 341, 68
476: \bibitem[Hughes \& Miller (1991)]{HM91} Hughes, P. A., \& Miller, L. 1991, in Beams and Jets in Astrophysics, ed. P. A. Hughes,
477: Cambridge Astrophysics Series, (Cambridge Univ. Press), 1
478: \bibitem[Jorstad et al.(2001)]{J01} Jorstad, S. G., et al. 2001, \apjs, 134, 181
479: \bibitem[Jorstad et al.(2007)]{J07} Jorstad, S. G., et al. 2007, \aj, 134, 799
480: \bibitem[Junkkarinen et al. (2004)]{JUNK04} Junkkarinen, V. T., et al. 2004, \apj, 614, 658
481: \bibitem[Lister \& Smith (2000)]{LS00} Lister, M. L., \& Smith, P. S. 2000, \apj, 541, 66
482: \bibitem[Marscher \& Gear (1985)]{MG85} Marscher, A. P., \& Gear, W. K. 1985, \apj, 298, 114
483: \bibitem[Marscher (1996)]{MAR96} Marscher, A. P. 1996,
484: in Blazar Continuum Variability, ed. H.R. Miller, J.R. Webb, \& J.C. Noble, ASP Conference Series, 110, 248
485: \bibitem[Mead et al. (1990)]{MEAD90} Mead, A. R. J., et al. 1990, \aaps, 83, 183
486: \bibitem[Miller \& Noble (1996)]{MN96} Miller, H. R., \& Noble, J. C. 1996, in Blazar Continuum Variability, ed. H.R. Miller, J.R. Webb, \& J.C. Noble, ASP Conference Series, 110, 17
487: \bibitem[Pacholczyk (1970)]{PAH70} Pacholczyk, A. G., 1970, Radio Astrophysics (W.H. Freeman and Company, San Francisco)
488: \bibitem[Piner et al. (2006)]{P06} Piner, B. G., Bhattarai, D., Edwards, P. G., \& Jones, D. L. 2006, \apj, 640, 196
489: \bibitem[Raiteri et al. (2001)]{R01} Raiteri, C. M., et al. 2001, \aap, 377, 396
490: \bibitem[Raiteri et al. (2005)]{R05} Raiteri, C. M., et al. 2005, \aap, 438, 39
491: \bibitem[Raiteri et al. (2006)]{R06} Raiteri, C. M., et al. 2006, \aap, 459, 731
492: \bibitem[Smith (1996)]{SMITH96} Smith, P. S. 1996,
493: in Blazar Continuum Variability, ed. H.R. Miller, J.R. Webb, \& J.C. Noble,
494: ASP Conference Series, 110, 135
495: \end{thebibliography}
496: %\input table1.tex
497:
498: \clearpage
499:
500: \begin{deluxetable}{lrlrlrlr}
501: \singlespace
502: \tablecolumns{8}
503: \tablecaption{Optical Data \label{TAB_1}}
504: \tabletypesize{\footnotesize}
505: \tablehead{
506: \colhead{MJD}&\colhead{$B$}&\colhead{MJD}&\colhead{$V$}&
507: \colhead{MJD}&\colhead{$R$}&\colhead{MJD}&\colhead{$I$}\\
508: \colhead{(1)}&\colhead{(2)}&\colhead{(3)}&\colhead{(4)}&\colhead{(5)}
509: &\colhead{(6)}&\colhead{(7)}&\colhead{(8)}
510: }
511: \startdata
512: 27.3865&20.26$\pm$0.17&27.3946&19.06$\pm$0.05&27.3784&18.43$\pm$0.03&27.3823&17.65$\pm$0.03 \\
513: 28.3267&19.71$\pm$0.17&28.3345&19.01$\pm$0.07&28.3489&18.44$\pm$0.04&28.3473&17.63$\pm$0.03 \\
514: 29.3514&19.95$\pm$0.17&29.3592&19.08$\pm$0.05&29.3749&18.45$\pm$0.04&29.3735&17.68$\pm$0.03 \\
515: \enddata
516: \tablecomments{MJD is JD-2454000; the table appears in full form electronically.}
517: \end{deluxetable}
518: %\input table2.tex
519: \begin{deluxetable}{lrlrlr}
520: \singlespace
521: \tablecolumns{6}
522: \tablecaption{Near-IR Data \label{TAB_2}}
523: \tabletypesize{\footnotesize}
524: \tablehead{
525: \colhead{MJD}&\colhead{$J$}&\colhead{MJD}&\colhead{$H$}&
526: \colhead{MJD}&\colhead{$K$}\\
527: \colhead{(1)}&\colhead{(2)}&\colhead{(3)}&\colhead{(4)}&\colhead{(5)}
528: &\colhead{(6)}
529: }
530: \startdata
531: 979.3865&17.24$\pm$0.11&979.5627&16.45$\pm$0.15&979.5666&15.31$\pm$0.19 \\
532: 980.3267&17.25$\pm$0.07&980.6206&16.32$\pm$0.09&980.6258&15.47$\pm$0.16 \\
533: 985.3514&17.27$\pm$0.06&985.5778&16.19$\pm$0.05&985.5828&15.70$\pm$0.16 \\
534: \enddata
535: \tablecomments{MJD is JD-2453000; the table appears in full form electronically.}
536: \end{deluxetable}
537: %\input table3.tex
538: \begin{deluxetable}{lrr}
539: \singlespace
540: \tablecolumns{3}
541: \tablecaption{Polarization Data in $R$ band\label{TAB_3}}
542: \tabletypesize{\footnotesize}
543: \tablehead{
544: \colhead{MJD}&\colhead{$p$,\%}&\colhead{$\Theta^\circ$} \\
545: \colhead{(1)}&\colhead{(2)}&\colhead{(3)}
546: }
547: \startdata
548: 77.3307&32.7$\pm$1.0&149.5$\pm$0.8 \\
549: 77.4282&27.7$\pm$1.4&152.2$\pm$1.3 \\
550: 77.4433&25.9$\pm$1.6&153.8$\pm$1.8 \\
551: \enddata
552: \tablecomments{Columns are as follows: (1) - JD-2454000;
553: (2) - degree of the polarization; (3) - position angle of the
554: polarization; the table appears in full form electronically.}
555: \end{deluxetable}
556:
557: \begin{deluxetable}{lrrrrrr}
558: %\singlespace
559: \tablecolumns{7}
560: \tablecaption{Multicolor Properties of the Variable Component \label{TAB_var}}
561: \tabletypesize{\footnotesize}
562: \tablehead{
563: \colhead{Band}&\colhead{$lg(\nu)$}&\colhead{$N$}&\colhead{$\tau_{\rm min}$,hr}
564: &\colhead{$DCF_{\rm peak}$}&\colhead{$\Delta t$,hr}&\colhead{$lg(F_\nu/F_R)$} \\
565: \colhead{(1)}&\colhead{(2)}&\colhead{(3)}&\colhead{(4)}&\colhead{(5)}&\colhead{(6)}&
566: \colhead{(7)}
567: }
568: \startdata
569: $B$&14.833&24&(5.27$^{+1.36}_{-0.90}$)&0.89$\pm$0.10&$+1.0^{+0.5}_{-1}$&$-$0.211$\pm$0.012 \\
570: $V$&14.736&24&(6.15$^{+0.40}_{-0.34}$)&0.95$\pm$0.11&$0.0^{+1}_{-0.5}$&$-$0.120$\pm$0.006 \\
571: $R$&14.670&66&1.73$^{+0.22}_{-0.29}$&1.0&0.0&0.0 \\
572: $I$&14.574&24&(4.81$^{+0.29}_{-0.25}$)&0.85$\pm$0.09&$0.0^{+1}_{-0.5}$&0.142$\pm$0.007 \\
573: $J$&14.387&55&2.91$^{+0.46}_{-0.34}$&0.98$\pm$0.02&$-3.5^{+1.5}_{-2.5}$&0.256$\pm$0.007 \\
574: $H$&14.262&55&2.19$^{+0.55}_{-0.37}$&0.98$\pm$0.03&$-3.5^{+1.5}_{-2.5}$&0.400$\pm$0.006 \\
575: $K$&14.140&54&3.01$^{+1.21}_{-0.67}$&0.97$\pm$0.03&$-3.5^{+1.5}_{-2.5}$&0.526$\pm$0.005 \\
576: $P$&14.670&54&0.31$^{+0.04}_{-0.04}$&0.74$\pm$0.08&$0.0^{+1}_{-1.5}$&\nodata \\
577: \enddata
578: \tablecomments{Columns are as follows: (1) - filter of photometry,
579: $P$ corresponds to the polarization data;
580: (2) - logarithm of effective frequency of the filter;
581: (3) - number of observations;
582: (4) - minimum timescale of variability, unreliable values are enclosed
583: in brackets; (5) - peak of discrete cross-correlation
584: function of the light curve at
585: frequency $\nu$ with respect to $R$ band light curve; in the case
586: of polarization data the peak is given for the $DCF$ between the $R$-band and
587: fractional polarization light curves; (6) - delay of the peak;
588: $+$ means that the light curve at frequency $\nu$
589: leads the $R$ band light curve, while $-$ is the opposite case; (7) - logarithm of
590: the contribution of the variable component in the SED at frequency $\nu$
591: relative to the contribution in $R$ band.}
592: \end{deluxetable}
593:
594: \begin{deluxetable}{lrrrr}
595: \singlespace
596: \tablecolumns{5}
597: \tablecaption{Polarization Properties of the Variable Component \label{TAB_Pvar}}
598: \tabletypesize{\footnotesize}
599: \tablehead{
600: \colhead{MJD}&\colhead{$r_{QI}$}&\colhead{$r_{UI}$}&\colhead{$p_{\rm var}$,\%}&\colhead{$\Theta_{\rm var},^\circ$} \\
601: \colhead{(1)}&\colhead{(2)}&\colhead{(3)}&\colhead{(4)}&\colhead{(5)}
602: }
603: \startdata
604: 77.447&0.93&$-$0.90&50.0$\pm$9.4&$-$30.4$\pm$5.4 \\
605: 78.373&$-$0.48&$-$0.98&36.7$\pm$3.4&$-$49.8$\pm$2.7 \\
606: 79.320&$-$0.58&$-$0.92&33.1$\pm$5.2&$-$53.9$\pm$4.5 \\
607: \enddata
608: \tablecomments{Columns are as follows: (1) - Julian date minus 2454000;
609: (2) - correlation coefficient between Stokes parameters $Q$ and $I$;
610: (3) - correlation coefficient between Stokes parameters $U$ and $I$;
611: (4) - degree of polarization; (5) - position angle of polarization.}
612: \end{deluxetable}
613:
614: \clearpage
615:
616: \begin{figure}
617: \epsscale{.8}
618: \plotone{f1.eps}
619: \caption{Light curves of AO~0235+16 in the optical and IR bands.}
620: \label{FDATA}
621: \end{figure}
622: \clearpage
623: \begin{figure}
624: \epsscale{0.8}
625: \plotone{f2.eps}
626: \caption{Simultaneous photometric and polarimetric data for AO 0235+16 at three
627: consecutive Julian dates during the outburst.}
628: \label{PDATA}
629: \end{figure}
630:
631: \begin{figure}
632: \epsscale{1.0}
633: \plotone{f3.eps}
634: \caption{Discrete cross-correlation function between light curves
635: in different bands with respect to the $R$-band light curve; a negative delay
636: means that $R$ band variations lead. {\it Bottom panel}: DCF between degree
637: of polarization and total flux density in $R$ band.
638: }
639: \label{cor}
640: \end{figure}
641:
642: \begin{figure}
643: \epsscale{1.0}
644: \plotone{f4.eps}
645: \caption{{\it Left panel}: Dependence between the degree of polarization and flux
646: in $R$ band. {\it Right panel}: Dependence between the degree of
647: polarization and polarization position angle; the dotted line indicates
648: direction of the parsec scale jet.}
649: \label{PTheta}
650: \end{figure}
651:
652:
653: \begin{figure}
654: \epsscale{0.5}
655: \plotone{f5.eps}
656: \caption{Comparison of absolute Stokes parameters for three
657: consecutive Julian dates during the outburst (a - 2454077, b - 2454078, c - 2454079); crosses correspond to the uncertainties of the parameters.}
658: \label{STOKES}
659: \end{figure}
660:
661: \begin{figure}
662: \epsscale{1.0}
663: \plotone{f6.eps}
664: \caption{Flux-flux diagrams for optical (a) and (b) IR wavelengths.}
665: \label{Fcompar}
666: \end{figure}
667:
668:
669: \begin{figure}
670: \epsscale{1.0}
671: \plotone{f7.eps}
672: \caption{The relative SEDs at IR and optical wavelengths (dashed lines) and
673: for the combined spectrum (solid line). The optical SED is relative to the $R$ band
674: flux. The IR SED is relative to the $J$ band flux. The combined SED is
675: constructed using a linear dependence between simultaneous observations in $J$ and $R$ bands.}
676: \label{Spectr}
677: \end{figure}
678:
679: \begin{figure}
680: \epsscale{1.0}
681: \plotone{f8.eps}
682: \caption{The derived values of Doppler factor, $\delta$, angle between the jet
683: axis and line of sight, $\Phi$, viewing angle of the shock, $\Psi$, and compession
684: factor of the shocked plasma, $\eta$.}
685: \label{model}
686: \end{figure}
687:
688: \end{document}
689:
690: