Multi-Wavelength Photometric and Polarimetric Observations of the Outburst of 3C 454.3 in Dec. 2009
Mahito Sasada, Makoto Uemura, Yasushi Fukazawa, Koji S. Kawabata, Ryosuke Itoh, Itsuki Sakon, Kenta Fujisawa, Akiko Kadota, Takashi Ohsugi, Michitoshi Yoshida, Hajimu Yasuda, Masayuki Yamanaka, Shuji Sato, Masaru Kino
aa r X i v : . [ a s t r o - ph . H E ] D ec PASJ:
Publ. Astron. Soc. Japan , 1– ?? , c (cid:13) Multi-wavelength Photometric and Polarimetric Observations of theOutburst of 3C 454.3 in Dec. 2009
Mahito
Sasada , Makoto Uemura , Yasushi Fukazawa , Koji S. Kawabata , Ryosuke Itoh , Itsuki Sakon ,Kenta Fujisawa , , Akiko Kadota , Takashi Ohsugi , Michitoshi Yoshida , Hajimu Yasuda , Masayuki Yamanaka , Shuji Sato , and Masaru Kino Department of Physical Science, Hiroshima University, Kagamiyama 1-3-1, Higashi-Hiroshima [email protected] Astrophysical Science Center, Hiroshima University, Kagamiyama 1-3-1, Higashi-Hiroshima 739-8526 Department of Astronomy, Graduate School of Science, the University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033 Graduate school of Science and Engineering, Yamaguchi University, 1677-1 Yoshida, Yamaguchi, Yamaguchi 753-8512 The Research Institute for Time Studies, Yamaguchi University, 1677-1 Yoshida, Yamaguchi, Yamaguchi 753-8511 Department of Physics, Nagoya University, Furo-cho, Chikusa-ku, Nagoya 464-8602 (Received 2011 April 18; accepted 2011 December 15)
Abstract
In December 2009, the bright blazar, 3C 454.3 exhibited a strong outburst in the optical, X-ray andgamma-ray regions. We performed photometric and polarimetric monitoring of this outburst in the opticaland near-infrared bands with TRISPEC and HOWPol attached to the Kanata telescope. We also observedthis outburst in the infrared band with
AKARI , and the radio band with the 32-m radio telescope ofYamaguchi University. The object was in an active state from JD 2455055 to 2455159. It was 1.3 magbrighter than its quiescent state before JD 2455055 in the optical band. After the end of the active statein JD 2455159, a prominent outburst was observed in all wavelengths. The outburst continued for twomonths. Our optical and near-infrared polarimetric observations revealed that the position angle of thepolarization (PA) apparently rotated clockwise by 240 degrees within 11 d in the active state (JD 2455063—2455074), and after this rotation, PA remained almost constant during our monitoring. In the outburststate, PA smoothly rotated counterclockwise by 350 degrees within 35 d (JD 2455157—2455192). Thus,we detected two distinct rotation events of polarization vector in opposite directions. We discuss these twoevents compared with the past rotation events observed in 2005, 2007 and 2008.
Key words:
BL Lacertae Objects: individual: 3C 454.3 — polarization — infrared: general
1. Introduction
Blazar is a class of active galactic nuclei, whose rela-tivistic jets are considered to be directed along the line ofsight (e.g. Blandford et al. 1979). Radiation of blazarshas three main properties. First, the blazars emit elec-tromagnetic radiation in a broad range from the radio togamma-ray bands. Their emission consists of two majorcomponents (e.g. Kubo et al. 1998). The low energy com-ponent is synchrotron radiation observed from the radioto the optical, sometimes extending to the X-ray bands.The high energy component is inverse-Compton scatter-ing from the X-ray to the gamma-ray bands. Second,blazars exhibit rapid and violent variability in all wave-length bands (Antonucci 1993). The variability has var-ious timescales from less than a day (e.g. Aharonianet al. 2007; Albert et al. 2007; Sasada et al. 2008; Raiteriet al. 2008b) to longer than years (e.g. Sillanp¨a¨a et al.1996). Third, blazars possess relativistic jets (e.g. Listeret al. 2009), which are responsible for high polarizationobserved at optical/near-infrared (NIR), and radio wave-lengths. Since the polarization can be a probe of the mag-netic field in the jet, polarimetric observations are impor-tant to study the structure of the jet. The temporal variation of the polarization vector iscomplex in general. Jones et al. (1985) reported that thepolarization behavior was erratic in blazars. On the otherhand, several papers reported that the polarization vec-tor exhibited systematic variations, for example, positivecorrelations between the flux and the degree of polariza-tion (PD) (e.g. Smith et al. 1986). Recently, Marscheret al. (2008) have reported that the polarization vectorin BL Lac smoothly rotated when the object was bright.From this result, they proposed that this rotation indi-cated an emission zone passing through a helical magneticfield in the jet. Abdo et al. (2010a) reported a rotationof polarization in 3C 279, and proposed that the rotationevent is attributed to a bent jet. However, there are onlya few rotation events which have been reported to date.3C 454.3 is one of the most famous blazars. The objectis classified as Flat Spectrum Radio Quasars (FSRQs),and its redshift is z = 0 .
859 (Jackson & Browne 1991).Although the object had been quiet in the optical banduntil 2001, the object has kept showing the active behav-ior since then (Villata et al. 2006). In 2005, the objectshowed an exceptional outburst. In this outburst, the ob-ject brightened from the radio to the gamma-ray bands(Fuhrmann et al. 2006; Pian et al. 2006; Giommi et al. M. Sasada et al. [Vol. ,2006 ; Villata et al. 2007). After this outburst, similaroutbursts were detected in 2007 and 2008. In 2005 and2007 outbursts, rotation events of the optical polariza-tion vector were reported by Jorstad et al. (2010) andSasada et al. (2010). In Dec. 2009, a prominent out-burst was reported for this object in the gamma-ray bandby
Fermi /LAT and AGILE (Striani et al. 2009a; Strianiet al. 2009b; Escande & Tanaka 2009; Striani et al. 2010;Pacciani et al. 2010; Ackermann et al. 2010), in the X-ray band by
INTEGRAL /IBIS (Vercellone et al. 2009), by
Swift /XRT (Sakamoto et al. 2009), by
Swift /BAT (Krimmet al. 2009), in the optical bands (Villata et al. 2009a;Bonning et al. 2009; Sasada et al. 2009). The objectwas the brightest source in the GeV-gamma-ray sky forover a week (Ackermann et al. 2010). The object showedflux variability over timescales less than three hours andvery mild spectral variability with an indication of gradualhardening preceding major flares. The minimum Dopplerfactor was 13, estimated by using these results. Bonnoliet al. (2011) also estimated the Doppler factors ∼
25 dur-ing the outburst by constructing a multi-wavelength spec-tral model.We performed monitor observations of 3C 454.3 fromMay 2009 to February 2010 in a multi-color photometricand polarimetric mode using the Kanata telescope. Wealso observed in the radio and infrared (IR) bands. In thispaper, we report on the behavior of the 2009 outburst inthese bands, and the detections of two rotation events inthe polarization vector. Directions of these two rotationswere different, suggesting a complex magnetic field in thejet. This paper is arranged as follows: In section 2, wepresent the observation method and analysis in the radio,infrared, optical and X-ray bands. In section 3, first wereport the light curves and the spectral indexes of theX-ray and optical bands. Then, we report the temporalbehavior of polarization in the optical and NIR bands.After that, we report the spectral energy distribution fromthe radio to optical band. In section 4, we discuss tworotation events which we detected, by comparing themwith the past rotation events observed in 2005, 2007 and2008. The conclusion is drawn in section 5.
2. Observation
We performed monitor observations of 3C 454.3using TRISPEC attached to the Cassegrain focusof the Kanata 1.5-m telescope at Higashi-HiroshimaObservatory. TRISPEC can perform photometric and po-larimetric observations in the optical and two NIR bands,simultaneously (Watanabe et al. 2005). In the observationof 3C 454.3, unfortunately, one of the two NIR arrays wasnot available due to a readout error. Therefore, we ob-served the object with the multi-color photometric andpolarimetric monitoring in the V and J bands. A unit ofthe observing sequence consisted of successive exposuresat four position angles of a half-wave plate; 0 ◦ , 45 ◦ , 22. ◦ ◦
5. A set of polarization parameters was derived from each set of the four exposures.We also observed in the multi-color photometric modein the R C and I C bands during the outburst of theobject using HOWPol (Hiroshima One-shot Wide-fieldPolarimeter; Kawabata et al. 2008) attached to a Nasmythfocus of the Kanata telescope. In this paper, we use multi-band photometric data obtained with HOWPol in § V and J bands, respectively. All im-ages were bias-subtracted and flat-fielded, before aperturephotometry. We performed differential photometry witha comparison star taken in the same frame of 3C 454.3.Its position is R.A.=22 h m . s
11, Dec.=+16 ◦ ′ . ′′ V =13.587, R C =13.035, I C =12.545 and J =11.858 (Gonz´alez − P´erez et al. 2001;Skrutskie et al. 2006). After the differential photometry,we calculated the flux, assuming that the 0 mag corre-sponds to the flux with 1.98, 1.42, 0.895 and 0.384 × − erg cm − s − in the V , R C , I C and J bands (FukugitaSimasaku & Ichikawa 1995; Bessell, Castelli & Plez 1998).We confirmed that the instrumental polarization wassmaller than 0.1 % in the V and J bands using the ob-servation of unpolarized standard stars. We, hence, ap-plied no correction for it. The instrumental depolariza-tion factors, α Vdep and α Jdep , was derived from the obser-vation using Glan-Taylor prism, to be α Vdep =0.827 and α Jdep =0.928. The observation was corrected for it. Thezero point of the position angle of polarization (PA) isdefined by the standard system (measured from northto east) by observing the polarized stars, HD 19820 andHD 25443 (Wolff, Nordsieck & Nook 1996).
We utilized the archival data of Swift/XRT for derivingthe X-ray light curve. The XRT observations were carriedout using the Photon Counting (PC) readout mode. TheXRT data were reduced using FTOOLS in the HEAsoftpackage (v6.6). We extracted the source event within aradius of 50” and background event within radii of 80—100” centered on the source. We used the XSPEC pack-age (v11.3) to fit the data. We applied an absorbed power-law model with Galactic absorption fixed at a value of N H = 1 . × cm − (wabs*powerlaw model in XSPEC)(Donnarumma et al. 2009). The NIR spectroscopic observations of 3C 454.3 werecarried out at 14:42:27 on 12 Dec., 15:25:22 on 13 Dec. and01:18:31 on 14 Dec. in 2009(UT) with the
AKARI satel-lite in the framework of the
AKARI
Open Time Observingprograms for the Phase 3-II “Blazar Variability in near-Infrared, Optical and Gamma-ray regions (BVIOG)” (PI:M. Sasada). All the observations were performed withthe spectroscopic mode (AOT IRCZ4; Onaka et al. 2010)in which the data were taken with the prism, NP (1.8–5.5 µ m; Ohyama et al. 2007), installed in the NIR chan-nel of the Infrared Camera (IRC; Onaka et al. 2007) ofthe AKARI satellite. Each of the data reduction proce-o. ] Multi-wavelength photopolarimetric monitoring of 3C 454.3 in Dec. 2009 3dure, including the subtraction of the detector dark cur-rent, correction for the high-energy ionizing particles ef-fects, the shift and co-addition of the exposure frames,and the wavelength calibration for NP data, follows thosein the IRC Spectroscopy toolkit for Phase 3 data Version20110114. In order to correct for the sensitivity changesduring the phase-3 of
AKARI mission to obtain the accu-rate flux level of the spectrum due to the seasonal tem-perature fluctuations of the detectors, we derived our ownsystem spectral response curve of NP by using the spectro-scopic datasets of a calibration standard KF06T2 collectedat the nearest epochs of our datasets. In this paper, weused only 3.0–5.0 µ m data because of a large uncertaintyof the flux calibration outside this region. The radio observation was carried out with theYamaguchi 32-m radio telescope at the center frequencyof 8.38 GHz and bandwidth of 400 MHz in the total powermode. The antenna temperatures were measured at theposition of 3C 454.3 and at four positions 2 arc-min (ahalf of FWHM) offset to positive and negative in both az-imuth and the elevation directions from the target so asto obtain the true antenna temperature by correcting thepointing error of the telescope. A flux calibrator 3C 48was observed at the same elevation of 3C 454.3. The fluxdensity of 3C 454.3 was determined from the ratio of theantenna temperatures of 3C 454.3 to 3C 48, and the fluxdensity of 3C 48 (3.34 Jy, Ott et al. 1994). The accuracyof the measured flux density is supported to be 5 % em-pirically. The observation was ensured to be 5 times fromOctober 21st to December 7th.
3. Result
Figure 1 shows the light curves in the X-ray and opti-cal bands, temporal variation of photon index Γ and the V − J color variation. The X-ray and optical light curvesshow that the flux of 3C 454.3 was variable, and we de-fine three states based on the light-curve structure. Thefirst state is a quiescent state from the start date of ourmonitoring to JD 2455055. The second state is an activestate from JD 2455055 to 2455159. And the third state isan outburst state after JD 2455159. The flux in the qui-escent state was less variable and faint both in the X-rayand optical bands compared with the flux after the activestate. The active state was characterized by several shortand small flares with duration of ∼
10 d and amplitudeof a factor < ∼
3. On JD 2455159, the object had suddenlybecome bright both in the X-ray and optical band, simul-taneously. In the decline phase of the outburst, we canestimate a decline rate, τ , assuming that the flux followsan exponential decay, that is, F ( t ) ∝ e − t/τ (B¨ottcher et al.2007). The decline rates were 14 ± ± F l u x ( ph c m - s - ) Γ ν F ν V ( e r g c m - s - ) V - J JD - 2450000
Fig. 1.
Light curves, temporal variation of photon index andcolor variation of 3C 454.3. From top to bottom, the panelsshow the light curve and temporal variation of photon indexΓ in the X-ray band at 1 keV, the light curve in the V bandand V − J color variation. radio observatory as mentioned in § ∼ JD 2455070, whileits variation amplitude is small. Raiteri et al. (2011) alsoanalyzed the same XRT data, and reported that no realchanges in Γ could be detected from 2008 to 2010. It isalso noteworthy that no prominent change in Γ was asso-ciated with the outburst state.In the quiescent state, V − J was about 1.6. The V − J color in the quiescent state was bluer than those in theactive and outburst states. This feature indicates a, so-called, redder-when-brighter trend. The same color be-havior was reported in past observations of 3C 454.3 (e.g.Raiteri et al. 2008a; Sasada et al. 2010). It is widelyaccepted that this feature appears because an underly-ing thermal emission in the UV band, called as Big BlueBump; BBB, is bluer than the variable synchrotron emis-sion (Raiteri et al. 2007). M. Sasada et al. [Vol. , ν F ν V ( x - e r g c m - s - ) a P D V ( % ) b PA V ( deg ) c -400-20002004950 5000 5050 5100 5150 5200 c o r . PA V ( deg ) JD - 2450000 d A B -400-20002005050 5060 5070 5080 c o r . PA ( deg ) JD - 2450000 i A ν F ν J ( x - e r g c m - s - ) e P D J ( % ) f PA J ( deg ) g c o r . PA J ( deg ) JD - 2450000 h A B -400-20002005120 5160 5200 c o r . PA ( deg ) JD - 2450000 j B Fig. 2.
Temporal variations of the flux and polarization parameters. The left and right panels from “a” to “h” show the eachvariation in the V and J bands. The top panels show the light curves of the object. The second, third and fourth panels showthe temporal variations of the PD and PA and the corrected PA. The bottom panels show temporal variations of the corrected PAfocused on the rotation “A” and “B”. The filled circle shows the polarization parameters in the V band, and the open and filledsquares show the polarization parameters in the J band. o. ] Multi-wavelength photopolarimetric monitoring of 3C 454.3 in Dec. 2009 5 Figure 2 shows the light curves and temporal variationsof the polarization parameters in the V and J bands. Thepanel “b”, “c”, “f” and “g” show the temporal variationof the PD and PA in the V and J bands. The PD in theactive and outburst states exhibited large variations com-pared with the PD in the quiescent state. The averagedPD V were 4.5, 6.0 and 8.9 % and the averaged PD J were7.3, 9.2 and 12.2 % in the quiescent, active and outburststates, respectively. In the outburst state, the maximaof the PD V and PD J were 22.0 ± ± J were higher than the av-eraged PD V in all states. The high PD in the J band ispartly due to a low contribution of the unpolarized fluxfrom the BBB component in the J band.We corrected the PA assuming that the temporal vari-ation in the PA is less than 90 ◦ between neighbor two ob-servations. We defined the variation as ∆ P A n = P A n +1 − P A n − q δ P A n +12 + δ P A n , where P A n +1 and P A n were the n+1- and n-th PA and δ P A n +1 and δ P A n werethe errors of n+1- and n-th PA. If ∆ P A n < − ◦ ( > +90 ◦ ),we add +180 ◦ ( − ◦ ) to P A n +1 . If | ∆ P A n | < ◦ , weperformed no correction of P A n +1 . The panel “d” and“h” show the corrected PA.In the panel “d”, two rotation events can be seen,“A” and “B”. The rotation event “A” occurred fromJD 2455063 to 2455074 when the object entered the activestate. The rotation event “B” occurred from JD 2455157to 2455192, during the outburst state. In the quiescentstate, there was no rotation event. Thus, we can considerthat these rotation events were associated with the ac-tivity of the object. After the rotation “A”, the PA wasconstant during the active state, at about 170 ◦ ± ◦ .These rotations can be confirmed also in the J -bandobservations, while the timings of the PA correction aredifferent in several points as shown in panel “h”. This ismostly due to large errors of PA ( δ P A ) in the J -band ob-servations. The PA correction mentioned above dependson δ P A . In addition, the data number of the J -bandobservation is smaller than that of the V -band one. Thisis partly due to mechanical errors of our NIR detector.Thus, the PA correction for the V -band data is more reli-able than that for the J -band one. For example, in panels“i” and “j”, we show the corrected PA around rotations“A” and “B”. The open squares denote the corrected PAof the J -band data. They apparently show different be-havior from the V -band PA (the filled circles). However,they becomes consistent if −
180 or −
360 deg is added tothe J -band PA. It demonstrates that the V - and J -banddata have the same behavior if the ±
180 deg ambiguityin PA is taken into account.Rotation rates of “A” and “B” were estimated as − ± ± − , calculated by a linear re-gression model of PA. If the rotation rate is positive, therotation direction is counterclockwise in the QU plane, orthe celestial sphere, and vice versa. Figure 3 shows thetemporal variation of the object in the Stokes QU planein the V band. The direction of “A” was clockwise and -0.4-0.20.00.20.4-0.4 -0.2 0.0 0.2 0.4 U ( x - e r g c m - s - ) Q (x 10 -11 erg cm -2 s -1 ) A -1.2-0.8-0.40.00.4-0.8 -0.4 0.0 0.4 0.8 1.2 U ( x - e r g c m - s - ) Q (x 10 -11 erg cm -2 s -1 ) B Fig. 3.
Behavior of the polarization Stokes parameters onthe QU planes during the rotation “A” and “B” in the V band. that of “B” was counterclockwise in the QU plane, as canbe seen in the top and bottom panels of figure 3. Thus,the directions of these two rotation were different.Figure 4 shows the light curve and temporal variationsof polarization parameters during the outburst state inthe V band. The solid line was the best-fitted linear func-tion for the corrected PA from JD 2455157 to 2455192.In the fourth panel, we show the residual PA from thelinear function. The PD was low during the early phaseof the outburst. After the outburst maximum, it becamehigh. The residual PA was largely deviated from zeroaround JD 2455170 (the forth panel of figure 4). Aroundthis deviation epoch, the V band magnitude was at themaximum (the top panel of figure 4) and the PD V wasminimum ( ∼ We also obtained the NIR spectroscopic data with
AKARI /IRC during the outburst state. In the top panelof figure 4, the arrows represent the observation epochswith
AKARI . The NIR fluxes and the shape of the spectrawere almost identical within the 1- σ error level among thethree observation epochs. The bottom panel of figure 4shows the averaged spectrum of 3C 454.3. The spectrumis dominated by featureless red continuum emission, in- M. Sasada et al. [Vol. , ν F ν V ( x - e r g c m - s - ) P ( % ) -400-2000 c o r . PA ( deg ) -1000100 5140 5160 5180 5200 5220 s ub ( deg ) JD - 24500000.050.100.150.203.0 3.5 4.0 4.5 5.0 f l u x den s i t y ( Jy ) Wavelength ( µ m) Fig. 4.
Temporal variation of the flux, polarization param-eters, the PA subtracted from the linear function fitted thedata from JD 2455157 to 2455192 in the V band and spec-trum in the NIR region. The arrows in the top panel are theepochs of the AKARI observations. The bottom panel showsthe averaged spectrum of three spectra. -12.5-12-11.5-11-10.5-10-9.5 10 11 12 13 14 15 l og ν F ν ( e r g c m - s - ) log ν (Hz) Fig. 5.
Spectral energy distribution from the radio to opticalbands during the outburst state. The solid line represents thebest fitted third-order log-polynomial function. dicating that the synchrotron radiation was dominant inthis wavelength range during the outburst state.We show the spectral energy distribution (SED) of theobject from the radio to optical regions during the out-burst state in figure 5. The NIR data points are the av-erage flux for three epochs obtained with
AKARI . As forthe optical data, the figure includes the observation withKanata which was obtained on JD 2455180, closest to the
AKARI observations. In the radio band, we use pub-lic data observed with the Submillimeter Array (SMA;1 mm on JD 2455181 and 850 µ m on JD 2455176 ), thethe University of Michigan Radio Astronomy Observatory(UMRAO; 4.8 on JD 2455169.5, 8.0 on JD 2455184.5 and14.5 GHz on JD 2455179.5) and our data observed withthe Yamaguchi radio telescope (8.38 GHz on JD 2455172).All radio data are obtained within 10 d of our AKARI observations. The SMA data was obtained as part ofthe SMA flux density monitoring program and used bypermission (Gurwell et al. 2007). The optical and NIRdata were corrected for the Galactic interstellar extinc-tion based on Schlegel et al. (1998). The extinction forthe
AKARI data were estimated by interpolating the ex-tinctions from 3 to 5 µ m obtained from Schlegel et al.(1998). The solid line represents the best fitted third-orderlog-polynomial function. The peak frequency of the syn-chrotron component was estimated as 7.6 × Hz. Thisis the same order of magnitude as the values reported inprevious studies (e.g. Raiteri et al. 2008b; Abdo et al.2010b). The optical and NIR data are smoothly con-nected, which suggests that at least the emission fromthe NIR to optical energy band can be explained by asynchrotron component.
4. Discussion
Our monitoring observations suggest that two rotationevents of polarization occurred during the active and out-burst states of 3C 454.3 in 2009. The features of theo. ] Multi-wavelength photopolarimetric monitoring of 3C 454.3 in Dec. 2009 7rotation “A” and “B” were summarized in table 1. Therotation rates, directions and periods of the rotations aredifferent between the rotation “A” and “B”.Two models have recently been proposed for the rota-tion of polarization; Marscher et al. (2008) and Abdo et al.(2010a). According to Marscher et al. (2008), a rotation ofthe polarization vector is a sign of helical magnetic field inthe jet. The directions of the rotation events in 3C 454.3are both clockwise and counterclockwise. In the case ofthe simple helical magnetic field in the jet, the directionof the rotation of polarization should be one-side. Thus,it needs more complex magnetic field structure in orderto explain the rotation events in 3C 454.3.Abdo et al. (2010a) reported a rotation event in 3C 279,and suggested a non-axisymmetric structure of the jet,implying a curved trajectory for the emitting material.This idea can explain rotations in both directions, andhence, the rotations in 3C 454.3. We can estimate thedistance, ∆ r , traveled by the emitting material during thewell-sampled rotation “B” in 2009. We calculate the traveldistance ∆ r as Γ c ∆ t , where Γ jet is the bulk Lorentzfactor, c is the speed of light and ∆ t is the duration ofthe rotation event (Abdo et al. 2010a). The duration ofthe rotation “B” was 35 d. In this case, we adopt a valuesof Γ jet = 19 . r ≈ . × cm. The travel distance of ourresult was the same order of magnitude compared withthat for 3C 279 reported in Abdo et al. (2010a). On theother hand, the duration of the rotation “A” was shorterthan that of the rotation “B”. Thus, the travel distanceof the rotation “A” could be shorter than that of “B” ifΓ jet in the active state is same or smaller than that in theoutburst state.Sasada et al. (2011) reported that there was a positivecorrelation between the amplitudes of the flux and PDof flares in 41 blazars. The large-amplitude variations inthe flux and PD were shown in the outburst state. Thus,there was a positive correlation between the amplitudes ofthe flux and PD in the outburst state of 3C 454.3, whichwas consistent with the positive correlation in blazar flaresfound by Sasada et al. (2011).The peak of the PD was delayed by 10 d from the peakof the flux in the outburst. This behavior might be ex-plained by the following simple geometrical effect scenario.Laing (1980) suggest that the PD is the highest when theline-of-sight is parallel to the shock plane in the co-movingframe. In this assumption, the observed flux from the jetis low because of a low Doppler factor. When the line-of-sight is perpendicular to the shock plane, the observed fluxis high because of a high Doppler factor, and PD is lowbecause the magnetic field direction is not aligned. Hence,in this scheme, if the shock comes toward us just after itsformation then the shock plane gradually inclines with re-spect to line-of-sight, the PD rises up after the flux peak.It should be noted that the Doppler factor is changed inthis scenario. The Doppler factor during the 2009 out-burst was estimated with multi-wavelength spectral mod-els, which indicate that the Doppler factor δ was almost ν F ν V ( x - e r g c m - s - ) -400-2000200 4680 4720 4760 4800 c o r . PA ( deg ) JD - 2450000
Fig. 6.
Temporal variations of the corrected PA in 2008. Thetop panel shows the light curve in the V band. The bottompanel shows the temporal variation of the corrected PA in thesame band. Filled circles show the rotation event in 2008. constant, δ ∼ − . F ν, obs , from the shocked region depends strongly on δ ,as F ν, obs ∝ δ α , where α is a spectral index. During theperiod of rotation “B” (JD 2455157–2455192), the max-imum and minimum V -band flux were 6.501 and 1.658 × − erg cm − s − , indicating the flux changed by afactor of 3.9. However, the amplitude of the flux variationshould be larger than that, because the J -band flux varia-tion was larger than that in the V band. If we explain allthe flux variation only by the variation in δ , δ should bechanged at least by a factor of 1.4, assuming the spectralindex, α = 1. This is, however, apparently inconsistentwith the result from SED modeling that it was constantwithin δ ∼ . .
5. Thus, it would be difficult to ex-plain the observation only by the variation in δ . Moredetailed and complex jet modeling would be needed forinterpreting the behaviors of the flux and polarization inthe outburst state. Other rotation events were also reported in active statesof 3C 454.3 in 2005 and 2007. Jorstad et al. (2010) re-ported that the optical outbursts in 2005 autumn and in2007 were accompanied by systematic rotation of the PA.Sasada et al. (2010) also reported a rotation event in the2007 outburst.In 2008, 3C 454.3 had an outburst in the radio, optical,X-ray and gamma-ray bands (Abdo et al. 2009 ; Villataet al. 2009b). We also monitored the object during the2008 outburst with Kanata/TRISPEC in multi-color pho-tometric and polarimetric mode. Figure 6 shows the lightcurve and temporal variation of the corrected PA in the V band in 2008 and 2009. The object reached its outburstmaximum on ∼ JD 2454664. The object stayed active evenafter the maximum, until ∼ JD 2454760, and then it re-turned to quiescence. In the lower panel of figure 6, wecan see a gradual decreasing trend of the PA just afterthe outburst maximum, which apparently continued until M. Sasada et al. [Vol. ,
Table 1.
Rotation rates in each rotation event.
Event Rate Total rotation period(deg d − ) (deg) JD − . ± . . ± . − . ± . − . ± . . ± . The 2005 and 2007 rotation events were reported byJorstad et al. (2010) and Sasada et al. (2010). the object returned to quiescence. In addition, there is asign of a rapid and short rotation of polarization betweenJD 2454680 and 2454691, as indicated with the filled circlein figure 6.To date, five rotation events of polarization were re-ported in 3C 454.3 during its active and outburst statesin 2005, 2007, 2008 and 2009. We summarize their rota-tion rates, total degree of the rotation in table 1. We alsoshow the periods of the rotation events, which we usedfor estimating the rotation rates. Three rotation rates arepositive values and two rates are negative. Thus, there arenot only counterclockwise, but also clockwise directions ofthe PA rotations. Another feature of the rotation eventsis that there seems to be two types of events; fast rotationevents (2007, 2008 and 2009 “A”) and slow rotation events(2005 and 2009 “B”). The fast ones have the absolute rateof the rotation larger than 20 deg d − , whereas the slowones have the value smaller than 10 deg d − . Time dura-tions of two types of rotations are different. The fast oneshave short durations, less than 11 d, and the slow rotation“B” in 2009 has a long duration, 35 d. It could partly bean observational bias because it is difficult to detect shortand slow rotation events. However, it is worth nothingthat there was not any long and fast rotation from 2007to 2009.One possibility is that the fast rotation is the result ofan erratic behavior of the polarization vector. Villforthet al. (2010) reported that the polarization vector can beseparated into two components: an optical polarizationcore and chaotic jet emission. If the chaotic polarizationcomponent with a short timescale moves around the originof the QU plane, a short fast rotation can occur. Ikejiriet al. (2011) also discussed that a rotation episode indi-cated only by a few data points cannot be distinguishedfrom results of a random walk in the QU plane.
5. Conclusion
We performed the photometric and polarimetricmonitoring of 3C 454.3 during 2009—2010. The objectshowed the active and outburst states in the opticaland near-infrared light curves and two rotation eventsin the temporal variations of polarization. In the pastand our polarimetric monitoring, five rotation eventswere reported in 3C 454.3. The rotations were observedin both clockwise and counterclockwise directions in the QU plane. The complex model of the structure ofthe magnetic field in the jet is needed to explain theserotation events.This work was partly supported by a Grand-in-Aidfrom the Ministry of Education, Culture, Sports, Science,and Technology of Japan (22540252,). This researchhas made use of data from the University of MichiganRadio Astronomy Observatory which has been supportedby the University of Michigan and by a series of grantsfrom the National Science Foundation, most recentlyAST-0607523. The Submillimeter Array is a joint projectbetween the Smithsonian Astrophysical Observatoryand the Academia Sinica Institute of Astronomy andAstrophysics and is funded by the Smithsonian Institutionand the Academia Sinica. M. Sasada and M. Yamanakahave been supported by the JSPS Research Fellowshipfor Young Scientists. References
Abdo, A. A. et al. 2009, ApJ, 699, 817Abdo, A. A. et al. 2010a, Nature, 463, 919Abdo, A. A. et al. 2010b, ApJ, 716, 30Ackermann, M., et al. 2010, ApJ, 721, 1383Aharonian, F., et al. 2007, ApJ, 664, L71Albert, J., et al. 2007, ApJ, 669, 862Antonucci, R. 1993, ARA&A, 31, 473Bessell, M. S., Castelli, F., & Plez, B. 1998, A&A, 333, 231Blandford, R. D., & K¨onigl, A. 1979, ApJ, 232, 34Bonning, E., et al. 2009, ATel, 2332, 1Bonnoli, G., Ghisellini, G., Foschini, L., Tavecchio, F., &Ghirlanda, G. 2011, MNRAS, 410, 368B¨ottcher, M., et al. 2007, ApJ, 670, 968Donnarumma, I., et al. 2009, ApJ, 707, 1115Escande, L., & Tanaka, Y. T. 2009, ATel, 2328, 1Fuhrmann, L., et al. 2006, A&A, 445, L1Fukugita, M., Shimasaku, K., & Ichikawa, T. 1995, PASP, 107,945Giommi, P., et al. 2006, A&A, 456, 911Gonz´alez − P´erez, J.N., Kidger, M.R., & Mart´in − Luis, F.2001, AJ, 122, 2055Gurwell, M. A., Peck, A. B., Hostler, S. R., Darrah, M. R., &Katz, C. A. 2007, in ASP Conf. Ser. 375, From Z-Machinesto ALMA: (Sub)Millimeter Spectroscopy of Galaxies, ed.A. J. Baker et al. (San Francisco, CA: ASP), 234Ikejiri, Y., et al.2011, PASJ, 63, 639Jackson, N., & Browne, W. A. 1991, MNRAS, 250, 414Jones, T. W., Rudnick, L., Fiedler, R. L., Aller, H. D., Aller,M. F., & Hodge, P. E. 1985, ApJ, 290, 627Jorstad, S.G., et al. 2010, ApJ, 715, 362Kawabata, K. S., et al.2008, in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series, Vol.7014, Society of Photo-Optical Instrumentation Engineers(SPIE) Conference SerieKrimm, H. A., et al.2009, ATel, 2330, 1Kubo, H., Takahashi, T., Madejski, G., Tashiro, M., Makino,F., Inoue, S. & Takahara, F. 1998, ApJ, 504, 693Laing, R. A. 1980, MNRAS, 193, 439Lister, M. L., Homan, D. C., Kadler, M., Kellermann, K. I.,Kovalev, Y. Y., Ros, E., Savolainen, T., Zensus, J. A. 2009,ApJ, 696, L22 o. ] Multi-wavelength photopolarimetric monitoring of 3C 454.3 in Dec. 2009 9o. ] Multi-wavelength photopolarimetric monitoring of 3C 454.3 in Dec. 2009 9