Kinematic Study of the Blazar S5 0716+714

0.69 MB·

by  U. Bach

Astronomy&Astrophysicsmanuscriptno.0388 February2,2008
(DOI:willbeinsertedbyhandlater)
Kinematic study of the blazar S5 0716+714
U.Bach1⋆,T.P.Krichbaum1,E.Ros1,S.Britzen1,2,W.W.Tian1,3,A.Kraus1,A.Witzel1,andJ.A.Zensus1
1 Max-Planck-Institutfu¨rRadioastronomie,AufdemHu¨gel69,53121Bonn,Germany
2 Landessternwarte,Ko¨nigstuhl17,69117Heidelberg,Germany
3 NationalAstronomicalObservatories,CAS,20ADatun,RoadChaoyang,Beijing100012,PRChina
5
0
Received5March2004;accepted3December2004
0
2
Abstract.Wepresenttheresultsofamulti-frequencystudyofthestructuralevolutionoftheVLBIjetintheBLLacobject
n
0716+714overthelast10years.WeshowVLBIimagesobtainedat5GHz,8.4GHz,15GHzand22GHz.Themilliarcsecond
a
J sourcestructureisbestdescribedbyaone-sidedcore-dominatedjetof 10maslength.Embeddedjetcomponentsmovesu-
∼
perluminallywithspeedsrangingfrom5cto16c(assumingz=0.3).SuchfastsuperluminalmotionisnottypicalofBLLac
1
1 objects,howeveritisstillintherangeofjetspeedstypicallyobservedinquasars(10cto20c).In0716+714, youngercom-
ponentsthatwereejectedmorerecentlyseemtomovesystematicallyslowerthantheoldercomponents.Thisandasystematic

2 positionanglevariationoftheinner(1mas)portionoftheVLBIjetsuggestsanatleastpartlygeometricoriginoftheobserved
v velocityvariations.TheobservedrapidmotionandthederivedLorentzfactorsarediscussedwithregardtotherapidIntra-Day
6 Variability(IDV)andtheγ-rayobservations,fromwhichveryhighDopplerfactorsareinferred.
0
4 Keywords.Galaxies:jets–BLLacertaeobjects:individual:S50716+714–Radiocontinuum:galaxies
2
1
4
0 1. Introduction dominated evolving jet extending several 10milliarcseconds
/ tothenorth(Eckartetal.1986,1987;Witzeletal.1988).The
h The S5 blazar 0716+714 is one of the most active BLLac
VLBI jet is misalignedwith respectto the VLA jetby 90
p ◦
objects. It is extremely variable on time-scales from hours ∼
- (e.g., Saikiaetal. 1987). In the literature the jet kinematics
o to months at all observed wavelengths from radio to X-rays. in0716+714arediscussedcontroversially.Thereexistseveral
r The redshift of 0716+714 is not yet known. However, opti-
t kinematic scenarios with motions ranging from 0.05masyr−1
s cal imaging of the underlying galaxy provides a redshift es-
a to1.1masyr−1.However,someoftheseearlierkinematicmod-
timateofz 0.3(Wagneretal.1996).ArecentX-raystudyby
v: Kadleretal≥.(2004)suggestsamuchlowerredshiftof0.1,but els are based on only two or three epochs separated by sev-
eral years, which sometimes could have led to ambiguities
i
X since this needs further confirmation a redshift of 0.3 is used
in the identification of fast components (Eckartetal. 1987;
throughoutthispaper.Intheradiobands0716+714isanintra-
r Witzeletal.1988;Schalinskietal.1992;Gabuzdaetal.1998;
a day variable(IDV)source (Witzeletal. 1986; Heeschenetal.
Perez-Torresetal. 2004). More recent studies based on more
1987). It exhibitsa flat radiospectrumat frequenciesupto at
datameasuredhigherjetspeeds(Tianetal.2001;Jorstadetal.
least 350GHz. Quirrenbachetal. (1991) found strongly cor-
2001b;Kellermannetal.2004),buttheresultswerestillincon-
related IDV between radio and optical bands. This and the
sistentbetweenthetwostudies.Thus,itisatpresentnotclear
correlated variations between X-ray and optical and the si-
if0716+714isasuperluminalsourceandandhowfastthemo-
multaneousvariationsbetweenopticalandradiostronglysug-
tionoftheVLBIjetcomponentsis.
gestanintrinsicoriginoftheintradayvariability(Wagneretal.
Inthispaper,wepresentanddiscussourresultsfromare-
1996;Wagner&Witzel1995).FromtheobservedIDVatcm-
analysis of the last 10 years of VLBI data on 0716+714,ob-
wavelengthsatypicalbrightnesstemperatureofT = 1015.5K
b tainedatfrequenciesbetween5GHzand22GHz.Ouranalysis
to1017Kisderived(Krausetal.2003;Wagneretal.1993).A
includes26observingepochslistedinTable1.Throughoutthis
Doppler factor of 15 to 50 would be needed to bring these
paper we will use a flat universe, with the following parame-
brightnesstemperaturesdowntotheinverse-Comptonlimitof
ters:AHubbleconstantof H = 71kms 1Mpc 1,apressure-
1012K. 0 − −
lessmattercontentofΩ =0.3andacosmologicalconstantof
m
Very Long Baseline Interferometry (VLBI) studies span-
Ω = 0.7.Withtheseconstantsanangularmotionof1mas/yr
λ
ning more than 20years at cm-wavelengths show a core-
correspondstoaspeedof18.8catz=0.3(6.6catz=0.1).
Sendoffprintrequeststo:U.Bach,e-mail:[email protected] InSect.2,webrieflydescribetheobservationsandthedata
⋆ Currentaddress:INAF-OsservatorioAstronomicodiTorino,Via reduction, the model fitting process and present the final im-
Osservatorio20,10025PinoTorinese,Italy ages.InSect.3wepresentthecross-identificationoftheindi-
2 U.Bachetal.:KinematicstudyoftheblazarS50716+714
vidualjetcomponentsandanewkinematicmodelforthejetin intervalsaslongasthewholeobservationaltimedowntomin-
0716+714.From this we derive the jet speed and orientation. utes.TheresultingmapsarepresentedinFigs.1&2.Here,the
WesummariseourresultsinSect.4. jetisclearlyvisibleto thenorthatP.A. 15 ,andisslightly
◦
∼
bent.Atthelowerfrequencieswecanfollowthejetuptoadis-
tanceof10masto15masfromthecore,whereasatthehigher
Table1.Observationlog.Listedaretheobservingepoch,fre-
frequenciesthejetisvisibleupto3mas.
quency ν, total flux density S (ν), beam size, beam position
tot
angle,peakfluxdensityS andthelowestcontourat3σof
peak
themapinFigs.1&2.
2.1.Modelfits
Epoch ν Stot Beam Speak 3σ After imaging in D we fitted circular Gaussian compo-
[GHz] [Jy] [mas×mas],[◦] [beJaym] [bmeaJmy] nents to the self-calibrated data to parameterise the source
1992.73a,(1) 5.0 0.67 0.8 1.0, 0 0.61 1.5 structureseen in theVLBI maps.We used onlycircularcom-
×
1992.85a 22.2 0.75 0.9 2.7, 55 0.64 3.0 ponents to reduce the number of free parameters and to sim-
× −
1993.71a 22.2 0.40 0.2 0.2, 35 0.31 2.0
× plifytheanalysis.AsummaryoftheGaussiancomponentpa-
1994.21 8.4 0.32 0.8 1.4, 80 0.26 0.7
× rameters from this model fitting is given in Table 2. For all
1994.21 22.2 0.34 0.3 0.5, 79 0.30 1.8
× modelfitsweusedthebrightestcomponentasareferenceand
1994.67(2) 15.3 0.46 0.5 0.7, 29 0.40 1.8
× − fixed its position to (0,0). The positions of the other compo-
1994.70a,(1) 5.0 0.36 0.8 1.2, 15 0.29 2.7
1995.15(3) 22.2 0.75 0.3×0.5,− 3 0.71 4.0 nentsweremeasuredrelativetothiscomponent,whichweas-
1995.31(3) 22.2 0.43 0.3×0.4, −39 0.40 4.2 sumedtobestationary(seeSect.3.1fordetails).Amaximum
×
1995.47(3) 22.2 0.20 0.3 0.6, 7 0.18 3.0 uncertaintyof15%inthefluxdensitywasestimatedfromthe
×
1995.65 8.4 0.31 0.9 1.0, 54 0.26 0.4 uncertaintiesoftheamplitudecalibrationandfromtheformal
×
1995.65 22.2 0.34 0.3 0.4, 40 0.28 1.8 errorsofthemodelfits.Thepositionerrorisgivenby∆r= σΘ
1996.34(3) 22.2 0.27 0.3×0.9,−48 0.22 3.2 (Fomalont1989),whereσistheresidualnoiseofthemapaf2tS·ePr
×
1996.53(2) 15.3 0.26 0.5 0.7, 15 0.22 0.7 thesubtractionofthemodel,θthefullwidthathalfmaximum
× −
1996.60(3) 22.2 0.25 0.4 0.4, 9 0.21 3.0
1996.63a,(1) 5.0 0.22 1.4×2.2, 41 0.18 1.3 (FWHM) of the component and Speak the peak flux density.
× Thisformulatendstounderestimatetheerror,if thepeakflux
1996.82(2) 15.3 0.26 0.4 0.7, 21 0.25 2.0
× density is very high or the width of the component is small.
1996.90(3) 22.2 0.29 0.4 0.5, 35 0.27 2.0
× − Thereforeweincludedanadditionalerrorarisingfromtheposi-
1997.03(4) 8.4 0.21 1.1 1.4, 7 0.18 0.0
1997.58(3) 22.2 0.98 0.4×0.5, 33 0.90 4.0 tionvariationsduringthemodelfittingprocedure.InFig.3,the
1997.93(5) 8.4 0.43 0.6×1.6,−33 0.38 0.5 core distances of individual VLBI components derived from
×
1999.41(5) 8.4 1.02 0.5 1.5, 7 0.92 0.5 ourmodelfitsatvariousfrequenciesareplottedagainsttime.
× −
1999.55(2) 15.3 1.25 0.5 0.8, 4 1.14 1.2
×
1999.89a,(1) 5.0 0.63 1.8 2.4, 60 0.54 0.7
2000.82b 5.0 0.55 0.8×0.9,−28 0.48 0.1 6
× −
2001.17(2) 15.3 0.65 0.4 0.8, 31 0.54 0.9 C11
× 5 C10
C9
Note:ThearrayusedwastheVLBA,unlessindicatedbyafootnote. C8
Epochsinboldfacedenoteowndata.a:Globalarray,b:VLBA+Eb C7
4 C6
References:(1):Britzenetal.(1999)&inprep.(2005),(2): C5
C4
K(2e0l0le1rbm),a(n4n):eFtaeyl.&(19C9h8a)r&lotZ(e2n0s0u0s),e(t5a)l:.R(2o0s0e2t)a,l(.3()2:0J0o1rs)tadetal. mas]3 CC32
r [ C1
lin. fit
2
2. Observationsanddataanalyses 1
All data were correlated in the standard manner using the
0
1993 1994 1995 1996 1997 1998 1999 2000 2001
MKIIIcorrelatorinBonnandtheVLBAcorrelatorinSocorro.
epoch
Part of the post-correlation analysis was done using NRAO’s
AstronomicalImageProcessingSystem(A)andtheCaltech Fig.3. Core separation as a function of time for the individ-
VLBI Analysis Programs (Pearson&Readhead 1984). Data ual modelfit components. Data from all frequenciesare com-
from other observers were provided as amplitude-calibrated bined. Possible frequency-dependent position shifts are less
and fringe fitted (u,v) FITS-files. The imaging of the source than0.1masandarenotcorrected.Thesolidlinesshowthelin-
includingphaseandamplitudeself-calibrationwasdoneusing earfits tothepathforeachcomponent.To showmoreclearly
the CLEAN algorithm (Ho¨gbom 1974) and SELFCAL pro- the well-defined components we do not show two of the far-
cedures in D (Shepherd 1994). The self-calibration was thestdata pointsof C1 at r = 11.2masandr = 11.9masand
doneinstepsofseveralphase-calibrationsfollowedbycareful onedatapointofC2atr=6.9mas.
amplitudecalibration.Duringtheiterationprocessthesolution
interval of the amplitude self-calibration was shortened from
U.Bachetal.:KinematicstudyoftheblazarS50716+714 3
Fig.1.Allcontourmapsof0716+714at5GHz(toprow),8GHz(middle)and15GHz(bottom).Themapsareconvolvedwith
circularbeamsof1.2masat5GHz,0.8masat8GHzand0.5masat15GHz.Totalfluxdensity,originalbeamsizeandthelevel
ofthelowestcontouratthreeσaregivenintheobservinglog(Table1).Contoursareincreasingbystepsoftwo.
3. Resultsanddiscussion The cross identification of moving VLBI componentsbe-
tweendifferenttimesandfrequenciesisnotunambiguousand
3.1.Identificationofthecomponents
dependson the dynamic range of the individualmaps and on
the time sampling.When we started ouranalysis we used the
Toinvestigatethekinematicsinthejetof0716+714,wecross-
followingscenariosasworkinghypotheses:
identified individual model components along the jet using
theirdistancefromtheVLBIcore,fluxdensityandsize.Since
1. Stationarycorewithstationaryorslow-movingjet,
the observations were not phase-referenced, the absolute po-
2. Stationarycorewithfastmovingjet,
sition information was lost, and it is impossible to tell which
3. Non-stationarycorewithfastorslow-movingjet,
components,if any,are stationary.To register the images and
4. Non-monotonicmotionofthejet.
totestthestabilityoftheVLBI-coreposition,wecomparedthe
coreseparationsofthejetcomponentsofeachepochwith re- Asmoredatabecameavailable,wecouldruleoutmostof
spect to the adjacent observations, but we could not find any these identification schemes, and we were left with a scheme
systematicpositionoffsets.Wealsocheckedforpositionshifts that assumes a stationary core and relatively fast component
between the jet components at higher and lower frequencies motion. The different schemes were tested, first separately at
duetoopacityeffects(e.g.,Lobanov1998).Inspectionofclose eachfrequencyandlaterwithallmodelfitscombined.
or simultaneously observed frequency pairs does not show a Supported by a graphical analysis, which is presented in
systematicfrequencyshiftbetweenhigherandlowerfrequen- Fig.3,wecouldobtainasatisfactoryscenarioforthekinemat-
cies larger than 0.1mas (position accuracy between 15GHz icsinthejetof0716+714.Thisscenarioconsistsof11identi-
and22GHzis<0.1mas).Therefore,weconcludethatthecore fiedsuperluminalcomponentsmovingawayfromthecore.As
position is stationary and that all componentpositions can be anexampleoftheprocedureusedforallcomponentsweshow
measuredwithrespecttothisbrightestcomponent. theidentificationschemefortheseparatefrequenciesforcom-
4 U.Bachetal.:KinematicstudyoftheblazarS50716+714
Fig.2.Allcontourmapsof0716+714at22GHz.The mapsareconvolvedwith a circularbeamof 0.3mas. Totalflux density,
originalbeamsizeandthelevelofthelowestcontouratthreeσaregivenintheobservinglog(Table1).Contoursareincreasing
bystepsoftwo.
ponentC5,C7,C8andC9inFig.4.Againthefrequencyshifts mogeneous time sampling, it is sometimes difficult to iden-
between the trajectories are typically smaller than 0.1mas to tifycomponentsoveralongtimeinterval.However,especially
0.2masandaresmallerthanthemeasurementuncertainty∆r. between 1993 and 1996, when we made many measurements
Table 3 summarises the angular separation rate µ of the spacedbyonlyafewmonths,ourproposedcomponentidenti-
VLBI componentsderived from the linear fits of r(t) and the ficationgivesthesimplestandmostreasonablefitto thedata.
corresponding apparent speed β = v /c for an assumed Since most of our data were obtained at high frequencies,
app app
redshiftof0.3. wheretheouterregionofthejetisfaintandalreadypartlyre-
solvedbytheinterferometerbeam,theparameterofthecorre-
sponding jet components(at r 4mas) are not as well con-
3.2.Thekinematicmodel ≥
strained as the inner jet components and therefore they have
larger positional errors. Despite this, there is still a very dis-
The components in this scenario move with 0.2masyr 1 to
−
tincttendencythattheoldercomponents(C1–C3),whichnow
0.6masyr 1 in theinnerpartof theVLBIjet(r 3mas)and
−
≤ arelocatedatlargecoreseparations,movefasterthanthecom-
with up to 0.9masyr 1 in the outer regions. Due to the inho-
−
U.Bachetal.:KinematicstudyoftheblazarS50716+714 5
1
s3 −8
5 GHz +0.19097 +0.066941s4 , with
4 8 GHz a #
15 GHz
22 GHz 1 Ω 1
fit 5 GHz s3 = − m and a= ,
3 ffiitt 81 5G GHHzz Ωm 1+z
C5
fit 22 GHz
as] whichisananalyticalfit(Pen1999)totheluminositydis-
m tance for flat cosmologies with a cosmological constant. For
r [2 C7
therangeof0.2 Ω 1.0,therelativeerrorofthedistance
m
≤ ≤
C8 islessthan0.4%foranygivenredshift.Using
1 C9
µd
β = L (2)
app c(1+z)
0
1993 1994 1995 1996 1997 1998 1999 2000 2001
epoch forthetransformationofapparentangularseparationrates
(µ)intospatialapparentspeeds(β )(Pearson&Zensus1987)
app
Fig.4.Coreseparationasafunctionoftimeforthecomponents
and assuming a redshift of 0.3 for 0716+714, one milliarc-
C5,C7,C8andC9separatedforthedifferentfrequencies.The
second corresponds to 4.4pc. The measured angular separa-
solidlinesshowthelinearfitstothepathforeachcomponent.
tionratesthencorrespondtospeedsof4.5c to 16.1c.Forthe
Shiftsbetweenthepathsatdifferentfrequenciesarevisible,but
components C3 to C9, for which the proper motion determi-
aretypicallysmallerthanthemeasureduncertainties.
nationisthemostconfident(oversevendatapoints),theaver-
agemotionisµ=(0.41 0.11)masyr 1whichcorrespondsto
−
±
ponentslocatedin theinnerjet(C4– C10).Thisbehaviouris βapp = 7.7 2.1.ApreliminaryanalysisbyTianetal.(2001),
±
alsovisibleinTable3,whichshowsacleartrendwithsystem- whousedonlypartofthedatapresentedhere,gaveverysimi-
atically decreasing speeds between component C1 (16c) and larspeeds.Arecentlypublishedkinematicanalysisofthedata
componentC10 (4.5c). The coarse time sampling for the in- from the VLBA 2cm Survey also measured motion of 10c
dividualjet componentsunfortunatelydoesnotallow usto fit to 12c in the jet of 0716+714 (Kellermannetal. 2004). The
acceleration to r(t). Future and more densely sampled VLBI speedsfoundinourpaperarelowerthanthe16cto21cfound
observationsarerequiredtoshowwhethertheinnerjetcompo- by Jorstadetal. (2001b)1, who observedthe source overonly
nentsindeedmovelinearly. a short time interval (3yr). The difference between their and
our results may be explainedby slightly different (and unfor-
tunatelynotunambiguous)mapparameterisationsusingdiffer-
Table3.Propermotionsinthejetof0716+714.Thenumberof
entGaussianmodelsorbynon-linearcomponentmotion,with
data pointswhich wereused forthe fit foreach componentis
possibleaccelerationforshorterperiodsoftime.
givenincolumn2.Thelastcolumngivestheback-extrapolated
With an average apparent jet speed of about 8c and
ejectiondatesofthecomponents,whichresultfromthelinear
likely higher speeds of up to 15c to 20c, 0716+714 dis-
fits.
plays considerably faster motion than other BLLac objects,
for which speeds of 5c are regarded as normal (e.g.,
Id. # µ[mas/yr] βapp(z=0.3) Ejectiondate Vermeulen&Cohen 1≤994; Gabuzdaetal. 2000; Rosetal.
C1 4 0.86 0.13 16.13 2.36 1986.7 1.3
± ± ± 2002). In this contextit appearsthat 0716+714is an extreme
C2 5 0.63 0.10 11.87 1.82 1988.5 1.0
± ± ± BLLac object with a jet speed (and Lorentz factor) much
C3 8 0.62 0.04 11.59 0.76 1989.7 0.3
C4 7 0.44±0.01 8.27±0.24 1989.8±0.1 higher than regarded typical for BLLac objects and close to
± ± ±
C5 7 0.43 0.02 8.16 0.29 1990.8 0.2 thespeedsof10cto20cseeninquasars.
± ± ±
C6 9 0.41 0.02 7.80 0.33 1992.3 0.1
± ± ±
C7 17 0.37 0.01 6.89 0.19 1993.2 0.1
± ± ± 3.3.Kinematicsandgeometryofthejet
C8 11 0.32 0.01 6.04 0.20 1994.4 0.1
± ± ±
C9 10 0.26 0.01 4.98 0.27 1995.1 0.1
± ± ± Using the measuredmotion,(11.6 0.8)c, of C3, the fastest,
C10 3 0.24 0.01 4.52 0.45 1996.5 0.2 ±
± ± ± best-constrainedjetcomponentinourmodel,wecanplacelim-
C11 2 0.29 0.02 5.48 0.55 1998.2 0.3
± ± ± itsonthejetspeedandorientationof0716+174.Adopting
βsinθ
Theluminositydistanced wascalculatedadoptingthefollow-
L β = (3)
ingrelation app 1 βcosθ
−
c 1 wefindthatthejetinclinationthatmaximisestheapparent
d = (1+z) η(1,Ω ) η( ,Ω ) , (1)
L H0 " m − 1+z m # speed,θmaxis4.9◦.Fromthat,onederivesaminimumLorentz
factor, γ = (1 + β2 )1/2 of 11.6, which corresponds to a
min app
where
1 Jorstad et al. originally published 11c to 15c using H =
0
1 s s2 100kms 1Mpc 1 andq = 0.1.Wecorrectedthesenumbersforthe
η(a,Ω ) = 2√s3+1 0.1540 +0.4304 − − 0
m "a4 − a3 a2 cosmologicalparametersusedinthispaper.
6 U.Bachetal.:KinematicstudyoftheblazarS50716+714
Dopplerfactorofδ =γ 1 (1 β cosθ ) 1 =11.7,where If we now try to explain the observed range of apparent
min m−in − min max −
componentspeedsthroughvariationsoftheviewingangle(un-
β = 1 γ 2 . For smaller viewing angles (θ 0) the
min q − m−in → dertheassumptionofaconstantLorentzfactoralongthejet),
Doppler factor approachesδ = 2γ, which yields δ = 11.7 to
we cannot reach slow apparent speeds of 4.5c without vi-
23.4 at viewing angles between 4.9 and 0 for component ∼
◦ ◦ olating the lower limit of the Dopplerfactor of 2.1 from syn-
C3.IncludingthehighervaluesfoundbyJorstadetal.(2001b)
chrotron self-Compton (SSC) models (Ghisellinietal. 1993).
andthehighestspeedderivedfromourdata(16.1c),aDoppler
Therefore, we can exclude θ > 5 and the only solution re-
boostingfactorofuptoδ 40appearspossible. ◦
≥ mainingis todecreasetheviewinganglewhichautomatically
leadstohigherDopplerfactors.ThegreyshadedareainFig.5
marks the allowed region of Lorentz factors and viewing an-
gles.ItbecomesobviousthatweneedatleastaLorentzfactor
ofγ 15andaviewingangleoftheVLBIjetofθ 2 toex-
◦
≈ ≈
plain the largerangeof observedapparentspeedsasan effect
of spatial jet bending. These values are consistent with those
derivedfromIDV.
3.4.Fluxdensityevolution
In Fig. 6 we show the evolution of the flux density of VLBI
components at 22GHz with their increasing separation from
thecore.Weshowthisfrequencybecausealargefractionofthe
datawereobtainedat22GHz(11epochs).Withtheexception
ofafewdatapoints,thecomponentsfadeastheytraveldown
thejet.Thisagreesqualitativelywiththetheoryofaconically
Fig.5. The Doppler factor versusthe apparentspeed for con- expanding jet (e.g., Blandford&Ko¨nigl 1979). As the VLBI
stantintrinsicLorentzfactor,γ,(solidlines)andconstantview- componentsexpandtheybecomeopticallythinandtheirspec-
inganglesθ(dottedlines).Therangeofmeasuredspeedsisin- trasteepen.ThisisillustratedforcomponentC7inFig.7.The
dicatedbythedashedlinesandthegreyshadedareamarksthe spectral index α is defined as S να. Knowing the turnover
∝
possiblevaluesforγandθ(seetextfordetails). frequencyandthe size of a jet componentallowsus to calcu-
latethestrengthofthemagneticfieldinthejet(e.g.,Marscher
1983)as
At cm-wavelengths, 0716+714 is a prominent intraday
θ4ν5δ
variable source, which shows amplitude variations of 5% B=10 5b(α) m gauss, (4)
−
to 20% on time-scales of 0.25d to 2d (Wagneretal. 1996; Sm2 (1+z)
Krausetal. 2003). Correlated radio-optical IDV observed whereb(α)isa tabulatedparameterdependentonthespectral
in this source (Quirrenbachetal. 1991; Wagneretal. 1996; index (Tab. 1 in Marscher 1983) with a value ranging from
Qianetal.1996)suggeststhatatleastsomefractionoftheob- 1.8 to 3.8 for optically thin emission (α = 0.25 to 1.0),
− −
servedrapidvariabilityhasanintrinsicsourceoriginandcan- S (in Jy) is the flux density at the turnover frequency ν
m m
not be attributed to refractive interstellar scintillation (RISS) (in GHz) and θ (in mas) is the size of the component. If we
alone, as is done for some other IDV sources (e.g., Rickett calculateBforeachofourmodelfitcomponents,withoutcor-
2001; Qianetal. 2001; Kedziora-Chudczeretal. 2001). The rectingfortheDopplerfactorδ,wegettheapparentmagnetic
presence of the observed broad-bandcorrelations of the vari- field along the jet. Since we do not have adequate spectra for
ability and recently detected IDV at 9mm wavelength,where all positions along the jet we will consider a decrease of the
RISSshouldnotplayadominantroleduetoitsλ−2dependence turnoverfrequencyalong the jet proportionalto rp, to correct
(cf.Krichbaumetal. 2002;Krausetal. 2003), furthersupport for the expansion of the component. The best results are ob-
anonnegligibleintrinsiccontributiontotheIDVin0716+714. tainedwhenwestartwithaturnoverfrequencyof4GHzforthe
Inthefollowingwe useanaveragetypicalbrightnesstemper- innerjet(r 0.2mas),thatisplausiblefromtheearlyspectrum
ature derived from the IDV observed at 6cm of 1015.5K to (1994)ofco≈mponentC7(Fig.7)thoughtheturnoverisnotac-
1017K.Slightlyhighervaluesofuptoafewtimes1018Kwere tually observed, and use p = 1.3, which is in good agree-
−
measured only occasionally. To bring this brightness temper- ment with the relativistic jet model (Marscher&Gear 1985).
atures down to the inverse Compton limit of 1012K, Doppler For all componentsalong the jet we derive magneticfields in
factors in the range of δ = (Tb/1012K)31 ≈ 15 to 50 are re- the range of 10−4δG to 10−7δG. Since the minimum mag-
quired.AdoptingtheseDopplerfactors,weobtainjetLorentz neticfieldstrengththatshouldbepresentduetheenergyden-
factorsofγ 8to25andviewinganglesofθ 1◦. sityofthecosmicmicrowavebackgroundisafewmicroGauss
≈ ≤
The relation betweenthe Dopplerfactor, intrinsicLorentz (e.g.,vanderLaan&Perola1969)the10 7δGalreadyimplies
−
factor,viewingangleandapparentspeedisillustratedinFig.5. alowerlimitoftheDopplerfactorofabout10.Atthisstagethe
The Doppler factors derived from IDV lie within the grey corecomponentisexcludedfromtheanalysisbecauseboththe
shadedarea. sourcesizeandtheturnoverfrequencyarepoorlyknown.
U.Bachetal.:KinematicstudyoftheblazarS50716+714 7
0.1 emission is relativistically beamed. In the latter case, the dif-
C10 ferentdependenceofthemagneticfieldontheDopplerfactor
CC0098 in Equation4(B δ)andin Equation8 (Bmin δ2/7α+1) can
CC0076 be used to calcula∝te the Doppler factor from Bm∝in/B = δ2/7α.
C05
C04 ComparingtheresultsfromalljetcomponentsyieldsDoppler
factors of 4 to 15 with a mean value of 9. These values are
y]
S [J 0.01 slightlylowerthanthosederivedfromthekinematics,butgiven
thatthecalculationsarebasedonmanyassumptionsthevalues
agreereasonably.
0.1
1994
1996
0.001 0.5 1 2 0.05 1997
r [mas]
Fig.6. Flux density of the VLBI componentsat 22GHz plot- α =-0.23±0.06
5/15
tdeodwvnerthsuesjecto.reseparation.Thecomponentsfadeastheytravel S [Jy]0.02 α5/15=-0.61±0.13 α15/22=-0.77±0.18
0.01 α15/22=-1.03±0.25
α =-0.78±0.16
8/22
Assumingequipartitionbetweentheenergyoftheparticles,
Ee,andtheenergyofthemagneticfield,EB,onecanalsoderive 0.005
theminimummagneticfieldfromthesynchrotronluminosity,
L. 5 10 20
ν [GHz]
ν2
L = 4πdL2 (1+z)Z S dν (5) Fig.7.SpectralevolutionofcomponentC7.
ν1
E = f(α,ν ,ν )LB 1.5, (6)
e 1 2 −
In Fig. 8 we show the long-term radio variability and the
where f(α,ν1,ν2) is a tabulated function (e.g. Pacholczyk spectral index of 0716+714over the time range in which the
1970)andν1andν2aretheupperandlowercutofffrequencies VLBI components were born. We show a combined data set
ofthesynchrotronspectrumwithtypicallyvaluesof107Hzto with flux density measurements at 5GHz and 15GHz from
1011Hz.Takingtheenergydensityofthemagneticfieldtobe the UMRAO flux density monitoring program (Alleretal.
1/8πB2 and assumingsphericalsymmetrythe totalenergyof 2003) and from our flux-density monitoring performed with
thesourceis the 100m radio telescope at Effelsberg (Pengetal. 2000). A
detailed discussion of the flux density variability using these
E = (1+k)E +E (7)
tot e B andotherdatawasrecentlydonebyRaiterietal.(2003).Here,
1
= (1+k) f(α,ν ,ν )LB 1.5+ R3B2, werestrictthediscussiontoapossiblecorrelationoftheradio
1 2 −
6 variabilitywiththe ejectionofnew jetcomponents.The ejec-
wherekistheenergyratiobetweentheheavyparticlesandthe tiondatesofnewVLBIcomponentsareindicatedbythegrey
electrons and R is the size of the component. The ratio k de- shadowedareas.Theirwidthscorrespondtotheuncertaintyin
pendsonthemechanismofgenerationoftherelativisticelec- the back-extrapolated ejection dates (see Table 3). Although
trons,whichisunknownatthepresent.Itcanrangefromk 1 these uncertainties are rather large and the time sampling of
foranelectron-positronplasmauptok 2000,ifmostof≈the thelightcurvesisnotalwaysdenseenough,aweakcorrelation
energyiscarriedintheprotons.Weused≈k 100inourcalcu- between the ejection of new componentsand the flares in the
lations,whichseemstobeareasonablevalu≈e(e.g.,Pacholczyk radiobandsis obvious.Eachofthe shadedareaseitherliesat
1970).GiventhatE B 1.5andE B2,E hasaminimum orshortlybeforethetimeofafluxdensityincreaseofatleast
e − B tot
andwefind ∝ ∝ oneofthetwoobservingbands.
The two outburstsin mid 1992and early 1995,which are
B = 9(1+k) f(α,ν ,ν )LR 3 2/7 (8) best seen in the 2cm band, are both surrounded by two new
min 1 2 −
2 ! components, but the time sampling is too sparse to draw any
= 5.37 1012(S ν d2 R 3)2/7. furtherconclusions.InthelowerpanelofFig.8wecalculated
· m m L − the spectral index between 5GHz and 15GHz from the two
with k = 100, f( 0.5,107,1011) = 1.6 107 andS in Jy, light curves. Since the light curveswere not measured simul-
m
− ·
ν in GHz and R in cm. Using z = 0.3 and the same values taneouslywesearchedfortheclosestpairsinthedatasetand
m
for S , ν and θ as in Equation 4 we obtain magnetic fields calculatedthespectralindexfromthesepairs.Themeansepa-
m m
between10 4Gand10 2Gforthejet.Thesevaluesarelarger rationbetweentwopointsisabouteightdays.Thegraphitself
− −
thanthosederivedbefore,whichmeansthateithertheparticle representsaneight-pointrunningaveragetoreducethenoisein
energydominatesandwedonothaveequipartitionorthatthe thedata.Thisfigureshowsanobviouscorrelationbetweenthe
8 U.Bachetal.:KinematicstudyoftheblazarS50716+714
spectrum. For the light curves this test yielded a 50% prob-
ability of our results occurring at random. The probability of
measuring a flattening of the spectrum in seven out of eight
components selected at random from our spectrum is 5.7%.
Generallyaresultisexceptedtobestaticallysignificantifthe
probabilityisbelow5%,whichmeansthatthecorrelationsof
the light curve with the ejection dates are likely to be a coin-
cidence,butgiventhatwemighthavemissedacomponentthe
correlationofthespectralindexismoresignificant.Acoordi-
nated and much denser sampled multi-frequencyflux density
andVLBImonitoringisnecessarytostudyinmoredetailsuch
outburst-ejectionrelations.
3.5.Connectionwithγ-rayflares
0716+714 was detected at γ-rays by the EGRET de-
tector on board the Compton Gamma Ray Observatory
(Hartmanetal.1999;Mattoxetal.1997).Sinceitisproposed
that the γ-ray emission, like the radio emission, is Doppler
boosted (Dermer&Schlickeiser 1994), the γ-ray detection of
0716+714yieldsfurtherevidenceforahighLorentzfactorin
thejet.
Recentstudiesabouttheconnectionbetweenγ-raysources
andsuperluminalVLBI-componentssuggestthatthesesources
havehigheraveragejetspeedsthansourceswhicharenotde-
tected atγ-rays(Jorstadetal. 2001a; Kellermannetal. 2004).
Moreover, Jorstadetal. (2001a) found a correlation between
the ejection of new VLBI componentsand the appearance of
γ-ray flares, suggesting that the radio and the γ-ray emission
Fig.8. The long term flux density variability of 0716+714 as
originate within the same shocked area in the relativistic jet.
measuredatEffelsbergandwith theMichiganradiotelescope
Theauthorsreportaγ-rayflareof0716+714in1992.2which
at5GHz(top)and15GHz(middle)andthespectralindex(bot-
appearstobecoincidentwithapeakinthelinearpolarization
tom),α .Theshadedareasmarktheejectiondatesofthe
5/15GHz
light-curve at 15GHz. Unfortunately, their 0716+714 VLBI
VLBIcomponentsandtheiruncertainties.
data does not cover this period, but with our larger sample
we found a new VLBI component (C6) with an extrapolated
ejectiondateof1992.3 0.1whichisingoodagreementwith
spectralindexvariationsandtheejectionofanewVLBIcom-
±
thereportedγ-rayflare.Althoughwe registerthreeadditional
ponent.Sevenoutofeightejections(exceptC9)ofnewVLBI
ejectionsupto1996and0716+714wasobservedseveraltimes
componentsareaccompaniedbyaflatteningofthesourcespec-
by EGRET in this period, no more flares were detected. But
trum. The flattening is in good agreement with an expanding
giventhattheγ-raymeasurementshavelargeuncertaintiesand
componentthat becomes optically thin first at the higher fre-
only 18 data points were recorded during these four years,
quencyandlateratthelowerfrequency(e.g.,Marscher&Gear
this seems not exceptional. Similar outburst ejection events
1985).
werealsofoundine.g,0528+134(Krichbaumetal.1995)and
We note that at the times of particularly dense time sam-
0836+710 (Otterbeinetal. 1998). Assuming that the correla-
pling (after 1994) a number of minor radio flares are visible,
tion between the new component,C6, and the 1992.2flare is
whicharenotrelatedtotheejectionofanyoftheknownVLBI
real, this supports the idea that the γ-ray emission originates
components. According to the light-house model and other
from the same regionfrom which we observethe radioemis-
related helical jet models (e.g., Camenzind&Krockenberger
sion(e.g.,Sikoraetal.(1994)).
1992; Rolandetal. 1994) it is possible that initial outbursts,
which are related to the ejection of new jet components are
followedbysecondaryfluxdensityvariations,whicharenotor 3.6.Variationoftheejectionpositionangleandcore
onlyareindirectlyrelatedtothecomponentejection.Ofcourse,
fluxdensity
it is also possible thatwe missed some jetcomponentsin our
infrequentlysampledVLBImonitoring. Thevisualinspectionofpositionangleoftheinnermostportion
Totestthesignificanceofthecorrelationsfoundbyeye,we oftheVLBIjetinthemapsofFigs.1&2indicatesavariation
placedeightfieldswithasimilarwidthrandomlyoverthelight oftheP.A.ofthejetnearthecorewithtime.Inordertoquantify
curvesbetween1989and1999anddeterminedhowmanyare this, we used theGaussian modelsandfitted a straightlineto
coincidentwith a flare in the lightcurveor a flatteningof the the first milliarcsecond of the jet. For each observing epoch,
U.Bachetal.:KinematicstudyoftheblazarS50716+714 9
the positionangle ofthis line providesa goodestimate of the 3.7.Redshiftdependenceoftheresults
direction of the inner portion of the VLBI jet. In Fig. 9 (top
Sinceallthederivedparametersofthekinematicsandgeome-
panel)weplottheP.A.groupedinoneyeartimebins.
try scale with thedistance to0716+714,knowingtheredshift
is of great importance. The distance to 0716+714 and there-
with the speeds of the jet components scale nearly linearly
20 withzattheselowredshifts.Thusaredshiftdecreasebyafac-
18
tor of three would also decrease the Doppler factor by three.
16
P.A. [°] 111024 rTisehdTeubcde∝epT(cid:16)e1n+zbdzye(cid:17)n26c(.e4e.)og.f.S,bianricdgeeh,ctrTneeasssetfδer3motmpheizsraw=tuor0ue.l3ddetiornicvzree=das0fer.o1tmhweoIdDuilsVd-
8 b b ∝
crepancy between the Doppler factor, which is needed to re-
1.5 duceT andtheDopplerfactorderivedfromthekinematics.An
b
y] increaseinredshiftwouldbringthe derivedparameterscloser
[JHz 1 together.
G
S22 0.5 Our assumption of the redshiftof z 0.3 is based on the
≥
non-detection of an underlying host galaxy by Wagneretal.
0
1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 (1996). BLLac host galaxies have typical absolute magni-
epoch
tudes of M 24 and an effective radius of r 10kpc
R c
≈ − ≈
Fig.9.Top:Positionangleofthecentral1masoftheVLBIjet (Jannuzietal. 1997 and ref. therein). The lower limit of
groupedinoneyeartimebins.Thedashedlineisasinusoidalfit Wagner et al. correspondsto an absolute magnitudeof M =
R
tothedata.Thepositionangleoftheinnerjet(ejectionangle) 20.4,whichwouldbeanunusuallyfaintBLLachostandso
−
varies with a period of 7.4 years. Bottom: One year bins is a veryconservativelower limit. An upperlimit forthe red-
∼
of the 22GHz core flux density. The dashed line represents a shift can be given from the lack of absorption by foreground
sinusoidalfitwithafixedperiodof7.4years. galaxies, which results in a redshift of z 0.5 (Wagner priv.
≤
comm.).AtpresentitisunclearifthetentativeX-raydetection
ofaspectrallinenear5.8keV(Kadleretal.2004)canbeiden-
tified with the FeK line. If so, the redshiftof the line would
Asinusoidalfittothedatarevealsaperiodof(7.4 1.5)yr. α
± indicate a distance of z = 0.10 0.04, much closer than ex-
Thefityieldsareducedχ2 of1.55,whichismuchbetterthan ±
r pected from the optical measurements. Possible ways out of
theχ2 of4.28,obtainedforalinearfit.Unfortunatelyourdata
r thisdifficultycouldinvokeablueshiftofthelineorsomeex-
coveronlyoneperiodofthevariabilityanditwillbeinterest-
ternalemissionprocessnotrelatedtothenucleusof0716+71.
ingtoseeifthetrendholdsinthefuture.Apossibleexplanation
AsolidconfirmationoftheX-raylineinfutureobservationsis
forsuchperiodicvariationwouldbeaprecessingfootpointof
urgentlyneeded.
the jet. This should also result in a flux density variation of
the VLBI core, and indeed, there is a tendency that the core
fluxdensityat22GHzisconsistentwithaperiodicvariationof 4. Conclusions
7yr(seeFig.9,bottompanel).RecentlyRaiterietal.(2003)
∼ We analysed 26 epochs of VLBI data at 4.9GHz, 8.4GHz,
publishedananalysisoftheopticalandradiofluxdensityvari-
15.3GHz, and 22.2GHz observed over 10yr, between 1992
abilityof0716+714andfoundaperiodicityof5.5yrto6.0yr
and 2001,and deriveda new kinematicscenario forthe jetin
fortheflux-densityvariationsintheradioregime.Thisiscom-
0716+714.In this scenariothe componentsmovewith appar-
patiblewithintheerrorbarswiththeperiodicityoftheP.A.in
entspeedsof 4.5c to16.1c (z = 0.3).Thesespeedsareatyp-
theinnerregionofthejet.Basedonthegeometricalconsider-
ically fast for BLLac objects, which typically exhibit speeds
ationsinSect.3.3,theapparentpeaktopeakvariationof 7
≈ ◦ of 5c(e.g.,Vermeulen&Cohen1994;Gabuzdaetal.2000;
of the ejection angle corresponds to a small change of only ≤
Rosetal. 2002). Since this analysis is based on much denser
7 sinθ 0.6 ofthejetdirection.
◦ ≈ ◦ sampled VLBI data than previous studies, we are convinced
Such small variations of a few degrees in the rest frame thatwecanruleoutthesomewhatslowerscenarioswhichwere
of the source in several years have recently been detected presentedpreviously(e.g.,Witzeletal. 1988; Schalinskietal.
also in other AGN, e.g. 3C345 (Birettaetal. 1986), 3C279 1992;Gabuzdaetal.1998).Wefindaperiodicvariationofthe
(Carraraetal. 1993), 3C273 (Krichbaumetal. 2001) and P.A. of the innermostjet which could be a sign of jet preces-
BLLac (Stirlingetal. 2003). The most common explanation sionona time-scaleof(7.4 1.5)yr.Ifthiscanbeconfirmed
±
is a precessing jet, which could be caused by a binary super- infutureobservationsthehigherspeedsfoundbyJorstadetal.
massive black hole system (Birettaetal. 1986; Hummeletal. (2001b) might be explained by non-linear motion of the jet
1992) or a warped accretion disc (Lai 2003 and ref. therein). components.
However, hydrodynamiceffects also can lead to a precession No correlation was found between the component ejec-
ofthejet(e.g.,Hughesetal.2002).Inthiscasetheprecession tionandradiofluxdensityflaresatthecm-wavelengths.There
isintroducedbyapotentiallystrongobliqueinternalshockthat seems to be a weak correlation between the ejection of new
arisesfromasymmetricperturbationoftheflow. componentsand the flattening of the radio spectrum. For one
10 U.Bachetal.:KinematicstudyoftheblazarS50716+714
of the new components (C6) we could find a reported γ-ray Hartman,R.C.,Bertsch,D.L.,Bloom,S.D.,etal.1999,ApJS,
flare andif the ejectionandthe flare are really connectedthis 123,79
wouldsupporttheideathattheγ-rayemissionoriginatesfrom Heeschen,D.S.,Krichbaum,T.,Schalinski,C.J.,&Witzel,A.
thesameregionfromwhichweobservetheradioemission. 1987,AJ,94,1493
From the componentmotion in the jet, we obtain a lower Ho¨gbom,J.A.1974,A&AS,15,417
limit for the Lorentz factor of 11.6 and a maximum angle to Hughes,P.A.,Miller,M.A.,&Duncan,G.C.2002,ApJ,572,
thelineofsightof4.9 .Toexplainthelargerangeofobserved 713
◦
apparent speeds as an effect of spatial jet bending, a Lorentz Hummel,C.A.,Schalinski,C.J.,Krichbaum,T.P.,etal.1992,
factorofγ>15andaviewingangleoftheVLBIjetofθ<2 A&A,257,489
◦
are more likely. Under these circumstances, the Doppler fac- Jannuzi,B.T.,Yanny,B.,&Impey,C.1997,ApJ,491,146
tor would be δ 20 to 30. Such high Doppler factors are Jorstad,S.G.,Marscher,A.P.,Mattox,J.R.,etal.2001a,ApJ,
≈
indeed required to explain the high apparent brightness tem- 556,738
peratures of up to 1017K derived from intraday variability at —.2001b,ApJS,134,181
cm-wavelengths. Kadler,M.,Kerp,J.,&Krichbaum,T.P.2004,A&A,submit-
ted
Acknowledgements. We thank S. Jorstad and A. Marscher for pro- Kedziora-Chudczer, L., Jauncey, D. L., Wieringa, M. A.,
viding their VLBA data for reanalysis. We also thank the group of Tzioumis,A.K.,&Bignall,H.E.2001,Ap&SS,278,113
theVLBA2cmSurveyandthegroupoftheCJFSurveyforprovid- Kellermann, K. I., Lister, M. L., Homan, D. C., et al. 2004,
ing their data. We appreciate the use of the flux density monitoring ApJ,609,539
datafromtheUMRAOdatabase.Wethanktheanonymousrefereefor Kellermann,K. I.,Vermeulen,R. C., Zensus, J. A., &Cohen,
helpfulcommentsandsuggestions.ThisworkmadeuseoftheVLBA,
M.H.1998,AJ,115,1295
whichisaninstrumentoftheNationalRadioAstronomyObservatory,
Kraus, A., Krichbaum, T. P., Wegner, R., et al. 2003, A&A,
afacilityoftheNationalScienceFoundation,operatedundercooper-
401,161
ativeagreementbyAssociatedUniversities,Inc.,theEuropeanVLBI
Krichbaum, T. P., Britzen, S., Standke, K. J., et al. 1995, in:
Network,whichisajointfacilityofEuropean,Chinese,SouthAfrican
‘QuasarsandActiveGalacticNuclei:HighResolutionRadio
andotherradioastronomyinstitutesfundedbytheirnationalresearch
councils.Thisworkisalsobasedonobservationswiththe100mradio Imaging’, ed. M.H. Cohen and K.I. Kellerman, Proc. Nat.
telescopeoftheMPIfR(Max-Planck-Institutfu¨rRadioastronomie)at Acad.Sci.,92,No.5,11377
Effelsberg. We gratefully acknowledge the VSOP Project, which is Krichbaum, T. P., Graham, D. A., Witzel, A., et al. 2001,
led by the Japanese Institute of Space and Astronautical Science in in ASP Conf. Ser. 250: Particles and Fields in Radio
cooperationwithmanyorganisationsandradiotelescopesaroundthe GalaxiesConference,ed.R.A.Laing&K.M.Blundell(San
world. Francisco,CA:ASP),184
Krichbaum, T. P., Kraus, A., Fuhrmann, L., Cimo`, G., &
Witzel, A. 2002, Publications of the Astronomical Society
References
ofAustralia,19,14
Aller,H.D.,Aller,M.F.,Latimer,G.E.,&Hughes,P.A.2003, Lai,D.2003,ApJ,591,L119
AmericanAstronomicalSocietyMeeting,202,18.01 Lobanov,A.P.1998,A&A,330,79
Biretta, J. A., Moore,R. L.,& Cohen,M. H. 1986,ApJ,308, Marscher,A.P.1983,ApJ,264,296
93 Marscher,A.P.&Gear,W.K.1985,ApJ,298,114
Blandford,R.D.&Ko¨nigl,A.1979,ApJ,232,34 Mattox, J. R., Schachter, J., Molnar, L., Hartman, R. C., &
Britzen, S., Vermeulen, R. C., Taylor, G. B., et al. 1999, in Patnaik,A.R.1997,ApJ,481,95
ASPConf.Ser.159:BLLacPhenomenon,ed.L.Takalo& Otterbein,K.,Krichbaum,T.P., Kraus,A., etal. 1998,A&A,
A.Sillanpa¨a¨(SanFrancisco,CA:ASP),431 334,489
Camenzind,M.&Krockenberger,M.1992,A&A,255,59 Pacholczyk, A. G. 1970, Radio astrophysics (San Francisco:
Carrara, E. A., Abraham, Z., Unwin, S. C., & Zensus, J. A. Freeman),97
1993,A&A,279,83 Pearson,T.J.&Readhead,A.C.S.1984,ARA&A,22,97
Dermer,C.D.&Schlickeiser,R.1994,ApJS,90,945 Pearson, T. J. & Zensus, J. A. 1987, in Superluminal Radio
Eckart,A.,Witzel,A.,Biermann,P.,etal.1986,A&A,168,17 Sources, ed. J. A. Zensus & T. J. Pearson (San Francisco,
—.1987,A&AS,67,121 CA:ASP),1
Fey,A.L.&Charlot,P.2000,ApJS,128,17 Pen,U.-L.1999,ApJS,120,49
Fomalont,E.B.1989,inASPConf.Ser.6:SynthesisImaging Peng, B., Kraus, A., Krichbaum, T. P., & Witzel, A. 2000,
inRadioAstronomy,ed.R.A.Perley,F.R.Schwab,&A.H. A&AS,145,1
Bridle(SanFrancisco,CA:ASP),213 Perez-Torres, M., Marcaide, J., Guirado, J., & Ros, E. 2004,
Gabuzda,D.C., Kovalev,Y.Y.,Krichbaum,T.P.,etal.1998, A&A,inpress
A&A,333,445 Qian, S., Li, X., Wegner, R., Witzel, A., & Krichbaum, T. P.
Gabuzda, D. C., Pushkarev, A. B., & Cawthorne, T. V. 2000, 1996,ChineseAstronomyandAstrophysics,20,15
MNRAS,319,1109 Qian,S.J.,Witzel,A.,Kraus,A.,Krichbaum,T.P.,&Zensus,
Ghisellini,G., Padovani,P.,Celotti, A.,& Maraschi,L. 1993, J.A.2001,A&A,367,770
ApJ,407,65