Coexistence of a non thermal jet and a complex ultra fast X-ray outflow in a moderately luminous AGN

0.3 MB·

by  M. Giroletti

Astronomy&Astrophysicsmanuscriptno.IRAS17020˙radio˙arxiv (cid:13)c ESO2017
January26,2017
Coexistence of a non thermal jet and a complex ultra fast X-ray
outflow in a moderately luminous AGN
M.Giroletti1,⋆,F.Panessa2,A.L.Longinotti3,,Y.Krongold4,M.Guainazzi5,6,E.Costantini7,M.Santos-Lleo5
1 INAFIstitutodiRadioastronomia,viaGobetti101,40129Bologna,Italy
2 INAFIstitutodiAstrofisicaePlanetologiaSpazialidiRoma,ViadelFossodelCavaliere100,00133Roma,Italy
3 CONACYT-InstitutoNacionaldeAstrof´ısica,O´pticayElectro´nica,LuisE.Erro1,Tonantzintla,Puebla,Me´xico,C.P.72840
4 InstitutodeAstronomia,UniversidadNacionalAutonomadeMexico,ApartadoPostal70264,04510MexicoD.F.,Mexico
7
5 ESAC,P.O.Box,78E-28691VillanuevadelaCan˜ada,Madrid,Spain
1
6 InstituteofSpaceandAstronauticalScience,3-1-1Yoshinodai,Chuo-ku,Sagamihara,Kanagawa,Japan
0
7 SRONNetherlandsInstituteforSpaceResearch,Sorbonnelaan2,3584CAUtrecht,TheNetherlands
2

n Received;accepted
a
J ABSTRACT

5
Context.RecentXMM-NewtonobservationshaverevealedthatIRAS17020+4544 isaveryunusual exampleofblackholewind-
2
producedfeedbackbyamoderatelyluminousAGNinaspiralgalaxy.

Aims.Sincethesourceisknownforbeingaradioemitter,weinvestigatedaboutthepresenceandthepropertiesofanon-thermal
]
A component.
Methods.WeobservedIRAS17020+4544withtheVeryLongBaselineArrayat5,8,15,and24GHzwithinamonthofthe2014
G
XMM-Newtonobservations.Wefurtheranalysedarchivaldatatakenin2000and2012.
. Results.Wedetectthesourceat5GHzandonshortbaselinesat8GHz.At15and24GHz,thesourceisbelowourbaselinesensitivity
h
forfringefitting,indicatingthelackofprominentcompactfeatures.Themorphologyisthatofanasymmetricdouble,withsignificant
p
diffuseemission.Thespectrumbetween5and8GHzisrathersteep(S(ν)∼ν−(1.0±0.2)).Ourre-analysisofthearchivaldataat5and8
-
o GHzprovidesresultsconsistentwiththenewobservations,suggestingthatfluxdensityandstructuralvariabilityarenotimportantin
r thissource.Weputalimitontheseparationspeedbetweenthemaincomponentsof<0.06c.
t Conclusions.IRAS17020+4544showsinterestingfeaturesofseveralclassesofobjects:itspropertiesaretypicalofcompactsteep
s
a spectrumsources,lowpowercompactsources,radio-emittingnarrowlineSeyfert1galaxies.However,itcannotbeclassifiedinany
[ ofthesecategories,remainingsofaraone-of-a-kindobject.

1 Keywords.galaxies:active–galaxies:Seyfert–galaxies:nuclei–galaxies:jets
v
8
9 1. Introduction spiral galaxy with black hole mass of M ∼ 5.9 × 106M
2 BH ⊙
(Wang&Lu 2001). We have observed it in January 2014 with
7 Ionized gas outflows from the nuclei of active galaxies (active
XMM-Newton,obtainingahigh-resolutionX-rayspectrum.Our
0 galactic nuclei, hereafter AGN) are a relatively common phe-
observations reveal a series of absorption lines corresponding
.
1 nomenon. They are of great interest since they transport mass to at least 5 outflowing componentswith velocity in the range
0 andenergytothehostgalaxyinaprocessknownasAGNfeed-
of 23,000-33,000km/s (Longinottietal. 2015). The energetics
7 back, which has fundamental impact on galaxy properties and
of the wind indicate that it might be able to produce feedback
1 evolution(DiMatteoetal.2005).Inthiscontext,ultra-fastout-
onthehostgalaxy.Evidenceforsuchphenomenonisalso pro-
: flows (UFOs, see e.g. Chartasetal. 2002; Poundsetal. 2003)
v videdbythepresenceoflargescalemoleculargasoutflowingat
i are particularly relevant because of their mass outflow rate lowvelocitythatisassociatedtothefastX-raywind(Longinotti
X (0.01−1M yr−1) and kinetic energy(1042−1045 erg s−1). A
⊙ etal. submitted),as predictedbyseveralfeedbackmodels(e.g.
r much less common phenomenon (less than 10% of AGNs) is
a Faucher-Giguere& Quataert 2012, Zubovas& King 2012).At
the presenceof a pair ofrelativistic jets ofnon-thermalplasma
the same time, the presence of brightand compactradio emis-
emerging from the central nucleus, which are generally de-
sionindicatesthatthissourceisaradioloud(RL)NLS1andthat
tectedin theformofcollimatedstructuresatradiowavelength.
italsohasaradiojet(Doietal.2007,2011;Gu&Chen2010).
In recent years, several works have studied the coexistence of
Inthepresentpaper,wepresentnewhighangularresolution,
fast outflows and relativistic jets and their possible interplay
multi-frequencyVeryLongBaselineInterferometry(VLBI)ra-
(see Tombesietal. 2014, and references therein). A connec-
dio observationsobtainednear simultaneouslywith the XMM-
tionbetweenthetwophenomenaisalsopredictedbymagneto-
Newton observations. We further report on the analysis of
hydrodynamic (MHD) models (e.g. Tchekhovskoyetal. 2011;
archivalVLBIdatasets,toprovideconstraintsontheproperties
Fukumuraetal.2014).
oftheejectionprocessinthissource.
IRAS17020+4544(locatedatR.A.17h03m30.3830s,Dec.
+45◦ 40′ 47′′.167,z = 0.0604)isanidealtargettofurtherex- The paper is organized as follows: we describe our new
observations and archival data analysis in Sect. 2; we present
plore this subject. It is a bright Narrow Line Seyfert 1 (NLS1)
the results in Sect. 3 and a review of the radio properties of
⋆ Email:[email protected] IRAS17020+4544fromtheliteratureinSect.4;finally,wedis-
1
M.Girolettietal.:RadiopropertiesofIRAS17020+4544
10 10
E B W
5 5
s s
d d
n n
o o
c c
se 0 se 0
c c
Ar Ar
Milli -5 Milli -5
-10 -10
15 10 5 0 -5 -10 -15 15 10 5 0 -5 -10 -15
MilliArc seconds MilliArc seconds
Fig.1.VLBAimagesofIRAS17020+4544on2014February12at5GHz(lefthandsidepanel)and8.4GHz(rightpanel),with
contourstracedat(−1,1,2,4,...,32)×the ∼ 3σnoiselevel,whichis 0.2and 0.6mJybeam−1 at5 and8GHz, respectively.The
halfpeakbeamwidthisshownintheupperleftcornerofeachpanel,andis4.6mas×1.4masinpositionangle(PA)−19◦and8.0
mas×3.4masinPA19◦at5and8GHz,respectively.Notetheunusuallylargerbeamat8GHz,duetothelackoffringedetections
onthelongestbaselinesatthisfrequency.Inthe5GHzimage,model-fitellipticalGaussiancomponentsareshownascrossesand
labelledE,B,W,fromeasttowest.
cuss them in Sect. 5. Throughout the paper, we use a ΛCDM 2.2.Archivaldata
cosmology with h = 0.705, Ω = 0.27, and Ω = 0.73
m Λ
2.2.1. 2000June&Augustdata
(Komatsuetal. 2009). With these values, the redshift of the
source correspondsto a luminositydistance of D = 269 Mpc
L The peak brightness and total flux density at 5 GHz obtained
and to an angular scale of 1mas=1.16pc; proper motion of 1 in our observationsare ∼ 5 times lower than those reported at
mas/yr correspondsto an apparent speed of 4.0c. The spectral
thesamefrequencybyGu&Chen(2010).Consideringthenon-
indexαisdefinedsuchthatS(ν)∼ν−αandpositionangles(P.A.)
detections at higher frequency, we are confident that the error
aredefinedpositivenorththrougheast.
onourmeasurementis notlargerthanthe typical,conservative
10% calibration uncertainty. Therefore, in order to investigate
possibleamplitudeandstructuralvariabilitywedownloadedand
re-analysedtheVLBAdatasetspresentedbyGu&Chen(2010).
2. Observationsanddatareduction ThesedatasetsareBM033DandBM033E,takenon2000June
16andAugust21,respectively.Bothdatasetswereobtainedwith
2.1.Newdata 2IFsof8MHzbandwidtheach,indualpolarization,foratotal
datarateof128Mbps.Thedelay,rate,andphasesolutionswere
We observed IRAS 17020+4544 with the NRAO Very Long determinedonthenearbycalibratorJ1713+4916andtransferred
BaselineArray(VLBA)on2014February12,20dayslaterthan tothetargetsource;otherstepsofthecalibrationwereidentical
theXMM-Newtonpointing.Weobservedthesourceat5,8,15, tothosedescribedforthenewobservations.Thetotalintegration
and24GHz,foranetintegrationtimeof10,20,60,and60min- timeonIRAS17020+4544wasabout100minutesperobserva-
utes, respectively. We recorded data with the new 2 Gbps rate tion.
offeredbytheVLBA,correspondingto8basebandchannelsof The images at the two 2000 epochs are consistent in terms
32MHzbandwidtheach,infullpolarization.Basedonthehigh of intensity and structure, so we combined the two visibility
sensitivityprovidedbythisobservingmodeandonthefluxden- datasetstofurtherimprovetheimagesensitivityandfidelity.We
sitypublishedbyGu&Chen(2010),wedidnotobserveaphase thenproducedafinalimagewithanadditionaliterationofphase
calibratorforphasereferencing. onlyandphaseandamplitudeself-calibration.
We carried out the data reduction following the usual pro-
cedures in AIPS, applying calibrations for the total electronic
2.2.2. 2012Jan&Febdata
contentoftheionosphereatν≤8GHz,theearthorientationpa-
rameters, the correlator digital sampling, the combined system We also found and downloaded two five minute scans on
temperature and gain curves, the parallactic angle, and the in- IRAS 17020+4544obtained in the context of a large observa-
strumentalsingle banddelays.We then fringefitted the data in tional VLBA project at 8 GHz. These observations took place
orderto removeresidualdelays, ratesand phasevariations;we on2012Jan12(VLBAexperimentcode:BC196ZQ)and2012
used as input modelan image producedfrom the same dataset Feb 8 (code BC201AC). In both experiments, 8 channels with
analysedbyGu&Chen(2010)butre-analysedbyourselves(see 16MHz bandwidtheach andsingle polarizationwererecorded
Sect. 2.2). At 5 GHz, we foundsolutions for most stations (all andcorrelated.Thesourcewasbracketedby1minutescanson
but Brewster); at 8 GHz we found solutions only for stations 1705+456,whichweusedtodeterminedelay,rate,andphaseso-
formingrelativelyshortbaselines;at15and24GHz,we could lutions.HancockandSaint-Croixdidnotproducedataineither
not find any solution. For the 5 and 8 GHz datasets, we then run because the calibrator was not detected (probablyresolved
proceededwiththe standarditerationsofhybridmapping,with out) on baselines to these stations. After the standard calibra-
phaseonlyself-calibrationatfirstandphaseandamplitudeself- tion, we split the data withoutaveragingin time nor frequency
calibrationinthefinalcycles. (otherthanwithineachoftheeightIFs),sincethephasetracking
2
M.Girolettietal.:RadiopropertiesofIRAS17020+4544
-2 -1 0
10
10 E A1 A2 W
5
s
d
5 on
nds sec 0
Arc seco 0 MilliArc -5
Milli -5
-10
-10
15 10 5 0 -5 -10 -15
MilliArc seconds
15 10 5 0 -5 -10 -15
MilliArc seconds
Fig.3. VLBA image of IRAS 17020+4544 on the combined
datasetfrom2000JuneandAugustobservationsat5GHz,with
Fig.2. Simultaneous 5-8 GHz spectral index VLBA image of
contours traced at (−1,1,2,4,...,32)× the ∼ 3σ noise level,
IRAS17020+4544on2014Feb12.Contoursshowtotalinten-
whichis0.27mJybeam−1.Thehalfpeakbeamwidthisshown
sityat5GHz,convolvedwiththe8GHzrestoringbeam;colours
in the upper left corner of each panel and it is 3.8 mas × 1.2
showthespectralindexvalues.
masinPA−21◦.Crossesandlettersindicateandlabelmodel-fit
ellipticalGaussiancomponents.
positionwasseveralhundredsmilliarcsecondfromtheposition
ofthe sourcedeterminedinthe otherphasereferencingexperi-
ment.Inthisway,wewereabletoproduceimagesnotaffected Theimagethenshowsasteepspectrumintegratedovertheeast-
bysmearingandindeedwedetectedthesourceinbothdatasets. erncomponentandaflatteroneinthewesternside.However,the
uncertaintyonthewesternside spectralindexislarger,andthe
easterncomponentiscertainlyaffectedbythepresenceofdiffuse
3. Results
emission,whichlikelybiastheresultstosteepervalues.Indeed,
if we only take the peak brightness of the component we find
3.1.Newdata
similar values, indicating a flat spectrum for the most compact
WeshowinFig.1thefinalimagesat5and8GHz.At5GHz,the emissionregion.Consideringtheemissiondetectedintheentire
sourceis extendedin east-westdirection,with a totalflux den- VLBIimages,thesimultaneousspectralindexisα=1.0±0.4.
sity of about23 mJy. A compactcomponentwith peak bright- Finally,weestimatetheupperlimittothebrightnessofany
ness8.9mJy beam−1issurroundedbydiffuseemissiononboth compact structure, based on the non detection of the source at
sides,brighterandmoreextendedonthewesternend.At8GHz, 24and15GHz,andonthelongestbaselinesat8GHz.The5σ
the overall morphologyis similar. However, due to the lack of VLBAsensitivityforasinglebaseline,channel,andpolarization
visibilities on the longestbaselines the final sensitivity and an- at8,15,and24GHzis12,28,and33mJybeam−1,respectively.
gularresolutionarepoorerthanat5GHz.Thesourcecanthenbe This assumes 10, 5, and 5 minutes integration times, respec-
mostbasicallydescribedasacompactasymmetricdouble,with tively,forastructurewithangularsizeof4masat8GHz(based
abrightereasterncomponent(9.0mJy beam−1peakbrightness) on the non detection on baselines longer than 45 Mλ), 17 mas
anda weaker,somewhatmorediffusewesternfeature.Thereis at 15 GHz and 11 mas at 24 GHz (based on the non detection
clearlyalso some extendedemission bridgingthe two features. on any baselines in the 15 and 24 GHz datasets). These corre-
Thetotalfluxdensityat8GHzis13mJy. spondtoupperlimitsonthebrightnesstemperatureof2.2×107
We carried out a model-fit in the image plane, in order to K,1.5×107K,and6.7×106K,atthethreefrequencies,respec-
measurethebrightnesstemperatureofthemaincomponent,i.e. tively.
theeasterncomponentlabelledE.Thebrightnesstemperatureis
obtainedfromthefollowingformula:
3.2.Archivaldata
1 c 2 S(1+z)
T = We show in Fig. 3 the image obtained from the reanalysis of
b 2k (cid:18)ν(cid:19) πab
B thecombineddatasetobtainedfromthe2000observations.The
wherek is theBoltzmannconstant,cis thespeedoflight, morphologyisoverallconsistentwithbothournewobservations
B
ν is the observing frequency, S is the total flux density of the and the images reported by Gu&Chen (2010): the brightest
component, a and b are the major and minor semi-axis of the componentis located at the eastern end, surroundedby diffuse
component.Inourcase,ν = 5.0GHz,z = 0.0604and,forcom- emission more extended in the western direction. Local peaks
ponentE,S = 14mJy,a = 2.5mas,b = 1.0mas.Theresulting are embeddedin this emission, with a final componentlocated
brightnesstemperatureisT =1.0×108K. about9maswestoftheimagepeak.Thetotalfluxdensityand
b
We show in Fig. 2 the first quasi-simultaneousspatially re- the peak brightness recovered (S ∼ 25 mJy and P ∼ 13
5 5
solvedspectralindeximageofthesource,basedonour5and8 mJybeam−1,respectively)areinmuchcloseragreementwiththe
GHzdata.Beforecombiningtheimages,wehaveconvolvedthe 2014datathanwiththevaluesreportedbyGu&Chen(2010).
5GHz imagewith thesame restoringbeamofthe8 GHzdata, We triedtoinvestigateontheoriginofthisdiscrepancybut
andclippedallthepixelsbelowthe3σnoiselevelineitherim- could not come up with a final explanation. We determined a
age.Asaresultofthisconvolution,thecompacteasternfeature fluxdensityof∼80mJyforthephasecalibrator,whichislower
becomes“contaminated”bysomeadditionalextendedemission. thanthe2.3and8GHzvaluesreportedintheVLBAcalibrator
3
M.Girolettietal.:RadiopropertiesofIRAS17020+4544
6 10 20 30 40 50
4
100
ds 2
n
o 50
c sec 0 onds
Ar ec 0
Milli -2 Arc s -50
-4
-100
-6
8 6 4 2 0 -2 -4 -6 -8
MilliArc seconds 200 150 100 50 0 -50 -100 -150 -200
Arc seconds
Fig.4.VLBA8GHzimageofIRAS17020+4544on2012Feb
8, with contours traced at (−1,1,2,4)× the ∼ 3σ noise level, Fig.5.OverlayoftheNVSS(contours)andFIRST(greyscale)
whichis1.0mJybeam−1.Thehalfpeakbeamwidthisshownin 1.4GHz VLA images. The NVSSHPBW is 45′′ and is shown
theupperleftcornerofeachpanelanditis2.8mas×1.1masin in the bottom left corner; the FIRST HPBW is 4.5′′. Contours
PA18◦. are traced at (−1,1,2,4,...) × 1.3 mJybeam−1; the grey scale
isshowninthetopwedgeandspanstherangebetween0.4and
55mJybeam−1.Thehalfpeakbeamwidthisshownintheupper
Epoch Component x y S5 leftcornerofeachpanelanditis2.8mas×1.1masinPA18◦.
(year) (mas) (mas) (mJy)
2000 E 0.0 0.0 17.9
A1 −1.3 −0.13 3.5 3.3.Modelfitsandpropermotion
A2 −3.9 −0.6 2.4
W −9.4 −0.9 2.0 We report in Table 1 the results of a model-fit in the image
2014 E 0.08 −0.05 14.1 planewithellipticalGaussiancomponentstothe2000and2014
B −3.1 −1.9 5.2 5 GHz images. We use four components to describe the 2000
W −9.1 −0.8 1.5 data (which have better coverage of the (u,v)-plane) and three
componentsforthe2014data.Inanycase,themainfeaturesat
Table 1. Results of model-fit to the 5 GHz images of thetwoedgesofthestructurecanbeeasilyidentified.Theirposi-
IRAS17020+4544.The coordinatesare referencedto the peak tionsareconsistenttobetterthanonebeamandthefluxdensities
position of the brightest componentin the 2000 combined im- arealsoconsistentwithinthecalibrationuncertainty.
ages. TheseparationbetweenE andW haschangedfrom9.4mas
in2002to9.2masin2014;sincethedifferenceislessthan1/5of
thebeam,weclaimthatbothcomponentsarestationary,andthat
theirseparationvelocitymustbesmallerthan0.015masyr−1,or
0.06c.
list(148mJyonshortspacingsand117mJyonlongspacingsat
2.3GHz,217and182mJyat8GHz);thismighthaveprompted 4. RadiopropertiesofIRAS 17020+4544
Gu&Chen(2010)torescaletheamplitudesupbyafactorofa
IRAS 17020+4544 is also known as B31702+457, since it
few,resultinginthehighfluxdensityforthetarget.However,the
was detected in the B3 survey at 408 MHz with the Bologna
fluxdensityof3C345observedinthesameexperimentdidnot
Northern Cross (Ficarraetal. 1985). As it is a B3 source, it is
revealanysignofbeingunderestimated,andlater5GHzobser-
quitebrightandindeeditisdetectedinmanyothersurveys:the
vations of J1713+4916found flux density values closer to our
6th Cambridgesurvey of radio sources(6C, Halesetal. 1988),
measurement;consideringalsoafairamountofvariabilitytypi-
the Texas survey of radio sources (Douglasetal. 1996), and
calforcompactsources,wedonotseeanyreasontomodifythe
theWesterborkNorthernSkySurvey(WENSS,Rengelinketal.
amplitudescaleforthe2000observationsofIRAS17020+4544.
1997) at low frequency; the NRAO VLA Sky Survey (NVSS,
Inthe20128GHzexperiments,wedetectaslightlyresolved Condonetal. 1998) and the Faint Images of the Radio Sky at
component,withanextensiontothenorth-westernsideandfaint Twenty Centimeters (FIRST, Beckeretal. 1995) at 1.4 GHz;
diffuse emission to the west (see Fig. 4). The peak brightness the 87GB and GB6 at 5 GHz (Gregory&Condon 1991;
is 7.4 and 5.9 mJybeam−1 in 2012 Jan and Feb, respectively. Gregoryetal. 1996). From its flux density and luminosity dis-
Giventheveryshortobservationtime, we cannotmakestrong tance, the radio power of IRAS 17020+4544 is P =
150MHz
claimsabouttheimagefidelityandanysignificantvariability.In 3.2 × 1024 WHz−1(νL = 4.8 × 1039 ergs−1) and P =
ν 1.4GHz
anycase, the totalcleanedflux densityis consistentamongthe 1.0×1024 WHz−1(νL = 1.5×1040 ergs−1),calculatedat150
ν
epochswithavalueofS8 = 11.5mJy,alsosimilartotheresult MHzfromthe6Candat1.4GHzfromtheNVSSvalues,respec-
ofour2014observations. tively.
Theastrometryofthe2000and2012observationsisnoten- The FIRST image (shown with grey tones in Fig. 5) re-
tirelyconsistent,witha ∼ 10masoffsetbetweenthe corepeak veals one main, bright, compact component, unresolved at the
at the two frequencies.The 8 GHz observationpositionsare in ∼ 5′′ angular scale, and a secondary, fainter component lo-
goodagreement(within<1mas)withthosereportedatthesame cated at 1.2′ (81 kpc projected at the IRAS 17020+4544 red-
frequencybyDoietal.(2007). shift), in PA −52◦, which is also visible in the NVSS image
4
M.Girolettietal.:RadiopropertiesofIRAS17020+4544
(contours in Fig. 5) as an extension of the main component. ratioofFIRST1.4GHzradiotoSDSSg-band.Suchvaluesputs
ThefluxdensityintheNVSSisonlyslightlylargerthaninthe IRAS17020+4544firmlyintheradioloudregime.
FIRST (121vs118mJy),indicatingthatthesourceiscompact Athighangularresolution,wecancompareournewandre-
at this angular resolution; the total flux density in the NVSS processedVLBA data with otherVLBI observations,in partic-
might also be slightly affected by the secondary component, ular the 8.4 GHz Japanese VLBI Network (JVN) observations
which is only partly resolved at the 45′′ angular resolution of byDoietal.(2007)andthe1.6VLBAdatabyDoietal.(2011).
the survey. The two components are in any case well sepa- Thesetwodatasetsshowafairlyresolvedmorphology,whichac-
ratedintheFIRST,whichcanbeusedtodeterminetheirbright- countsforoveronehalfofthetotalfluxdensitymeasuredatthe
ness ratio (118 : 6, or ∼ 20). We are thus confident that any samefrequencyby,e.g.,theVLA.Inparticular,S =79.9±4.0
1.6
flux density measured at lower resolution is only slightly af- mJyandS = 18.5±2.6mJy.Theresultingnonsimultaneous
8.4
fected by the presence of this component. A visual inspection spectralindexisα= 0.84±0.21,consistentwiththeonefound
of the SDSS images doesnot revealany obviousopticalcoun- with the simultaneousdata. A furthernon detectionat 22 GHz
terpart,soitremainsuncertainwhetherthisisabackgroundob- atthe < 9 mJy level,entirelyconsistentwith ourresult, is also
ject or a large-scale feature associated to IRAS 17020+4544 reportedbyDoietal.(2016).
itself. We note that kiloparsec scale emission is not uncom-
mon in NLS1s, having been discovered in 10 objects so far
(Anto´netal. 2008; Doietal. 2012, 2015; Gliozzietal. 2010;
5. Discussion
Richards&Lister 2015; Whalenetal. 2006). However, with ∼
80 kpc (projected) linear size, IRAS 17020+4544 would be Basedontheinformationgatheredbythenew,thearchival,and
the largest NLS1 known so far. If the third weaker component the literature data on IRAS 17020+4544, we determine that it
located further north-west were also physically associated to isajettednon-thermalsource.Thesteepoverallspectrumon∼
IRAS 17020+4544, the size would become even larger. Deep, kpc(α = 0.9,Snellenetal.2004)and∼pc(α = 1.0,thiswork)
lowfrequencyobservationswith afew arcsecondangularreso- scalesindicatesthattheradioemissionhasasynchrotronorigin.
lutionwillbenecessarytoclarifythisissue. Thetotalradioluminosity(P = 1.0×1024 WHz−1)and,
1.4GHz
Inadditiontothese1.4GHzdata,Snellenetal.(2004)con- aboveall,thebrightnesstemperatureoftheVLBImaincompo-
sidered a numberof additionalmulti-frequencyinterferometric nent(Tb =1.0×108K)arefartoolargeforstarburstprocesses,
and single dish datasets between 150 MHz and 22 GHz. They andconstraintheradioemissiontohavepredominantlyanuclear
reporta spectral index of α = −0.88, slightly flattening at low origin.The differencebetween the VLBI and total flux density
frequency,andprovideconstraintsontheturnoverfrequencyand revealsthepresenceofradioemissiononpc-to-kpcscales,most
peakfluxdensityofν < 150MHzandS > 380mJy,re- likely fuelled by the central region. Taken together, these ele-
peak peak
spectively.Alowfrequencyturnoverisconfirmedbytheflatter mentsindicatethattheblackholeinIRAS17020+4544ispow-
(α=−0.23)spectralindexbetween325and1400MHzreported eringanon-thermalplasmajet,inadditiontotheX-rayoutflow
by Massaroetal. (2014) and the non detection at 74 MHz in ofgasstudiedbyLonginottietal.(2015).
the VLA Low-Frequency Survey (VLSS, whose typical point- Fromthisstartingpoint,we can thencantryto classify the
sourcedetectionlimit is0.7Jy beam−1, Cohenetal. 2007). On sourceinafewdifferentways,complementaryratherthancon-
the basis of its spectrum and linear size, Snellenetal. (2004) flicting.AsalreadyproposedbySnellenetal. (2004), thecom-
classifythesourceasacandidatelowpowercompactsteepspec- pact structure and steep integrated radio spectrum suggest that
trum(CSS)sourceandincludeitinthe“compactradiosources IRAS17020+4544isacandidateCSSsource.Differentlyfrom
atlowredshift”(CORALZ)sample1. compactflat-spectrumsources,suchasblazars,CSSsourcesare
To construct their spectra, Snellenetal. (2004) considered orientedontheplaneofthesky,sotheircompactnessisintrin-
data taken at the same frequencyin differentepochs, thus pro- sic rather than due to projection effects, as highlighted by the
vidingusalsovariabilityinformation:onlysmalldeviationsare lackofaprominentcompactflat-spectrumcore.CSSsourcesare
present,suggestingthatvariabilityintheradioisnotsignificant; thoughttobetheearlystagesofevolutionofthemoreextended,
ifwetakeintoaccountpossiblecalibrationuncertaintiesanddif- classical FR2 radio galaxies, as indicated by their young kine-
ferentangularresolutioninthevariousobservations,theamount maticandspectralageestimates(Fantietal.1995;Murgiaetal.
of real source variability has to be very limited, if any. In this 1999;Giroletti&Polatidis2009).Theparsecscalestructurepre-
sense,themostremarkablefindingisthecomparisonofthesin- sented in this work is in agreement with the lack of a promi-
gledish(87GBandGB6)andinterferometric(VLA)5GHzflux nent core and of strong beaming effects; however, the absence
density values,with S = 26±6 mJy, S = 29±4 mJy, of componentmotionovera 14 yearstime scale doesnotindi-
87GB GB6
andS = 40±4mJy.Itisnotclearhowreliablearethesin- cate that this source is in a state of rapid growth,which would
VLA
gledishmeasurements,sincetheyarenearthefluxdensitylimit be typical of CSS sources. Moreover, the radio luminosity of
of the catalogs(25 and 18 mJy for the 87GB and the GB6, re- IRAS 17020+4544 is also significantly (∼ 10−4×) lower than
spectively).Asthe87GBdatawerereportedbymanyauthorsto thatofthearchetypicalCSSsources,suggestingthatitcouldlie
estimatetheradioloudness(R),iftherealfluxdensityislarger inadifferentevolutionarypath.
than26mJy,thenalsotherealRcouldbeaccordinglyhigherby Girolettietal.(2005)andKunert-Bajraszewskaetal.(2010)
∆logR ∼ +0.2.On theissue ofradioloudness,itisinteresting haveindeedintroducedtheclassoflow-powercompact(LPC)or
to note that the values of R reported in the literature vary sig- low-luminosity compact (LLC) objects, which similarly to the
nificantlyindifferentworks,typicallybasedontheassumptions CSS sources are thought to be intrinsically compact, yet they
on the opticalspectrumused to estimate the B magnitude.The presentlowerluminosities(P <1026WHz−1)andslower
1.4GHz
mostrecentvalueislogR=2.17(Doietal.2016),basedonthe or null advance velocities. These sources might be short-lived
objectsortheprogenitorsoflowluminosityFR1radiosources.
1 However,laterhighangularresolutionobservationsofthefullsam- In terms of radio properties (morphology, spectrum, power),
ple(deVriesetal.2009)donotincludethesource,withoutanyreason IRAS17020+4544fitswellwithinthisclass;however,itspecu-
giveninthepaper liaropticalpropertiesdonotmatchthetypicalfeaturesofaFR1
5
M.Girolettietal.:RadiopropertiesofIRAS17020+4544
progenitor:the host is a barred spiral galaxy (Ohtaetal. 2007) IRAS 17020+4544 and the source proximity would in princi-
andtheopticalspectrumcharacterizesthesourceasaNLS1. plepermitustostudythisissuewithadequatelinearresolution.
The presence of radio emission in NLS1 has attracted a However,the radioluminositylogνL = 40.2([ergs−1], calcu-
ν
lot of interest in recent times, especially after the discovery lated at ν = 1.4GHz) is comparatively large in comparison to
of gamma-rayemission from a growing number of RL NLS1s thebolometricluminosity,confirmingthatIRAS17020+4544is
(Abdoetal. 2009a,b; D’Ammandoetal. 2012, 2015). Gamma- aradio-loudsourceandthereforequiteexceptionalwithrespect
rayemissionin NLS1sisascribedto thepresenceofa beamed to the radio-quiet population studied by Zakamska&Greene
relativisticjet,similartowhathappensinblazars,althoughsome (2014).
similarities with CSS sources in spectrum and size have been
claimed, e.g. in PKS 2004-447(Galloetal. 2006; Orientietal.
6. Summary
2015; Schulzetal. 2015) and SDSSJ143244.91+301435.3
(Caccianigaetal. 2014). Evidence for beaming in RL NLS1s – Theradioluminosity(logL = 24.5WHz−1)istypicalof
150
in general (not only in gamma-ray NLS1s) has been found in FR1 radio galaxy, but the structure is compact, i.e. < kpc
several objects, in the form of dominant cores, high bright- scale. As already hinted at by other works (Snellenetal.
ness temperatures (above 1010 K, e.g. Girolettietal. 2011; 2004),itmeetsthe criteriafora lowpowercompactsource
Wajimaetal. 2014; Guetal. 2015), and occasionally the pres- (LPC,Girolettietal.2005).
enceofsuperluminalmotion(D’Ammandoetal.2012).Ourre- – The integrated steep spectrum (α = −0.9) clearly indicates
sultsonthemorphologyandbrightnesstemperatureindicatethat thattheradioemissionhasasynchrotronorigin,i.e.thepres-
IRAS 17020+4544 is different from this sub-set of the NLS1 enceofrelativisticparticlesandmagneticfields.
population;yet,itremainsalsomarkedlyseparatedfromthebulk – On parsec scales, the images at 5 and 8 GHz indicate that
oftheradioquietNLS1population,whichhavemuchlowerra- thesourceisclearlyextendedinanamorphousyetpredomi-
diopowers(aslowas P1.4GHz ∼ 1019 WHz−1)andareseldom nantlylinearstructure,indicatingthepresenceofajetorout-
detectedwithVLBI(Giroletti&Panessa2009;Doietal.2013). flowonscalesofabout10parsecs;alsothenondetectionsat
OneintriguingpossibilityisthatIRAS17020+4544belongsto 15and24GHz,whencomparedwiththeVLAfluxdensities
the elusive population of misaligned RL NLS1s (Bertonetal. at the same frequency, indicate that the source structure is
2015). heavilyresolved.
Fromamulti-wavelengthperspective,IRAS17020+4544is – Within these diffuse emission, there is a main, brighter
alsohighlyremarkableforitsX-rayproperties.WhileitsX-ray component, which one would naturally identify with the
and bolometric luminosities are not exceptional (L0.3−10keV ∼ “core”, which however is not characterized by a classical
1.5×1044 ergs−1, LBOL ∼ 5.2×1044 ergs−1), the recent high flat/invertedspectrum.
spectralresolution(E/∆E ∼ 1000)XMM-Newtonobservations – The lack of a prominentflat spectrum componentsuggests
of Longinottietal. (2015), obtained within one month of our thattheamountofbeaminginanypossiblecoreisnegligible,
VLBAdata,haverevealedauniquesub-relativisticoutflow.The i.e.theplasmadoesnothaveabulkrelativisticmotionorit
spectrum shows a series of absorption lines corresponding to ismovingat a verylargeangleto the line of sight- also in
atleastfiveabsorptioncomponentswith velocitiesin therange agreementwiththelackofvariability.
of 23,000–33,000kms−1; the associated column densities ob- – Thelackofmotionover15yearssuggeststhattheseparation
served in the X-ray wind span from 5×1020 to ∼ 1024 cm−2. speedislowerthan0.2mas/14years,i.e.<0.06c.
The mass and energy output estimated by these authors indi-
catethatthehighercolumndensitywindiscapableofexpelling All these properties locate IRAS 17020+4544 as a unique
sufficient gas to produce feedback onto the host galaxy, even object, not easily classifiable as any standard radio emitting
if the source is ∼ 100 times less luminous than quasars where AGN. In addition IRAS 17020+4544 is an optimal laboratory
this phenomenon is commonly observed (Nardinietal. 2015; to study in detail AGN feedback. The coexistence of an ultra-
Tombesietal. 2015). Whether the presence of radio emission fastX-raywindandamolecularlargescaleoutflowindicatethat
issomehowrelatedtothispeculiarwindisaquestionworthex- thisprocessisinactionviaaninducedshock.Thepeculiarradio
ploring. In a similar case of coexistence of fast X-ray outflow propertiesarelikelyrelatedtothisphenomenon.Futureradioob-
andradiojet,albeitinahighradiopowersource,Tombesietal. servationsupto24GHzinphasereferencemodewillallowusto
(2012)reportedthatthepressureestimatedbytheX-raycolumn obtainhigherfidelitytotalintensityandspectralimages,useful
densityiscomparabletothepressureoftheradioinnerjet,there- toinvestigatetheroleoftheradioemissioninthiscontext.
forelendingsupporttotheideathattheX-rayoutflowmayhelp
tocollimatethejet.InthecaseofIRAS17020+4544 higherres- Acknowledgements. Weacknowledge financialcontribution fromgrantPRIN-
olutionVLBAobservationswillprovideabetterestimateofthe INAF-2014.ThisresearchhasmadeuseofNASA’sAstrophysicsDataSystem
Bibliographic Services. This research has made use of the NASA/IPAC
spatial scale of the jetted structure and its properties, shedding
Extragalactic Database (NED) which is operated by the Jet Propulsion
lightonthepossiblewind-jetconnection. Laboratory, California Institute of Technology, under contract with the
The mutual interactionbetween AGN winds and jets is the NationalAeronauticsandSpaceAdministration.TheNationalRadioAstronomy
subjectofmuchdebate.Zakamska&Greene(2014)foundacor- ObservatoryisafacilityoftheNationalScienceFoundationoperatedunderco-
operativeagreementbyAssociatedUniversities,Inc.
relationbetweenradiopowerandgaskinematics(inparticular,
[OIII]linevelocitywidth)inasampleofluminoustype-2radio
quietquasars.Theyconsideredseveralscenariostointerpretthis
correlation,andconcludedthateither(i)radiativelydrivenwinds
propagateintotheinterstellarmediumofthehostgalaxy,gener- References
ating shock fronts which in turn accelerate particles producing
Abdo,A.A.,Ackermann,M.,Ajello,M.,etal.2009a,ApJ,699,976
radioemissionor(ii)mechanicalenergyofrelativisticjetsheats
Abdo,A.A.,Ackermann,M.,Ajello,M.,etal.2009b,ApJ,707,L142
overpressuredcocoons,whichthenlaunchwindsofionisedgas
Anto´n,S.,Browne,I.W.A.,&Marcha˜,M.J.2008,A&A,490,583
(Mullaneyetal. 2013). The detection of the X-ray outflow in Becker,R.H.,White,R.L.,&Helfand,D.J.1995,ApJ,450,559
6
M.Girolettietal.:RadiopropertiesofIRAS17020+4544
Berton,M.,Foschini,L.,Ciroi,S.,etal.2015,A&A,578,A28
Caccianiga,A.,Anto´n,S.,Ballo,L.,etal.2014,MNRAS,441,172
Chartas,G.,Brandt,W.N.,Gallagher,S.C.,&Garmire,G.P.2002,ApJ,579,
169
Cohen,A.S.,Lane,W.M.,Cotton,W.D.,etal.2007,AJ,134,1245
Condon,J.J.,Cotton,W.D.,Greisen,E.W.,etal.1998,AJ,115,1693
D’Ammando,F.,Orienti,M.,Finke,J.,etal.2012,MNRAS,426,317
D’Ammando,F.,Orienti,M.,Larsson,J.,&Giroletti,M.2015,MNRAS,452,
520
deVries,N.,Snellen, I.A.G.,Schilizzi, R.T.,Mack,K.-H.,&Kaiser, C.R.
2009,A&A,498,641
DiMatteo,T.,Springel,V.,&Hernquist,L.2005,Natur,433,604
Doi,A.,Asada,K.,&Nagai,H.2011,ApJ,738,126
Doi,A.,Fujisawa,K.,Inoue,M.,etal.2007,PASJ,59,703
Doi,A.,Nagira,H.,Kawakatu,N.,Kino,M.,Nagai,H.,&Asada,K.2012,ApJ,
760,41
Doi,A.,Asada,K.,Fujisawa,K.,Nagai,H.,Hagiwara,Y.,Wajima,K.,&Inoue,
M.2013,ApJ,765,69
Doi,A.,Wajima,K.,Hagiwara,Y.,&Inoue,M.2015,ApJ,798,L30
Doi,A.,Oyama,T.,Kono,Y.,etal.2016,PASJ,68,73
Douglas,J.N.,Bash,F.N.,Bozyan,F.A.,Torrence,G.W.,&Wolfe,C.1996,
AJ,111,1945
Fanti,C.,Fanti,R.,Dallacasa,D.,Schilizzi,R.T.,Spencer,R.E.,&Stanghellini,
C.1995,A&A,302,317
Faucher-Gigue`re,C.-A.,&Quataert,E.2012,MNRAS,425,605
Ficarra,A.,Grueff,G.,&Tomassetti,G.1985,A&AS,59,255
Fukumura,K.,Tombesi,F.,Kazanas,D.,etal.2014,ApJ,780,120
Gallo,L.C.,Edwards,P.G.,Ferrero,E.,etal.2006,MNRAS,370,245
Giroletti,M.,Giovannini,G.,&Taylor,G.B.2005,A&A,441,89
Giroletti,M.,&Panessa,F.2009,ApJ,706,L260
Giroletti,M.,&Polatidis,A.2009,AN,330,193
Giroletti,M.,Paragi,Z.,Bignall,H.,etal.2011,A&A,528,L11
Gliozzi, M., Papadakis, I. E., Grupe, D., Brinkmann, W. P., Raeth, C., &
Kedziora-Chudczer,L.2010,ApJ,717,1243
Gregory,P.C.,&Condon,J.J.1991,ApJS,75,1011
Gregory,P.C.,Scott,W.K.,Douglas,K.,&Condon,J.J.1996,ApJS,103,427
Gu,M.,&Chen,Y.2010,AJ,139,2612
Gu,M.,Chen,Y.,Komossa,S.,Yuan,W.,Shen,Z.,Wajima,K.,Zhou,H.,&
Zensus,J.A.2015,ApJS,221,3
Hales,S.E.G.,Baldwin,J.E.,&Warner,P.J.1988,MNRAS,234,919
Longinotti,A.L.,Krongold,Y.,Guainazzi,M.,etal.2015,ApJ,813,L39
Komatsu,E.,Dunkley,J.,Nolta,M.R.,etal.2009,ApJS,180,330
Kunert-Bajraszewska,M.,Gawron´ski,M.P.,Labiano,A.,&Siemiginowska,A.
2010,MNRAS,408,2261
Massaro,F.,Giroletti,M.,D’Abrusco,R.,etal.2014,ApJS,213,3
Mullaney,J.R.,Alexander,D.M.,Fine,S.,etal.2013,MNRAS,433,622
Murgia,M.,Fanti,C.,Fanti,R.,Gregorini,L.,Klein,U.,Mack,K.-H.,&Vigotti,
M.1999,A&A,345,769
Nardini,E.,Reeves,J.N.,Gofford,J.,etal.2015,Sci,347,860
Ohta,K.,Aoki,K.,Kawaguchi,T.,&Kiuchi,G.2007,ApJS,169,1
Orienti,M.,D’Ammando,F.,Larsson,J.,Finke,J.,Giroletti,M.,Dallacasa,D.,
Isacsson,T.,&StobyHoglund,J.2015,MNRAS,453,4037
Pounds,K.A.,Reeves,J.N.,King,A.R.,etal.2003,MNRAS,345,705
Rengelink,R.B.,Tang,Y.,deBruyn,A.G.,etal.1997,A&AS,124,
Richards,J.L.,&Lister,M.L.2015,ApJ,800,L8
Schulz,R.,Kreikenbohm,A.,Kadler,M.,etal.2015,arXiv,arXiv:1511.02631
Snellen,I.A.G.,Mack,K.-H.,Schilizzi,R.T.,&Tschager,W.2004,MNRAS,
348,227
Tchekhovskoy,A.,Narayan,R.,&McKinney,J.C.2011,MNRAS,418,L79
Tombesi,F.,Sambruna,R.M.,Marscher,A.P.,etal.2012,MNRAS,424,754
Tombesi,F.,Tazaki,F.,Mushotzky,R.F.,etal.2014,MNRAS,443,2154
Tombesi,F.,Mele´ndez,M.,Veilleux,S.,etal.2015,Nature,519,436
Zubovas,K.,&King,A.2012,ApJ,745,L34
Wajima,K.,Fujisawa,K.,Hayashida,M.,Isobe,N.,Ishida,T.,&Yonekura,Y.
2014,ApJ,781,75
Wang,T.,&Lu,Y.2001,A&A,377,52
Whalen,D.J.,Laurent-Muehleisen,S.A.,Moran,E.C.,&Becker,R.H.2006,
AJ,131,1948
Zakamska,N.L.,&Greene,J.E.2014,MNRAS,442,784
7
10 20 30 40 50
100
50
s
d
n
o
c 0
e
s

c
r
A
-50
-100
200 150 100 50 0 -50 -100 -150 -200
Arc seconds