Mon.Not.R.Astron.Soc.000,000–000 (0000) Printed9May2012 (MNLATEXstylefilev2.2) Comparison of ejection events in the jet and accretion disc outflows in 3C 111 2 1 F. Tombesi1,2(cid:2), R. M. Sambruna3, A. P. Marscher4, S. G. Jorstad4,5, C. S. Reynolds2 0 and A. Markowitz6,7,8 2 y 1X-ray Astrophysics Laboratory and CRESST, NASA/Goddard Space Flight Center, Greenbelt, MD 20771, USA a 2Department of Astronomy, Universityof Maryland, College Park, MD 20742, USA M 3Department of Physics and Astronomy, MS 3F3, 4400 UniversityDrive, George Mason University,Fairfax, VA 22030, USA 4Institute for Astrophysical Research, Boston University,725 Commonwealth Ave., Boston, MA 02215, USA 5Astronomical Institute, St. Petersburg State University, UniversitetskijPr. 28, Petrodvorets, 198504 St. Petersburg, Russia 8 6Centerfor Astrophysics and Space Sciences, Universityof California, San Diego, Mail Code 0424, La Jolla, CA 92093-0424, USA 7Dr. Karl Remeis-Sternwarte and Erlangen Centre forAstroparticle Physics, Frederic-Alexander Universita¨t Erlangen-Nu¨rnberg, 7 ] E Sternwartstraße, 96049 Bamberg, Germany 8Alexander von Humboldt Fellow H . h p Accepted ???.Received???;inoriginalform??? - o r t ABSTRACT s a [ We present a comparisonof the parameters of accretion disc outflows and the jet of the broad-line radio galaxy 3C 111 on sub-pc scales. We make use of published 1 X-ray observations of ultra-fast outflows (UFOs) and new 43 GHz VLBA images to v trackthejetknotsejection.Wefindthatthesuperluminaljetcoexistswiththemildly 4 relativistic outflows on sub-pc scales, possibly indicating a transverse stratificationof 3 aglobalflow.The twoareroughlyinpressureequilibrium,with the UFOspotentially 7 providing additional support for the initial jet collimation. The UFOs are much more 1 . massivethanthe jet,but theirkinetic powerisprobablyaboutanorderofmagnitude 5 lower, at least for the observations considered here. However, their momentum flux 0 is equivalent and both of them are powerful enough to exert a concurrent feedback 2 impact on the surrounding environment. A link between these components is natu- 1 rally predicted in the context of MHD models for jet/outflow formation. However, : v giventhe highradiationthroughputofAGNs,radiationpressureshouldalsobe taken i X into account. From the comparison with the long-term 2–10 keV RXTE light curve we find that the UFOs are preferentially detected during periods of increasing flux. r a We also find the possibility to place the UFOs within the known X-ray dips-jet ejec- tion cycles, which has been shown to be a strong proof of the disc-jet connection, in analogue with stellar-mass black holes. However, given the limited number of obser- vations presently available, these relations are only tentative and additional spectral monitoring is needed to test them conclusively. Key words: accretion,accretiondiscs–blackholephysics–galaxies:active–galax- ies: jets – X-rays: galaxies – radio continuum: general 1 INTRODUCTION galaxy and surrounding environment? New insights on the characteristics and importance of winds/outflows in radio- One of the most enduring open questions surrounding ac- quiet AGNs have been recently obtained thanks to deep tive galactic nuclei (AGNs) concerns the relation between XMM-Newton,Chandra andSuzaku observations.Inpartic- accretionandejectionprocesses,i.e.,whatistheconnection ular,thedetectionofblue-shiftedFeXXV-XXVIabsorption between the black hole, the accretion disc and the forma- lines in the X-ray spectra of several sources demonstrated tion of outflows and jets? Then, a related question is, what the presence of massive, highly ionized and mildly/nearly is the feedback impact of AGN outflows/jets on the host relativisticaccretiondiskoutflows(e.g.,Chartasetal.2002, 2003; Pounds et al. 2003; Dadina et al. 2005; Markowitz et al. 2006; Braito et al. 2007; Cappi et al. 2009; Reeves et (cid:2) E-mail:[email protected] 2 F. Tombesi et al. al. 2009; Chartas et al. 2009a; Giustini et al. 2011; Dauser probetheinneraccretion discthroughX-rays.3C 111isX- et al. 2011; Gofford et al. 2011; Lobban et al. 2011). More- ray bright (2–10 keV flux of ∼2−8×10−11 erg s−1 cm−2) over, a systematic spectral analysis and photo-ionization and shows Seyfert-like properties. It is also one of the modellingperformedbyTombesietal.(2010a;2011a;2012) two BLRGs, the other being 3C 120, recently detected on a sample of 42 Seyfert galaxies observed with XMM- in γ-rays with FERMI (Kataoka et al. 2011; Grandi et Newton demonstratedthattheseultra-fastoutflows(UFOs) al. 2012). For the central black hole of 3C 111 we con- arerathercommon phenomena,beingpresentinmorethan sideramass1 oflogMBH=8.1±0.5 M(cid:3),takingintoaccount 40% of the sources, and confirmed the claims that these the maximum and minimum values derived by Ch11 using UFOsare indeed powerful enough to potentially play a sig- Hα measurements. The Eddington luminosity is therefore nificant role in the AGNcosmological feedback. LEdd=1.3×1038(MBH/M(cid:3))(cid:3)2×1046 erg s−1. In radio-loud AGNs, relativistic jets are routinely ob- In this work we will focus on the comparison between served at radio, optical and X-rays. However, the presence ejectioneventsintheradiojetandthosefromtheaccretion ofdiscoutflowsintheseobjectshasrecentlyemergedthanks discof3C111,representedbytheUFOs,andthesearchfor to X-ray spectroscopy. For instance, Tombesi et al. (2010b) a possible link between these two. The paper is structured reported the discovery of highly ionized and massive gas asfollows.In§2weestimatetheparametersoftheUFOsin outflowing with mildly relativistic velocities ∼0.1c, consis- 3C111usingthepublisheddata.In§3weextendtheworkof tentwithUFOs,in3/5Broad-LineRadioGalaxies(BLRGs) Ch11andestimatetheparametersoftheinnerradiojetfrom observedwithSuzaku,namely3C111,3C120and3C390.3. the VLBA images. In § 4 we compare the characteristics of TheUFOin3C111wasdetectedinalongobservationper- the UFOs and the jet and discuss the possibility to place formed in August 2008 (see Table 1) and a follow-up study also the UFOs in the context of the source variability and wasthenperformedinSeptember2010tostudyitsvariabil- the known jet ejection cycles, with conclusions following in ity through three Suzaku observations spaced by ∼7 days § 5. Throughout this paper we adopt a Hubble constant of (Tombesi et al. 2011b). A systematic 4–10 keV spectral H0=70 km s−1 Mpc−1 (Spergel et al. 2003). analysis revealed the presence of an ionized Fe K emission lineinthefirstobservation,indicativeofreflection/emission from the accretion disc, and blue-shifted Fe K absorption in the second one, when the flux was 30% higher, indicat- 2 OBSERVATIONS OF UFOS ing the presence of a highly ionized and massive outflow 2.1 Parameters from X-ray spectroscopy withvelocity∼0.1c(seeTable1).Thelocation ofthemate- rial was constrained at (cid:2)0.006 pc ((cid:2)500rs, rs=2GMBH/c2) Table1reportstheparametersofthepublishedSuzaku and from the black hole through the ∼7 days variability time- XMM-Newton observations of 3C 111 in which a search for scale.Thisprovidedthefirstdirectevidenceforanaccretion UFOs has been performed. The first observation refers to disc-outflowconnection in anAGNand isconsistent with a Tombesietal.(2010b),thesecondtoBalloetal.(2011)and picture in which a disruption/over-ionization of the inner thelastthreetoTombesietal.(2011b).Thecolumndensity, disc is followed by the ejection of an outflow from ∼100rs. ionization and outflow velocity are reported. We estimate Then, this is possibly accelerated through radiation and/or the lower/upper limits of the location, mass outflow rate magnetic forces tothe observed velocity of ∼0.1c. and kinetic power of the UFOs following the approach of Chatterjee et al. (2011; hereafter Ch11) recently re- Tombesi et al. (2012). ported the results of an extensive multi-frequency monitor- An estimate of the minimum distance can be derived ing campaign on 3C 111 at X-ray (2–10 keV), optical (R from the radius at which the observed velocity corresponds band), and radio (14.5, 37 and 230 GHz) wave bands, as to the escape velocity, rmin = 2GMBH/vo2ut. However, we well as multi-epoch imaging with the Very Long Baseline note that when deriving this quantity we do not take into Array (VLBA) at 43 GHz, between 2004 and 2010. They accountthepossibleadditionalacceleration oftheflow,but find that major X-ray dips are systematically followed by assume that it is ejected at the observed velocity. Instead, anincreaseoftheradiocorefluxandtheappearanceofnew in order to derive a firm estimate of the distance from the jetknotsintheVLBAimagesafter∼2–3months.Newknots definition of the ionization parameter ξ =Lion/nr2 (Tarter areejected∼1–2timesperyearwithtypicalapparentspeeds et al. 1969) we would need an estimate of the density of of∼3–5c. Thisshowstheexistenceofaconnectionbetween thematerial n, which can not beobtained with the present theradiativestateneartheblackhole,wheretheX-raysare data. However, the observed short time-scale variability produced, and events in the jet, providing a solid proof of of the UFOs (e.g., Braito et al. 2007; Cappi et al. 2009; thedisc-jet connection in thisradio-loud AGN.These com- Tombesi et al. 2010a, 2011b; Giustini et al. 2011) suggests plexcycles,wherebysomeinstabilityappearstodisruptthe that these absorbers are compact and that their thickness innerregionsoftheaccretiondiscandthentriggerspowerful is less than the distance to the source, Δr/r<1. Therefore, mass/energy ejections provide an observational clue to the we can derive a lower limit of the density of the material origin ofradiojets.Thisbehaviourhasparallels inGalactic as n=NH/Δr>NH/r, where NH is the line-of-sight column microquasars (e.g., Neilsen & Lee 2009; Fender et al. 2009) and equivalent results have been obtained also for another BLRG,3C120(Marscheretal.2002;Chatterjeeetal.2009). 1 WenotethatMarchesinietal.(2004)estimatedalargerblack TheBLRG3C111(z=0.0485)isoneofthebesttargets hole mass of logMBH∼9.5 M(cid:3) assuming the bulge luminosity for these studies. In the radio it is an FR II with a blazar- relation.Thediscrepancyisprobablymainlyduetothedifferent like behaviour. The jet lies at θ=18.1◦±5.0◦ to our line of extinction adopted and we consider the Ch11 one to be more sight (Jorstad et al. 2005), allowing us to simultaneously reliable. Jet and disc outflows in 3C 111 3 Table 1.Suzaku andXMM-Newton observations of3C111andparametersofthedetected UFOs. Sat Tobs UFO logLion logNH logξ vout r M˙out logE˙K (ergs−1) (cm−2) (ergs−1 cm) (c) (pc) (M(cid:3) yr−1) (ergs−1) 1∗ S 2008.65 u7 44.4 >23.0 5.0±0.3 0.041±0.003 0.003-0.02 0.1-0.6 42.8-43.5 2 X 2009.13 44.7 ... ... ... ... ... ... 3 S 2010.67 44.7 ... ... ... ... ... ... 4∗ S 2010.69 u10 44.9 22.9±0.2 4.3±0.1 0.106±0.006 0.001-0.006 0.1-0.8 43.4-44.3 5 S 2010.71 44.9 ... ... ... ... ... ... Note. Columns: observation number; satellite, X for XMM-Newton and S for Suzaku; observation start date; UFOidentifier;absorptioncorrected luminositybetween 1–1000Ryd(1Ryd=13.6eV); columndensity; ionization parameter;observedvelocity;location;massoutflow rate;kineticpower.∗ Observationswithdetected UFOs. density.Then,substitutingthisintheexpressionfortheion- than u7, we note that many of their other characteristics izationparameterwecanestimateanupperlimitonthedis- (NH, r,M˙out and E˙K) are consistent with each other. tance of the absorber from the source as rmax =Lion/ξNH. FromtherelationLbol(cid:3)10Lion ergs−1 (e.g.,McKernan The material can not be farther away than this given the etal. 2007) thebolometric luminosity isLbol(cid:3)1045 ergs−1, observedionizationparameter.Wenotethatthisexpression whichcorrespondstoanEddingtonratioofLbol/LEdd(cid:3)0.1. contains theimplicit assumption that the ionizing source is Considering a radiative efficiency of η(cid:3)0.1 (Davis & Laor seen as a point source by the absorber. The validity of this 2011), the accretion rate is M˙acc=Lbol/ηc2(cid:3)0.5 M(cid:3) yr−1, supposition is supportedfrom thefact that theX-rayemit- which is comparable to the outflow rate derived for the tingregion in AGNsisconstrained byX-rayvariability and UFOs. micro-lensing observations to be within a few rs from the black hole (e.g., Chartas et al. 2009b), instead the UFOs consideredherearealwaysatdistances(cid:3)100rs (seetextbe- 2.2 Variability and the X-ray light curve low and Table 1). We use the expression for the mass outflow rate de- InFig. 1weshow the2.4–10 keVfluxlight curveof3C 111 rived by Krongold et al. (2007), which is more appropriate observed with the Rossi X-ray Timing Explorer (RXTE) forabiconicalwind-likegeometry:M˙out (cid:3)1.2πmpNHvoutr. from the beginning of 2008 to mid 2011. The typical expo- This formula has also the important advantage of already suretimeis1–2ksandthesamplingoftheobservationswas implicitly taking into account the covering and filling fac- 2–3 times per week. We adopted the same data reduction tors. This is due to the fact that it considers only the net proceduresasexplainedinCh11.Weobservefivemajordips flowofmass,directlyallowingforclumpingintheflow.How- in the light curve and, adopting the same nomenclature of ever,wenotethatconsideringaclumpinessfactorof∼Δr/r Ch11,weidentifytheminTable2with“d”andtherelative we obtain equivalent results using the usual spherical ap- number. proximation (Tombesi et al. 2010b, 2011b) and a covering We can use this light curve to check for possible re- fraction of ∼0.3, consistent with observations (Tombesi et lations between the source flux variability and that of the al.2010a,b).Thekineticpowercanconsequentlybederived UFOs.Therearethreepossiblecausesfortheobservedvari- as E˙K= 21M˙outvo2ut. ability of UFOs. First, it could be an intermittent on/off Substitutingtherelativeparameters,theUFOobserved variability, i.e. the UFOsare not a continuous phenomenon inAugust2008,u7inTable1,isconstrainedatadistance2of and they are ejected only at certain times. In this case, the (cid:3)0.003–0.02 pc ((cid:3)200–1000rs) from the central black hole, lackofdetectionisduetotheabsenceofaUFOatthetime with a mass outflowrate of M˙out(cid:3)0.1–0.6 M(cid:3) yr−1 and ki- of theobservation. Second,even if a UFO is present during netic power of E˙K(cid:3)6×1042–3×1043 erg s−1. The UFO thetime of theobservation, therecould be some additional relative to theSeptember2010 observation, u10 in Table 1, absorbervariabilityduetoinhomogeneities inthestructure is located at (cid:3)0.001–0.15 pc((cid:3)80–12,000rs),it has an out- and density of the flow (e.g., turbulence, clumpiness) and flow rate of M˙out(cid:3)0.1–10 M(cid:3) yr−1 and a kinetic power of transversemotion.Thiseffectisexpectedtooccuronmuch E˙K(cid:3)2×1043–4×1045 erg s−1. However, considering the shorter time-scales than the first one, on intervals of the variability on ∼7 days time-scales (Tombesi et al. 2011b), order of a few hours (e.g., Braito et al. 2007; Giustini et wecanfurtherconstrainitsdistanceto(cid:2)0.006pc((cid:2)500rs). al.2011). Thisis also expectedtobechaotic andnot corre- Consequently, the upper limits on the mass outflow rate latedwithastateofthesource.Third,theabsenceofaUFO and kinetic power for u10 reduce to (cid:2)0.8 M(cid:3) yr−1 and could be masked by an insufficient signal-to-noise (S/N) of (cid:2)2×1044 erg s−1, respectively. Even though u10 is faster thedata. The UFOs have been clearly detected in 2/5 observa- tions,thustheirfrequencyofdetectionis∼40%.Inparticu- 2 AssumingthehigherblackholemassestimateofMarchesiniet lar,theF-testandMonteCarloprobabilitiesfortheabsorp- al. (2004), as inTombesi et al. (2010a), the observed velocity of tion lines detected in both the first (Tombesi et al. 2010b) u7 would be lower than the escape velocity at the estimated lo- andfourthobservations(Tombesietal.2011b)inTable1are cation.However,usingtherefinedlowerblackholemassestimate ofCh11(see§1)andconsideringalsothepossibilityofadditional >99%.Inaddition,asanalternativetestonthesignificance acceleration, we are confident enough that also u7 escapes the ofthelines(e.g.,Vaughan&Uttley2008), wenotethatthe system. ratio between the equivalent width and the relative 1σ er- 4 F. Tombesi et al. Figure 1.Long-term 2.4–10keV fluxRXTElightcurveof3C111between 2008andmid2011. Thevertical solid/dotted linesreferto thedetection/non-detection ofUFOsintheSuzaku andXMM-Newton spectra.ThedetectionsofUFOsaremarkedwith“u”.Thedates relativetotheX-raydipsandtheappearanceofnewjetknotsintheVLBAimagesaremarkedwith“d”and“k”,respectively. rors3 is(cid:3)4forallthecases.AsstatedbyBalloetal.(2011), trends. In order to distinguish between periods of rising or the non-detection of a UFO in the second observation in steady/decreasing flux we consider the difference between Table1isnotduetoalowS/N.Thesameconclusion isde- the fluxes ∼3 days after and before the observations. This rivedalsoforthethirdandfifthobservationsbyTombesiet is equivalent to the minimum time interval of the RXTE al.(2011b).Infact,the2–10 keVS/Nintheseobservations light curve and also allows to oversample by a factor of ∼2 was (cid:3)50, 190, 105, 110 and 105 for the first, second, third, the typical variability time-scale of the UFO in 3C 111 of fourthandfifthobservation,respectively.Therefore,thisin- about ∼7 days (Tombesi et al. 2010b, 2011b). If the differ- dicates that a UFO was not present along the line-of-sight ence is positive it indicates a rising flux, instead if null or duringtheseobservationsoritwassoweakthatitcouldnot negative it indicates a steady/decreasing flux.We findthat bedetected even in high S/N spectra. the first and fourth observations in Table 1, the ones with We then performed a test in order to check for a detectedUFOs,happenedduringperiodsofincreasingflux4. possible relation between the source X-ray flux from the Instead,thenon-detectionsin thesecond and fifthobserva- RXTE light curve shown in Fig. 1 and the detections/non- tionsoccurredduringintervalsofdecreasingflux.Following detectionsofUFOs(markedwithsolid/dottedverticallines, thiscriterion,thenon-detectioninthethirdobservationoc- respectively). We can see that there seems to be no depen- curredinanintervalofsteady/decreasingfluxtoo.However, dence on the absolute flux of the source, the UFOs being wenotethatthislattercaseislessstringentbecauseithap- detected/non-detected both in high/low flux states. How- pened very close to a sudden spike in flux and we adopt ever, there could be a relation with the flux variability 4 We note that the observation of the UFO u7 occurred at the beginning of a period of rising flux, just after the major X-ray 3 Wenotethattheequivalentwidtherrorsoftheabsorptionlines dipd7.Iftherisingperiodisrelatedwiththeaccelerationofthe reported in Table 3 of Tombesi et al. (2010b) for 3C 111 are at outflow, this mightexplainwhythe velocityof u7ismuchlower the90%level,insteadtothe1σlevelindicatedinthetablenotes. thanu10,whichinsteadwasdetectedclosetoamaximuminflux. Jet and disc outflows in 3C 111 5 tions (Ch11), new radio jet knots systematically appear a Table2.TimesofX-raydips,observationsofUFOsandappear- anceofradioknots. fewmonthsaftermajorX-raydips.Thisisvalidalsoforthe two new detected ones, k10 and k11, which appear about Dip TXmin UFO Tufo Knot Tknot βapp ∼3 months after the relative X-ray dips d10 and d11. We are confident that the dips d10 and d11 are indeed related d7 2008.51 u7 2008.65 k7 2008.83±0.07 4.54±0.38 totheknotsk10andk11forseveralreasons:thereisnosig- d8 2008.98 ... ... k8 2009.07±0.08 4.07±0.43 d9 2009.26 ... ... k9 2009.29±0.04 4.33±0.66 nificantdetectionofanewknotejectionintheradioimages d10 2010.57 u10 2010.69 k10 2010.85±0.02 5.66±0.09 between2010.60and2010.80,thetimeintervalbetweend10 d11 2010.78 ... ... k11 2011.01±0.07 5.22±0.35 andd11isequivalenttothatbetweenk10andk11andalso both knots appear in the radio images with an equivalent Note.βapp istheapparentspeedoftheradioknotsinunitsofc. delay of about 3 months after the relative dips. In general, thedelay is distributedbetween 0.03 yrand0.34 yr,with a a conservative approach not considering it in the following meanvalueof0.15±0.08yr(Ch11).Asalreadydiscussedby discussion. Ch11,consideringtheapparentspeedsof∼4–5c,anaverage From Fig. 1, we derive that overall the UFOs seem to delay from theX-ray dipsof ∼0.15 yrand an inclination of bepreferentiallydetectedduringintervalsofincreasingflux. ∼18◦, we can derive that the typical distance traveled by Inordertoestimatethestatisticalconfidenceofthepossible the jet knots before appearing out from the 43 GHz core is relationbetweentheUFOsandtheperiodsofrisingfluxwe d∼0.6pc. testedthenullhypothesisthatUFOsarenotdetectedduring Considering thejet knotsk7andk10,theiractualbulk phasesofascendingfluxbutonlyinsteadyordecreasingin- velocity(v=βc)canbeestimatedfromtheapparentvelocity tervals.Thishypothesisissatisfiedinnoneofthefourcases (vapp=βappc) adopting an inclination to the line of sight of described before, yielding a probability of <1/4. Therefore, θ∼18◦ (Jorstad etal.2005), βapp =βsinθ(1−βcosθ)−1.We conservatively, we can say that the statistical probability obtainvk7(cid:3)0.9(cid:2)82candvk10(cid:3)0.995candtherelativeLorentz of the claim that UFOs are preferentially observed during factors(Γ=1/ 1−β2)areΓk7∼5.3andΓk10(cid:3)10fork7and phasesofrisingfluxisP=1−(1/4)(cid:3)75%.Giventhelimited k10, respectively. The knot k10 is faster than k7 and their numberofobservationsavailable,westressthatthestatisti- parameters are reported in Table 2. calsignificanceofthisrelationisonlymarginalanditshould Assuming equipartition and using the formula (A3) in be regarded only as an indication. However, we note that a Jorstad&Marscher(2004)wederiveanestimateofthemag- similarbehaviorwasobservedalsoinothersourcesshowing neticfieldofB(cid:3)0.1Gauss,consistentwiththetypicalvalues UFOs(e.g., Braito et al. 2007; Giustini et al. 2011). at∼sub-pcscales(e.g.,O’Sullivan&Gabuzda2009).Then, it is possible to roughly quantify the jet kinetic power as E˙K,j(cid:3)2(B2/8π)(πd2)c∼3×1044 ergs−1,wheredisthepre- viouslyestimateddistanceoftheknotsfromtheblackhole. 3 RADIO OBSERVATIONS OF THE JET ON However,ifweincludealsothepossibleadditionaltermdue SUB-PC SCALES totherest-massenergyoftheprotons,thekineticpowercan 3C 111 is actively monitored with the VLBA at 43 GHz reach values up to ∼1045 erg s−1. These estimates are con- at roughly monthly intervals by the blazar group at the sistentwiththetypicaljetpowerofradiogalaxiesestimated Boston University.Herewepresenta temporal extension of from the associated radio lobes of E˙K,j∼1044–1045 erg s−1 the VLBA analysis of Ch11 (see their Fig. 6) from 2008 up (e.g., Rawlings & Saunders 1991). Subsequently, from the tomid2011.ThesequenceofVLBAimagesshowninFig.2 relationE˙K,j(cid:3)(1/2)M˙out,jc2(Γ−1)wecanalsocalculatethe providesadynamicviewoftheinnerjetbetweenNovember massfluxratethatisfunneledintothejet. Considering the 2010andSeptember2011atanangularresolution∼0.1mil- average Γ∼7 of the jet knots in Table 2, we obtain a mass liarcseconds(mas),correspondingto∼0.094pc.TheVLBA outflow rate of M˙out,j(cid:3)0.0005–0.005 M(cid:3) yr−1. datahavebeenprocessedinthesamemannerasdescribedin Ch11.Wecanclearlyobservethepresenceoftwonewradio jet knots,each characterized byitsfluxdensity,FWHMdi- 4 DISCUSSION ameterandpositionrelativetothecore.Timesof“ejection” aredefinedastheextrapolatedtimeofcoincidenceofamov- Inthispaperwefocusonacomparison betweentheparam- ing knot with the position of the 43 GHz core. We use the eters of the jet and accretion disc outflows, also referred as position vs. time data to determine the projected direction UFOs, observed in 3C 111. This is the first time that such on the sky of the inner jet, as well as the apparent speeds astudyisperformedforanAGN.Thetypicalkineticpower and ejection times of new superluminal knots. Continuing of the UFOs reported in Table 1 is E˙K∼1043–1044 erg s−1. with thenomenclature adopted byCh11, we haveknot k10 Thisisaboutoneorderofmagnitudelowerthanthekinetic appearing from the 43 GHz core at 2010.85±0.02 and k11 power of the jet. However, if we take into account the un- at 2011.01±0.07, respectively.Bothof themhaveapparent certainty in the determination of the jet kinetic power and superluminal velocities of 5.66±0.09c and 5.22±0.35c, re- the possibility that the jet knots experienced an additional spectively.Thepropermotionoftheseknotscanbedirectly acceleration with respect to the UFOs, it is plausible that followed in Fig. 2 for almost one year. theirvaluesareactuallycomparable.Infact,theUFOswere The dates relative to the X-ray dips and jet knots ap- detectedatdistancesofonly∼0.001–0.02 pcfromtheblack pearance between 2008 and 2011 are marked in the RXTE hole, well within the∼0.6 pcscale of the43GHz core, after light curve in Fig. 1 by arrows. We see that, in line with which new jet knots are observable in theVLBA images. the reported correlation between X-ray dips and jet ejec- The mass outflow rate of the UFOs of M˙out∼0.1– 6 F. Tombesi et al. 1 M(cid:3) yr−1 is much higher than that estimated for the jet M˙out,j∼0.0005–0.005 M(cid:3) yr−1. Moreover, even if their ki- netic power is different, their momentum flux is actually equivalent.Thisisduetothelineardependenceofthisquan- tity on the outflow velocity. The ratio of the kinetic power with respect to the bolometric luminosity corresponds to ∼1–10% and ∼10–100% for the UFOs and the jet, respec- tively.Therefore,bothofthemareabletoexertaconcurrent feedbackimpactonthesurroundingenvironment(e.g.,Hop- kins&Elvis2010;Gasparietal.2011).However,aninterest- ing point to make here is that the powerful and collimated jettendsto“drill”outofthegalaxyanddeliverenergyand momentumtolargedistances.Ontheotherhand,theslower, widerangleand massiveUFOsmay bemuchmoreeffective atimpactingthestructuresofthehostgalaxy(e.g.,Tombesi et al. 2012). In fact, from the fraction of sources with de- tectedUFOsTombesietal.(2010a, b)estimatedthatthese absorberscoverasignificantfractionofthesolidangle,∼0.4. Thiscorrespondstoanopeningangleoftheoutflowwithre- specttothepolaraxisof∼60◦.Instead,Jorstadetal.(2005) derivedanintrinsichalfopeningangleofthejetin3C111of only ∼3◦. This indicates that the jet covers only a fraction of ∼0.001 of thesolid angle, which correspondsto less than 1% of that of theUFOs. The lower limit of the density of the material in the UFOs can be roughly derived as n=NH/Δr(cid:3)NH/r. This is valid for compact absorbers (Δr/r(cid:2)1) and is supported by the detection of short time-scale variability (e.g., Braito et al. 2007; Cappi et al. 2009; Tombesi et al. 2010a, 2011b). For both u7 and u10 we obtain n(cid:3)107 cm−3. From the photo-ionization code Xstar used for the modelling of the Fe XXV–XXVI absorption lines in Tombesi et al. (2010b, 2011a, 2011b) we obtain a typical temperature of the plasma of T∼106–107 K. Therefore, the lower limit on the gas thermal pressure in the UFOs is Pth(cid:3)nkT(cid:3)0.001– 0.01 dyne cm−2. This can be even higher considering the significant turbulent velocities of ∼1,000 km/s observed for UFOs (Tombesi et al. 2011a, 2011b). Using the estimate of the jet kinetic power of E˙K,j∼1044–1045 erg s−1 and as- suming a quasi-instantaneous, symmetrical injection of en- ergy, we can derive a crude lower limit of the jet ram pres- sure at the location of the UFOs of Pth,j(cid:3)E˙K/r3(cid:3)0.001– 0.01 dyne cm−2. These two estimates are comparable, sug- gesting that accretion disc outflows in the form of UFOs mayactuallyhelpcollimatetheinnerjet.Thisalsosuggests that the initial jet material would encounter more radial thanverticalresistance,providingapreferentialdirectionof propagationandthiscouldleadtoa“nozzle-like”geometri- calconfiguration, which would again helpcollimate/confine theinnerjet (e.g., Blandford & Rees 1974). Inthisregard,wenotethatrecentdetailedobservations oftheinnerradiojetofM87,theclosestpowerfulradio-loud AGNtous,revealedthatthejetformationisalreadytaking placeatdistancesdownto∼10–20rs fromthesuper-massive black hole (Hada et al. 2011). Moreover, in the inner few ∼100rs region the jet is seen opening widely, at an angle of∼60◦,andhavingaparaboloidalshape.Then,itsqueezes Figure 2. Sequence of VLBA images at 43 GHz during 2010– downto∼5◦onlyatdistancesof∼105rs,afterwhichittran- sitsintoaconicalshapeandbecomesafreestream(Asada& 2011.Theglobalpeakofthemapis1.13Jy/beamwiththebeam size of 0.32×0.16 mas2 at PA=-10◦, the contours levels start at Nakamura2012).Thissuggeststhatthejetisprobablysub- 0.25%ofthepeakandincreasebyafactorof2.Thelinesdenote jecttoaninitialcollimation bytheexternalgas.Thesimul- thepropermotionoftheradiojetknots k10andk11. taneouspresenceatsub-pcscalesofthesetwocomponentsin Jet and disc outflows in 3C 111 7 rough pressure equilibrium: a inner, highly relativistic jet, tion in Table 1. This observation occurred during a short and an outer, more massive, mildly relativistic plasma, is steady/decreasing flux period after the X-ray dip d10 and overallconsistentwiththepictureofatransversestratifica- noclearUFOwasdetected.Instead,ablue-shiftedFeXXVI tionof theflow(e.g., Ghisellini etal. 2005; Xieetal. 2012). absorptionlineindicativeofaUFOwasdetectedinthesuc- Theline-of-sight inclinationangleof3C111isinfact∼18◦ cessivefourthobservation,∼7daysafter,duringaclearpe- (Jorstad et al. 2005). Therefore, this eventuality should be riodsofascendingflux.The2–10keVS/Nofthethirdobser- consideredwhenperformingnumericalsimulationsofthejet vation((cid:3)105)wasenoughtodetectanabsorptionlinewith formation. thesameEW as in thefourth (S/N(cid:3)110) if present and its Theoretically, thecomplex coupling between radiation, constancy between these two observations was excluded at magnetic fields and matter that should be considered to the99.9%. Therefore, thelack of detection of aUFOin the properly explain the formation of outflows/winds from ac- thirdobservation pointstotheconclusion that it was prob- cretion discs has not been accurately solved yet. However, ablynotpresentalongtheline-of-sight atthattime,butwe simulations show that disc outflows are ubiquitously pro- can not rule out also the possibility that the non-detection ducedandcanbeacceleratedtomildlyrelativisticvelocities was due to strong inhomogeneities/turbulence in the flow. by radiation and/or magnetic forces (Blandford & Payne As already stated in Sec. 2.2, we note that the identifica- 1982;King&Pounds2003;Proga&Kallman2004;Ohsuga tion of the third observation with a steady/decreasing flux et al. 2009; Fukumuraet al. 2010). Moreover, several MHD state is less clear because it happened close to a sudden jet models predict that the jet production is accompanied spike in flux and we adopted a conservative approach not by the formation of broad, mildly relativistic disc outflows considering it in the discussion of the possible relation be- (Blandford & Payne 1982; Kato et al. 2004; McKinney & tween the UFOs and period of increasing flux. Tombesi et Gammie 2004; McKinney 2006; Tchekhovskoy et al. 2011), al.(2011b)interpretedtheseobservationsastheevidenceof potentially providing a direct link between these compo- aconnectionbetweenthermal/structuralinstabilitiesinthe nents. However, even if MHD models alone could, in prin- disc, possibly related to an ADAF state (Wang, Cheng & ciple, already explain the formation of the jet and mildly Li2012),andtheformation ofpowerfulwinds/outflows.In- relativistic disc outflows, the high radiation throughput of terestingly,evenifthelaunchingmechanism(s)oftheUFOs AGNscannotbeneglectedandradiationpressuremustplay might not be the same as for the jet, the fact that these an important role as well (King & Pounds 2003; Everett & observations occurred within two clear dip-ejection cycles Ballantyne2004;Everett2005;Ohsugaetal.2009;Ram´ırez suggests that they could be related, in the sense that the &Tombesi2012).Thecomparisonwiththesimilaroutflows strongdisc/radiative instabilities thatareknown toaccom- detected in radio-quiet AGNs,which however are known to pany the ejection of a new jet knot could then also trigger harbor weak jets as well (e.g., Giroletti & Panessa 2009; or boost the production of disc outflows (Livio et al. 2003; Maitra et al. 2011), may help to clarify this point. This is Xieetal.2012).Wenotethatasimilarqualitativebehavior, postponed toa future work. whereby the ejection of jet knots during X-ray dips is fol- Comparing the periods of X-ray dips and the observa- lowedbyanincreasedproductionofdiscoutflowsduringthe tions of UFOs and jet knots in Fig. 1 and Table 2 we note successiverising/flaring periods,hasbeenrecentlyreported that UFO u7 was observed between the X-ray dip d7 and byNeilsen et al. (2012) regarding theGalactic microquasar the jet knot k7. Similarly, u10 was observed between d10 GRS 1915+105 duringthe β state. and k10. This evidence is intriguing and may suggest the The main point that we would like to make here is placement of UFOs within the known dip-ejection cycles, to introduce the notion that UFOs could preferentially ap- which is the most solid observational proof of the disc-jet pearduringphasesofincreasingfluxandpossiblyatcertain connection (Marscher et al. 2002; Chatterjee et al. 2009, times of disc-jet activity. However, given the very limited 2011).AsalreadydiscussedbyCh11,theX-raydip-ejection number of observations, the statistical significance of these connectionsuggeststhatadecreaseintheX-rayproduction relations is only tentative and additional observations are is linked to an increase in the speed of the jet flow, caus- neededtotest thisproperly5.Nonetheless,wecan certainly ingashockfronttoeventuallyform andmovedownstream. concludethattherearenowplentyoftheoreticalandobser- Theactualphysicalcause ofthisrelation iscurrentlystill a vational evidences that mass and energy loss in theform of matterofconsiderablespeculation, however,ithassimilari- winds/outflows from the accretion disc are likely to be the tieswiththestellar-massblackholes(e.g.,Livioetal.2003; norm rather than the exception and models attempting to Neilsen & Lee 2009; King et al. 2012), for which more de- tailed studieshavebeenperformed.Thermal instabilitiesin theinneraccretiondisccancauseastatetransitionbetween a radiatively efficient to a low-efficient accretion mode dur- 5 In this regard, we are submitting a large monitoring cam- ing the X-ray dips, such as in the ADAF/ADIOS regimes paignfocusedon3C390.3,anotherBLRGwithaUFOdetection (Narayan & Yi 1995; Blandford & Begelman 1999), which (Tombesi et al. 2010b) and showing the typical jet dip-ejection predicts that a sizeable fraction of the accretion power is cycles. The choice fell on this target because it is continuously not radiated away and instead is released in the form of a visiblebyallmainX-rayobservatoriesallyearround.Thelength ofthecampaignis20monthsandwerequestaSwift monitoring jet or outflow. In particular, the jet production mechanism every ∼10 days, a long 100ks Suzaku and XMM-Newton obser- has been demonstrated to be more efficient for both high vationevery2months,along100ksChandra HETGobservation blackholespinsandthick,ADAF-like,inneraccretiondisks every 4 months and a parallel VLBA monitoring at intervals of (Meier 2001; Nemmen et al. 2007). 3 months. If approved, this will allow to have enough observa- WenotethatTombesietal.(2011b)detectedanionized tions toconclusivelytest the relationbetween UFOsandthe jet Fe K disc line in the X-ray spectrum of the third observa- dip-ejectioncycles. 8 F. Tombesi et al. explain the link between the jet and the accretion process Chartas, G., Kochanek, C. S., Dai, X., Poindexter, S., should takethese components intoaccount. Garmire, G. 2009b, ApJ, 693, 174 Chatterjee, R., et al. 2011, ApJ, 734, 43 Chatterjee, R., et al. 2009, ApJ, 704, 1689 5 CONCLUSIONS Dadina, M., Cappi, M., Malaguti, G., Ponti, G., de Rosa, A.2005, A&A,442, 461 In this paper we compare the characteristics of the X-ray Dauser, T., Svoboda, J., Schartel, N., et al. 2011, detectedaccretiondiscoutflows,alsoreferredasUFOs,and arXiv:1112.1796 the radio jet in the broad-line radio galaxy 3C 111. This is Everett, J. E., Ballantyne, D. R.2004, ApJ,615, L13 the first time that such a study is performed for an AGN. Everett, J. E. 2005, ApJ, 631, 689 We find that these two outflows, an inner relativistic one Fender, R. P., Homan, J., Belloni, T. M. 2009, MNRAS, andanotherbroaderandmildly-relativistic, coexist onsub- 396, 1370 pcscales,possiblyinagreementwithatransversestratifica- Gaspari, M., Melioli, C., Brighenti, F., D’Ercole, A. 2011, tionofaglobalflow.Thetwoarealsopotentiallyinpressure MNRAS,411, 349 equilibrium,providingadditionalsupportforthecollimation Ghisellini, G., Tavecchio, F., Chiaberge, M. 2005, A&A, ofthejet.Thediscoutflowsaremuchmoremassivethanthe 432, 401 jetbutprobablytheirkineticpowerislower.However,their Giroletti, M., Panessa, F. 2009, ApJ, 706, L260 momentumfluxiscomparableandbothofthemarecapable Giustini, M., Cappi, M., Chartas, G., et al. 2011, A&A, to exert a concurrent feedback impact on the surrounding 536, A49 environment.Even though a link between these component Gofford, J., Reeves, J. N., Turner, T. J., et al. 2011, MN- isalreadynaturallypredictedbyMHDjet/outflowssimula- RAS,414, 3307 tions,wenotethatalsoradiationpressuremustbetakeninto Grandi, P., Torresi, E., Stanghellini, C. 2012, account for a more realistic modelling. The comparison of arXiv:1204.2268 thedetectiontimesofUFOswiththelong-termRXTElight Hada, K., Doi, A., Kino, M., et al. 2011, Nature,477, 185 curve suggests that they are preferentially observed during Hopkins, P.F., Elvis, M. 2010, MNRAS,401, 7 periodsofincreasingfluxandtheinvestigationoftheVLBA Jorstad, S. G., Marscher, A. P.2004, ApJ, 614, 615 images points to the possibility of placing also these events Jorstad, S. G., et al. 2005, AJ, 130, 1418 withintheknownX-raydip-ejectioncycles,whichisthedi- Kataoka, J., Stawarz, L(cid:6)., Takahashi, Y., et al. 2011, ApJ, rect evidenceof thedisc-jet interaction. However, given the 740, 29 limited numberof observations, these possible relations are Kato, Y., Mineshige, S., Shibata,K. 2004, ApJ,605, 307 onlytentativeandadditionalobservationsareneededtotest King, A. R.,Pounds, K. A.2003, MNRAS,345, 657 them properly. If confirmed, this could provide a new win- King, A. L., Miller, J. M., Raymond, J., et al. 2012, ApJ, dowforX-rayspectroscopytostudyprocessesrelatedtothe 746, L20 jet activity on scales even smaller than the∼0.1 milliarcsec Krongold,Y.,Nicastro,F.,Elvis,M.,etal.2007,ApJ,659, reachable with VLBA images. 1022 Livio, M., Pringle, J. E., King, A. R.2003, ApJ,593, 184 Lobban, A. P., Reeves, J. N., Miller, L., et al. 2011, MN- ACKNOWLEDGMENTS RAS,414, 1965 Meier, D. L. 2001, ApJ, 548, L9 FT thank K. Fukumura, R. Nemmen, D. Kazanas, R. F. Maitra, D.,Miller, J.M., Markoff,S.,King,A.2011, ApJ, Mushotzky, A. Tchekhovskoy and S. Dalena for useful dis- 735, 107 cussions. The research at Boston University was funded in Marchesini, D., et al. 2004, MNRAS,351, 733 part by National Science Foundation grant AST-0907893. Markowitz, A., Reeves, J. N., Braito, V. 2006, ApJ, 646, The authors thank the anonymous referee for suggestions 783 that led to improvementsin thepaper. Marscher, A. P., Jorstad, S. G., G´omez, J.-L., et al. 2002, Nature,417, 625 McKernan,B.,Yaqoob,T.,Reynolds,C.S.2007,MNRAS, REFERENCES 379, 1359 Asada, K., Nakamura,M. 2012, ApJ, 745, L28 McKinney,J. C., Gammie, C. F. 2004, ApJ, 611, 977 Ballo, L., Braito, V., Reeves, J. N., Sambruna, R. M., McKinney,J. C. 2006, MNRAS,368, 1561 Tombesi, F. 2011, MNRAS,418, 2367 Narayan, R., Yi,I. 1995, ApJ,444, 231 Blandford,R.D.,Begelman,M.C.1999,MNRAS,303,L1 Neilsen, J., Lee, J. C. 2009, Nature,458, 481 Blandford, R. D., Payne,D. G. 1982, MNRAS,199, 883 Neilsen,J.,Petschek,A.J.,Lee,J.C.2012, MNRAS,421, Blandford, R. D., Rees, M. J. 1974, MNRAS,169, 395 502 Braito, V., et al. 2007, ApJ, 670, 978 Nemmen, R. S., Bower, R. G., Babul, A., Storchi- Cappi, M., et al. 2009, A&A,504, 401 Bergmann, T. 2007, MNRAS,377, 1652 Chartas, G., Brandt, W. N., Gallagher, S. C., Garmire, Ohsuga,K.,Mineshige,S.,Mori,M.,Kato,Y.2009,PASJ, G. P. 2002, ApJ, 579, 169 61, L7 Chartas, G., Brandt, W. N., Gallagher, S. C. 2003, ApJ, O’Sullivan, S.P., Gabuzda, D.C. 2009, MNRAS,400, 26 595, 85 Pounds, K. A., Reeves, J. N., King, A. R., Page, K. L., Chartas, G., Saez, C., Brandt, W. N., Giustini, M., O’Brien, P. T., Turner,M. J. L. 2003, MNRAS,345, 705 Garmire, G. P. 2009a, ApJ,706, 644 Proga, D.,Kallman, T. R.2004, ApJ,616, 688 Jet and disc outflows in 3C 111 9 Reeves, J. N., et al. 2009, ApJ, 701, 493 Ram´ırez, J. M., Tombesi, F. 2012, MNRAS,419, L64 Rawlings, S.,Saunders, R.1991, Nature, 349, 138 Spergel, D. N., Verde, L., Peiris, H. V., et al. 2003, ApJS, 148, 175 Tarter, C. B., Tucker, W. H., Salpeter, E. E. 1969, ApJ, 156, 943 Tchekhovskoy,A.,Narayan,R.,McKinney,J.C.2011,MN- RAS,418, L79 Tombesi,F.,Cappi,M.,Reeves,J.N.,Palumbo,G.G.C., Yaqoob,T.,Braito,V.,Dadina,M.2010a,A&A,521,A57 Tombesi,F.,Sambruna,R.M.,Reeves,J.N.,etal.2010b, ApJ,719, 700 Tombesi, F., Cappi, M., Reeves, J. N., et al. 2011a, ApJ, 742, 44 Tombesi, F., Sambruna, R. M., Reeves, J. N., Reynolds, C. S., Braito, V. 2011b, MNRAS,418, L89 Tombesi, F., Cappi, M., Reeves, J. N., & Braito, V. 2012, MNRAS,L413 Vaughan, S., Uttley,P. 2008, MNRAS,390, 421 Wang, J.-M., Cheng, C., Li, Y.-R.2012, arXiv:1202.0059 Xie,W.,Lei,W.H.,Zou,Y.C.,etal.2012,arXiv:1202.5024