Black Holes in 4 Nearby Radio Galaxies
Jeremy Mould, Tony Readhead, Garret Cotter, David Batt, Mark Durre'
aa r X i v : . [ a s t r o - ph . GA ] O c t BLACK HOLES IN 4 NEARBY RADIO GALAXIES
Jeremy Mould , Tony Readhead , Garret Cotter , David Batt and Mark Durr´e [email protected] Received ; accepted Centre for Astrophysics and Supercomputing, Swinburne University of Technology,Melbourne, Victoria 3122, Australia School of Physics, University of Melbourne, Parkville, Victoria 3010, Australia California Institute of Technology, CA91125, USA Department of Physics, University of Oxford, Denys, UK 2 –
ABSTRACT
We study the velocity dispersion profiles of the nuclei of NGC 1326, 2685,5273 and 5838 in the CO first overtone band. There is evidence for a black hole(BH) in NGC 1326 and 5838. Gas is seen flowing out of the nuclear region ofNGC 5273. We put upper limits on the nuclear BHs responsible for its activityand that of NGC 2685.
Subject headings: infrared: general — active galactic nuclei – galaxies: elliptical –radiosources – black holes
1. Introduction
Understanding activity in galactic nuclei requires high spatial resolution. Kormendy &Richstone (1995) have outlined the techniques for quantifying the supermassive black holesthat power active galactic nuclei (AGN). Our strategy (Mould et al 2012) is good seeinginfrared spectroscopy of AGN in a volume limited sample, followed by adaptive opticsspectroscopy on large aperture telescopes. In this paper we present Palomar TripleSpecspectra of a number of nearby radiogalaxies of early type.NGC 1326 is a ring barred S0 galaxy in the Fornax cluster with circumnuclear starformation (Buta et al 2000). Our second galaxy is a Hubble Atlas polar ring galaxy, anS0 Seyfert 2. Schinnerer & Scoville (2002) detected four giant molecular cloud associationswithin the polar ring in NGC 2685 (the Helix) with of order 10 M ⊙ of molecular hydrogen.Dust has been detected with Spitzer in our third S0 galaxy, NGC 5273, totalling 2.5 × M ⊙ by Martini et al (2013). NGC5838 has a nuclear star cluster of 5 × M ⊙ (Scott &Graham 2013).
2. Sample and observations
We have drawn our radiogalaxy sample from Brown et al (2011), further limiting thedistance to 20 Mpc in order to have 100 pc resolution in 1 ′′ seeing. Observations of NGC1326, 2685, 5273 & 5838 were obtained on the Hale Telescope in 2011 and 2012. Obtainingour Palomar TripleSpec spectra was described by Mould et al (2012) and data reduction wasoutlined by Batt et al (2014, Paper I). We very briefly recap this here. The spectrographhas resolution of 2600 with a 1 ′′ slit, and observations were made with the nucleus in twoslit positions ABBA in 4 × f xcor wasused for this purpose with the Gemini library stellar template HD2490 interpolated to thesame resolution.Table 1 gives the radial position of the extracted spectrum in column (1), the pixelshift between that and the template in column (2), the peak height of the cross-correlationin column (3) and the FWHM of the fit to the cross-correlation in column (4). The units ofcolumns (2–4) are pixels. 5 –Table 1. Raw crosscorrelation data NGC 2685 position pixel shift peak fwhm position pixel shift peak fwhm(arcsec) [2] [3] [4] (arcsec) [2] [3] [4]0 -189.96 0.23 15.4 0 -190.55 0.21 12.160 -190.68 0.26 14.96 0 -190.7 0.25 15.10.9 -190.5 0.26 14.2 1.63 -187.09 0.24 14.80.73 -188.4 0.27 13.9 0.79 -188.99 0.18 14.980.79 -190.65 0.26 13.41 0.84 -189.96 0.2 12.70.79 -191.01 0.2 12.3
NGC 5838 position peak height fwhm shift position peak height fwhm shift(arcsec) [2] [3] [4] (arcsec) [2] [3] [4]0 0.366 27.4 -0.76 0 0.519 29.1 -0.670.316 0.329 27.1 -1.31 0.632 0.301 13.1 -0.10.948 0.333 14.2 -0.6 1.264 0.368 10.2 -0.411.58 0.331 10 -0.27 1.896 0.314 8.18 -0.082.212 0.322 11.4 -0.28 0.316 0.325 13.1 0.550.632 0.391 8.35 -0.03 0.948 0.398 8.99 0.2491.264 0.476 10.6 0.01 1.58 0.462 12.3 0.41.896 0.484 18.6 0.89 2.212 0.503 14.3 1.010.316 0.397 13.7 -0.13 0.632 0.398 14.6 -0.4
NGC 5273 position peak fwhm position peak fwhm(arcsec) [2] [3] (arcsec) [2] [3]0 0.28 28.36 0 0.22 15.60 0.28 23.22 0 0.22 20.631.57 0.13 21.09 0.73 0.26 17.121.99 0.23 16.76 1.09 0.23 24.181.58 0.12 11.94 0.79 0.28 20.571.69 0.12 14.22 0.9 0.2 11.311.18 0.18 21.12 1.97 0.16 7.220.54 0.19 22.96 1.33 0.1 21.680.84 0.29 27.98 1.63 0.13 15.21.07 0.24 24.97 1.86 0.19 14.78
NGC 1326 position peak fwhm position peak fwhm0 0.4 20.61 0 0.39 20.560 0.4 20.13 0 0.42 20.840.78 0.43 22.4 0.56 0.41 21.94
3. Kinematics and dynamics3.1. NGC 1326
NGC 1326 has been imaged by the Hubble Space Telescope (Figure 1) and itsultraviolet light distribution is displayed in the radial profile from the IRAF STSDASsurface photometry task ellipse in Figure 2.The Jeans equation allows us to predict the velocity dispersion profile σ (r) correspondingto this light distribution, assuming spherically distributed stars on isotropic orbits. To dothis, we need the logarithmic derivatives with respect to radius of the density and velocitydispersion profiles. The former is obtained numerically using an Abell transform, the latterby calculating the (small) slope of the velocity dispersion data. The visual mass to lightratio is a free parameter in this model and we fit it to the data at r >
80 pc, finding M/L= 6.5 in solar units, a normal value for a stellar population not dominated by dark matter.TripleSpec line width values were normalized in the same way as in Paper I.The addition of a 1 × M ⊙ black hole modifies the mass distribution and σ (r). It isa better fit to the data than the solid line in the lower part of Figure 2. The no BH modelis ruled out with 70% confidence based on χ . NGC 2685 is a polar ring galaxy, known as ‘the spindle’. The HST nuclear image isreproduced in Figure 3 and the light distribution has been fitted with a ‘nuker profile’(Lauer et al 2007). The profile appears in Figure 4, the model fit, and χ per degree offreedom implies that M • > × with less than 20% probability. This is consistent withBeifiori et al (2009), who find an upper limit M • < × M ⊙ . The innermost datapoint 8 –has been located, not at zero radius as Table 1 would imply, but at the effective light centreof the zero radius observation taking account of seeing. We fitted a nuker profile to archival HST WFPC2 PC data (Figure 5), obtaining( α, β, γ ) = (1.8, 1.8, 0.75) and normalized the profile to the surface photometry of Mu˜nozMarin et al (2007) with r b = 50 pc. Figure 6 is the model fit, and χ per degree of freedomimplies that M • > with less than 25% probability. We assumed the Tonry et al (2001)surface brightness fluctuations distance of m-M = 31.09 ± L ⊙ ) x-raygas (Liu 2011) is a P-Cyg profile of cooler (kT ∼
30 eV) neutral gas with a terminal outflowvelocity of 750 km/sec. This object will repay IFU study of its circumnuclear gas andmodelling to determine the outflow rate.
Calculation of a predicted stellar velocity dispersion profile was described for galaxieswith nuker profiles in Paper I. NGC 5838 has such a profile (Lauer et al 2007). Figure 7shows the nucleus of NGC 5838 and Figure 8 shows a fit with M/L = 30 and a black holeof 1 × M ⊙ . Note that Lauer et al assume V-H = 2.39 in converting NICMOS datato visual magnitudes. We also adopted their distance of 22.2 Mpc. The no BH model isrejected with 98% confidence. 9 –
4. Summary
We summarize our findings in Table 2. In two cases we have SMBH detections; in twocases we have upper limits on the SMBH mass. Our upper limit for NGC 5273 is consistentwith the result from reverberation mapping of 4.7 ± × M ⊙ by Bentz et al 2014.Figure 9 shows our 4 radio galaxies in their Magorrian diagram. NGC 5838 is plotted at σ = 290 km/sec (McElroy 1995). Table 2: Black hole masses
Name Type Distance M V M/L SMBHNGC (Mpc) M ⊙ N1326 SB0+ 20.5 –21.05 6.5 1 × N2685 SB0+ 14.3 –19.72 1.3 < ∗ × N5273 S0 20 –20.1 1 † < × N5838 S0- 22.2 –20.51 30 1 × *1.1 (Beifiori et al 2009) † UV M/LWe thank our referee for comments that improved the paper. We are grateful forthe support of the Australian Research Council through DP140100435. GC acknowledgessupport from STFC grant ST/K005596/1. Spectra were extracted using a version of theSpextool program modified for the Palomar TripleSpec Spectrograph (Cushing et al. 2004,; M. Cushing, private communication 2011). We acknowledge the Hubble Legacy Archive,a facility of STScI, which is operated by AURA for the National Aeronautics and SpaceAdministration (NASA). This research has made use of the NASA/IPAC Extragalactic 10 –Database (NED) which is operated by the Jet Propulsion Laboratory, California Institute ofTechnology, under contract with NASA. This research has also made use of IRAF, softwarewritten by NOAO and data products from the Gemini Observatory, which are operated byAURA under a cooperative agreement with NSF. David Batt was a summer student atSwinburne University while this work was carried out. 11 –Table 1—Continued
12 –
REFERENCES
Batt, D. et al 2014, in prep.Bentz, M. et al 2014, ApJ, in press, astro-ph 1409.5794Buta, R. 2000, AJ, 120, 1289Beifiori, A. et al 2009, ApJ, 692, 856Brown, M. et al 2011, ApJ, 731, L41Cushing, M., Vacca, W., & Rayner, J. 2004, PASP, 116, 362Durr´e, M. & Mould, J. 2014, ApJ, 784, 79Kormendy & Richstone 1995, ARAA, 33, 581Lauer, T. et al 2007, ApJ, 664, 226Liu, J. 2011, ApJS, 192,10Martini, P. et al 2013, ApJ, 766, 121McElroy, D. 1995, ApJS, 100, 105Mould, J. et al 2012, ApJS, 203, 14Mu˜noz Mar´ın, V. et al 2007, AJ, 134, 648Schinnerer, E. & Scoville, N. 2002, ApJ, 577, L103Scott, N. & Graham, A. 2013, ApJ, 763, 76Tonry, J. et al 2001, ApJ, 546, 681This manuscript was prepared with the AAS L A TEX macros v5.2. 13 –Fig. 1.— The HST WFPC2 PC image of the nucleus of NGC 1326, fov 36 ′′ , orientation Nup and E to the left. The filters in RGB order are F814W/F439W/F255W.Fig. 2.— top UV radial surface brightness profile of NGC 1326 converted to V band by sub-tracting the NED U-V colour 1.15; bottom
E and W velocity dispersion profile distinguishedby the symbols. The solid line is a fit to the data with M/L = 6.5. The dashed line supposesthe presence of a 10 M ⊙ black hole. Velocity dispersion uncertainties are a similar siz tothe plotting symbols. 14 –Fig. 3.— The HST ACS HRC image of the nucleus of NGC2685. The filter is F330W.Fig. 4.— top NGC 2685’s nuker profile; bottom a fit to the TripleSpec data without andwith (dashed line) a 1.2 × M ⊙ SMBH. 15 –Fig. 5.— The HST WFPC2 PC image of the nucleus of NGC5273. The filters in RGB orderare F547M/F300W/F218W. 16 –Fig. 6.— top the two data points are from Mu˜noz Marin et al 2007; bottom the dashedline has a SMBH of 1 × M ⊙ . Error bars are proportional to the cross correlation peakheights in Table 1. 17 –Fig. 7.— The HST WFPC2 PC image of the nucleus of NGC5838. The blue and red filtersare F450W and F814W. NASA: Hubble Legacy Archive.Fig. 8.— The lower figure compares a fit to NGC 5838 with M/L = 30 (solid line) andadding a SMBH with mass 1 × M ⊙⊙