Supermassive Black Holes with Adaptive Optics


Motivation Supermassive Black Holes with AO First Results

Motivation

The importance of the MBH-sigma relation

It is now commonly accepted that supermassive black holes (SMBHs) are found in the centres of galaxies and that their mass accretion history is linked to the formation and evolution of their hosts (see de Zeeuw 2004). The most striking evidence for this link is the discovery that black-hole mass (MBH) is tightly correlated with the velocity dispersion sigma of the stellar bulge (Ferrarese & Merritt 2000; Gebhardt et al. 2000). The importance of the MBH~sigma relation is that it connects two quantities on vastly different scales, namely the SMBH on ~10-5 pc scales, with the bulge potential on ~103 pc scales. It provides a powerful constraint on models of galaxy assembly (e.g. Haehnelt & Kauffmann 2000). It can also be used to provide reasonably accurate estimates of MBH from ground-based measurements of sigma.

Demographic Limitations of SMBHs

The universality of this relationship, however, is not yet established. Of the 31 MBH measurements listed in the compilation of Tremaine et al. (2002), 18 are for ellipticals, 9 for lenticulars, and 4 for spirals (Figure 1). The MBH~sigma relation is therefore still based on a limited sample of galaxies, biased towards early types. Several questions remain unanswered. Is the relation the same for ellipticals and spirals? Does it hold for all disk galaxies, barred and unbarred? Furthermore, what is the intrinsic scatter of the relation? Clearly, the tighter the MBH~sigma relation the greater the challenge for theorists and the greater its usefulness as an MBH estimator. Even the slope of the MBH~sigma relation is still disputed (Ferrarese & Ford 2004).

Figure 1: The MBH~sigma relation (dashed line, according to Tremaine et al. (2002), along with the MBH measurements usedto derive it, coded according to the adopted tracer for the gravitational potential (left) or to the different type of SMBHs host galaxies (right). Notice the shortage of galaxies at both end of the MBH~sigma relation.

Underlying Reasons of these Limitations

There is a simple reason why the sample of galaxies with good MBH measurements is small and biased. In order to measure MBH reliably, one needs to be able to resolve the kinematics of stars or gas on spatial scales DBH~2GMBH/sigma2, a crude estimate of the diameter of the SMBH's sphere of influence. Even for reasonably massive SMBHs in nearby galaxies DBH is a few tenths of an arcsec, so that almost all existing measurements of MBH rely on kinematics obtained from the Hubble Space Telescope (HST). HST, however, is a 2.4m telescope, making it prohibitive to obtain high-S/N stellar kinematics of the relatively dim central regions of giant ellipticals (e.g. Ferrarese 2003). It has no near-IR spectroscopic facilities to probe the dusty nuclear regions of spiral galaxies (e.g. Carollo et al. 1997). These problems could be circumvented by using the gas kinematics, although not all galaxies have central emission and even when present, gas very rarely show clean circular disk-like motions (Sarzi et al., 2001; Ho et al., 2002).

Given the limitations of HST and that gas kinematics is seldom regular, it is not surprising that the current SMBH census is deficient in galaxies with MBH≤5x107Msun (most likely to be in dusty spiral bulges) and MBH≥109Msun (only possible with HST if there is a clean gas disk). With such short leverage in sigma, it is clear why the slope of the MBH~sigma relation has not yet been firmly established.

Completing the Demography of SMBHs with Adaptive Optics

The long-awaited advent of adaptive-optics assisted near-IR spectroscopic facilities on 8m-class telescopes (e.g. NAOS-CONICA and SINFONI at VLT; and NIRI-ALTAIR at Gemini), finally gave us the opportunity to complete once and for all the SMBHs census in nearby galaxies.
The ability to correct for the atmospheric turbulence and reach spatial resolution similar to HST, combined with the larger collecting power of ground-based telescopes and near-IR spectrographs, indeed makes it possible to measure MBH in precisely the kinds of galaxies that are currently underrepresented in the MBH~sigma relation, namely giant ellipticals and spiral galaxies.

For adaptive-optics correction to work, however, it is necessary to have a bright reference point-source very close to the science target. Until very recently this was only possible by using natural guide stars. Unfortunately, this considerably limits the number of potential targets, since only for very few galaxies there is a sufficiently bright star in the Milky Way that happen to be, in projection, close enough to galaxy nucleus.
To search for reference stars we have have both searched the HST archive and cross-correlated the all-sky 2MASS point and extended source catalogues. The 2MASS survey provides near-IR images to compute exposure times, wherase the HST images let us to check for central isophote twists indicating triaxial structures that would complicate the interpretation and to xplore our MBH sensitivity limit.
We have already identified sufficient targets to probe both extremes of the MBH~sigma relation, and Figure 2 shows our five best targets from the HST archive selection only. Figure 2 also shows their position in the MBH~sigma plane.

Figure 2: HST images of our 5 best AO targets, and their position in the the MBH~sigma plane, should they follow exactly the MBH~sigma relation (central plot).

With VLT, we have already observed the giant elliptical NGC1399, and are in the process of completing our observations for the spiral NGC3705 and the lenticular NGC2865, which were severely hampered by bad weather conditions.
With Gemini we have also started to observe the other giant elliptical in our sample, NGC1265. Figure 3 shows, quite spectacularly, how well the NIRI-Altair near-IR image compares to HST images at shorter wavelengths, The quality of the AO correction can be appreciated by noticing the similar witdh of the reference stars in both images. Note also how in the near-IR the nucleus of this galaxy is more clearly visible, due to a much reduced dust attenuation. See also this web-page for a preliminary analysis of these data.

Figure 3: Gemini AO-corrected NIRI-Altair Ks-band image (right) compared with an I-band HST WFPC2 image. The two images should have a comparable spatial resolution given than Gemini is larger than HST but that the NIRI-Altair images were taken at a longer wavelength than the WFPC2 images.

First Results
A SMBH in NGC1399
A Nuclear Eccentric Disk in NGC1399?

A 1.2x109 Msun SMBH in NGC1399

In Houghton et al. (2006) we have recently published our NACO AO-assisted observations for the central stellar kinematics of NGC1399 and, using state-of-the-art modelling techniques, detected a 1.2x109 Msun SMBH. This is an unprecedented achievement in an external galaxy. Figure 4 and 5 show, respectively, the NACO kinematics in the very central regions of NGC1399 and the confidence level for stellar mass-to-light ratio and MBH, the only parameters in our dynamical models.
Furthermore, the MBH we found for NGC1399 is consistent with the expectation of the MBH-sigma relation, at the very high-values end of the MBH-sigma plane.

Figure 4 (left): Stellar surface brightness, velocity, and velocity dispersions from the NACO long-slit observations of NGC 1399. Previously published low-resolution data are shown as orange triangles (Longo et al. 1994) and blue squares (Graham et al. 1998). The PA of the slit was 5.06º so that the positive radius is approximately northward. Note the counter rotation in the central ~0''.3, and the central minimum in velocity dispersion.

Figure 5 (right): Chi2 confidence contours in the mass-to-light ratio (V-band) and MBH around the location of the best values for these parameters. Successive contour level show DChi2=1 intervals.


A Nuclear Eccentric Disk in NGC1399?

Our data also revealed the presence of an intriguing elongated nuclear structure (Figure 6). The observed feature is unlikely to be a double nucleus resulting from the presence of a black-hole binary (BHB). Dynamical friction causes a BHB to harden and reach an unresolvable separation relatively quickly (in 106 - 107 yr, Milosavljević & Merritt 2001), implying that in NGC1399 we would have catched the BHB evolution at a very special time. The possibility of a nuclear bar can also rejected, bearing in mind that bars are quickly destroyed by central mass concentrations (in few dynamical time-scales, Hasan & Norman 1990) and that the observed structure is deeply embedded in the SMBH sphere of influence (r~0.''3). Instead, the flat structure of Figure 6 is more likely to be an eccentric disk, since the structure is elongated preferentially towards one direction (SE) and because eccentric disks can survive only in a Keplerian potential (Tremaine 1995). The presence of an eccentric disk is also supported by the finding in our NAOS-CONICA data of two kinematical signatures of a dynamically cold structure, a central decrease of the velocity dispersion and a central velocity gradient (Figure 4).

Figure 6: NACO image of the nucleus of Galaxy NGC 1399. Isophotal ellipses are overplotted in blue, whereas the white horizontal lines show the location of the slit. Note the elongation of the nucleus to the SE. Each pixel represents 27mas.

Despite these indications, the quality of our NACO image is admittedly poor, since it was taken in only 40s. To confirm the presence of a nuclear eccentric disk we have recently been awarded VLT time.
The finding of a nuclear eccentric disk in a giant elliptical like NGC1399 could be a very important result. Nuclear eccentric disks could be a very common feature in the nuclei of early-type galaxies (Lauer et al. 2005). Their presence, combined with a limited spatial resolution, could indeed explain the finding that a considerable fraction of galaxies host off-centered nuclei.
Yet, only one nuclear eccentric disk has so far been found, in our closest galactic neighbour the Andromeda galaxy. The elongated structure observed in Figure 6 could therefore be the first eccentric disk ever found in an early-type galaxy like NGC1399, and support the claim of Lauer et al. Furthemore the finding of an eccentric disk will allow to obtain an independent constraint on MBH, as in the case of M31 (Peiris & Tremaine 2003; Salow & Statler 2004), which will be very important for a galaxy like NGC1399 that is sitting in a rather extreme position of the MBH-sigma plane plane.

Telescope Time Allocated to our Program

Collaborators in this project are Roger Davies, Niranjan Thatte, Ryan Houghton, John Magorrian, Davor Krajnović, James Binney (Oxford), Marc Sarzi (Hertfordshire), and also Tim de Zeeuw and Michele Cappellari (Leiden).

References

Back home
Last modified: Mon Jul 24 10:26:33 GMT 2005