A Massive Object in the Core of NGC 5055 ?

Sébastien Blais-Ouellette

Département de physique and Observatoire du mont Mégantic,
Université de Montréal, C.P. 6128, Succ. centre ville,
Montréal, Québec, Canada. H3C 3J7
Electronic-mail: blaisous@astro.umontreal.ca

Philippe Amram

IGRAP,Observatoire de Marseille,
2 Place Le Verrier,
F-13248 Marseille Cedex 04, France
Electronic-mail: amram@observatoire.cnrs-mrs.fr

and

Claude Carighan

Département de physique and Observatoire du mont Mégantic,
Université de Montréal, C.P. 6128, Succ. centre ville,
Montréal, Québec, Canada. H3C 3J7
Electronic-mail: claude@astro.umontreal.ca

Abstract:

In a global kinematical study of NGC 5055 using high resolution Fabry-Perot interferometry, intriguing spectral line profiles have been observed in the center of the galaxy. These profiles seem to indicate a rapidly rotating disk with a radius near 365 pc and tilted 50 $\deg$ with respect to the major axis of the galaxy. In the hypothesis of a massive dark object, a naive keplerian estimate gives a mass between 10^7.2 to $10^{7.5} M_\odot$.Unfortunately the limited spectral domain of the Fabry-Perot leaves some ambiguity on the exact movement and velocity of this H$\alpha$ emission. 2-D spectroscopy with a larger spectral range (eg.: TIGRE or low resolution Fabry-Perot) is thus required.

Introduction

  It is now well established that many if not all galaxies hide a massive object in their central region (Kormendy 95). Presence of such objects are usally deduced from the kinematics and photometry of the core of these galaxies. These Massive Dark Objects (MDO), tought to be black holes, produce normally a high velocity dispersion or a rapid rotation around... nothing (van der Marel et al. 1997).

With its high spectral and spatial resolution, the Fabry-Perot interformeter is well suited for the kinematical study of extended objects like spiral galaxies. NGC 5055 (M63) is a bright Sbc galaxy classified as a LINER in which we wanted to study the detailed kinematical structure of the H$\alpha$ emission. In the process, our attention have been caught by the very central part of the galaxy...

Observations

  The Fabry-Perot observations of the H$\alpha$emission line were obtained in March 1998 at the Canada-France-Hawaii Telescope (CFHT). The Fabry-Perot etalon (CFHT1) was installed in the CFHT's Multi-Object Spectrograph (MOS). A narrow-band filter ($\Delta \lambda$ = 10Å), centered at $\lambda_0$ = 6574Å (nearly at the systemic velocity of NGC 5055, V$_{sys} \approx 504$ kms^-1), was placed in front of the etalon. The available field with no vignetting was $\approx$ 8.7$\times$8.7, with 0.34 pix ^-1. The free spectral range of 5.66Å (258 kms^-1) was scanned in 28 channels, giving a resolution per channel of 0.2Å (9.2 kms^-1). Integration time was 565 seconds per channel.

After reduction (see Amram 91 for details), we ended up with a 3-D data set with x,y and $\lambda$ as axis. Velocity maps are then obtained using the intensity weighted mean of the H$\alpha$ peak to determined its $\lambda$position thus the radial velocity for each pixel.

Kinematics of the Center

Globally, the galaxy rotates smoothly and without noticible asymmetry although some redder flux seems to be missing. This could be due to a possible blueshift of the passband of the (old) filter. The H$\alpha$line is normally symmetrical and well defined where the flux is sufficient.

When we get to the central 5 arcsecond ($\sim$ 110 kpc), things are changing radically. In a region where H$\alpha$is normally rare, two bright spots are visible each side of the exact photometric center of the galaxy (Figure 1). Even more interesting are the antisymmetrical appearance of the spectrum of the two blobs (Figure 2). When looking at a Fabry-Perot spectrum, one has to keep in mind the intrinsic ambiguity relative to which interference order we are looking at. If the filter is wide enough, two or more orders can even be superimposed (by slice of 5.67 Å in this case). This also means that there is a continuity between the two sides of the spectrum.

 

Figure 1:   Isophotes of the integrated flux superimposed on the velocity field. Line of sight velocity can be 346 or 604 for the south-eastern (blue) blob and 653 or 395 for the north-western (red).

 

Figure 2:   Velocity field with typical profile in each region.

If one looks at the profiles in the two spots, one can clearly see a peak with a long wing on one side and a sharp cut-off on the other. In between an almost symmetrical profile, probably a combination of the profiles from both sides. Because of the ``wraparound'' in the spectrum, it is very difficult to fix the level of the continuum so that absorption features cannot be rejected.

For this central region, velocities have been fixed at the position of the peak of each pixel to avoid being sensitive to the asymmetric morphology of the peak. To relieve some degeneracy of the different order of interference, it as been decided to take one spot being redshifted from the systemic velocity and the other spot being blueshifted. Two possibilities remain. One gives a peak velocity of 653 kms^-1 for the north-western blob and 346 kms^-1 for the south-eastern one. The other assumes a possible (somewhat) conterrotating disk with peak velocity of 395 kms^-1 in the north-west and 604 kms^-1 in the south-east. Separation between these velocities are about one arcsecond (37 pc).

A naive edge-on keplerian model would give for these rotating velocities between 100 and 150 kms^-1 at 18.5 kpc from the center of rotation, a MDO mass between 10^7.2 and $10^{7.5} M_{\odot}$.

Conclusion

The Fabry-Perot data presented here were optimised for the observation of a large, moderately rotating galaxy. There is then no surprise if many sources of errors and ambiguities are present when one try to extract valuable information from a few tens of pixel in a dynamically very active region.

Obviously, the ambiguity on the real observed wavelength is very annoying but managable at the cost of a supplementary hypothesis of a rotation around the systemic velocity. More disturbing is the superposition of many order of interference since it is forbidding us to fix the real continuum level and rule out an absorption effect that could cause the observed profiles.

On the other side, the symmetrical shapes of the profile clearly indicates that it is not a systematic error like a drift or a photometric variation. The high H$\alpha$fluxes involved is also a sign that we are in presence of a quite big amount of energy compatible with the presence of a MDO.

Overall, this study shows the interest and the necessity of more adapted observations using integral field spectroscopy where one could trade some field of view for a larger spectral domain and a similar resolution than this high resolution Fabry-Perot data.