Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 335, 449-462 (1998)

Previous Section Next Section Title Page Table of Contents

6. Discussion

The model of stellar populations constructed in this paper for the galaxy M 81 is based on a large body of observational data. Surface photometry along the major and minor axis in the UBVRI colours, rotation velocities, line-of-sight velocity dispersion data, distributions of gas and young stars, distribution and kinematics of globular clusters and satellite galaxies have been taken into account for decomposing the galaxy into subsystems and for the determination of population parameters.

On the basis of a smaller set of observational data several models were constructed earlier.

Monnet & Simien (1977) constructed a two-component model for M 81 ([FORMULA] bulge + exponential disk) on the basis of Brandt et al (1972) B-colour photometry and Rots (1975) HI velocities. The [FORMULA] ratio for the bulge was determined to lie between 5.6 and 11, for the disk between 10 and 14.

A different two-component model (Brandt spherical model for the bulge + exponential disk with a central hole) was constructed by Rohlfs & Kreitschmann (1980). As observational data they used the surface photometry from Brandt et al. (1972) and Schweizer (1976), the rotation velocities from Goad (1976), Rots (1975) and their own measurements. The structural parameters of the components were determined by fitting the model calculations with the observed rotation curve. The [FORMULA] ratios were derived by comparison with the light distribution in BV-colours along the major axis. Their model represents more adequately the observed rotational velocities for [FORMULA] 6 kpc. The [FORMULA] derived from this model were between 3 and 20 for the bulge, between 6 and 11 for the disk.

Kent (1987) used his own r-colour photometry and rotation velocities measured by Rots (1975) and reprocessed by Visser (1980). In his model photometry along major and minor axis was used. The disk was assumed infinitely thin. Near the centre Kent decided not to use rotation velocities measured by Goad (1976) emphasizing that corrections for noncircular motions are very uncertain. For that reason the [FORMULA] for the bulge was also uncertain. For the disk [FORMULA] 3.7 (corresponding to [FORMULA]).

When modelling the spiral structure of M 81 Lowe et al. (1994) constructed also a mass distribution model. They used the same observational data as Kent (central kinematics was interpreted slightly differently), but in addition also the morphology of spiral arms. The resulting [FORMULA] was the same for the bulge and the disk [FORMULA] 4.0 (giving [FORMULA] 5.2).

It is seen that within the errors and especially when taking into account model differences all these values coincide with those of our model. We would like to point out two differences.

Firstly, difference in determining the mass of the spheroidal component. In our model the masses of the inner spheroid components (core and bulge) were determined mainly on the basis of the velocity dispersion data. In the models referred above the bulge mass was determined on the basis of the inner part of the rotation curve. Thus, the differences in calculated bulge [FORMULA] ratios are not surprising.

In studying the UV morphology of the central parts of M 81 (the core and the bulge in our notation) Reichen et al. (1995) noted variations in major axis position angle at [FORMULA] ([FORMULA]  kpc). Thus it is possible that the core/bulge may have a triaxial symmetry. Together with noncircular motions it may explain why the observed rotational velocities at innermost points deviate from the model calculations. According to Devereux et al. (1995) the existence of shock waves can be used in explaining also specific [FORMULA] morphology and hot dust temperature in the bulge of M 81.

Secondly, our model includes a DM component - the corona (only the model by Lowe et al. (1994) includes also DM). On the basis of the rotation curve only it is not possible to determine all three free parameters of the DM distribution. Due to the limited extent of the rotation curve only the central density of the DM component can be determined. For this reason, we used additional data on the distribution and kinematics of satellite galaxies of M 81. Thus the DM distribution parameters are restricted with sufficient precision.

Our model gives for the mean line-of-sight velocity dispersion of globular clusters [FORMULA] 130 km/s. This is less than measured by Perelmuter et al. (1995) [FORMULA]  km/s. Because globular clusters lie in the gravitational potential of the whole galaxy, increasing of the mass of the halo or the mass of the DM corona within reasonable limits does not reduce the discrepancy significantly. In order to have the measured dispersion the mass within [FORMULA]  kpc (outer extent of globular clusters population) must be [FORMULA]. This mass gives the circular velocity at 20 kpc approximately 220 km/s. Such a high value is in conflict with the observed rotation curve. For this reason we were satisfied with the dispersion calculated from our model.

The model without DM component has the following deficiencies:

  1. it is in conflict with the M 81 group velocity dispersion;
  2. it is in conflict with the rotation curve for [FORMULA] 16 kpc;
  3. calculated velocity dispersion of globular clusters is 100 km/s which is in quite serious conflict with the measured value.

For this reason we conclude that the best fit with observational data is obtained in models which include a DM component.

Rather high axial ratio [FORMULA] may indicate that our old disk is a mixture of an old thin disk and a thick disk (see Wyse & Gilmore 1995). Indirect argument for the presence of a thick disk is also the chemical composition gradient of HII regions measured by Henry & Howard (1995). Unfortunately, in this study we cannot give pro or contra arguments about the thick disk in M 81.

Finally, we would like to discuss the nucleus. The mass of the nucleus was determined on the basis of the velocity dispersion 250 km/s measured by Bower et al. (1996) (see Sect. 3.1). The point mass in the centre of our model has [FORMULA], corresponding to the [FORMULA] of the stellar nucleus. This BH mass is significantly larger than it was estimated by Ho et al. (1996) [FORMULA]. In principle, these dispersion measurements may be handled as being only preliminary results. However, even when we accept the `standard' M 81 central velocity dispersion (McElroy 1995) 170 km/s within [FORMULA] the [FORMULA] ratio of the stellar nucleus, corresponding to the point mass [FORMULA], is [FORMULA]. This value also disagrees with [FORMULA] resulting from chemical evolution models of stellar populations and, we rejected it. Thus we decided in our final model to start with [FORMULA]. The BH mass in our model is in quite good agreement with the general trend of BH mass versus luminosity of the spheroid (Ford et al. 1998).

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1998

Online publication: June 18, 1998