Astron. Astrophys. 359, 907-931 (2000)

6. Conclusions

We have observed 20 galactic globular clusters with multi-color Schmidt plates and films on wide fields. Field and cluster stars are sorted in the color-magnitude plane. A star-count analysis is performed on the color selected stars to study the overdensities that can be attributed to the stars stripped from the globular clusters by tidal shocks (disk/bulge) as well as from internal dynamical evolution. We use the wavelet transform in order to enhance the weak tidal structures at large scales and in order to filter the high background noise at low galactic latitude. After highlighting the observational biases resulting from dust extinction and background galaxy clustering at faint magnitudes, we reach the following conclusions:

• All the clusters observed, which do not suffer from strong observational biases, present tidal tails, tracing their dynamical evolution in the Galaxy (evaporation, tidal shocking, tidal torquing, and bulge shocking).

• The clusters in the following sub-sample (viz. NGC 104, NGC 288, NGC 2298, NGC 5139, NGC 5904, NGC 6535, and NGC 6809) exhibit tidal extensions resulting from a shock, i.e. tails aligned with the tidal field gradient.

• The clusters in another sub-sample (viz. NGC 1261, NGC 1851, NGC 1904, NGC 5694, NGC 5824, NGC 6205, NGC 7492, Pal 5, and Pal 12) present extensions which are likely tracing the orbital path of the cluster with various degrees of mass loss.

• NGC 7492 is a striking case because of its very small extension and its high destruction rate driven by the galaxy as computed by GO97. Its dynamical "twin" for such an evolution, namely Pal 12, exhibits, on the contrary, a large extension tracing its orbital path, with a possible shock which happened more than 350 Myr.

• The velocity diffusion of the stripped stars is tentatively estimated, in one case (viz. NGC 6254), to be similar to the cluster velocity dispersion.

• Thanks to the relatively small distance of NGC 5139 and its high release of unbound stars during its current disk shocking, we estimate the mass loss to be between 0.6 and 1% of the cluster total mass, taking into account a possible mass segregation in the cluster halo. This mass loss rate is consistent with our estimates from N-body simulations (CLM99).

• The second break in the surface density slope, apart from the break at the tidal radius (cf. the case of NGC 6254) could be an indicator of some recent gravitational shocks, with the indicator displaying a range of values between -0.9 and -2. The latter is likely overestimated because of the uncorrected crowding towards the clusters.

The use of better quality data, e.g. wide-field CCD observations, combined with the present star-count method will allow in a near future to get rid easily of the observational biases and to obtain a better color selection thanks to a more accurate photometry. These observations will provide more precise observational estimates of the mass loss rates for different regimes of galaxy-driven cluster evolution. With the help of numerical simulations and accurate proper motions, it will be possible to constrain efficiently the parameters describing the galactic potential (disk scale-height, surface density, bulge size). In case of a flat dark matter halo (Pfenniger et al. 1994), the tidal shock on the globular clusters would be enhanced, depending on the surface density and the scale-height of this dark matter halo. Pal 2 is a good candidate to probe such dark matter halo flattening, because of its small distance to the galactic plane (Z = -2.2 kpc) and its relatively large distance to the galactic center (21.6 kpc), placing this cluster in a region where the tidal shocking by the disk only is expected to be low.

© European Southern Observatory (ESO) 2000

Online publication: July 13, 2000
helpdesk.link@springer.de