Astron. Astrophys. 324, 80-90 (1997)

2. Observations and data reduction

2.1. Sample and observations

For a statistical study of the influence of galaxy interactions on the z -structure of the disks of involved galaxies, we selected a sample of apparently edge-on spiral galaxies belonging to strongly interacting systems. Our sample consists of 24 interacting systems containing at least one edge-on galaxy. The sample is relatively complete since we included practically all known interacting systems that are suitable in angular diameter and could be observed during our observational run (see Paper I). All interacting systems (as far as 7 isolated galaxies) were observed at the OHP 1.2 m telescope in the B, V and I passbands. General photometric results of the observations (including isophotal maps of all objects) are presented in Paper I.

2.2. Reduction

We considered photometric cuts (4-6 typically) along minor axes of the sample galaxies at different galactocentric distances. For the galaxies with warped disks the directions of the minor axes were determined locally as a perpendicular to local major axis direction. The positions of the galaxy planes were determined by averaging the vertical profiles and under assumption of symmetrical light distributions with respect to the planes. In general the vertical cuts of interacting galaxies look quite symmetric (especially in the I passband). Comparing asymmetry of the averaged profiles of interacting galaxies with asymmetry of the minor axis profiles in our sample of isolated spirals, we concluded that interacting and normal galaxies are distributed approximately in the same range of inclinations with respect to the line of sight. Following Guthrie (1992), we found that our sample isolated galaxies are, on average, within from edge-on orientation and, therefore, concluded that most interacting edge-on galaxies are in the same range also. Let us note also that according to van der Kruit & Searle (1981a) and Barteldrees & Dettmar (1994) moderate ( - ) deviations from edge-on orientation do not significantly change the slope of vertical surface brightness distribution. Moreover, the control sample of edge-on isolated galaxies is also contaminated by not exactly edge-on objects, so this systematic effect is somewhat compensated when both samples - interacting and non-interacting - are compared.

By statistically studying the disk thickness of edge-on interacting galaxies, we did not consider question about best fitting function for each galaxy and fitted all the averaged vertical profiles by standard law (van der Kruit & Searle 1981a), where z is the distance from the galaxy plane and is the scale height. (At large z a comparison between value and exponential scale height is possible via 2 .) We choose this function in order to have the largest possible comparison sample of normal galaxies with uniformly determined scale heights. We found that galaxies with published values are predominant in the literature (e.g., van der Kruit & Searle 1981a,b and 1982a,b (vKS); Barteldrees & Dettmar 1994 (BD)). It should be noted also that " -distribution" gives quite satisfactory (within ) approximation for most of the sample. This can be understood, since this is the distribution of a self-gravitating isothermal layer of stars; and here the stellar component is representing most of the mass (the gas mass is negligible, and the dark halo mass is very small within the optical disk), and there is only small departures from z-isothermality (e.g. van der Kruit 1988). Close to the plane, the stars are cooler, but the vertical dispersion at worst can be represented by a vertical density profile (instead of a , cf Bottema 1993).

We excluded central regions of the galaxies from our study in order to avoid the bulge light contribution to the fitted profiles. After inspection of the radial surface brightness distribution, we considered vertical cuts at radii along the major axis where the bulge influence is negligible. From the other side, in order to have reasonable photometric profiles with amplitude (difference between central surface brightness and faintest level of the cut) larger than we excluded faint outer regions of the galaxies. We find that our vertical cuts are distributed, on average, between 1 and 2.4 exponential scalelengths h (see below) of the galaxy disks in the I passband.

For some of our sample galaxies the seeing is rather bad (larger than 2 ). The seeing effects do not strongly affect the slopes of the surface brightness profiles (see, for instance, Nieto et al. 1990) and cannot noticeably influence our results. To check the size of this effect, we compared scale height values determined from the original images of the galaxies and after Lucy-Richardson restoration (for this we used standard MIDAS routine, the PSFs were constructed from the star images in the corresponding frames). We found that original frames give about 20% systematically larger values of scale heights for the thinnest (in comparison with star images) galaxies. For most of our objects the effect of seeing is insignificant. Therefore, we corrected values of some flat galaxies (with about or less 2 ) for this effect (correction of at most 20%).

To study the structure of galaxies in the radial direction, we extracted major axis profiles of all galaxies. Excluding central bulge-dominated regions, we determined exponential scalelengths by fitting outer parts of the profiles. The B band radial cuts of our sample interacting galaxies often look peculiar and asymmetric and does not provide reasonable fit. Therefore, we will use the I band derived scalelengths in the following discussion (photometric profiles in the I filter are significantly smoother and more regular). Let us note also that our exponential fit of the sample galaxies does not show any systematical difference with literature data. For instance, average central surface brightness of the disks in the B band () is close to the canonical Freeman's (1970) value. The mean B to I scalelength ratio () is in good agreement with de Grijs & van der Kruit (1996) (dGvK) value ().

Fig. 1a and b. The scale height distributions for the sample galaxies as a function of radius along the major axis. Both axes are in kpc (in arsec for Arp 208 and UGC 11230). Circles represent the data in the I passband, squares - V, stars - B. The figures orientations is such that the distance increases from S to N or from E to W. We use general name of interacting system for the systems with one edge-on galaxy and give more detailed name for the objects with less clear identification (see Fig. 1 in Paper I).

Fig. 1c. (continued)

© European Southern Observatory (ESO) 1997

Online publication: May 26, 1998

helpdesk.link@springer.de