3 Observations and data reduction

3.1 Optical observations

Due to the large number of galaxies needed for reliable statistics the optical observations were obtained with different telescopes and during several observing runs between February 1996 and June 1998. The following telescopes were used:

1.54 m Danish and 61 cm Bochum telescope on La Silla, Chile; 42-inch telescope at Lowell Observatory, Flagstaff; 1.23 m telescope at Calar Alto Observatory, Spain; 1.06 m telescope at Hoher List Observatory, Germany. The used passbands were Johnson R and Thuan & Gunn r, with a central wavelength at $\lambda_{\rm c} = 652$ nm and 670 nm, bandpass $\Delta \lambda = 162$ nm and 103 nm, respectively.

All observing runs and the used telescope/detector characteristics are listed in Table 2. The seeing conditions are given as averaged FWHM values.

Details of the optical observations are listed in Table 4. Additionally for most of the sample galaxies the total blue surface brightness, $B_{\rm T}$, and the morphological type T (revised Hubble type, according to Lauberts & Valentijn [1989], Table 1) are given.

3.2 Near infrared observations

The near infrared observations were obtained with the MAGIC camera of the MPIA attached to the 2.2 m and 1.23 m telescopes of the Calar Alto Observatory, Spain, and with the IRAC2b camera on the ESO/MPIA 2.2 m telescope of the European Southern Observatory (ESO), Chile, respectively. Both the MAGIC and IRAC2b cameras are equipped with a Rockwell $256 \times 256$ $\rm pixel^{2}$ NICMOS3 HgCdTe array. The data were acquired during several observing runs between February 1996 and June 1998.

For all runs we used the K filter (central wavelength $\lambda_{\rm c} = 2.20 ~\mbox{$\space \mu \rm m $ }$, bandpass $\Delta \lambda = 0.40 ~\mbox{$\space \mu \rm m $ }$), or the K' filter ( $\lambda_{\rm c} = 2.15 ~\mbox{$\space \mu \rm m $ }, \Delta \lambda = 0.32 ~\mbox{$\space \mu \rm m $ }$). For some objects images in the H filter were obtained ( $\lambda_{\rm c} = 1.65 ~\mbox{$\space \mu \rm m $ }, \Delta \lambda = 0.30 ~\mbox{$\space \mu \rm m $ }$).

We recorded object and sky frames alternately, with typical single integration times of $10 \times 5$ s per sequence, and spatially separated by $\approx 6'$. To eliminate bad pixels, the telescope "on source''-positions were dithered by a few arcseconds between subsequent exposures. Since many of the sample galaxies are faint objects it was required to reach at least 30 - 40 min integration time "on source'', which corresponds to $\approx 9 - 12$ cycle repetitions and a resulting observing time of $\approx 2.5 - 3$ hours. Therefore, only one third of our total optical galaxy sample was observed in the near infrared.

Since our study is aiming at an investigation of the structure and geometrical properties of galactic disks, we could make use of observations obtained under non-photometric conditions. For that reason the optical and near infrared contour maps of the sample galaxies shown in Figs. 4 and 5 are not flux calibrated.

As for the optical observations, the telescope/detector characteristics as well as a list of all galaxies observed in the near infrared are given in Tables 2 and 4, respectively.

0.8mm

 

 

Table 2: Observing dates and telescope/detector characteristics, given for optical and near infrared observations. Columns: (1) Observing dates; (2) Telescopes used: 2.2 CA = 2.2 m Calar Alto, 2.2 ESO = 2 .2 m ESO/MPIA, 1.5 DA = 1.54 m Danish, 1.2 CA = 1.23 m Calar Alto, 1.1 LO = 1.1 m Lowell, 1.0 HL = 1.06 m Hoher List, 0.6 BO = 0.6 m Bochum; (3) Field of view; (4) Pixel scale; (5) Detector used; (6) Average seeing conditions (FWHM)

Observing		Telescope	Field	Scale	Detector	Seeing
dates			[']	$[''/\rm pix]$		['']
(1)		(2)	(3)	(4)	(5)	(6)
Optical
Mar.	1996	1.1 LO	4.9	0.729 $\mbox{$\space ^{\rm a)} $ }$	TI 800	2.0
May	1996	1.0 HL	27.0	0.825	LORAL 2048	3.0
Jun.	1996	1.2 CA	8.6	0.503	TEK 1024	2.0
Aug.	1996	1.0 HL	27.0	0.825	LORAL 2048	3.0
Sep.	1996	1.2 CA	8.6	0.503	TEK 1024	2.0
Dec.	1996	0.6 BO	4.8	0.496	TH 7882	1.8
Jan.	1997	1.2 CA	8.6	0.503	TEK 1024	2.0
Feb.	1997	1.0 HL	6.8	0.800 $\mbox{$\space ^{\rm a)} $ }$	LORAL 2048	3.2
Apr.	1997	1.5 DA	13.7	0.390	LORAL 2048	1.5
May	1997	1.1 LO	9.9	0.730 $\mbox{$\space ^{\rm a)} $ }$	SITe 2k	2.0
Sep.	1997	1.0 HL	27.0	0.825	LORAL 2048	3.0
Jun.	1998	1.5 DA	13.7	0.390	LORAL 2048	1.6
Near Infrared
Mar.	1996	1.2 CA	5.1	1.200	NICMOS3	1.8
Sep.	1996	2.2 CA	2.7	0.642	NICMOS3	1.8
Jan.	1997	2.2 CA	2.7	0.642	NICMOS3	1.5
Apr.	1997	2.2 ESO	3.0	0.708	NICMOS3	1.6
Feb.	1998	2.2 CA	2.7	0.642	NICMOS3	1.8
May	1998	2.2 ESO	3.0	0.708	NICMOS3	1.5

$\mbox{$\space ^{\rm a)} $ }$ Pixel binning $2 \times 2$.

3.3 Data reduction

The optical data were reduced using the MIDAS software package, developed by ESO. Following the standard reduction procedures (bias subtraction, flat fielding with sky flats) the remaining gradients in the background of galaxies that covered a major fraction of the field of view were removed using a two-dimensional polynomial. For some of the frames that were affected by bad columns these columns were also removed by the standard MIDAS fitting routine. In order to increase the signal-to-noise ratio S/N (important for an investigation of faint disk features) the images of fainter objects were binned ( $2 \times 2$).

Finally, the frames were rotated in such a way that the galaxy planes are in a horizontal position (assuming symmetrical light distribution of the vertical disk profiles). It should be stressed that the remaining small rotation error - which is typically $\pm 0.4 \hbox{$^\circ$ }$for most of the relatively uniform disks of non-interacting galaxies - can be considered by the disk fitting routine (Sect. 4). For galactic disks affected by strong disk warping (mostly interacting objects) precise rotation was more difficult. For these cases, only the inner parts of the galaxy disk were considered to determine the rotation.

For standard reduction of the near infrared data, the IRAF software package was used. In particular, the sky frame subtraction, the flat fielding, and the combining of the flat-fielded images was done with the ARNICA (Arcetri Near-Infrared Camera) add-on package. The sky frames used for the flat field subtraction were obtained from a set of the nearest frames in time, filtered by a median. The median filtering also removed the stars in the sky frames. To produce a final source frame, the reduced, flat-fielded images were combined using the ARNICA standard "mosaic'' task (this task includes both median filtering, and centering of frames by stars in the field).

Due to the small detector size - the resulting field of view was $\approx 3'$ on average (Table 2) - images of larger fields were produced by mosaicing. Since this is a time-expensive procedure, it was only applied in order to obtain images of some larger objects. For precise adjustment of each of the frames we used either stars in the field or the sharp central bulge regions of the galaxies themselves.

Image rotation was applied as explained for the optical images.

Isophote maps of the complete samples of interacting/merging and non-interacting galaxies, respectively, are shown in Figs. 4 and 5 of this paper.