Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 360, 85-91 (2000)

Previous Section Next Section Title Page Table of Contents

3. The OHP observations and data reduction

3.1. Observations

In the course of their morphological study of E-S0 galaxies, Michard and Marchal (1994) obtained in 1990/94 numerous frames of ellipticals at the 120cm telescope of the Observatoire de Haute-Provence. Among these, two series were of interest for the present work, as adding usable complements to the bulk of the data.

The first group contains 10 objects observed in UBVRI, i.e. NGC596 (U not usable), 1052, 2300 (R missing), 2768, 3377, 3379, 3610, 3613, 4125, 4564. In the second, there are 9 more galaxies observed in B and V only, i.e. NGC2974, 3605, 3608, 3640, 3665, 4278, 4494, 4636, 5322. Pic du Midi observations for NGC821 were added.

In the runs of 1990/92 the CCD camera with RCA targets had a field of view of about 4x7 arcmin with a pixel of 0.84 arcsec. In the 1993/94 runs, a new Tek512 CCD allowed a field of 7x7 arcmin for a pixel size of 0.76 arcsec. Since the seeing at the OHP is usually in the [FORMULA] range, the sampling is just adequate.

Finally, a number of frames obtained by P. Prugniel were also used. These are available from the HYPERCAT, an on-line data base created and maintained by himself and colleagues (http://www-obs.univ-lyon1.fr/hypercat/). These show NGC1600, 1700 (observed twice), 2300, 3379, 3610, 4486. UBVRI are available for the first three objects, U is missing for the others. Part of these frames, taken in 1995, had the same characteristics than our 1993/94 material, while the latter ones (1997), were taken with a Tek1024 CCD giving a field of near 11x11 arcmin and a pixel size of 0.68 arcsec. The thinned Tek CCD's were affected by the "red halo" phenomenon, and the V-I data proved unreliable.

3.2. Data reduction

Our reduction procedure entails the following steps:

  1. The images of a colour set are prepared by substracting a background, extracting parasitic objects, calibrating against the aperture photometry data, and applying a cosmetic treatment.

  2. The best image in a colour set is analysed according to the principles of Carter (1978). The object is described by a set of isophotes 0.1 mag distant from one another, each represented by a "Reference ellipse" and a series of harmonic coefficients detailing the deviations from the ellipse.

  3. The frames are calibrated against the results of aperture photometry, using whenever possible the homogeneous data by Poulain (1988) or Poulain & Nieto (1994). After calibration, the R and I bands are in Cousins's system, although an i filter according to Gunn was used in the observations.

  4. A procedure of PSF matching is applied to lessen the effects of "differential seeing", that is the errors upon colour measurements resulting from the different PSF's of the two used frames. A discussion of these, and references to previous work, may be found in Paper IV of this series (Michard 1999). In the same Paper IV and in Paper V (Michard & Poulain 2000), our technics of PSF matching are described. In the present work we often tried to match the various frames in a colour set to the PSF of the best one , usually in the R or i band: this implies deconvolution by narrow correcting functions. When the two PSF's in a pair differ too much, so that the deconvolution might give rise to "ringing", the match was obtained by convolution of the best frame.

  5. The effect of the red halo phenomenon upon the V-I colours from the 1994/97 frames strongly depends upon the size and flattening of the studied object. It may be corrected, in principle , by a compensating convolution of the two frames, the V one with the I band PSF and conversely. For various technical reasons this procedure is not fully successful, and part of the V-I data was rejected.

  6. Isophotal colours [FORMULA] are then measured, using the previously determined set of isophotal contours to locate corresponding points in the two frames of a selected pair. A fit of the colour against [FORMULA] (where r is the mean isophotal radius) is then performed in a carefully selected range of r, where the data are supposedly devoid of large systematic errors. It provides the reference colour at a specified r and the gradient [FORMULA] noted [FORMULA].

The adopted procedure offers the possibility to introduce small corrections to the initially adopted sky-background values, in order to cancel out the characteristic deviations from linearity that occur in the plot of [FORMULA] in the presence of background errors. This procedure may easily be criticized, since it introduces a perhaps unwanted constraint upon the determination of colour gradients (see also 2.2.1).

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 2000

Online publication: July 27, 2000