SpringerLink
Forum Springer Astron. Astrophys.
Forum Whats New Search Orders


Astron. Astrophys. 364, 26-42 (2000)

Previous Section Next Section Title Page Table of Contents

3. Data reduction and analysis

We have started our analysis from the pipeline-reduced HST datasets retrieved directly from the archive. Typically, for each field, the observations consist of a sequence of frames slightly shifted with respect to each other (shifts range from a few pixels to a few tenths), often strongly affected (in particular the WFPC2 ones) by cosmic ray hits. To obtain a single frame per object and to remove at the same time cosmic rays and residual bad pixels, we have performed a few further reduction steps to align and combine the images available for each field; a few tests carried out on field stars have shown that the procedure adopted to this purpose, outlined below, does not significantly degrade the quality of the PSF in the output images, even in the case of relatively large shifts. All the steps have been performed using routines from the IRAF-STSDAS data-reduction packages.

The relative shifts between two different exposures of the same field have been evaluated measuring the position of the peak in the cross-correlation of the two images. We have subsequently aligned all the frames using the DRIZZLE task (Fruchter & Hook 1998), preserving the original scale in the output images. We have chosen not to resample the data on a smaller scale, since this yields an actual improvement of the spatial resolution in the final combined image only if the frames to be combined are shifted by non integer amounts, and if such shifts sample homogeneously the sub-pixel scale, which usually does not happen for our data. Also, as we will explain in the next section, we compare our data with model distributions convolved with a theoretical PSF, and the effects of the resampling on the PSF are rather difficult to quantify. In the end, a better accuracy in the evaluation of the PSF shape more than compensates a possible little loss in spatial resolution. The shifted frames are combined in a following stage, rather than by DRIZZLE itself, to achieve a more efficient cosmic-ray rejection. In the case of the HDFS, the public F160W image has been used, without any further processing.

Using the ELLIPSE task in IRAF, a radial surface brightness profile has been extracted for all the galaxies except one very irregular object (number 21 in Table 1). The final images of the sample galaxies are shown in Fig. 2 on a logarithmic scale, together with the respective radial brightness profiles. The map of object 21 is shown in Fig. 3: we identify the ERO with the irregular object in the center of the map, but we do not exclude that the two close components observed may be part of a single interacting system.

[FIGURE] Fig. 2. For each galaxy we show the final image in a logarithmic scale, and a plot with the radial surface brightness profile. The orientation of each map is with north up and east to the left, and the scale on the two axes is in arcsec; the filter is specified in the plot panel. Here, the dots and the solid line represent respectively the surface brightness profile of the galaxy and of the best fit model; both profiles are computed as the average intensity of the distributions along the same set of elliptical contours. For objects n. 5 and 30 a satisfactory fit could not be obtained. Magnitudes are computed in the AB system.

[FIGURE] Fig. 2. (continued)

[FIGURE] Fig. 2. (continued)

[FIGURE] Fig. 2. (continued)

[FIGURE] Fig. 3. We identify object n. 21 with the one in the center of the map (the filter is F814W). We did not obtain a fit for this galaxy due to its disturbed morphology, and to the closeness of a bright companion.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 2000

Online publication: December 15, 2000
helpdesk.link@springer.de