3 Observations and data reduction

 

3.1 Nançay $H{\scriptstyle I}$ observations

During the period November 1997 - November 1998 we obtained, with the Nançay decimetric radio telescope 21-cm, HI line spectra of 19 Hydra cluster galaxies, including 16 objects from our primary list (see Table 2), as well as 3 spirals close to the sample objects. The remaining 4 VLA sources from our primary list were considered too weak for detection at Nançay, given the data published by McMahon (1993). Separate observations were also made with pointing centers around the HI source without an optical counterpart, H1032-2819, in order to confirm the weak VLA detection.

The Nançay telescope is a meridian transit-type instrument with an effective collecting area of roughly 7000 m² (equivalent to a 94-m parabolic dish). Due to the elongated geometry of the telescope, at 21-cm wavelength it has a half-power beam width of $3\hbox{$.\mkern-4mu^\prime$}6$ E-W $\times$ 22' N-S at the declination of the Hydra cluster. Tracking was generally limited to about 45 minutes per source. Typical system temperatures were $\sim$40 K.

We obtained our observations in total power (position-switching) mode using consecutive pairs of two-minute on- and two-minute off-source integrations. Off-source integrations were taken at approximately 20' E of the target position. The 1024 channel autocorrelator was divided into two pairs of cross-polarized (H and V) receiver banks, each with 512 channels and a 6.4 MHz wide bandpass. This yielded a channel spacing of 2.64 km s^-1, for an effective velocity resolution of $\sim$3.3 km s^-1 at 21 cm. The center frequencies of the two banks were tuned to the known redshifted HI frequency of the target, as measured at the VLA by McMahon (1993). Total integration times were about 5 to 6 hours for most galaxies, but shorter for some of the stronger spirals and longer for the faintest objects.

We reduced our HI spectra using the standard Nançay spectral line reduction packages available at the Nançay site. With this software we subtracted baselines (generally third order polynomials), averaged the two receiver polarizations, and applied a declination-dependent conversion factor to convert from units of $T_{\rm sys}$ to flux density in mJy. The $T_{\rm sys}$-to-mJy conversion factor is determined via a standard declination-dependent calibration relation established by the Nançay staff through regular monitoring of strong continuum sources. This procedure yields a calibration accuracy of $\sim$10%. In addition, we applied a flux scaling factor of 1.26 to our spectra based on statistical comparisons of recent Nançay data on strong, compact spiral galaxies unresolved by the telescope beam (Matthews et al. 1998) with past observations made at Nançay and elsewhere.

3.2 Optical imaging

We carried out optical imaging observations in February 1997 with the Danish 1.5 meter telescope at La Silla observatory. We used the DFOSC optical camera which offers a large field of view of $13\hbox{$.\mkern-4mu^\prime$}7 \times 13\hbox{$.\mkern-4mu^\prime$}7$. The detector was a LORAL 2k $\times$ 2k CCD with a pixel size of $0\hbox{$.\!\!^{\prime\prime}$}39$.

  

Figure 1: Locations in the Hydra I cluster of the fields imaged with the DFOSC CCD camera. Each of the 16 rectangles covers the final field of view of the co-added I-band images. The final area covered by our imaging survey may thus be easily visualized. The numbers refer to the field ID of the finding charts shown in Figs. 2-8. The circle indicates the location of the cluster core as indicated by the center of the X-ray gas distribution and the position of the giant elliptical NGC 3311

  

Figure 2: Identification chart of Field 1. The chart is composed of the co-added DFOSC I-band images obtained at the location of Field 1. The galaxies detected in HI with the VLA (McMahon 1993) are identified with black labels. Several HI-poor galaxies that appear to be companions to the HI-detected dwarf galaxies are also indicated with white labels. The bar at the bottom indicates a scale of 2$\hbox{$^\prime$}$ or 26 kpc at the adopted distance of Hydra I. North is to the top and East to the left

  

Figure 3: Identification chart of Field 2 (see Fig. 2 for details)

Images through the Bessel B and Gunn I filters were obtained for sixteen individual positions in 7 distinct fields of Hydra I containing the 20 HI sources of the primary list. The location of these fields in the cluster are indicated in Fig. 1, the field number of each source is given in Col. 5 of Table 1 and the identification charts of the 7 fields are presented in Figs. 2-8. For most objects, B-band observations consisted of two individual exposures of 900 s each, whereas I-band observations consisted of 4 individual exposures of 450 s each. However, due to time constraints, some fields were observed with a lower total exposure time, as indicated in Table 3. Offsets of 10$^{\prime\prime}$ for the observations in the B-band and 30$^{\prime\prime}$ for the I-band were performed between each exposure. Several standard stars from the fields of Landolt (1992) were observed throughout the nights. Weather conditions were photometric.

Standard data reduction was performed using the CCDRED package in IRAF. A median image had to be produced from adjacent I-band images, by which the original images were divided in order to remove the prominent fringes seen in the I-band. The reduced frames were then co-added with a shift-and-add method. The flux calibration has been performed from photometric transformation equations, whose coefficients, i.e., zero-points, color and extinction terms, were determined from our standard stars observations. The astrometry of all images has been carried out with guide stars from the USNO catalog, queried via the ESO SKYCAT tool. The images were corrected for distortions during the same process. The seeing in the I-band varied between 1$.\!\!^{\prime\prime}$1 and 1$.\!\!^{\prime\prime}$5 with a median value of 1$.\!\!^{\prime\prime}$4.

  

Figure 4: Identification chart of Field 3 (see Fig. 2 for details)

  

Figure 5: Identification chart of Field 4 (see Fig. 2 for details)

  

Figure 6: Identification chart of Field 5 (see Fig. 2 for details)

  

Figure 7: Identification chart of Field 6 (see Fig. 2 for details)

  

Figure 8: Identification chart of Field 7 (see Fig. 2 for details)

3.3 Near-infrared imaging

Near-infrared images in the K' band were obtained in March 1998 with the IRAC2-B instrument on the ESO/MPI 2.2 m telescope using a 256 $\times$ 256 NICMOS3 array with a $2\hbox{$.\mkern-4mu^\prime$}1$ $\times$ $2\hbox{$.\mkern-4mu^\prime$}1$ field of view and a pixel size of $0''\hbox{$.\!\!^{\prime\prime}$}5$. The observations were made during full moon, causing strong reflections in the telescope that gave rise to important background variations in the frames which could not always be corrected for. The accuracy of the photometric measurements is largely limited by this problem, especially for low surface brightness objects. Different observing templates with various sets of offsets were used, with the aim of maximizing the exposure time on target. Sky images have either been extracted from the target images ("On'' positions) or from "empty'' adjacent fields ("Off'' positions) depending on the object's shape and size. Typically, Off images had to be taken for objects more extended that 1/5$^{\rm th}$ of the detector field of view. The exposure time was basically adjusted according to the surface brightness of the galaxies. For various reasons, the actual total exposure times listed in Table 3 may depart from the initial estimates. They range between 5 and 60 minutes. The highest values were obtained in the favorable cases where the galaxy was small enough to be kept always on the detector. A total of 12 out of 15 galaxies from the final list were observed in the K' band.

Data reduction was carried out semi-automatically with IRAF scripts written by Duc and Lidman. All images were flat-fielded with an illumination-corrected dome flat and then sky-subtracted. Sky images were obtained by averaging dithered frames with a min-max rejection algorithm to get rid of infrared sources. The sky-subtracted object images were first registered automatically using the rough telescope offset information available as FITS header parameters and then more precisely using the positions of isolated common reference stars. We skipped all images showing large-scale, strong background variations, mainly due to reflections from the moon. The remaining frames were shifted and added with the task IMCOMBINE in IRAF, applying a sigma clipping rejection parameter.

Faint infrared standard stars from the NICMOS project (Persson et al. 1998) were used for photometric calibration. An extinction coefficient was fitted but no color corrections were applied. Weather conditions varied throughout the observing run from partly cloudy to photometric. They produced some systematic errors in the photometry that are difficult to estimate but that might be as high as 0.5 mag.