Astron. Astrophys. 349, 411-423 (1999)

2. Observations and data reduction

2.1. The HI data

The observations were made with the NRAO's Very Large Array (VLA) ¹, for a description see Napier et al. (1983). The field was centered on NGC 4548 (Table 1). We observed in December 20 1994 for 235 minutes with the C configuration and in March 29 1995 for 90 minutes with the D configuration. A Hanning smoothing was applied on-line to the initial frequency channels, yielding 63 channels covering a total velocity width of 630 km s^-1. The velocity channels are centered on v500 km s^-1. The data were calibrated using the standard VLA reduction programs (AIPS). A CLEANed image of all strong continuum sources were made and afterwards directly subtracted from the UV data cube. At the end a linear interpolation of the UV data points with respect to the frequency channels using the first and last 10 channels was made in order to subtract the continuum. The resulting image was CLEANed with a 2020" FWHM beam. We ended up with a r.m.s. noise of =0.4 mJy/beam in one 10 km s^-1 channel, or =4.7 cm^-2 expressed in column density.

Table 1. The parameters of NGC 4548. Col. (1) and (2): 1950 celestial coordinates. Col. (3): morphological type (de Vaucouleurs et al. 1973). Col. (4): heliocentric velocity, in km s^-1. Col. (5): HI flux, in Jy km s^-1. Col. (6): HI deficiency (Cayatte et al. 1994).

2.2. The CO data

NGC 4548 was observed in 1994 with the IRAM 30 m telescope at Pico Veleta (Granada, Spain). The beam size of the telescope is 22" at 115 GHz [¹²CO(J=1-0)], which corresponds to 1.8 kpc at a distance of 17 Mpc. Weather conditions were good, with typical zenith opacities of 0.25-0.45. The pointing accuracy was checked hourly by broad band continuum observations of the nearby source 3C273; the average error was 3" rms. We used a SIS receiver in single sideband mode with T_rec=140-270 K and T_sys=500-800 K (in scale) at the elevation of the source. Two filter banks of 512 contiguous 1 MHz channels provided a velocity resolution of 2.6 km s^-1 and a total velocity coverage of 1330 km s^-1. We used a wobbler switching procedure, with a wobbler throw of 4´ in azimuth. Each 8 minute scan began by a chopper wheel calibration on a load at ambient temperature and a cold load. The total integration time on each position was between 16 and 54 minutes on+off (i.e. half time on source) depending on the intensity of the signal, yielding a rms noise level of 15-30 mK (in the T_mb scale) after boxcar velocity smoothing to 20.8 km s^-1.

NGC 4548 was observed at the nominal central coordinates and at different position offsets of 40" (Fig. 1). The galaxy was observed in 25 different positions and detected in 23 of them in the CO(1-0) line.

Fig. 1. The 12CO(1-0) spectra plotted on a - map centered on the nucleus of NGC 4548. The map's axes are offsets in arcseconds from this center. The velocity range of each spectrum is 0-1000 km s-1 and the brightness temperature range is -0.08-+0.16 K.

The data were reduced with the CLASS package (Forveille et al. 1990). The baselines were generally flat owing to the use of the wobbler, allowing us to subtract only linear baselines. The antenna temperature () was corrected for telescope and atmospheric losses by the calibration procedure. We adopt the main beam scale for the antenna temperature, with for the CO(1-0) line. The integrated emission is given by: I(CO)=T_mbdv K km s^-1.

2.3. The H data

The H image (Fig. 3) was obtained by J.R. Roy and P. Martin with the 1.6m of Observatoire du Mont Mégantic employing a f/8 f/3.5 focal reducer, in May 1988. The exposure time was 5 times 2000 sec with a 6577/10 filter. The FWHM of the filter (10 Å = 457 km s^-1) covers the whole velocity range of the HI data. NGC 4548 was observed with a RCA chip 360 512 pixels (scale of 1.1 arcsec/pix).

The images where reduced using the software package IRAF following the procedures described in Belley & Roy (1992) and in Martin & Roy (1992). Two sources of uncertainties relevant to the present analysis must be pointed out. First the accuracy of monochromatic flux measurement in the inner regions depends on how well the relatively bright stellar continuum in the central regions is subtracted; the final result is somewhat uncertain because the continuum filter used is about 400 Å to the red side of H. A first order subtraction is usually done using relative scaling of the flux of several stars in the field. The resulting monochromatic image is that where the base level H flux in interarm regions (apart the presence of interarm HII regions) shows net zero flux in the monochromatic image; this is achieved after careful steps of trial and error. Previous to this, subtraction of the sky background has been done. This brings up the second source of uncertainties, which is the accuracy of the flat fielding procedure. This is quite critical because it affects the measurements of mean H surface brightness in the faint outer regions. Uncertainties arise because of the restricted field and of the difficulties of eliminating large-scale variations of the the background illumination to better than 2-3% in monochromatic imagery done with our focal reducer. To circumvent this, we eliminated the regions of weak surface brightness by applying conservatively a high cut-off to the data. We kept only individual pixels which had number of counts (in ADU) greater than the rms value above the background in the annulus were the total number of counts was not longer seen to increase with increasing galactocentric distance.

Finally, the images were calibrated using the aperture spectrophotometry done by McCall (1982).

2.4. The optical data

A CCD B image of NGC 4548 (Fig. 2) was obtained at the NOAO 0.9 m telescope at Kitt Peak ², Arizona, in June 1995. We used a 2048x2048 T2KA detector (2 e^-/ADUs) in the f/13.5 configuration, which gives a field of view of with a pixel size of 0.4". The galaxy was observed through the Kitt Peak B Harris filter in a 15 min exposure during non photometric conditions. The image was reduced using standard procedures, including bias correction, dome flat fielding and cosmic ray removal. The seeing was 2".

© European Southern Observatory (ESO) 1999

Online publication: September 2, 1999
helpdesk.link@springer.de