2. Observations and data reduction
2.1. Pico Veleta 30-m telescope
The single dish observations were carried out in July 1994 using the IRAM 30-m telescope at Pico Veleta (Spain). Spectra of IRAS 20126+4104 were obtained in position switching with a reference position at offset or using the wobbler (i.e. a nutating secondary) with a beam-throw of and a phase duration of 2 sec. The calibration was performed with the chopper wheel method with both hot (sky and absorber) and cold (nitrogen) loads. The calibration was verified on well-known molecular sources and found to be accurate to within 20%. The absolute pointing was good to . The telescope half power beam width (HPBW) ranges from to depending on the frequency. The conversion factor from main beam brightness temperature () to flux density is 4.7 Jy K-1. The backends were two filterbanks spanning 512 MHz and providing a frequency resolution of 1 MHz, a filterbank with 25.6 MHz bandwidth and 100 kHz resolution, and an autocorrelator used in various high spectral resolution configurations. Two point raster maps were done, one with sampling, the other with . All maps are centred on the position of the H2 O masers (Tofani et al. 1995), namely , . The integration time per point was 2 min and the maps were repeated up to three times and then averaged. The pointing and the focus were verified regularly (at least every hour) on nearby continuum sources. A total of 7 molecular species and 11 rotational transitions were mapped, namely: HCO (1-0), CS(3-2), C34 S(2-1), C34 S(3-2), C34 S(5-4), HCN(1-0), 13 CO(2-1), CH3 CN(8-7), CH3 CN(12-11), CH3 OH(3-2), and CH3 OH(5-4).
2.2. Plateau de Bure interferometer
The interferometric observations were carried out between January and March 1995 using the IRAM four element array at Plateau de Bure (France). A description of the instrument is given by Guilloteau et al. (1992). The four 15 m antennas were equipped with 82-116 GHz SIS receivers yielding DSB receiver temperatures around 40 K. The receivers were tuned double side-band. The facility correlator was centred at 98.020 GHz. We show the frequency set-up of the correlator and list the observed spectral lines in Table 1. One of the six units in the correlator with bandwidth 160 MHz was centred on a frequency range free of lines and used to measure the continuum flux. The spectral resolution of the correlator is a factor greater than the channel spacings given in Table 1. The phase calibrations for our data were carried out by means of frequent (5 minutes every 20 minutes) observations of nearby point sources.
Table 1. Frequency set-up for detected molecular lines with the PdBI. DSB tuning was used
The bandpass calibration was carried out in the baseline-based manner using 2013+370, which was also used for the instrumental (amplitude and phase) calibration. The flux of the primary calibrators was bootstrapped from IRAM monitoring measurements and used to derive the absolute flux density scale. Table 2 gives a list of the main parameters for our PdBI observations.
Table 2. Instrumental parameters for the IRAM PdBI observations
Continuum subtraction was performed in the U-V plane by using the integral over the two 160 MHz units. Finally, channel maps were produced for all the lines both using natural and uniform weighting. The conversion factor from flux to brightness temperature in the synthesised beam is K (Jy/beam)-1 for maps obtained with uniform weighting, and K (Jy/beam)-1 for those obtained with natural weighting.
2.3. TIRGO telescope
The NIR observations were obtained on the September 1994, using the Arcetri near infrared camera (ARNICA) mounted at the 1.5m TIRGO 1 telescope. ARNICA is equipped with a NICMOS3 infrared array, the scale on the detector at the TIRGO is , see Lisi et al. (1996) and Hunt et al. (1996a) for a complete description of the instrument and its performances at TIRGO.
2.3.1. Broad band observations
The broad band observations were obtained with the standard J, H, and K NIR filters ( m, m, and m respectively). The field was observed using a "dithering" technique: several images of the source were taken in each band, slightly moving the telescope. All the data reduction and analysis have been performed using the IRAF 2 and ARNICA 3 (Hunt et al. 1994) software packages. A flat field image for each band was obtained as median average of the observed frames. After flat fielding, the images have been registered and coadded into a mosaic frame. Photometric calibration was achieved by observing a set of the ARNICA standard stars (Hunt et al. 1996b). The photometric calibration is accurate to about 6%. Point source photometry was performed using the DAOPHOT package, the limiting magnitudes achieved in each band are ( in a aperture): 17.3, 16.7, and 16.2, in J, H, and K respectively. Accurate astrometry (rms ) was performed using a table of optical stars extracted from the Digitized Sky Survey, made available by the Space Telescope Science Institute.
2.3.2. Narrow band observations
The narrow band images were obtained with an interferential filter centered on the H2 vibrational line at m and with a resolution of , at a working temperature of K (Vanzi et al. 1996). The observing and data reduction techniques were the same as for the broad band observations, but the useful field of view is about half in the case of narrow band images (see Vanzi et al. 1996). Continuum subtraction and calibration were performed using the K band image, assuming that a number of field stars present in both mosaics do not have detectable line emission. The noise achieved in the final, continuum subtracted, image is .
2.4. NOT telescope
During Summer 1996 ARNICA has been mounted at the 2.5-m Nordic Optical Telescope (NOT) on La Palma. The plate scale on the detector was pix. We observed IRAS 20126+4104 on the August at K and H2. The observing technique used was the same than at TIRGO, but, due to the smaller pixel size, the final mosaics cover a smaller area. Photometric calibration was achieved by observing the standard star AS37 (Hunt et al. 1996b). The limiting magnitude in the final K-band mosaic is of the order of 17.2 (3 in aperture), the noise in the continuum subtracted H2 image is .
© European Southern Observatory (ESO) 1997
Online publication: April 28, 1998