Astron. Astrophys. 357, 651-660 (2000)

2. Observations and data reduction

We have carried out observations of several molecular lines - HCO⁺ (J=1-0), SO (J=2₂-1₁), H¹³CN (J=1-0), SiO (v = 1, J=2-1), and NS (, J=5/2-3/2, parity-e) - and the 3 mm continuum in OH 231.8+4.2, using the IRAM interferometer (Guilloteau et al., 1992) located at Plateau de Bure (French Alps). The observations, except for NS, were made between September 1996 and January 1997, and in February 1999. NS was detected while carrying out a ¹²CO mosaic in this source, during the 1996-1997 and 1997-1998 winter periods (details of the observation will be given in Paper III).

Five or four antennas were used (depending on availability), combining compact and extended array configurations - D (four antennas), C2, B1, and B2 - to obtain a good coverage of the uv-plane. The projected baselines ranged from 15 m to 285 m. Just the 3 mm receivers, operated in double side band, were used. The units of the cross-correlator were placed on the lines of interest and set to bandwidths of 20, 40 and 80 MHz with channel spacings of 0.078, 0.156 and 0.625 MHz (leading to spectral resolutions of 0.3, 0.5, and 2 km s^-1, respectively). The channels of the different correlator units (in the lower and upper side bands) in which there was no line emission, were combined to obtain a continuum map. The total frequency band used for the continuum is 210 MHz.

Data calibration was performed following the standard procedure using the IRAM/GAG software package. The visibilities were flux, amplitude, and bandpass calibrated using continuum point sources (quasars) or sources with known spatial structure (CRL 618). Amplitude calibrators were observed at regular time intervals (approximately every 20 minutes). Bandpass calibrators were observed once at the beginning of every run. Antenna-based correction factors were obtained and interpolated in time to obtain those adequate to our source.

The calibration of the visibility phase (except for NS, see Paper III) was done relative to that of the OH 231.8+4.2 SiO maser. This line originates in the close stellar surroundings (at distances of a few cm in AGB stars, see e.g. Diamond et al. 1994), so it can be considered as a point-like source given the large distance to the source (Sect. 1) and the resulting beam size in these observations (see below). We have used just the central 6 km s^-1 of the maser line (Sect. 3.3) to phase-reference our visibilities. The advantage of self-calibration over the use of external phase calibrators is that no interpolation in time of the phase correction factors is needed. This procedure thus reduces the phase noise and yields maps with higher dynamic range. Baseline-based phase calibration has been performed, obtaining phase corrections for each baseline and for every single moment in which OH 231.8+4.2 was observed. Also, a smooth function fitting the variation with time of the antenna-based phase correction factors has been calculated and applied to our data. The final results using both procedures were found to be in good agreement. On the other hand, phase calibration has been also done using an external (point-like) calibrator to test reliability of the self-calibration procedure. In particular, we have checked that no spurious structures appear in the self-calibrated maps and that the total fluxes obtained in both cases are compatible. The absolute coordinates of the SiO maser determined from the standard (external) phase calibration are R.A. = , Dec. = (J2000). The origin in our maps corresponds to that position. The absolute positional error (excluding possible systematic errors) is estimated to be .

The visibilities were Fourier transformed (using natural weighting) and then CLEANed using the Clark method. The primary beam correction has not been applied to our data. This fact does not substantially affect the intensities of the maps (that are relatively compact compared to the primary beam, 60") except for HCO⁺, for which an underestimation of the flux 15 is expected in the northern clumps at 15" from the center. The continuum emission was subtracted from the molecular emission maps before the CLEANing was done. Note that the observed lines are not expected to be optically thick (an estimate of the HCO⁺ line opacity is given in Sect. 3.1.2). For NS, the velocity maps are not shown due to the high noise level. In this case, the continuum has been subtracted directly from the spectrum. The resulting maps and/or spectra for the different lines and the 3 mm continuum are presented in Sect. 3. The gaussian `clean' beam adopted for image restoration is indicated for each map in their respective figure captions.

We have complemented our HCO⁺ interferometric observations with zero-spacing H¹³CO⁺ (J = 1-0) and H¹²CO⁺ (J = 3-2) data obtained toward the central position of OH 231.8+4.2. The H¹³CO⁺ (J = 1-0) spectrum was obtained with the 30 m IRAM radiotelescope at Pico de Veleta (Granada, Spain) in July 1997. The HCO⁺ (J = 3-2) transition was observed with the 10.4 m radiotelescope of the Caltech Submillimeter Observatory at Mauna Kea (Hawaii, USA) in November 1998. The FWHM of the beam was in both cases 27". The two spectra are shown in Fig. 4 (Sect. 3).

© European Southern Observatory (ESO) 2000

Online publication: June 5, 2000
helpdesk.link@springer.de