Astron. Astrophys. 323, L17-L20 (1997)

2. Experiments and Data Analysis

The 1.4 mm VLBI observations between Pico Veleta and a single antenna on Plateau de Bure were made on March 1-2, 1995. Because of degrading weather conditions on Plateau de Bure, only data of March 2, 0h00-08h00 UT, could be used for correlation. The observational setup was similar to the earlier 1.4 mm experiment performed in December 1994 (see Greve et al. 1995 for a description). On Plateau de Bure we used again the hydrogen maser from the CNRS (France) and the VLBA terminal from the Yebes observatory (CAY, Spain).

The data were recorded with the MK III VLBI system (Rogers et al. 1983) in mode A at a tape recording speed of 270 inches/s (double speed). This allows 6.5 min continuous VLBI scans with a bandwidth of 112 MHz (ie. 14 bands at 8 MHz each) at frequencies between 215530.99 MHz and 215642.99 MHz. The observations were made in left circular polarization (LCP). We recorded 3-4 scans per hour, using the time between VLBI scans for focus checks, pointing and calibration (determination of , , opacity ).

We combined the 1.4 mm VLBI observations with 3 mm (86.25 GHz) snap shot VLBI observations of a few strong sources (3C 273, 3C 279, NRAO 530) for a check of the systems and easier fringe search at the correlator. The GPS satellite network was used at both observatories to check the masers and to obtain accurate time information.

The data were correlated at the MK III VLBI correlator of the MPIfR at Bonn using a pre-integration time of 2 s. The data were then fringe fitted (Program: FRINGE), edited, and calibrated using the standard correlator software (eg. Alef 1989), the CalTech VLBI package (Pearson, 1991) and Difmap (Shepherd et al. 1994). To reduce the detection threshold, the fringe fitting of weak scans was performed with search windows reduced in size and centered around rate and delay values extrapolated from neighbouring stronger detections (cf. Krichbaum et al. 1992 for details of the method). In order to confirm weaker detections found with FRINGE, we applied also the new method of incoherent fringe search (Programs: AVERAGE, SEARCH) using the HOPS-package (Rogers et al. 1995): both methods gave consistent results (see Table 2).

Phase fluctuations due to the atmosphere severely limit the phase coherence of our data. Coherence times were determined for all detected sources and ranged - depending on elevation and time - between 4 s and 12 s, with a typical value of 8 s which we adopted for the final segmentation. A comparison of the phase fluctuations seen at 215 GHz and 86 GHz in VLBI scans taken under good atmospheric conditions (water vapor mm H₂ O, ) shows 2-3 times stronger fluctuations at the higher frequency, consistent with expectations. Phase stability measurements of the LO-chains and inspection of the fringe rate spectrum and visibility phases (showing variations not exceeding 3-4 turns in 6.5 min) leads to the exclusion of severe degradations of the signal on timescales ( s) shorter than the segmentation time.

The data were amplitude calibrated using regular system- and antenna temperature measurements from pointing scans across the program- and calibrator sources.

At Pico Veleta, the calibration measurements were made in the 3 and 1 mm bands with SSB tuned SIS receivers ( 10 dB sideband rejection). The chopper wheel calibrated pointing cross-scans gave , and the single-dish flux densities of the sources, using the planets and standard H II regions as calibrators. At 1.4 mm the accuracy of the measurements is 20 , the elevation dependent antenna gain curve is accurate to 10 above elevation, and 20 at lower elevations.

At Plateau de Bure, only the 3 mm SIS receiver was SSB tuned (rejection 15 dB), while the 1.4 mm SIS receiver was DSB tuned (accurate within 30 ). For the 1.4 mm VLBI-observations we used interlaced 3 mm observations made with the connected array for antenna pointing at 1.4 mm and to extrapolate to 1.4 mm wavelength, using the known antenna characteristics and the IRAM atmospheric model (Cernicharo 1985). While the 3 mm data of are accurate to 20 , the data extrapolated to 1.4 mm are estimated to be accurate to within 50-70 . The adopted sensitivity of 40 Jy/K ( 5 Jy/K) was later confirmed by absolute flux calibration measurements with the interferometer in the 1 mm band. In Table 1 we summarize the typical values of the aperture efficiency (col. 3), effective antenna gain (col. 4), and the range of (col. 5) and zenith opacity (col. 6) during the observations.

The total (single dish) flux densities of the sources at 86 GHz and 215 GHz were measured at Pico Veleta on March 1 and on March 3-4, 1995, shortly before and after the experiment (H. Ungerechts, priv. comm.) using Mars and compact H II regions as flux density calibrators.

Table 1. Antenna Characteristics at 215 GHz

Table 2. Summary of Detections at 215 GHz on the baseline Pico Veleta - Plateau de Bure

© European Southern Observatory (ESO) 1997

Online publication: June 5, 1998

helpdesk.link@springer.de