next previous
Up: New detections of

2. Observations

All observations were made with the 13.7 m telescope of Qinghai Station, Purple Mountain Observatory, located at Delingha, Qinghai Province, West China (tex2html_wrap_inline828, 3200 m above sea level).

The antenna of the telescope is a radome-enclosed classic Cassegrain type. The surface accuracy of the main dish is 0.2 mm. At the water maser frequency (22.23508 GHz was used for the rest frequency of the line), the HPBW of the telescope is tex2html_wrap_inline830, the antenna aperture efficiency measured by continuum observations of the planet Venus is 0.48, including the absorption of the radome, equivalent to a flux temperature ratio of 38 Jy K-1. The pointing accuracy of the telescope calibrated and checked by five point observations of strong continuum point sources is 20''.

The 22 GHz receiver front end is composed of a cooled Schottky mixer and a 1.4 GHz FET IF amplifier together with a phase-locked Gunn local oscillator. The system temperature of the telescope is around 300 K. The spectral back end is a high resolution 1024-channel AOS with a channel bandwidth of 12 KHz and equivalent velocity resolution of tex2html_wrap_inline836.

The noise diode is a secondary temperature calibrator calibrated at the beginning of each observing run by hot and cold black bodies. The stability of the noise diode output was checked regularly by observations of the continuum calibration sources.

Unless the sources had rich spectral features and wide frequency coverage (for which position switching mode was used), most observations were made in frequency switching mode. The typical integration time was 10 minutes on Signal and Reference respectively. For survey observations, several scans of spectral data were usually accumulated.

Before observing the source, calibration observation (switching on and off the noise diode) was usually done in the vicinity of the source position to get the gain curve of the AOS: tex2html_wrap_inline838, where TC is the calibration temperature of the temperature calibrator, tex2html_wrap_inline840 and tex2html_wrap_inline842 are the channel outputs of AOS when the noise diode is on and off respectively. The system noise and the passband effect (including the response of AOS itself) are removed, and the temperature scale is given by the standard relation: T(i)=G(i)*(S(i)-R(i))/R(i).

At 22 GHz, the elevation dependence of the antenna gain is not apparent, so the temperature scale was converted directly to the flux scale by multiplying the flux temperature ratio of 38 Jy K-1. Usually the atmospheric attenuation at zenith (ATTN) was determined by the continuum sky dip observation, and the effect of atmospheric absorption was corrected by multiplying a factor of tex2html_wrap_inline848. We estimate that the final error in flux scale of our spectral data is about 20%.

All spectral data are stored on tape and disk in a format conforming to the POPS package specifications. POPS is the data reduction package developed by NRAO (Tuscon), kindly provided to Purple Mountain Observatory and replanted to the PDP-11 data reduction computer by the engineers of the observatory.

Owing to the limited memory size of the original computer for data reduction, it is difficult to process the whole frame of 1024 AOS data points by using POPS. Therefore, a program is run on the computer to convert the format of the data file from POPS to Drawspec, and transfer the data to a personal computer. The data are then reprocessed with Drawspec. Drawspec is a compact data reduction package running on PC, developed by Dr. Liszt (NRAO) and kindly provided by the author. Possible systematic errors of velocity have been corrected before data reduction, and the velocity accuracy of the data is estimated to be better than tex2html_wrap_inline850. The data reduction with Drawspec consists essentially of 3 steps. First a polynomial fit to the baseline is removed in each scan of the spectrum and all related scans added together, the rms noise is computed for the final accumulated spectrum. Second, a suitable number of gaussians are fitted to the lines to evaluate the peak intensity and the central velocity of the strongest component. Finally, the integrated intensity of the gaussian components is computed.


next previous
Up: New detections of

Copyright by the European Southern Observatory (ESO)
This email address is being protected from spambots. You need JavaScript enabled to view it.