2 Observations

The Nançay radio telescope is a meridian instrument. It can observe sources at declination $\delta > -39\hbox{$^\circ$ }$ and can track them for about $\pm 1.1 / \cos \delta$ hour. For observations at 18 cm wavelength, its RA $\times$ dec beam is $3.5\hbox{$^\prime$ }\times 19\hbox{$^\prime$ }$at declinations lower than $30\hbox{$^\circ$ }$. Its sensitivity for point sources is 0.9 K Jy^-1 at $\delta \approx 0\hbox{$^\circ$ }$ and its variation with declination is given in Fig. 1 of Gérard et al. (1989). The system temperature was typically 45 K (at $\delta \approx 0\hbox{$^\circ$ }$) at the end of the observing period considered here.

Since 1982, all cometary observations have been made using a 1024-channel autocorrelator split into $4 \times 256$-channel banks, allowing simultaneous observations of the OH 18-cm main lines at 1667 and 1665 MHz (except for occasional observations of the satellite lines at 1612 and 1721 MHz), and two polarizations (right- and left-circular polarizations, except for comets Bowell 1982 I and 6P/Churyumov-Gerasimenko 1982 VII which were observed in vertical and horizontal linear polarizations). In 1983, the Nançay radio telescope was closed for upgrading (installation of a new carriage house, of new computers and operating software). As a consequence, comets Sugano-Saigusa-Fujikawa 1983 V and IRAS-Araki-Alcock 1983 VII, which respectively passed at only 0.063 and 0.033 AU from the Earth, unfortunately could not be observed.

The autocorrelator is set to a 782 Hz channel separation, corresponding to 0.14 km s^-1. The effective resolution is 0.28 km s^-1 after Hanning smoothing. A frequency-switching mode is used, with a frequency shift of 100 kHz. A $\sqrt{2}$ factor in the signal-to-noise ratio is gained after folding the spectra. The spectral coverage is about $\pm$8 km s^-1, which is enough for observing cometary spectra with a good determination of the baselines since the cometary line widths are typically 2 km s^-1. The comets are tracked in position and velocity using an ephemeris computed from the latest available orbital elements (which, except otherwise stated, are those distributed by the Central Bureau for Astronomical Telegrams). The occasional errors which resulted from inaccurate ephemeris are discussed in the Appendix.

A check of the telescope performance and backend setup was done by frequently (usually every other day) observing the OH lines of the galactic source W12. The radial velocity scale was accurate within $\pm 0.02$ km s^-1.

Interferences of external origin were very rare during the whole period 1973-1999 because most of the cometary observations were made in the main lines at 1665 and 1667 MHz. The protected radio astronomy band has a primary status over the whole range 1660.0 MHz-1670.0 MHz while the radial velocity of the nucleus with respect to the telescope never shifts the rest frequency by more than 0.5 MHz. Strong interferences of internal origin, however, were recorded in early 1982 which ruined the 1665 MHz observations (e.g., comet Bowell 1982 I) when a new receiver front-end was installed with improper local oscillator setting.