During the observing periods both comets were mainly visible in day time, which usually limits the chance for observation and the quality of the data. This became true for Comet Hyakutake; within 10 simultaneous observing periods between the Heinrich-Hertz-Telescope and Pico Veleta there was only one period with marginal observing conditions at 860 GHz and no period with usable data at 250 GHz and/or 345 GHz at both sites simultaneously. Atmospheric conditions for Comet Hale-Bopp were much better. During this time 3, 7, and 10 usable observations were obtained at 860, 345 and 250 GHz, respectively, resulting in a good spectral index and "light" curve determination. The instrumental parameters for both observing sessions are collected in Table 1.
Table 1. Instrumental parameters for observations of Comet Hyakutake in 1996 and Comet Hale-Bopp in 1997.
2.1. IRAM 30m telescope (Pico Veleta)
The MPIfR 19-channel bolometer at 250 GHz was used for multiple scan, ON-OFF, and ON-THE-FLY map observations. These observing techniques have been used e.g. by Altenhoff et al. (1994), the ON-THE-FLY maps were analysed by the NIC program, documented by Broguiere et al. (1996), and MOPSI, written by Zylka (1997). All three methods gave consistent results.
The 19-channel bolometer is a hexagonally close packed, diffraction limited array operating at 0.3 K. A common filter defines a bandpass centered at 250 GHz. In the first period Ceres was used as a calibrator and e.g. 1413+135 as a pointing and reference source; in the second period NGC7027 was the calibrator and BL Lac and 3C345 were the main pointing sources. Multiple scans through the comets at the beginning of each observing period showed that the (main) signal could be well represented by a Gaussian shape and that the measured flux density per beam from scans or ON-OFF together with the measured beam broadening would allow a good estimate for the integrated flux density. The derivation of the beam broadening will be discussed later. The allocated observing time per day was typically only 2 hours; consequently the time for calibration and sky dip measurements was reduced to the bare minnimum.
In the main part of this paper either the measured flux density per beam is used or the integrated intensity, calculated with the beam broadening to describe the observed time variations (light curves) and the spectral energy distribution of both comets; later it will be estimated how good the Gaussian approximation is.
2.2. SMTO Heinrich-Hertz-Telescope (Mt. Graham)
The observations were done with the four colour bolometer, consisting of four diffraction limited bolometers in a common 3He cryostat. The effective frequencies are 250 GHz, 345 GHz, 670 GHz, and 860 GHz. Only one bolometer can be coupled to the telescope at a time, but because all four are operational, change over to a different frequency takes only a few minutes.
We used beam switching, chopping horizontally with the subreflector at a rate of 2 Hz. Since atmospheric transmission is very similar at the 670 GHz and 860 GHz bands, only the latter was used for a better spectral index determination. Some essential instrumental parameters are listed in Table 1. The half power beamwidths , reported here for these bolometers, are bigger than expected for the respective frequencies, possibly as a result of under-illumination of the main reflector. The calibration takes this error into account. The only negative consequence is a reduction in sensitivity.
For the cometary observations the ON-OFF technique was used to obtain an optimal signal to noise ratio. It is similar to that used at the 30m telescope. The size correction for the total flux density was derived from the beam broadening by the comet, measured at 250 GHz with the 30m telescope. In the first period the calibration was related to Mars and Uranus, in the second period mainly to K3-50, NGC7538, and W3OH with the scale of Sandell (1994).
2.3. IRAM interferometer (Plateau de Bure)
The Plateau de Bure interferometer was used in configuration 4C2 for Comet Hyakutake and in configuration 5C1 (5C2 last period) for Comet Hale-Bopp. Simultaneously with our observations spectroscopic measurements were done by Dr. Bockelée-Morvan and associates, to whom the choice of frequencies was left for their molecular line search. The effective baselines ranged from 25 to 180 m. For these observations 3 units of the correlator were assigned to the 90 GHz and 230 GHz receivers, respectively. One of each was at narrow bandwidth for molecular observations. The other two pairs were used at 160 MHz adjacent to the line frequencies.
The observing cycle for Hale-Bopp was: pointing, focusing, 4 min of crosscorrelation on the phase calibrator, one autocorrelation ON-OFF on the comet, and three 17 to 20 min of crosscorrelation on the comet; the cycle was finished by another autocorrelation ON-OFF on the comet. For Hyakutake the cycle did not contain the autocorrelation ON-OFFs. The telescopes tracked the comet to second order. For this, we provided precise ephemerides and up to second order derivatives as well as geocentric radial velocity each 60 min for Hale-Bopp, each 10 min for Hyakutake. The velocity was kept constant for each of the observations in the cycle at the value at the beginning of this observation.
Calibrator for Comet Hyakutake was 1413+135, and the bandpass was measured on 3C273. Calibrator for Comet Hale-Bopp was BL Lac (2200+420). MWC349 was used to determine the flux density of BL Lac at the frequency used. 250 GHz fluxes were also provided from the 30m telescope and showed no variation during the days concerned, For bandpass calibration 0415+373, 3C273, and NRAO530 were measured.
The evaluation was done with the standard program CLIC (Lucas 1996). The amplitude calibration is based on for BL LAC and for 1413+135, where is in GHz. To produce the uv- tables all lines seen in the ON-OFFs and additionally all lines listed in the line catalog, supplied to us by Bockelée-Morvan, were blanked out. The uv-tables were processed and cleaned with the standard program GRAPHIC/MAPPING (Guilloteau et al. 1997).
2.4. NRAO V.L.A. (Socorro)
The VLA consists of 27 25-m antennas spread out in a Y shape on the plains of San Augustin, New Mexico. It was used to observe Comet Hyakutake at a frequency near 22 GHz (K-band). The VLA electronics system allows 2 independent frequency ranges to be received, which were centered at 22.4351 and 22.8351 GHz. Stokes LL and RR were measured independently then combined into total intensity (Stokes I). The total equivalent bandwidth after combining the 2 Stokes and 2 frequency bands is about 190 MHz.
For the observations, the following observing strategy was employed. Time variable phases and amplitudes were adjusted by monitoring a nearby point source calibrator. Absolute flux densities were tied to a relatively strong point source with presumed known flux density. The absolute calibration scale should be good to about 5-10 %. Every hour, a special pointing observation at X-band (3.5 cm) was done to adjust the pointing of the individual antennas. The opacity and system temperatures were determined by the TIP procedure. Data reduction was performed in the AIPS processing package. After initial data inspection and flagging, calibration solutions for the nearby calibrator were transferred to the comet data, and these data were used to make maps of the sky brightness distribution.
Comet Hyakutake was observed on 1996 March 25 and 29. The VLA was in the C configuration at the time, with physical separation of antennas from about 35 m to about 3.4 km. All 27 antennas were tuned to K-band on both dates. The comet was observed for 3 hours on March 25 and 2 hours on March 29. The nearby calibrators used were 1436+636 on the 25th and 0217+738 on the 29th. The absolute calibrator was 3C286, with assumed flux densities of 2.496 and 2.500 Jy in the two frequency bands. Weather was good during both observing periods with zenith opacities near 5 %, and total system temperatures near 150 K.
2.5. MPIfR 100m telescope (Effelsberg)
The new 32 GHz multibeam system at the secondary focus has three horns with two channels, each with a bandwidth of 2 GHz. The signals of the total power receivers can be used for "software beam switching" to surpress atmospheric noise. This receiver system was still in the commissioning phase during this observation, so the performance was not fully optimized. Of the five scheduled observing periods, three were lost due to adverse weather conditions.
Observations were done with multiple scans across the two near beams; by folding the scan around the center between both beams one can optimize the integration on source and minimize the signal of confusing sources, which the telescope is crossing while following the comet. As another safegard against confusion the path of the comet was compared with the 4.85 GHz sky survey of Gregory & Condon (1991) with a limiting sensitivity of 25 mJy. During observations the comet did not cross any source in the catalog. This double beam observing method has been used successfully by e.g. Altenhoff et al. (1983) for the observation of Comet IRAS-Araki-Alcock. The observed signals were calibrated against NGC7027, using the scale of Ott et al. (1994).
The ephemerides for the Plateau de Bure Interferometer and the VLA were based on the calculations by P. Rocher (Bdl), and the orbital elements by D.K. Yeomans (JPL). Similarly in the first period the ephemerides at the other sites were based on the most recent orbital elements of Yeomans (solutions 14 to 28) by adjustment of the osculating epoch to the date of observation by taking into account the perturbations of the planets (DE200). In the second period the orbital elements of Marsden (1996, 1997) were used to calculate the ephemerides. For Comet Hyakutake both sets of orbital elements resulted in nearly identical ephemerides, for Comet Hale-Bopp there were small but significant differences between the ephemerides and also between the ephemerides and the observed position of the comet.
© European Southern Observatory (ESO) 1999
Online publication: August 13, 199