2. Observations and results
The observations were carried out from August 25 through September 24, 1993 using the 13.7 m telescope at the Qinghai Station of Purple Mountain Observatory. This antenna is located in the Gobi desert, a very dry and arid region at 3200 m above sea level in western China. It is a classic Cassegrain telescope and has a pointing accuracy better than 20" and a HPBW 4.2 at 22 GHz. It is front-end operated with a cooled schottky mixer and a 1.4 GHz FET IF amplifier. The system temperature is about 1500 K. The local oscillator is a phase locked Gunn diode. A high resolution 1024 channels AOS is employed as a back-end. We checked the AOS performance daily during our observations. The measurements suggested that the AOS operated with an average channel separation of 12.10 0.02 kHz and a frequency resolution of 22.11.8 kHz (Zheng & Lei 1998).
Observations were made in position-switching mode and a 120 K noise diode was used as a second calibrator. To check the stability of the noise diode a number of continuum sources were observed. During the period of observations the weather conditions at the site were excellent. The atmospheric opacity was about 0.05 in the zenith direction. The telescope sensitivity was 38 Jy K-1, and the absolute calibration for flux density was about 20%.
In order to eliminate any gain dependence effect of the radio telescope all the observations were carried out at the same sidereal time, i.e. at the same parallactic angle and elevation angles. To discriminate between the instrumental effects and the real changes in the flux density the monitoring program included 14 sources selected with different spectral features. The spectrum of W3(OH) water maser emission shows three components at VLSR= -47.5, -50.3 and -52.8 km s-1 (see Fig. 1). The rms noise level (1) in the plot is about 1.0 K, obtained in 240 s integration. The peak antenna temperature, the radial velocity, the linewidth and the line area for each component was fitted by a one-dimensional Gaussian model. In order to eliminate the effects of uncalibrated gain fluctuations, the strongest component at VLSR=-50.3 km s-1 was selected as a reference feature. The light curve of the reference feature and amplitude ratio curves to the reference feature for the other two components are plotted in Fig. 2. The -50.3 km component was the most intense of the three components during the whole observing period, and its peak intensity was about 1830 Jy (see Fig. 1). The flux density of the -52.8 km component decayed linearly, and was still decreasing when our observations ceased. The period for the feature to reach half the initial intensity was about 19 days. The flux density of the component at -47.5 km showed no significant variations or, if there was any variation, it was of the same order of magnitude as the rms value. The time-variations of the W3(OH) spectra are displayed in Fig. 3 for selected scans. The time variability of the -52.8 km component is apparent.
In Fig. 4 the relation between the flux density, F, and the line width, V, of the -52.8 km component is plotted. The data were fitted by a straight line V . This relationship is very close to the results obtained by Mattila et al. (1985).
The spectrum of the water maser emission in the NGC 6334C region, shown in Fig. 5, manifests three water components at -81.2, -84.4 and -86.8 km . The other two bumps in the line wing may result from the combination of these three components. As mentioned above, the strongest component at VLSR=-81.2 km s-1 was selected as a reference feature. Ratios of amplitude to the reference feature for two other components as well as the light curve of the peak component are plotted in Fig. 6.
After a phase of slow climbing of about 15 days, the flux density of the -84.4 km component jumped from 1090 to 2200 Jy. The timescale was about 4 days. We found that the event happened in all three components (see Fig. 6). Unfortunately, we did not have continuous records of an hour or shorter time interval for the source. Therefore, a positive assessment for instrumental effects or intrinsic changes may not be determined completely. If the event in these components is true, one alternative explanation is an external pumping event, such as clump-clump collision, which will be discussed in detail below. The external event affects the three components simultaneously, but with different degrees of impact. The time-variations of the maser source are displayed in Fig. 7.
© European Southern Observatory (ESO) 2000
Online publication: December 15, 2000