next previous
Up: A survey of SiO


Subsections

3 Observations and data analysis

 

3.1 SEST observations

The observations of the southern sources ($\delta < 0^\circ $) were made during two periods in October 1995 and in April 1996, with the SEST[*] on La Silla in Chile. We measured the (v=0,J=2-1) and (v=0,J=3-2) transitions of SiO simultaneously, using the 80 - 116GHz and 135 - 165GHz SIS receivers. The backend was a 2000 channel acousto-optical spectrometer with a bandwidth of 86MHz. The AOS was split in two bands, each of 43 MHz. The precise transition frequencies, the half-power beam widths (HPBW), and the antenna ($\eta_{\rm A}$) and beam efficiences ($\eta_{\rm MB}$) of the telescope are listed in Table 1. Also the Jy/K conversion factors assuming a point source, and the velocity resolutions corresponding to the channel separation at the frequencies 87 and 130 GHz are listed there. Further details of the equipment can be found in Booth et al. (1989) or in the SEST manual (see http://www.ls.eso.org/lasilla/Telescopes/SEST).

Typical system temperatures at 87 and 130 GHz were 130 and 160K, respectively. The observations were made in the dual beam switching mode, with a beam throw of $11^\prime_\cdot5$. The pointing was checked by observing circumstellar SiO maser sources, and the accuracy was found to be better than $3\hbox{$^{\prime\prime}$}$.


  
Table 1: Antenna and beam efficiences, the Jy/K conversion factor for a point source, and spectrometer resolutions for the 15-m SEST and the 20-m Onsala telescope at the observed frequencies

\begin{tabular}
{\vert lr\vert ccccc\vert cccccc\vert}\hline 
 & & \multicolumn{...
 ...0 & 40 & 0.56 & 0.68 & 28 & 0.10 & - & - & - & - & - & - \\  \hline\end{tabular}

3.2 Onsala observations

The northern sources ($\delta \gt 0^\circ$) were observed in January 1995 and in March 1996 with the 20-m radio telescope of Onsala Space Observatory in Sweden. The SiO(v=0,J=2-1) line was measured using a 3mm SIS receiver. As backends we used both the autocorrelator with a resolution of 50kHz, and the 256 channel filterbank of resolution of 250kHz. The antenna parameters and the velocity resolutions of the spectrometers are listed in Table 1. For most of the sources the signal-to-noise ratio in the autocorrelator spectra was not high enough to study the line profiles at low intensity levels. On the other hand, the relatively low resolution of the filter-bank (0.86kms-1 at the line frequency) is sufficient as the lines are typically broad. Therefore the analysis is based on the filter-bank spectra alone.

The first observing run was hampered by snow and rain, and consequently the system temperature varied between about 240-1000K. During the second run, the typical system temperatures were between 300-400K. The observing mode was dual beam switching with a beam throw of $11\hbox{$.\mkern-4mu^\prime$}5$. The focus and pointing was checked by observing SiO masers in late-type stars. The pointing accuracy was found to be better than 4$\hbox{$^{\prime\prime}$}$.

3.3 Determination of the line parameters

Because many of the spectra are asymmetric, gaussian fits are not applicable. Linear baselines were first subtracted from the spectra. Due to spectral noise it is difficult to determine the minimum and maximum velocities ($V_{\rm min}$, $V_{\rm max}$ in Tables B.1 and B2) of the low-intensity line wings. Therefore the spectra were first smoothed by averaging adjacent channels in order to reduce the noise level.

The SEST spectra were smoothed by averaging 8 adjacent channels resulting in a velocity resolution of 1.2 and 0.79kms-1 for the J=2-1 and J=3-2 transitions, respectively. In the Onsala spectra two adjacent channels were averaged and the resulting velocity resolution is 1.73kms-1. The smoothing lowers noise but also reduces the accuracy of the determined velocities. However, compared with the typical full widths of the lines as defined below, the obtained spectral resolution is reasonable.

The peak antenna temperatures and the corresponding velocities ($T_{\rm
A}^{\star}$ and $V_{\rm peak}$ in Tables B.1 and B.2) were determined directly from the smoothed spectra. The full velocity range of detectable emission, which hereafter will be called the full width (FW), was then determined as the range where the channel values are greater than twice the RMS noise of the smoothed spectra. Due to smoothing the selected intensity threshold corresponds to $T_{\rm rms}/\sqrt{2}$ for the SEST observations and $T_{\rm rms} {\cdot} \sqrt{2}$ for the Onsala observations, where $T_{\rm
rms}$ is the RMS noise of the original spectra indicated in Tables B.1 and B.2. The area, mean velocity and width (the columns Area, $V_{\rm mean}$ and Width in Tables B.1 and B.2, respectively) of the emission within the range $V_{\rm min}-V_{\rm max}$ were calculated using the original spectra. The mean velocity and width are defined as the first and second moments of the channel values.


next previous
Up: A survey of SiO

Copyright The European Southern Observatory (ESO)