The telescope was conceived as a tool for the study of the Galactic Hi, which is known to have structure on all angular scales at which observations have been attempted, from tens of degrees (Hartmann & Burton 1997) to tens of milli-arcseconds (Faison et al. 1999). An angular resolution of 1 arcmin was the target, an order of magnitude improvement over the resolution available with the largest single-antenna radiotelescopes, while retaining sensitivity to intermediate and larger structures. This required a finely sampled aperture plane and short interferometer baselines, which in turn dictated the use of small antennas. It was also clear at the outset that it would be necessary to incorporate single-antenna data into the Synthesis Telescope images (see Sect. 2.3) and to mosaic individual images together to extend the field of view.
A telescope which can successfully image Hi emission will also be suitable for imaging Galactic continuum emission, which similarly displays a wide range of structural sizes. Continuum emission at decimetre wavelengths originates in the ionized and relativistic components of the ISM. The latter gives rise to a small percentage of linear polarization in the emission, and the ability to measure polarization was included in the telescope to give information on magnetic fields at the point of origin and on the magneto-ionic medium along the line of sight.
With only seven antennas, the number of instantaneous baselines is just 21; many more baselines are required to meet the desired imaging criteria. The chosen configuration was therefore an east-west aperture synthesis telescope with movable elements. Earth rotation varies the baseline vector during an observation, and the movable antennas are relocated until the required set of baselines has been completely sampled in consecutive observations. (The normal observing strategy for the telescope is described in Sect. 8).
The task of designing an array configuration to meet the above objectives
was constrained by the existing
four-antenna array, which had two movable
antennas on 300 m of rail, along with two fixed antennas, one at the west
end of the rail and one m to the east. With the possibility of
adding only three additional antennas, the final configuration (Fig. 1)
was dictated by the needs of both continuum and Hi-line imaging
described above.
The diameter of the antennas used in the array is 8.5 m (for further
details see Sect. 3). The baseline increment is L=4.286 m, about half
an antenna diameter. This provides the necessary fine baseline sampling,
and places the first grating response at an angular radius where the
primary beam of the antennas is at a very low level, so that grating
responses to objects within the usable field of view lie outside the field
(although responses to strong sources outside the usable field can still
appear within it). Antenna shadowing and the risk of mechanical
interference limit the minimum usable baseline to
m,
which is sufficient to allow full representation in the images of
structures as large as 40 arcmin at 1420 MHz and 2.3
at 408 MHz. A
maximum baseline of
m provides the desired 1-arcmin
resolution limit at 1420 MHz (at 408 MHz a resolution of 3.4 arcmin is
achieved). Table 1 lists the telescope specifications.
Operating frequencies: | 1420 MHz | |
408 MHz | ||
Number of antennas: | 7 | |
Antenna diameter: | five 8.53 m, two 9.14 m | |
Maximum baseline: | 617.18 m | |
Minimum baseline: | 12.86 m | |
Baseline increment: | 4.286 m | |
Visibility averaging time: | 90 s | |
Field of view: | 1420 MHz | 2.65![]() |
408 MHz | 8.22![]() |
|
Angular resolution: | 1420 MHz |
![]() ![]() |
408 MHz |
![]() ![]() |
|
Radius of first grating ring: | 1420 MHz | 2.82![]() |
408 MHz | 9.82![]() |
|
Polarization imaging: | 1420 MHz | Stokes I, Q, and U |
408 MHz | One hand of circular | |
(usually RHCP) | ||
System temperature: | 1420 MHz | 60 K |
408 MHz | 105 K +
![]() |
|
Continuum bandwidth: | 1420 MHz | 30 MHz |
408 MHz | 3.5 MHz | |
Tuning range for Hi: | -1100 to +3000 km s-1 | |
Spectrometer frequency coverage (B): | 0.125, 0.25, 0.5, 1.0, 2.0, 4.0 MHz | |
Number of spectrometer channels: | 256 | |
Velocity coverage for Hi: | 211B km s-1 | |
Channel separation: | 0.824B km s-1 | |
Channel width: | 1.32B km s-1 | |
Noise on 1-channel spectral map: |
![]() |
The three movable antennas are mounted on motorized platforms that ride on
a precise, machined track that is dimensionally stable and accurately
surveyed. Station markers along the track are used in positioning the
antennas, with a one-station move taking about 5 min. The antennas are
usually moved so that they are 12L apart, with the separation between
antennas 1 and 2 (s in Fig. 1) ranging from 3L to 14L. By this
means, complete coverage of baselines from 3L to 141L (plus 144L) is
obtained with 12 settings of the movable antennas. The time required to
synthesize a complete aperture is therefore
hours. The
telescope is almost always used in this mode. An exception is solar
imaging (Burke & Tapping 1995); for this purpose a set of
spacings has been selected which allows the formation of a satisfactory
image of the Sun (in the absence of rapid variation such as burst activity)
in one 12-hour observation.
For Hi imaging, only the 12 baselines between the four fixed and
three movable antennas are used. Hi-line images are typically of
limited dynamic range, determined by the ratio of the maximum brightness
temperature (rarely more than 120 K) to the expected noise level
( K). However, antenna-based calibration parameters, determined
from continuum measurements using all 21 baselines, are routinely applied
to Hi visibilities.
The short baselines available on the Synthesis Telescope permit accurate imaging of quite large features, but it is still necessary to incorporate data from single-antenna radio telescopes into the images. The DRAO 26-m Telescope is usually used to provide the necessary information on the largest Hi structures. Single-antenna continuum data at 1420 MHz are obtained from the Effelsberg surveys of Kallas & Reich (1980), Reich et al. (1990), and Reich et al. (1997) or from the Stockert surveys of Reich (1982) and Reich & Reich (1986). When a long-wavelength channel was planned for the Synthesis Telescope in the 1980s, the frequency of 408 MHz was chosen because of the existence of a complementary, well-calibrated, single-antenna survey of the whole sky at that frequency (Haslam et al. 1982).
Copyright The European Southern Observatory (ESO)