M.A. Holdaway and S.J.E. Radford
National Radio Astronomy Observatory
949 N. Cherry Ave.
Tucson, AZ 85721-0655
email: mholdawa@nrao.edu, sradford@nrao.edu
February 15, 1998
We illustrate various options for placement of a second site testing interferometer on the Chajnantor site, mentioning the various scientific benefits and technical drawbacks to each option.
Introduction
We have been monitoring the atmospheric phase stability of the Chajnantor site in Chile since May, 1995. With superior phase stability and 225 GHz opacity, it seems likely that the MMA, the LMSA, and the LSA projects will all converge upon the Chajnantor area. As each project has a substantial investment in site testing equipment, including interferometers to monitor the atmospheric phase, it seemed like a good idea to explore different wasy of utilizing multiple site testing interferometers on the same site. We present seven different concepts for using two interferometers on the Chajnantor site. The concepts which we prefer are presented first.
Figure: Option 1: The interferometers are collocated, but pointing
to different satellites. As the velocity aloft can be measured from
one site testing interferometer, when the wind direction is parallel
with the E-W baseline (as it usually is), the time delay of turbulence
crossing from one interferometer's line of site to the other's provides
us a means of determining the height of the turbulence. This will
also provide us with a means of estimating the evolution of the
turbulence on very short time scales (ie, a few seconds). The
technical drawback is that there may not be an appropriate second
satellite. This is the only option presented that relies upon a
second satellite.
Figure: The geometry of Option 1 for the Intersat 601 satellite
at elevation and azimuth (which we observe from
Chajnantor) and the Intersat 605 satellite at elevation and
azimuth. Under this geometry, accurate determination of the
phase fluctuation delay between the two interferometers plus an
accurate determination of the velocity aloft should allow us to determine
the elevation of the turbulent water vapor.
Figure: Option 2: The interferometers are colinear, and one antenna
from each interferometer is collocated. The atmospheric phase on the
two collocated antennas will be the same, so adding the correlated
phase of interferometer a to that of interferometer b will result
in the phase one would get from an interferometer made up of the two
non-collocated antennas. In the case illustrated, this gives us two
300 m baselines and one 600 m baselines. We stress that the phase
difference on the third baseline does not require any special
electronics, connections, or sub-second synchronization of the
correlating computers; the computers should be synchronized to within
a second to permit the calculation of the third baseline's phase
offline from the two interferometers' data files. If two baselines of
identical length are used, we can monitor the velocity of the
turbulent structure and the evolution of the turbulence on 300 m size
scales as it passes over the interferometer. Evolution of the
turbulence is important for paired array phase calibration
techniques that monitor the phase screen as it passes over a compact
array. Another option worth considering is a non-redundant version
with baselines of, say, 300 and 150 m, with an inferred third baseline
of 450 m. This would require shortening the cables on the second
interferometer.
Figure: Option 2a: Conceptually the same as Option 2, this option
uses the same trick to get a third baseline which is shorter
than 300 m. This option can monitor the velocity of the turbulent
structure and the evolution of the turbulence on 150 m size scales as
it passes over the interferometer. The 150 m size is important for
the most compact arrays. Another option worth considering is a
non-redundant version with baselines of 300 and 200 m, with an
inferred third baseline of 100 m.
Figure: Option 3: Two 300 m baselines monitor the turbulence on
different parts of the site. A comparison between the NRAO Chajnantor
interferometer and the NRO interferometer 10 km distant at Pampa la
Bola has shown that the phase stability can vary by a factor of 5-10.
Placing the second interferometer 1-3 km from the existing one on the
Chajnantor site would provide interesting data.
Figure: Option 4: Placing the interferometers in different orientations
provides information on anisotropic turbulence. However, other
interferometer arrays such as the VLA can study this effect already.
Typically, phase fluctuations on baselines perpendicular to the wind
agree with fluctuations on parallel baselines to 20 or 30%.
Figure: Option 5: Placing the antennas for the interferometer pairs
10-20 m apart would provide a test of the paired antenna calibration
method. The orientation of the vector between the paired antennas is
not specified, but if the interferometers were colinear parallel to
the prevailing wind direction, we could also monitor evolution of the
turbulence on short scales. At this point, no array is seriously
considering paired antenna calibration.
Figure: Option 6: The second interferometer could be installed with
a much longer baseline, say 1-3 km, to provide information on the outer
scale of turbulence. However, since it is well recognized that the phase
problem must be solved on the shortest scales (ie, seconds), the long time
scale or long spatial scale turbulence will be removed automatically
and is not a particularly important issue. Furthermore, some engineering
work would be required to install a phase-stable transmission line
on the much longer baseline.
Figure: Option 7: All antennas could be combined into a single interferometric
array, producing six correlated baselines. This provides an
independent verification that the measured spatial phase structure
function indeed agrees with the spatial phase structure function
inferred from the temporal phase fluctuations on a single baseline.
However, this comparison has been done with the five element Nobeyama
phase monitor system, and can be done at astronomical arrays such as
the VLA as well. Furthermore, it would require re-engineering the
interferometers.
MMA Memo 196: Options for Placement of a Second Site Test Interferometer on Chajnantor
Kate Weatherall