[mmaimcal] Survey requirements

[Simon Radford]

Comments, anyone?


Chajnantor survey requirements

To plan the actual ALMA configurations will require iteration between
the conceptual configuration design and the field identification of
locations where it is feasible to put antenna foundations, build roads,
etc. The MMA project book states this detailed planning should begin
during the upcoming year. Here I outline the resulting requirements for
surveying the site and a recommended technique.

ALMA configurations

The current ALMA goal is sixty-four antennae each 12 m in diameter.
There will be five configurations ranging from a compact configuration
about 140 m in diameter (50% filling factor) to an extended array 10 km
in diameter. To optimize the array configurations, the relative
locations of the antenna stations must be known with some precision.
Strictly speaking, the required precision varies with both the number of
antennae and with the configuration size. As a practical matter,
however, it is most expedient to adopt a uniform precision, say 10% of
the antenna diameter or about 1 m, that meets the most stringent
requirements.

Different configurations present different layout challenges.
Optimization of the compact configuration, for example, is very
sensitive to perturbations in the antenna positions. But this
configuration is perhaps the easiest to lay out on the ground. It is
about the size of a sports field or running track (140 m diameter).
There are several relatively flat areas of this size on Chajnantor. We
need only choose one of these areas, make it as flat as needed, and
install the antenna foundations where desired. Even the next array,
about 430 m diameter, can probably be laid out in this manner. 

The extended configurations, on the other hand, are more tolerant of
perturbations in the antenna positions, but must be laid out with more
attention to the topographic details of the site. Soil conditions and
local topography vary across the site and road feasibility, cable
lengths, etc., must also be considered.

Surveying the extended configurations will take a considerable amount of
field work. For the 1225 m configuration, the typical separation is
about 60 m and we might expect to survey two or three stations per hour,
completing the entire configuration in a week. For the 10 km
configuration, the travel time between stations will be longer and we
should probably count two weeks for a survey.

Ground control

Configuration layout requires setting up a survey network on Chajnantor.
Prior to surveying the target antenna stations, etc., several benchmarks
should be established around the site. These project benchmarks should
be spaced about 5 km apart so every target location lies between two
benchmarks. A secondary objective is registration of these project
(secondary) benchmarks with the Chilean geodetic network. This requires
identification of the nearest official (primary) benchmarks. We may also
be able to use benchmarks installed for construction of the road or the
gas pipeline. Landmarks, such as mountain peaks, etc., should also be
surveyed for map and aerial photo registration, etc.

Survey techniques

Modern surveying is based on the Global Positioning System (GPS). In
increasing order of sophistication, precision, and expense, GPS
techniques include navigation, mapping, static survey, and real time
kinematic survey receivers.

Inexpensive (< $200) and readily available, navigation systems receive
the Standard Positioning Service signals (L1 C/A code) and offer a
horizontal accuracy of +- 15 m (1 sigma) in each dimension. This is
inadequate for configuration layout on Chajnantor. Substantially better
accuracy can be achieved by averaging repeated measurements for long
times (MMA memo 261), although this is impractical for surveying. 

Differential techniques, where the GPS signals received at a benchmark
and at the target location are compared, provide higher accuracy.
Mapping receivers achieve real-time sub-meter accuracy using
differential code phase corrections broadcast from either terrestrial
stations or (commercial) communications satellites. A typical mapping
system costs about $12000 and requires a $2000/year subscription to
satellite correction service for use at Chajnantor.

Surveying systems use two or more receivers with at least one on a known
benchmark and the others at the target location. These systems use the
carrier phase as well as the code phase. In static surveying, there is
no real time communication between the receivers, but data are recorded
simultaneously and the data are postprocessed. The postprocessing
software runs on a standard PC. A single frequency (L1), two receiver,
static survey system costs about $12000. This has a horizontal accuracy
of 0.5 cm + 1 ppm x baseline length (rms) with a 15 minute integration.
Additional receivers can be placed on benchmarks to improve the accuracy
because the redundant baselines provide a closure check. Accuracy on
long baselines can be improved by longer integration times. Dual
frequency (L1/L2) systems are somewhat more accurate, but
correspondingly more expensive.

Real time kinematic (RTK) receivers use radio modems to transmit
differential carrier phase corrections from a reference receiver,
stationed on a benchmark, to a roving receiver at the target location.
These systems cost $35000 or more. The accuracy is comparable to a
static system, but the results are available in real time. Typical
modems are narrow band, 410-470 MHz systems with a line of sight range
of about 10 km or spread spectrum, 902-928 MHz units with a shorter
range. In the US, 410-470 MHz systems require an FCC license (and there
was a flap over co-channel interference last year that lead to a
temporary suspension of these licenses). The licensing requirements in
Chile are unknown. RTK systems can also be used for static surveys
beyond the range of the radio modems.

Current measurements

To date, many measurements of various points on Chajnantor have been
made with navigation receivers, but these have not been very systematic.
A long integration (4 months) with a navigation receiver determined the
position of the MMA equipment container within 16 cm. Later in 1999, an
Onsala group plan to install for an extended period a geodetic GPS
receiver for atmospheric studies. As a byproduct of this work, the
global position of the receiver will be established to within a few mm.

Recommendation

With an accuracy of 1 cm or so, a carrier phase correction, static
survey system would meet our requirements for configuration layout, even
in the hands of amateur surveyors (astronomers). A code phase
correction, mapping system is unattractive because it is less accurate,
is no less expensive, and requires an annual subscription to the
correction service. We do not require the higher accuracy provided by a
dual frequency, RTK system, so the added complexity and (especially)
expense of a this technique seem unjustified. 

Static surveying is a well established technique used for many civil
engineering projects (e. g., Shinkel 1998). Because the survey results
are not known until after postprocessing, the technique is particularly
well suited for determining the positions of preexisting landmarks or
for establishing new benchmarks where the exact location of the
benchmark is unimportant. 

For configuration layout, however, we want to determine the physical
location of unmarked points corresponding to the model configuration.
Hence an additional survey step will be necessary. First, the
approximate location of a target antenna station, etc., is determined
with a GPS navigation receiver and the area is inspected. Then two
positions about 10 m to either side of the target are marked and
surveyed with the static GPS system. After the survey data are
processed, the exact position of the target station is established by
simple geometric construction (intersection of arcs) from the two
surveyed markers. Finally, the target position can be surveyed if
necessary.




References

MMA Project Book, Chapter 15, Configurations
http://www.tuc.nrao.edu/~demerson/project_book/chap15/chap15.html

MMA Memo 261,
http://www.mma.nrao.edu/memos/html-memos/abstracts/abs261.html

K. Shinkel, Conducting Highway Surveys the NYSDoT Way, 1998 February GPS
World, http://www.gpsworld.com/columns/0298column/0298column.html