Ground-Penetrating Radar Investigations
of the Proposed DomeCARA Tunnel Route and
Utilities at South Pole Station, Antarctica
STEVEN A. ARCONE, WAYNE TOBIASSON AND ALLAN J. DELANEY
INTRODUCTION
station. The surveys performed near the main sta-
tion are discussed first, as they allowed establish-
Currently, a new facility is being built at South
ment of the snow dielectric properties needed for
Pole Station, Antarctica, for infrared (IR) and
estimating the depth of the objects in the tunnel
millimeter wave (MMW) observatories operated
route survey.
by the Center for Astrophysical Research in Ant-
The tunnel route was surveyed using a GPR
arctica (CARA). This new research site is located
antenna transducer, operating at 400-MHz pulse
about 200 m west of the skiway. Sensitivity to IR
center frequency, that was either pulled by hand
or towed by vehicle along closely spaced, parallel
absence of atmospheric water vapor and by very
transects covering a swath up to 60 m wide. The
low telescope temperatures that reduce ambient
buried facility surveys were done at 100- and 400-
IR noise; consequently, winter operations are nec-
MHz center frequencies. Some of the results were
essary. A tunnel has been proposed to connect the
used to determine snow dielectric properties,
CARA site with the main station. In it, personnel
which were needed to transform the GPR echo
could move safely between the two places, which
time delays into depth scales. As of this writing,
are about 1 km apart, even during the dark, cold
there has been no drilling or excavation to verify
winter-over period. A machine has been built at
the features of interest on the GPR records. Verifi-
CRREL to excavate the tunnel, but the tunneler
cation would be valuable.
may be damaged if it encounters a large metal or
GPR has been a common geophysical tool for
wood object. Thus, the tunnel route must avoid
subsurface exploration in snow and ice (e.g.,
any debris that has ended up in the snow during
Kovacs and Gow 1975, Kovacs et al. 1982, Arcone
the 35 years that people have been present at the
and Delaney 1987, Arcone 1991) and frozen soils
South Pole.
(e.g., Annan and Davis 1976, Kovacs and Morey
Commonly, magnetometry and low-frequency
1979, Delaney et al. 1991, Arcone 1992) because of
magnetic induction methods are used for metal
the ease at which radio waves penetrate these
detection. However, the nonmetallic nature of
materials. Commercially available, low-power,
some of this debris, the 10-m depth of interest and
low-gain GPR systems are able to penetrate sev-
the need to locate the objects accurately makes
eral meters in fine-grained permafrost, tens of
Ground-Penetrating Radar (GPR) the preferred
meters in frozen sands and gravels, and well over
method.
a hundred meters in polar snow and ice. Vertical
The objectives of our GPR surveys were 1) to
resolution is determined by the bandwidth center
locate any objects buried in the snow along the
proposed route of the tunnel, 2) to determine the
ing near 500 MHz, resolution is generally about
radar's ability to delineate known objects associ-
24 cm in dense snow, 20 cm in ice and less for
ated with the buried facilities near the main sta-
dielectrically denser materials such as perma-
tion, and 3) to assess the possible hazard of sew-
frost. This exceptional resolution has led GPR to
age sumps to the new water well near the main
become a popular device for characterizing utili-