Figure 3. 400-MHz antenna transducer unit in tow behind a Spryte tracked
vehicle. The transducer contains the antennas and transmit and receive electronics.
No reflections from the Spryte were apparent in the records. Spryte speed was gener-
ally about 5 m/s. Also shown is the two-way radar antenna directivity patterns in firn,
both perpendicular (⊥) and parallel () to the profile line direction. The perpendicular
Data recording, filtering and processing
to be surveyed and then slowly dragging the an-
All data were recorded in scans of either 512 or
tennas by hand from diamond to diamond.
1024 8-bit words for later processing by dedicated
Transects along the proposed tunnel route were
software (Galinovsky and Levin 1990). Horizontal
done by slowly dragging the antennas over the
(i.e., over distance or equivalently, many scans)
surface behind a Spryte tracked vehicle (Fig. 3) at
frequency filtering mitigated near-surface ringing
813 km/hr (58 miles/hr) along similarly
from the direct coupling between antennas, and
marked lines. Differences in record length ascrib-
vertical (i.e., over time, or equivalently, along one
able to changes in towing speed between dia-
scan) filtering mitigated dc gain variations result-
monds or flagged bamboo poles were compen-
ing from inconsistent antennaground coupling.
sated for by computer processing. All surveys
Data are displayed as consecutive scans in a dis-
were performed with the antennas polarized per-
tance-vs.-echo-time arrangement (known as a
pendicular to the transect direction so that the
"time section") (Fig. 4) for which several schemes
stronger amplitude H-plane pattern was in the
for displaying signal amplitude are available.
line of the transect. The spacing of the tunnel lines
Only selected sections of the tunnel survey will be
sometimes varied by a few meters to avoid sur-
shown owing to the large amount of data.
face obstacles. Thus, the exact position of targets
Translation of echo time delay t from a target
located is not known. Event markers were placed
into the depth d of the target is generally based on
electronically on the tunnel records at known po-
the simple echo delay formula
sitions along each traverse. The transect line spac-
ing averaged 2.8 m, slightly more than the 2.5-m
d = ct/2n
(1)
width of the vehicle's tracks.
4
Previous Page
