therefore linear interpolation is used to estimate laser elevations
at the coffee can site. Results showed that laser elevations
agreed to within 0.10 m of the coffee can elevation.
7. TOPOGRAPHIC MAPS
Large portions of ice streams B1, B2, C, and E have been
surveyed. Topographic maps have been made for all four ice
streams with varying uncertainties. An example is shown in
Figure 6 for Ice Stream C. Ice surface contours compare very
well with an earlier radar study (Retzlaff et al., 1993), although
the older maps have uncertainties that are too large to be useful
for detection of changes in surface elevation. The laser derived
surface map is superimposed over bed elevations taken from
radar measurements (Fig. 6) to show how bed features affect the
ice in terms of velocity vectors, surface elevations, and slope.
8. GLACIOLOGIC INTERPRETATION
Ice Stream C was chosen as a likely location for rapid ice
thickness change because the downstream portion is nearly
stagnant while the upstream portion is still active (Fig. 6).
Within the laser grid, the ice velocities go from 25 m/a to 3 m/a
(Whillans and Van der Veen, 1993 and Joughin et al., 1999).
As more ice moves into the region, ice thicknesses are expected
to increase by 0.50-1.0 m/a (Whillans and Van der Veen, 1993
and Joughin et al., 1999). Figure 6 also shows that surface
slopes are greater in the zone where velocity decreases the
quickest. This is an impossible scenario for steady-state
conditions and could be evidence that Ice Stream C is being
reactivated. It seems sensible that a build-up of ice will cause
the ice to begin streaming again.
The pattern of ice thickness, ice velocity, and surface slope
observed on Ice Stream C are unlike what is normally expected.
Ice thickness decreases by a factor of two from the upstream
portion of the grid to the downstream portion mostly as a result
of increased bed elevations. The velocity decreases by a factor
of four for the same region and the surface slope increases by a
factor of 2. In an ordinary glacial setting, the velocity would
increase. with increased surface slope and decreased thickness.
The ice seems to be piling up as it reaches the bedrock ridges.
This could be due to a lack of subglacial water at the
downstream end. If the upstream portion of the ice stream is
sliding over clay-rich sedimentary rocks, basal friction would be
very small for two reasons. First, ice thicknesses are great
enough in this region to induce pressure melting despite cold
temperatures. Second, clay rich materials have very low
permeabilities so water is not able to escape. This scenario has
been shown to exist beneath other fast flowing ice streams
(Anandakrishnan and Bentley, 1993; Engelhardt et al. 1990).
As for the downstream portion of Ice Stream C,
Anandakrishnan and Bentley (1993) found basal seismic noises
that were not sensed beneath fast flowing ice streams. The
interpretation was that Ice Stream C must have experienced a
dewatering and loss of dilatency in the lubricating till layer.
Without well-distributed water the ice can become frozen to the
bed, forming ‘sticky’ spots, or in this case an entire ‘sticky’
International Archives of Photogrammetry and Remote Sensing, Vol. 32, Part 3W14, La Jolla, CA, 9-11 Nov. 1999
region. The future course of the growing bulge will determine
the behavior of ice stream C and will illustrate the effect of
glaciologic processes, presumably the advancement of a
lubricated bed. One can expect a major change of the studied
region as this thickening continues.
-530+
-5404
-5504 -700m
-800m
-900m
-5604
-1000m
-1100m
-570+ =
ESO -
>
-5908 =
-6004 a
Coffee-can
-6103 "
3 Ice Velocity e
and direction
6208
Arrow Scale: m/a
-660 -640 -620 -600 -580 -560
X (km)
Fig. 6. Laser derived elevation contours (blue) of the surface of Ice
Stream C superimposed over bed elevations from airborne radar surveys
(Taken from Retzlaff et al., 1993). Red arrows show ice flow direction
and magnitude (Taken from Whillans and Van der Veen, 1993, and
Hamilton, pers. comm.). The yellow diamond is a coffee-can site
whose known elevation is used to validate laser derived elevations
(Hamilton, pers. comm.). Dashed black line shows the track of the
laser surveys.
9. CONCLUSIONS
At the onset of this project it was expected that uncertainties
would be at least 1-meter. The large uncertainties were
expected to come from the GPS solutions because the GPS
satellites that cover Antarctica are at low angles making it
difficult to resolve errors in the vertical. Uncertainties in the 10
to 30 centimeter range were found instead. Most elevations
from crossing flight lines compare within 20 cm during the
same GPS survey and within 30 cm for crosses involving two
separate GPS surveys. This level of precision is acceptable, but
could probably be improved with better GPS surveying
techniques. These would include static initialization at the
beginning and end of each survey, beginning and ending flights
at the same base camp, and more surveys conducted on the
snow surface for bias determination. Nevertheless, this level of
precision will allow changes in ice sheet elevations to be
detected in the 2 year time interval for most of the study region.
Laser altimetry is shown here to be a valuable tool for mass
balance measurement for all of Antarctica. It is also shown to
be a useful tool in locating unusual topographic features that
can lead to a greater understanding of glacier mechanics.
International
ACK
We would like to than
GPS surveys, Dorot:
processing the GPS dat
and helping with proce
is provided by Nati
9615114.
[Adalgeirsdóttir et al.,
K. Harrison, W., 1996
Harding Icefield, Alask:
[Anandakrishnan and B
Bentley, C., 1993. Mi
and C, West Antarctica:
of Glaciology, 39 (133),
[Bindschadler and Vor
Vornberger, P., 1998. (
since 1963 from decla:
279, p. 689-692.
[Csatho et al., 1996] C
Krabill, W., 1996. Rem
altimetry. Internation:
Remote Sensing. XXXI
[Echelmeyer et al, 19
Larsen, C., Sapiano, J.,
Adalgeirsdottir, G., Sor
profiling of glaciers: A c
(142), pp. 538-547.
[Engelhardt, 1990] Eng
Fahnestock, M., 1990. I
moving Antarctic ice stre
[Garvin and Williams, 1€
Geodetic airborne laser
Skeidarárjókull, Icelanc
Greenland. Annals of Gl:
[Hamilton et al., 1998]
PJ., 1998. First point
change in Antarctica. An
[Joughin et al., 1999] J
Price, S., Morse, D., Hi
1999, Tributaries of
RADARSAT interferome
