isformation 
M with the 
istics, and 
on offset of 
o after the 
1e different 
rmations: 
084m 
343 m 
828m 
8 o 
869 
30° 
9 
100 
ver stable 
005 aerial 
‘hing. 
uncertainty 
on. In the 
-day aerial 
he scale is 
1 imperfect 
> matching 
he derived 
edundancy 
points, as 
in a least 
terrain, at 
EM before 
toric DEM 
ranslations 
! the lower 
TER data. 
it could be 
here. An 
observed. 
Antarctic 
e between 
e vertical 
OWS steep 
relief gradients. However, a particular advantage of one 
technique over the other was not found. 
5.1. Glacier change assessment 
The successful DEM co-registration allowed glacier change 
measurements to be made. Figure 5 shows an elevation profile 
along the glacier front of Leonardo glacier. Substantial surface 
lowering can be observed while the glacier front remains 
relatively stable. The lowering is higher near the front and 
reduces further upstream. Maximum elevation changes of up to 
50 m are observable which translates into an annual lowering of 
~1.5 m/yr over the period of 33 years (1968-2001). Similar 
results were found for Nemo glacier where a lowering of up to 
30 m (~0.8 m/yr) at the glacier front was observed (1969-2005). 
While the front of Leonardo glacier remains relatively stable 
over the observed period (see Figure 6), the front of Nemo 
glaciers shows significant retreat of ~165 m between 1969 and 
2005 (not shown here). The frontal retreat was measured 
perpendicular to the glacier front in the ortho-imagery for both 
   
       
  
  
epochs. 
300 s 
w 1968 
—— 2001 
A 
200 = Upstream 
9 
® 
> 
0 
! 190 - 
Front 
0 i 1 A — T. 
0 500 1000 1500 2000 2500 
Distance [m] 
Figure 5: Surface elevation profiles (dotted line in Figure 6) for 
Leonardo glacier, western AP (61.91* W, 64.68* S) for 1968 
and 2001 
Leonardo Glacier 1968 - 2001 
  
——- Frontline 1968 
^ Frontline 2001 
~ Profile 
"Volume sampling area 
  
Figure 6: Satellite image of Leonardo glacier front in 2001 
(dotted line represents profile in Figure 5) 
DEM differencing also allowed the measurement of volumetric 
changes using ArcGIS. À glacier wide mass balance assessment 
was not possible because both glaciers were not completely 
covered by the imagery. However, it was possible to derive a 
specific surface mass balance, assuming an ice density of 
0.917 kg/m’, for an area along the front of Leonardo and Nemo 
glacier. For Leonardo glacier (Figure 6) a mass loss of 25.1 = 
3.5 million m? w.e. over an area of 3.6 km? (1968-2001) was 
determined. The respective mass loss for Nemo glacier was 5.6 
+ 0.6 million m? over an area of 1.5 km? (1969-2005). It should 
be noted, that higher elevation mass gains, due to increased 
accumulation of snow (Pritchard and Vaughan, 2007) might 
outbalance the observed loss. However, this aspect is yet to be 
examined. The overall trend from the results presented here is a 
mass loss along the glacier fronts. 
6. DISCUSSION 
The observed lowering is in agreement with observations of 
regional warming and glacier retreat in the AP over the last 
decades (Cook et al., 2005). The frontal lowering at Leonardo 
glacier (-64.7°S) is higher than at Nemo glacier (~67.7°S). If 
the observed glacier surface lowering is caused by atmospheric 
warming these differences could be explained by changes in the 
mean air and sea surface temperatures across the western AP 
(Meredith and King, 2005; Cook and Vaughan, 2010). The 
recently observed acceleration of glaciers in the AP (Pritchard 
and Vaughan, 2007) is likely to be connected to frontal thinning, 
as observed here. 
Although, this study allowed the measurement of multi-decadal 
glacier change since the 1960s, the temporal resolution is too 
low for the assessment of changes between the decades till now. 
The glaciers studied here only represent a small section of the 
roughly 400 glaciers in the western AP. A better spatial and 
temporal resolution across the whole region is needed to answer 
the following questions: 
e When did the surface lowering begin to be significant? 
e Is the lowering constant over time or has it changed at 
different rates? 
e Is the lowering evident for all glaciers in the AP and 
up to what altitude does the lowering occur? 
e — What is the spatial pattern of change across the AP 
and how does it fit with climatic observations? 
These questions could be answered with further exploration of 
the USGS and BAS image archives. It should be noted, that 
parts of these archive images are not directly suitable for DEM 
extraction e.g. due to extensive cloud cover or high image 
brightness caused by reflection of ice and snow. Currently, 
within this research, more image frames are being processed 
and analysed, with further results to be reported in due course. 
ASTER provides a suitable source of reference data for the 
comparison with the historic datasets if higher resolution data is 
not directly available. However, attention should be paid to 
upcoming data sources with higher resolution such as TanDEM- 
X, which could further extend the time series. Glacier wide