500 
Insufficient data 
  
  
    
  
for contouring 68°57’ 68°55’ 
1 / o ’ 
© A 70°:30S 
to e Insufficient data 
for contouring 
© 
© 
xr 
70931 
Insufficient data 
for contouring 
  
x7 
Insufficient data 
for contouring 
70°32’ 
  
  
Figure 3 A map of a part of the Colbert Mountains, Alexander Island, Antarctica. The contour interval is 25 m. Exposed 
rock is shown by grey shading. 
BLOCK ADJUSTMENT 
The block adjustment was carried out using the BLOKK 
suite of programs provided with the CARTO Instruments 
AP190 analytical stereo-plotter. Several combinations of 
the control points were used to find the optimum selection. 
The best results have an RMS error of x60 m horizontally, 
w 
  
  
  
  
  
58°57 98:551 68:53" 70°32 
  
Figure 4 A map of the same region as fig. 3, compiled from 
satellite imagery at 1:250,000 scale in 1978. 
and +15 m vertically. The various combinations of points 
used are shown in Fig. 2, and the corresponding errors in 
Table 1. 
The lowest errors were obtained from test 1, which only 
contained 6 control points. Although this is more than the 
minimum number required to control the block, the 
distribution of control points is weak, particularly in the 
lower left quadrant. Test 2 used all the available control 
and thus improved the distribution of points. Unfortunately, 
the error is much higher, suggesting that some of the 
control points are in error. Test 3 was designed to 
investigate the accuracy of the points derived from the 
shape-from-shading algorithm. This gave very good results 
but the distribution of control is unacceptable since the left- 
hand end of the block completely lacks control. However, 
the low error figures, especially in the vertical dimension, 
shows that the shape-from-shading algorithm has proved 
reliable. Test 4 attempted to add control to the left-hand 
side of the block and the poor result shows that some of 
the points in the lower left corner must have large errors. 
This is the only test where the vertical error is 
unacceptable. Test 5 used all the points with 
measurements in three dimensions, and the result is 
comparable with that from test 3; it gives a better 
distribution of control but the lower left corner is still 
unsupported. Tests 6 and 7 represent attempts to 
424 
strengthen 
apparently | 
The overall 
control in 
achieved in 
with the le 
countries, 
region. A fu 
measureme 
The lower e 
use. 
The output 
AP190 ana 
models of 
extracted c 
crevasses : 
detail for st 
elevations ç 
shading alc 
contours on 
The map c 
25 m, can 
compiled a 
detailed line 
Dr. J. Sieve 
provided the 
Mrs J.W. ' 
much helpf 
University o 
project, whi 
Knight, J.A. 
Antarctic € 
Antarctic SL 
36 pages [| 
Mcintyre, | 
altimeter: In 
p. 1775-17¢ 
Rapley, C.( 
Birks, A.R. 
Cooper, A.F 
Huckle, H.t 
Milne, K., | 
Scott, R.F., 
study of sat 
surfaces: E 
Renner, R. 
network ov 
Survey Bull