scale of photograph 1 : 40 000 
Lei from aerial mapping scale of map 1 : 20000 ao 
contourline-interval 20 m 120 
scale of map 1 : 10000 on 
from plane-table surve 
P Y 1contourline-interval 10 m 80 
  
Fig. 5. Comparison of profiles from photogrammetric map with profiles from larger-scale map 
prepared later by terrestrial methods. 
Coverage: heavy tropical jungle. 
scale of photograph 1 : 40 000 
umm from aerial mapping scale ot map 1 : 20000 
contourline-interval 20 m 
    
scale of map 1: 10000 
contourline-interval 10 m 
100 À from plane-table ser] 
   
80 80 
60 
40 
  
1 km 2 
Fig. 6. Comparison of profiles from photogrammetric map with profiles from larger-scale map 
prepared later by terrestrial methods. 
Coverage: heavy tropical jungle. 
system and the area is covered by thick jungle. 
The errors shown in the profile from the photogrammetric map as compared 
with that from terrestrial mapping are therefore a combination of the above dis- 
crepancies and the photogrammetric errors. 
Comparison of the profiles shows that the deviations are not, as a rule, 
greater than half the distance between the contour lines. Here and there greater 
deviation is present, but that may be attributed to lack of ground control data or 
very irregular height of vegetation. In general, however, 
the errors have not exceeded what is permissible. 
Use of the Map. 
Taking the photogrammetric maps as a basis for calculations, a few alter- 
native locations for the road can be projected, which must then be reconnoitred 
on the ground in order to enable a final decision to be taken on the best location. 
This is followed by more detailed mapping of a narrow strip along the definite 
location, felling of trees and land clearance, and the survey of preliminary 
longitudinal and cross-sectional profiles. The definite project can then be made, 
and axis and profiles finally staked out. Where the contour maps described above 
7 
it has been found that