on 
he 
p- 
e- 
de 
en 
ea 
1e 
rd 
VE 
W 
11 
  
200mm“ =" 4 ^ ÁN e 
a—— 9 
T . 
| Imm error 
A r M - - ~ S "u A 
  
  
  
Fig. 6. Error-vectors in a PMRT-triangulation with "theoretical" templets having 
"practical" errors and performed according to the author's method. Seven strips 
with 12 templets each. Average distance between fixed controi point studs in the 
corners 1,5 meter. Mean radial error 0,24 mm, maximum 0,60 mm. 
some individual results may be of special interest as e. g. the follo- 
wing: 
In area-triangulation the standard error in distances 125—500 mm in 
lenght varied between + 0,1 and + 0,2 mm, which clearly shows the 
very low "local error". The typical "systematic" errors with whirl pat- 
tern can also be seen in fig. 6. 
When using a ground control density of 2,7 points/100 sqkm, which 
is normal for multiplex-triangulation in Sweden, and assuming a picture 
scale 1:20 000 and a triangulation scale 1:16 000 the mean error for the 
triangulation with the theoretical but not error-free templets was less 
than + 0,1 mm. 
Parallel tests with American star templets showed that their accuracy, 
stability and the assembly adjustability were much lower and that they 
could only be recommended for approximate work. 
The MRT and PMRT are typical area- or block-methods from tech- 
nical, economical and accuracy viewpoint. But in order to get some com- 
parison with other methods, mechanical strip radial triangulation tests 
were performed with ten templets parallel with numerical strip radial 
triangulation tests with and without rigorous adjustment and parallel 
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
   
   
  
  
  
  
  
  
  
ei a v eb ry HO