ich means that 
must than be 
vn the strip in 
1g-line or simi- 
termined with- 
every plotting 
orous, but still 
line interval is 
1g lenses place 
'ACy In separa- 
iphs are taken 
J0 meters) and 
30—65 sqkm), 
— and I do not 
ig — the prob- 
te obstacle for 
> we must ob- 
or many activi- 
ind 
es normally is 
the determina- 
1g and orienta- 
as inaccessible 
artics but than 
it high altitude 
be substituted 
surements. 
asized that — 
ping — photo- 
n many cases 
:strial levelling 
d for planning 
strial levelling 
ening there is 
tal accuracy i$ 
m the geodesy, 
letic points of 
:oncerned, this 
] purposes. In 
ially in block- 
  
  
  
Ce 
mum Je ie ZI so tug se 104 Se [777777700089 mmm 
| | | I 
| | I I 
‘ ~ — t = = = ^ — ~ — 
F : r ! 
| | ; 
| : Pm VA - c s. 
* . . 
| \ I 
| ~ - 1 = ‘ -— — N V 
} Pd | 
| | | 
4 - . I | i x | 
| food | 
Da À le oe ig wom rm nt 
Fig. 1. Suppose an area 5 X 5 km taken from a district mapped by two different 
methods and suppose further that check surveys at 36 evenly distributed points 
in both cases show a standard co-ordinate error of *= 10 m. The map to the left 
has only random errors varying in size according to the Gaussian frequency law 
as shown with vectors. The map to the right has random errors equal to one half 
of those of the left map but in addition a systematic error /\ y — + 8 m. The 
maximum co-ordinate error becomes 30 m for the left and 16 m for the right 
map; distances of 1.000 m have a standard and maximum error 11 m and 30 m 
for the left and + 5 m and 15 for the right map. This illustrates the importance 
of knowing the character of the errors. 
methods, such coordinate-errors do not give the best information for 
the majority of map users. The block-methods — two-or three-dimen- 
sional — are characterized by correlation which, if the fixed geodetic 
controls are widely separated, causes regional displacements without 
discrepancies. For most map users such regional errors, of course of 
limited magnitude, are of no or limited disadvantage; it is the local er- 
ror within his region of interest (e. g. 5 X 5 km) he needs to know. 
In fig. 1 this is illustrated. 
The general discussion above has been included in order to put the 
reader in a zero-position before the following concentrate of my study 
1s presented.*) The two-dimensional radial triangulation methods, how- 
ever, are not introduced here as a substitute for the three-dimensional 
stereotriangulation methods, but their advantages — and disadvantages 
— are to be earnestly considered without a step-by-step comparison 
with those methods more general accepted for precision work. 
The very limited research work in and instrument design for radial 
triangulation during the last ten or fifteen years could be incorrectly 
understood as this group of methods be of limited value. There are, 
however, other reasons for this lack of interest for radial triangulation 
  
*) P. O. Fagerholm: A study of Mechanical Radial Triangulation and some 
Related Problems in Modern Mapping. Stockholm 1952. 
  
  
  
  
  
  
  
  
  
i en