13
for this two Wild Radial Triangulators RT. For the whole method of computing from
radial triangulation data up to the adjustment of the block coordinates, KLM Aerocarto
N.V., in cooperation with IBM, developed an electronic computing programme in four
stages. IBM used for these computations their machines IBM 1401 and IBM 7090.
Area A as well as area B have been divided into a number of blocks bounded along the
outlines by the minimum number of 10-12 terrestrial points, necessary for the adjustment
of each block of about 150 sections which consisted of either 1, 2 or several models.
At the moment the adjustment of the analytical aerial triangulation of area A has been
finished and the mean square errors reached in the blocks are as follows:
Scale of photography 1 : 51,000 RC9
Scale of mapping 1 : 25,000 Aii = 5m
minimum
m xy = abs. 49 ¡xm
rel. 27 gm
m-xy
maximum
abs. 94 gm
rel. 60 ¡am
The adjustment of about three quarters of the blocks, carried out by radial triangulation
are completed:
The mean square errors in the handled blocks are:
Scale of photography 1 : 60,000 RC9
Scale of mosaics 1 : 10,000
minimum
m xy = abs. 31 ¡am
rel. 28 ¡am
maximum
m xy = abs. 57 ¡am
rel. 58 ¡am
The block adjustment of the planimetric coordinates for the mapping orders (scale 1 : 1,000)
in the Netherlands after August 1962 also has been executed by the new developed two
phases computing programme for computing machine IBM 7090.
At handling a scale of photography 1 : 5,000 Zeiss Vertical 18/18 cm, the mean square
errors reached were as follows:
minimum
m xy = abs. 24 ¡am
rel. 20 ¡am
maximum
m xy = abs. 52 ¡am
rel. 52 ¡am
Regarding precision of radial triangulation of super wide angle photography, taken with
Aeroflex stabilized mount a study of closing errors made available by KLM Aerocarto to
Prof. Roelofs was analyzed statistically. One of the staff members of the Geodetic Institute,
Mr. Timmerman, carried out a study about “The influence of instruments adjusting errors
in numerical radial triangulation”. Although the closing errors of the quadrangles showed
a normal distribution pattern, there was an average shift of the point of symmetry of
0.00024 (log values). This shift can be due to small adjustment errors. Timmerman found
a small value of 6 microns for the eccentricity of the arms of the floating marks. The
standard value of the closing errors was 0.00033 (log values). It could be derived that this
closing error corresponds with the following small sources of errors:
Monocular observation on pricked points 6 microns
Zero setting of x-parallax 4 microns
Irregular film deformation 5 microns
Variation in eccentricity of the floating mark arms 2 microns
Mean square error in the observation of a direction 20 dmgr.