»Topometer-Polar* program (P. Vetterli, Fribourg) revealed inpermissible discrepancies.
It was therefore necessary to repeat the survey with a Geodimeter on a single, local
network in which the new traverse points were computed and adjusted. The mean
square error of the coordinates of the passpoints can be taken as + 2 cm, correspon-
ding to the required accuracy of the track points. The passpoints were levelled with
an accuracy of approximately + 1 cm.
5. Plotting
Shortly after the flight glass diapositives, without electronic dodging, were prepared
from the negatives of the RC 5a. Plotting was done at a model scale of 1:1000 and
a Z of approximately 428 mm in the Wild A 7 Autograph of Dr. Schlund, in Zurich,
equipped with special measuring marks of 0,02 mm diameter. Every point was mea-
sured twice, independently, and recorded on the Remington Card Punch connected to
the Wild EK 5 Electric Coordinate Printer. The mean square difference between the
two sets of model coordinates was + 0,005 mm in X and Y, corresponding approxi-
mately to 1,8 um in the negative. The model transformation was performed on the
means of the two sets of coordinates. Of approximately 1700 signalized points 88 %
could be measured and plotted. The remaining 12 % were covered by goods waggons
and were measured in a terrestrial survey.
Simultaneously with the coordinate registration the points were plotted at a scale of
1:500 (2 x model scale) and the remaining details were then plotted. (See Fig. 5.)
The orientation of the models gave mean square residual errors of approximately
0,1 mm in position in the plan (corresponding to 5 cm in the terrain) and approxi-
mately + 2 cm in height.
6. Coordinate transformation
The transformation of the model coordinates into terrestrial coordinates and the
adjustyment of the coordinates of the passpoints onto the terrain coordinates were
performed model by model with the FORTRAN IV program ,,Aero in a large com-
puter UNIVAC 1107. The theory of this program is fully described in the Doctoral
Thesis of Dr. Schlund to ETH, Zurich, entitled: ,,Adjustment of Spatial Aerial
Triangulation with the help of an Electronic Automatic Computer“. A further pro-
gram was used to mean the two coordinates of points observed in two neighbouring
models and to produce a definitive coordinate list.