3
All three of the field workers found that a scale of 1:15.000 normal angle was the
easiest to mark. Two of them also preferred a scale of 1:15.000 wide angle. Notes
about the time used do not give any basis on which to draw conclusions about the
relationship between the different scales. The field work for each series took about
two weeks. Each worker marked about 5 boundary points in the hour. This in
cluded marking and moving between the points.
The series where signals were used (the control series), were marked in the field
at the end, as a fifth marking of the same boundaries — by the same workers in
the field. Where the signal could not be seen, a mark was made on some clear
detail, and the boundary point measured from it, and a sketch made.
The accuracy of the field control.
Accuracy of the plotting of the signal points has been more closely examined by
photogrammetric and terrestrial measurements of visible signals.
For 23 visible signals in stereo-models, from 3-5 different series, 96 definitions were
made. The mean error for the definition of a visible signal was worked out from
the deviation of each signal’s mean value. This gave a mean point error:
m p = + 0.45 m
Numerical and graphical coordinate definition has been examined for 50 points on
a map in the scale 1:5.000, and gave a mean point error:
m p = 0.31 ± 0.24
Finally, the ground plan error has been examined because of mistakes in the
elevation setting of measuring points in the machine. This can be due to the
boundary point being marked on the top of a tree, or because the plotter cannot
decide if the prick is on a tree top or on the ground. This man error varied from
0.5 m to + 1.5 m, respectively at points 2 cm and 9.5 cm from the photo’s base.
The conclusion may therefore be drawn that the coordination value of the points
is sufficiently accurate for control of the method, where boundary points are not
signalled.
Plotting.
Both the control series and each experimental series was plotted by both com
putations and graphically. This turned out to be of great help, when preparing
the material and when classifying mistakes.
The plotting was done with a Wild Autograf A8 by a plotter with another worker
writing the coordinates and marking points on the map. Electrical recording work
would have saved much time and eliminated errors in copying.
In the plotting of boundary points where there are no visible signals, the prick
marks made on the photograph in the field have to be transferred to the model in