but only those with a crown diameter above a certain minimum. The 
minimum level for crown diameters to be counted was fixed at 5 metres 
for the photo interpretation test of Freiburg; asymmetric crowns were 
measured across. In photos of a scale of 1:10,000, taken in 1971, this 
corresponds 0,5 mm; in photos 1:5,000 of 1972 it was 1 mm. In addition, 
some trees were counted, which have a narrow crown width by nature, if 
they were taller than about 15 metres (some conifer species and pyramid 
shaped poplars). Finally, also very big bushes of more than 5 metres 
height and 5 metres diameter were counted, because they do fulfill 
almost the same functions as trees and, because they cannot be replaced 
like big trees. Crown diameters were measured using a transparent rule 
which was put over the transparent aerial photo on the light-table; in 
cases of doubt a measuring magnifier could be used for an exact measure 
ment. A hand-operated counting-instrument proved to make the counting 
more reliable. In order to estimate the accuracy of the results of the 
counting this work was carried out parallel by two interpreters; no 
remarkable differences could be noticed, except in those few areas, 
where unusually many trees had a diameter of crowns in a range of more 
or less 5 metres. No remarkable error could be found for the next 
larger units, the statistical districts, because all positive and nega 
tive deviations for the blocks were levelled up. The differences of 
scale which occur within a single photo and within a photo flight result 
ing from different elevations of terrain could be neglected for the test 
area. 
The counting of trees did not take very much time; on an average 
it took 5-8 minutes for 1 hectare urban area. A further reduction of 
working time is possible, if in those blocks which contain forest parts 
or parks and a very high number of trees, no accurate counting is 
carried out but only an estimation; as these areas are recognized 
immediately to be no problem areas at all, such a simplification may be 
allowed; at least some parallel calculations of the code index "inhabi 
tants per tree" using for comparison both counted numbers (80 - 250) 
and an average estimation of 100 trees for each unit did not shift the 
resulting index very much. All small trees with a crown width of less 
than 5 metres were neglected with the tree counting, but were surveyed 
as a part of the "green area". 
ad b) An accurate ground-survey of all vegetation within a city is principally 
possible, but could never be carried out, because it would be extremely 
time consuming and expensive. Using photogrammetric methods and stereo 
plotters mapping and measuring of areas could be done very precisely, 
but would cause fairly high expenses, too. The evaluation of the aerial 
photos of Freiburg was done without any sophisticated instrumentation 
and without the need of skilled interpreters, but just under a magnifier 
A transparent dot-grid and the already mentioned counting-instrument 
were the only used tools. As a first step all dots representing the 
area to be evaluated, were counted. As the second step only those dots 
were counted, which happened to represent any form of green vegetation. 
As a result, the second figure divided by the first gives a good 
approximation for the portion of green area within the interpretation