unit. To make the dot grid capable of being seen at a glance and to 
enable easy counting, the dots on the grid had to be arranged in 
straight lines, though this arrangement is likely to cause systematic 
errors of representation in urban areas. The dots were arranged in a 
triangular instead of a rectangular pattern (distance of dots was 3 mm 
for photos 1:5,000 and 1,5 mm for 1:10,000), in order to avoid major 
bias which might result from correspondence of the rectangular dot grid 
pattern and the occasionally rectangular patterns of urban settlements. 
Furthermore, the position of the dot grid over the aerial photograph 
was chosen randomly to avoid frequent parallelity of dot-lines and 
streets. This part of the interpretation was carried out at least twice, 
each time working with a ‘different position of the interpretation-grid 
over the photo. The required working time for the described procedure 
was about 15 minutes for 1 hectare on an average. 
There are two possibilities for the interpreter to control his 
own results: first the described procedure itself, because it includes 
a twofold interpretation; second the chance to detect major counting 
errors or the interpretation of units within wrong boundary lines, by 
comparison of the known area of the statistical block with the approxi 
mate size of the area which can be derived from the number of dots 
representing the unit (each dot represents an area of almost 1 ha). 
Any major difference between the known size of the area and the estima 
tion must not occur — or something is wrong. 
The "portion of green" which is obtained by the described method 
includes all forms of vegetation: parks, turf areas, gardens, bushes, 
some agricultural areas, wasteland and also tree crowns, which partially 
were already counted for the tree-census. The big trees are thus repre 
sented twice. This proceeding, however, does not only simplify and 
make easy the interpretation work, so that it can be done by everybody 
without much training, but can also be based on and explained by the 
following considerations: though it is difficult and not always pos 
sible 'at all to find out from photo interpretation whether there is any 
vegetation under a tree crown or canopy or not, the existence of big 
trees does at least prove, that there are soil conditions in the par 
ticular area which would allow growth of any vegetation. If there is 
actually no additional vegetation under tree crowns, either due to 
shadow or due to human influences — in both possible cases it may not 
be wrong to assume that the tree's ecological functions do include the 
potential functions of any additional vegetation. 
Principally it is possible to interpret more detailed information 
about urban vegetation from the aerial photographs — for example, to 
split up the expression "total green area" into several classification 
groups, even if the same method of interpretation is applied. The 
diligence and ability of a human interpreter, however, is already 
stressed very much by this described simple interpretation task, if no 
additional instruments are used. Increasing demand for information 
from an interpretation means also: less accuracy, more errors or much 
more training and working time. To do the described simple inventory