in 9th Inter- 
nment, Ann Arbor, 
s of Multi- 
'erence of the 
Aerial 
> Sensing for 
he Endangered 
> of Germany, 
STATISTICAL LAND USE DATA BASED ON THE INTERPRETATION OF 
AERIAL PHOTOGRAPHS AND SAMPLING METHODS 
Heinz Trachsler 
Institute for National, Regional an Local Planning, 
Swiss Federal Institute of Technology Zurich 
1. Introduction 
Detailed information about the existing spatial organization 
is an important postulate at the beginning of every planning 
process. As land use is a significant part of this spatial 
organization the planner in-charge also has to collect and 
process data of the actual land use. This will be an important 
basis for securing a harmonious utilization of the land in 
the future (1l). 
The main sources for land use data are statistics and thematic 
maps. Switzerland, at the moment does not dispose of any large 
scale land use maos, the only land use map available has a 
scale of 1:500'000 (2) and is therefore not suited for 
planning purposes on the local and regional level. Concer- 
ning statistical data the situation is also not satisfactory. 
The official land use statistic (3) which is published on the 
community level is derived from the cadastral survey. This 
survey however is not yet accomplished for the whole country 
and also has not been updated regularly in the past years. 
For this reason the statistical data for many communities are 
not up to date anymore and they are not referring to the same 
period of time. The land use statistic therefore is only of 
limited use for planning purposes, particularly no reliable 
figures on changes in land use can be obtained. Due to this 
reasons other possibilities for obtaining data on a nationwide 
land use inventory are investigated. It is obvious, that the 
methods considered will have to be faster and less expensive 
than a modification and acceleration of the mentioned 
cadastral survey. Collecting the land use information by 
visual interpretation of aerial photographs and determining 
the surfaces of the various land use categories by sampling 
methods turned out to be a realistic approach. 
The first part of this paper deals with collecting land use 
data by sampling methods. Afterwards different possibilities 
for sampling with aerial photographs are presented and finally 
test studies with three different methods for data acquisition 
are discussed.