IAPPING 
lel 
and nature 
n center in 
method for 
hotographs 
GIS model 
teen years 
rly suitable 
Croatia, the 
making the 
he town of 
wo regions 
the biggest 
vironments. 
should be 
3 for the 
gy, namely 
or uniform 
the remote 
o the urban 
continuous” 
m, on the 
concerning 
ote sensing 
rried out in 
objective of 
eliable and 
ages under 
e mapping 
lity, which, 
able. 
CIR aerial 
ken black- 
indsat TM 
here is not 
re, for the 
e  biotope 
any by the 
Task Force for biotope mapping in populated areas” have been 
used. 
The biotope mapping process includes 4 basic operation 
. In the first phase, the data on the investigated area, the 
available literature, the maps (BM 1:5000), the existing black- 
white and CIR aerial photographs, as well as Landsat TM 
imagery have been collected. In the second phase, the biotope 
identification key was verified in the field to determine the 
criteria for biotope identification on the aerial photographs, and 
the photointerpretation key has been constructed. In the third 
phase, the biotopes were separated on the scanned and 
georeferenced aerial photographs using the key for the 
"Biotope mapping in pupulated areas" (Schulte et al., 1993). In 
the fourth phase, after the vectorization in the ARC/INFO ver. 
7.0.3., all existing data necessary for the biotope mapping have 
been entered into the database, and the biotope map has been 
made. 
All satellite imagery processing such as necessary for this work 
has been made using the ER Mapper 5.0 program package 
installed on the DEC Alpha AXP 2000 working station. For 
research purposes, a part of the scene on the Landsat TM 
imagery taken in 1995 has been used. The scene was 
georeferenced into the Gauss - Kruger projection with a pixel 
Table 1. Key for the "Biotope mapping in pupulated areas” 
size of 30 x 30 m. By means of the vizual interpretation on the 
satellite imagery, "training" areas for 10 main biotope types 
have been selected. The imagery classification was made using 
the maximum similarity method. It has been found out that on 
the Landsat TM imagery 5 biotope types could be identified in 
the city of Zagreb. 
After the accuracy of this classification has been appraised, 
that was 85%, the interpreted imagery has been transformed 
into a vector theme through the ARC/INFO grid format and the 
biotope map has been made. The photographs were integrated 
into the GIS model on the basis of the ODK in a scale 1:5000. 
The GIS is a basic prerequisite for modern planning, recording 
and management of natural and built resources. This 
requirement is particularly pronounced in urban environment, 
where systems are very layered and branched. 
4. RESULTS 
Based on the AP's photointerpretation, and according to the 
used methodology, the researched area of the city of Zagreb has 
been classified into 8 main biotope types, subtypes and 
categories such as shown in table 1. 
  
Type Subtype Category 
  
2.1.1 Closed blocks 
2.1.2 Open blocks 
2.Mixed housing area biotope 
2.1 Blocks of houses 
2.2 Blocks of edge houses 
2.3 Terraced houses 
2.4 High multi-storey houses 
2.5 Row of detached houses 
2.7 Public buildings with free 
surfaces 
2.2.1 Closed blocks 
2.2.3 Open blocks 
2.7.1 Public buildings (schools, 
offices) 
2.7.2 Public buildings with large 
green surfaces (intensive 
cultivation) 
2.7.3 Public buildings with large 
green surfaces (extensive 
cultivation) 
  
4. Industry and handicrafts area 
biotope 
4.1 Industrial areas 
4.2 Handicrafts areas 
4.1.1 Very sealed (asphalt) 
4.1.2 Slightly sealed 
4.2.3 Slightly sealed 
  
5. Green area biotope 
5.1 Public green areas and parks 
5.2. Sport and recreation positions 
5.1.1 Intensively cultivated 
5.1.2 Extensively cultivated 
5.2.1 Bathing place (pools) 
5.2.2 Playgrounds (stadium) 
5.2.5 Tennis courts 
  
  
  
6. Water biotope 6.8 Lakes - 
7. Road biotope 7.2 Roads 7.2.4 Large parking places 
8. Agricultural land biotope 8.2 Green areas - under grass 8.2.11 Lawns 
  
9. Forest, schrub biotope 
9.1 Deciduous trees 
  
12. Uncultivated (unused) land 
biotope 
  
  
12.1 Uncultivated lands without 
vegetation 
12.2 Young uncultivated lands (with 
predominant one - two year old 
pecies) 
12.4 Old uncultivated lands (with 
predominant shrub 
communities). 
  
  
  
According to the classified biotope types, the digital maps 
for the researched area of the city of Zagreb for the year 
Intemational Archives of Photogrammetry and 
Remote Sensing. Vol. XXXII, Part 7, Budapest, 1998 
1982 (Fig. 1) and for the year 1997 (Fig. 2) as well as the 
map of changes have been made (Fig. 3). By interpretation 
241