99, 
lly, 
on, 
ich 
ler. 
ind 
nar 
od 
atic 
Iso 
ion 
nal 
rea 
on 
ith 
an 
n- 
he 
Banzhaf, Ellen 
  
other one taken is located in the old “Bundesländer”. The first mentioned region is the one between Halle and Leipzig, 
and is example-giving to be investigated for the following reasons: 
+ it underlies high dynamics with respect to land-use development, 
+ it shows a strong growth of peri-urban settlements, 
* itis regarded as a region formerly polluted by industrial plants and mining industries, and presently as a region put 
pressure on by traffic, commercial sites, new settlements, 
+ the configuration of its nature space potential presently shows a deficit in comparison to other regions, and it offers 
relatively little recovery potential, 
+ with respect to heterogenity it has suffered of a high loss over centuries compared to the former grown culture 
landscape. 
The region between the City of Halle and the City of Leipzig can be described as a suburban area intensively used for 
business parks, new settlement quarters, and recreational spaces. Apart from its historical meaning as an agricultural 
region and the ongoing of agricultural land use this area contains a multiple use that underlie competitive economic and 
ecological constraints. Since the reunification of east and west Germany took place in 1990 this suburban region has 
been suffering a lot from a lack of regional planning, especially during the first years, and is now about to get 
restrictions for further landscape fragmentation by planning institutions. A major problem for planning authorities is 
that Halle belongs to Saxony-Anhalt and Leipzig is part of Saxony, so in-between this multipurposely used region the 
border of two different ,,Lánder" (federal states) cuts through the responsibilities of each particular planning authority. 
Thus regional planning does not work on the whole area but stops at the administrative border. This is where analysed 
satellite data offer a very useful instrument to show both ,,Länder“ the necessity of concerted actions in order to 
recognize the state in which this region is in and to be able to regulate the process. Furthermore, this highly dynamic 
region with competitive land use types offers a splendid possibility to test the suitability of landscape structure 
indicators for landscape monitoring. 
Based on this region, a method will be developed and will be checked for its transference to other example-giving 
regions. In particular, structural developments of the peri-urban cultural space between Leipzig and Halle are pointed 
out. Parallels and deviations of such typical structures will be derived in a similar cultural landscape. 
4.2 Remote Sensing Data 
In this first project phase the following satellite image data have been analysed. 
Tab. 1 List of Utilized Satellite Images 
  
  
  
  
  
  
  
  
  
Date of Acquisition Sensor Path / Row Total RMS Error 
07.07.1989 Landsat TM - 5 193/24 0.3 
13.09.1999 Landsat TM - 7 193/24 0.3 
08.08.1998 IRS - 1C / LISS-III 27732 0.4 
21.06.1998 SPOT-XS 058 / 246 0.3 
  
  
5 FIRST RESULTS 
5.1 Change Detection Classification Based on Landsat TM-5 and TM-7 
The area between the two conurbations is classified for the years 1989 and 1999 in order to differentiate the 
development of human interventions right after the political change has taken place. As IRS data have only been 
available for Germany since 1996 and as the two classifications need to be compared as precisely as possible both dates 
are classified on the basis of Landsat-TM data. Several preprocessing steps are necessary on the images before the 
classification could take place. The older image is taken from Landsat-5-TM (75 July 1989) and contains bad lines that 
need to be eliminated and veils of clouds that are diminished applying the histogram minimum method. The latter image 
could be gained from Landsat-7-TM, dated on the 4 September 1999, and does not contain any disturbances. 
Atmospheric conditions of the two images are corrected by means of the histogram minimum method. In this 
preprocessing step histograms of all TM bands are computed for the full image, which generally contains some areas of 
low reflectance (e.g. clear water, deep shadows or exposures of dark basalt). These pixels will have values very close to 
zero in the short wave infrared band ( TM band 4). If the histograms of TM bands 1 to 3 are plotted they will generally 
be seen to be offset progressively towards the higher grey levels. The lowest pixel values in the histograms of these 
  
International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B7. Amsterdam 2000. 121