grass areas with high and low content of plants and
shrubbery rich and bushy rich areas were classified.
The verification of the results with the help of contin-
gency tests has shown an amazing high accuracy of classi-
fication (90 %). Such a verification on the basis of the
actual training sites does not guarantee an independent
evaluation for the accuracy (Will, 1996). Only after hav-
ing verified additional mapped areas details are available.
A visual verification has shown a good correspondence
but it must be quantified in the next step. That is why
more verifying areas are fixed at the moment to enable a
more objective evaluation of the accuracy of classifica-
tion.
5. CONCLUSION
The high potential of airborne scanner data for the moni-
toring of dominating stocks of relatively small-structured
units of vegetation could be demonstrated by the classifi-
cation of vegetation stocks in the testing area. This way of
exhaustive monitoring of sand-dry lawn stocks according
to their units of vegetation will lead to an enormous in-
formation gathering. The level of spectral and spatial high
resolution airborne scanner data cannot be used exhaus-
tively because it requires too much money and effort. But
it should be available for specific highly ecological rele-
vant questions. Satellite data and generalised classes
enable the extensive area observation of ,,mining succes-
sive landscapes".
It was possible to determine a set of spectral endmember
(various Tertiary and Quaternary sediments, Lignite) for
the particular areas as an interim result of the evaluation
of spectral data. Further research will investigate the
existence of mixed units of vegetation by spectral mixing
procedures (Glaesser et al. 1997).
The reached statistically reliable results of the classifica-
tion are an excellent database for the monitoring of this
extremely damaged area. In combination of digital terrain
models, multispectral and multisensoral remote sensing
methods and GIS-data processing an integrated tool for
monitoring is available.
ACKNOWLEDGEMENT
The research has been funded by the German Federal
Ministry of Research and Technology and the Ministry of
Environment and Agriculture of Federal State of Saxony-
Anhalt. The CASI flight has been sponsored by the
ITRES company, CANADA. We could also use observa-
tional areas of the Geobotanical Institute of the Martin
Luther University and those of the Faculty of Agricul-
tural, Ecotrophology and Regional Planning of the Tech-
nical College Anhalt as training sites.
REFERENCE
Glaesser, C., Birger, J., Herrmann, B. & R. Lippold 1997.
Assessment and Monitoring of Lignite Opencast Mining
in Eastern Germany. Proceedings of the third Internatio-
nal Airborne Remote sensing Conference and Exhibition,
Kopenhagen, p 772-779.
Mahn, E.-G., Schmiedeknecht, A. & S.Tischew 1995.
Modelluntersuchungen zur Renaturierung von Bergbau-
folgelandschaften auf der Basis spontaner und gelenkter
Sukzessionen am Beispiel des Braunkohlentagebaus
Goitsche. AbschluBbericht, FA-Nr. MLU 1170A/0023,
Halle.
Will, H. 1996. Fernerkundung und Weideókologie in der
subhumiden Savannenzone NW-Benins. In: Karlsruher
Schriften zur Geographie und Geoókologie, Bd.4,
Karlsruhe.
74 International Archives of Photogrammetry and Remote Sensing. Vol. XXXII, Part 7, Budapest, 1998
