5. CONCLUSION AND OUTLOOK 
The results of the different analysis methods allow the 
following conclusions: 
- Transformation and calibration from NDVI sealing degree 
value is a less costly method to produce sealing maps and 
deliver sufficiently good results. 
- The evaluation of soil sealing causes some unexpected 
problems using the temperature characteristics of urban 
surface materials. In general, sealed areas are 
characterized by a great heating at day time, a strong 
cooling at night. Surface temperatures are highly 
dependent on material. Partly temperature values of sealed 
areas are similar to values of trees, whose tops show 
unexpected high values. 
It has not yet been reached to separate sealed and 
unsealed areas completely by transforming data in an 
index, though separation of traffic areas is possible. 
- The land-use classification does not deliver a quantitative 
registration of sealing degree values but differentiates 
surface cover types. The most important types such as 
grass, meadow, trees/shrubs and water can be classified 
exactly. More difficulties appear when classifying buildings 
because of their immense variation in roof materials. In this 
case it is necessary to classify each roof material which is 
costly, and, finally, to merge it into a "building class". Partly 
a spectral overlap of the classes "traffic area" and "building" 
is noticeable. 
- Incorrectness of rectification cause severe problems in the 
multitemporal classification. Not at last for this reason it 
would be desirable for future projects to rectify using the 
exact positions of flight track. 
The different sealing maps shall be controlled by means of 
reference material. Therefore it is planned to overlay a 
building layer obtained by a topographical map (1 : 10 000) 
and to determine the classification error for the "building" 
class. Furthermore, the produced maps need to be examined 
with sealing degree values checked in the terrestrial test 
sites. 
Literature: 
Amann, V., 1986. Multipektraler Linescanner Daedalus AADS 
1268 (ATM), Leitfaden für den Datennutzer, Deutsche 
Forschungsanstalt für Luft- und Raumfahrt (DLR), 
Oberpfaffenhofen. 
Baumgartner, A., Mayer, H. Bründl, W. Noack, E.M., 1985. 
Stadtklima Bayern - AbschluBbericht zum Teilprogramm 
"Thermalkartierungen", München. 
Daedalus International Newsletter, Daedalus Enterprises, 
1995. Special Compendium, Airborne Multispectral 
Applications...Worldwide, Ann Arbor, USA. 
GoBmann, H., 1982. Grundlegende Probleme der 
Thermalbildauswertung über Landflächen. 
Akademie für Raumforschung und Landesplanung, Band 62 
Thermalbilder für die Stadt- und Landesplanung, Hannover. 
492 
Beiträge der ' 
Heber, B. Lehmann, |, 1993. Stadtstrukturelle 
Orientierungswerte für die Bodenversiegelung in 
Wohngebieten. IOR - Schriften 05, Institut für Ökologische 
Raumentwicklung e.V., Dresden. 
Jansa, J., 1983. Rektifizierung von Multispektral 
Scanneraufnahmen, Geowissenschaftliche Mitteilungen 
(Heft 24), TU Wien. 
Kaufmann, V., 1989. Thermalkartierung Graz 1986, Graz 
Kraus, K., Schneider, W., 1990. Fernerkundung Band 1 
und 2, Bonn. 
Meinel, G.; Netzband, M., 1996. Erarbeitung von 
Ubersichtskarten zur Versiegelung, Forschungsbericht, 
Dresden. 
Sukopp, H., Wittig, R., 1993. Stadtókologie, Stuttgart. 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B7. Vienna 1996