Titelaufnahme

Titel
Spatial and statistical analysis of thermal satellite imagery for extraction of coal fire related anomalies / Jianzhong Zhang
VerfasserZhang, Jianzhong
Begutachter / BegutachterinVan Dijk, Paul ; Wagner, Wolfgang
Erschienen2004
UmfangXVI, 166 S. : Ill., graph. Darst.
HochschulschriftWien, Techn. Univ., Diss., 2004
SpracheEnglisch
Bibl. ReferenzOeBB
DokumenttypDissertation
URNurn:nbn:at:at-ubtuw:1-12531 Persistent Identifier (URN)
Zugriffsbeschränkung
 Das Werk ist frei verfügbar
Dateien
Spatial and statistical analysis of thermal satellite imagery for extraction of coal fire related anomalies [6.82 mb]
Links
Nachweis
Klassifikation
Zusammenfassung (Deutsch)

In der Arbeit wird eine Methode vorgestellt, um Kohlefeuer aus thermischen Anomalien in Landsat-7 Kanal 6 Daten zu bestimmen.

Zusammenfassung (Englisch)

Coal fires not only cause losses of natural resources, but also cause severe environmental problems. The study undertaken in this thesis focuses on the development of a practical approach for the extraction of coal fire related thermal anomalies at the land surface over large areas using remote sensing data sets. An overview of the theory and case studies on detecting coal fires using remote sensing techniques is given in a literature review. From field measurements, it is known that the temperature variance caused by uneven solar heating can mask thermal anomalies related to coal fires. Predawn is the optimum time for coal fire detection by using thermal remote sensing techniques. Through analyses of the statistical characteristics of the thermal anomalies in different scenes of night-time Landsat-7 ETM+ band 6 images, it is shown that the mean and the standard deviation values of a coal fire related thermal anomaly on ETM+ images are higher than its background. A practical approach for the extraction of coal fire related thermal anomalies in a large area using Landsat-7 band 6 data was developed in this study. Using the thermal anomalies automatically extracted from Landsat-7 ETM+ thermal data based on this algorithm, plus the land cover information derived from the multi-spectral non-thermal bands, the areas with a high potential for coal fire occurrence can be defined.