UNIGIS Abschlussarbeiten

Der krönende Abschluss eines UNIGIS MSc Studiums ist sicherlich die Master Thesis. Mit ihr belegen unsere MSc-AbsolventInnen, dass sie den akademischen Grad "Master of Science (Geographical Information Science & Systems)" zu Recht führen.  Im UNIGIS professional Studiengang muss keine Abschlussarbeit verfasst werden. Dennoch nehmen einige Studierende die Möglichkeit war, ein Geoinformatikprojekt durchzuführen und entsprechend zu dokumentieren.

Sie sind auf der Suche nach aktueller Literatur zu Geoinformatik-Themen?
Hier finden sie die mitunter preisgekrönten Abschlussarbeiten unserer AbsolventInnen!

Simon Baier [02-2015]:

Brandflächenklassifizierung in Riau, Sumatra, unter Verwendung mittelauflösender Fernerkundungsdaten

Diese Arbeit ist online verfügbar: Download

Fire menace has always been a matter in Sumatra during the annual dry periods. Though, over the past two decades these fires accumulated. The large-scale fires have a global influence on nature, human health, economy and climate. In order to be able to face future fire disasters, an accurate knowledge of position and size of fires is important. Due to large dimensions and the fact that field data can be determined only with considerable effort the use of remote sensing to solve this problem is obvious. The underlying thesis examines whether medium resolution satellite data from Landsat 7 ETM+ and Landsat 8 OLI are adequate applications for qualitative as well as quantitative detection of burnt areas. Two pixel-based and one object-oriented methods are compared among each other, observing their classification accuracy and practicability. Data from the province Riau is used in order to illustrate findings. Firstly, an object orientated method is used in order to map land cover, as this is important for estimating emissions. Subsequently three different approaches are taken to reach successful burnscar mapping. The pixel-based methods are expected to yield appropriate outcomes whilst kept easy, whereas the object-oriented classification is more sophisticated. Results show that the land use classification is completed with an overall accuracy of 91.2 %. Using these methods for burnt area mapping overall accuracies between 67.3 % and 85.6 % exist. It is found that the object-oriented approach provides the most accurate results. Comparing outcomes with MODIS and NPP satellite hotspot data shows large congruence in detected fires. Estimated burn emissions for the fire season in 2013 are located between 8806.8 kt and 10122.8 kt of carbon, depending on the method which is used. However, the object oriented method is not able to compete with the pixel-based methods in the processing speed of the analyst. Especially when using large scale data, the pixel-based methods are proved superior. Concluding, Landsat 7 ETM+ and Landsat 8 OLI data are found suitable for detecting burnt areas in tropical regions, as demonstrated for the province of Riau.

Zurck weitere Arbeiten ...