Faculty of Mathematics and Natural Sciences - Earth Observation Lab

Write your thesis in the Earth Observation Lab

Scene: Beskid Mountains, Landsat satellite images of three point in time (2003, 2010, 2015) ,

Scene: Beskid Mountains, from left to right: Landsat-5 2003, Landsat-5 2010 and Landsat-8 2015 satellite images. Fig.: Image courtesy of the U.S. Geological Survey


You specialized or plan to specialize in remote sensing and are now seeking answers on how to write your Bachelor's or Master's thesis in the Earth Observation Lab?

You will find the answers to your questions such as

  • Is the Earth Observation Lab the right place for me and my thesis?
  • How to find a topic for a thesis?

and the entire workflow on how to write a thesis in our guidelines:

"How to write a thesis in the Earth Observation Lab"


Below you will find several good practice examples.


Theses carried out at the Earth Observation Lab

Type Title Abstract Experts
M.Sc. thesis Simulating maize yields in Brandenburg with the Scalable Crop Yield Mapper In this Master thesis the Scalable Crop Yield Mapper (SCYM) was used to simulate 2017 silage maize yield at pixel and field scale in Brandengburg, Germany. The results of the study helped to understand the feasability of the SCYM and offered a good grounding for further research on the subject. Patrick Hostert, Daniel Müller
B.Sc. thesis Using Landsat time series for mapping winter and spring crops in Southeastern Anatolia In this Bachelor's thesis, 140 Landsat images were pre-processed in Google Earth Engine to generate spectral-temporal metrics for mapping winter and spring crops in Southeastern Anatolia in 2015. Patrick Hostert, Dirk Pflugmacher, advisor: Philippe Rufin
M.Sc. thesis Characterizing Spring Phenology of Broadleaf Forests Across germany Combining Landsat/Sentinel Time Series and Phenological Models In this thesis, combined Landsat and Sentinel-2 time series have been used to estimate spring phenology across Germany. Katja compared the results to ground observations and mechanistic phenology models. Thereby the study showed that dense time series can help us to better understand vegetation phenology across large spatial scales. Dirk Pflugmacher, Cornelius Senf
B.Sc. thesis Potential of a multi seasonal spectral mixture analysis using Landsat imagery for detecting urbanization patterns in Ouagadougou, Burkina Faso This Bachelor's Thesis evaluates a method of a multi seasonal spectral mixture analysis on Landsat imagery for generating sub-pixel information. This method is particularly adequate in regions featuring dry and rain seasons. Franz applies it to detect gradual urbanization processes in Ouagadougou, Burkina Faso. Sebastian van der Linden, Jonas Ø. Nielsen


Seminar papers carried out at the Earth Observation Lab

Type Title Abstract Supervisor(s)
seminar paper Detecting permafrost landscape changes in northern Alaska and the Mackenzie River Delta with Landsat Data This seminar paper used the Landsat archive to identify change processes in the North American Arctic coast employing a trend analysis process based on research by Nitze and Grosse (2016). In addition, the extent of these processes were estimated based on the trend data and a general change classification was undertaken. Philippe Rufin, Patrick Hostert
seminar paper Forest Structure and Land Use Change in the Bieszczady Mountains This seminar paper classifies the development of forest populations in the border triangle of Poland, Slovakia and Ukraine since 1860. Using field measurements, species compositions and forest structure were analysed in eight forest patches complemented by historic maps. Tobias Kümmerle, Jan Knorn, Patrick Culbert
seminar paper Räumliche Interpolation des
durchschnittlichen Jahresniederschlages im
This work discusses approaches for precipitation modelling in the Harz mountain range using inversed distance weighting interpolation. It gives a very structured insight in applied modelling in natural sciences including a detailed reflection of the work flow. It combines different data sets and adapts its method to regional particularities. Sebastian van der Linden, Reinhard Kleßen, Akpona Okujeni
seminar paper Mapping impervious urban surfaces using regression analysis on synthetic hyperspectral EnMAP data This seminar project works on an algorithm for mapping urban imperviousness based on multiple regression analysis using R. It uses simulated data of the future hyperspectral EnMAP satellite trying to model sub-pixel information by synthetically mixed training data of reference surfaces. Akpona Okujeni, Cornelius Senf
seminar paper Grassland intensity in the Polish and Slovakian Carpathians - A methodological classification approach This project study develops a classification method for highly heterogeneous grassland surfaces in the Polish and Slovakian Carpathians using field and remote sensing approaches. It makes use of sophisticated methods of remote sensing and spatial statistics without neglecting political questions of land reform and agrarian policy. Patrick Hostert, Jan Knorn
seminar paper Qualitative Veränderung der Baumbestände in der Jungfernheide (Brandenburg, Deutschland) über die Jahre 1777, 1857 und 2013 This seminar paper reconstructs historic land use patterns in the state of Brandenburg. It uses elaborate methods of geo referencing and analyses changes in tree populations since 1777. Christoph implements historic geodata into GIS and finds out that regional populations of beeches and pines decreased considerably within the last 250 years. Jan Knorn, Richard Jäger