Page Content
to Navigation
Dr. Michael Förster
- © M. Förster
Senior Scientist
Phone: +49 (0)30 / 314 - 72 79 8
Email: michael.foerster(at)tu-berlin.de
Room: EB 236b
Consultation hour: by arrangement
| Date and place of birth: 1975 (Burgstädt, Saxony, Germany) |
| 2018 | Visiting Scientist at the Joint Research Center (JRC) in Ispra, Italy (Bioeconomy Unit) |
| 2012 | Visiting Scientist at the University Utrecht, Netherlands (Department of Physical Geography) |
| 2010 | Visiting Scientist at the European Academy Bolzano (EURAC), Italy (Institute for Applied Remote Sensing) |
| since 2009 | Post-doctoral Research Fellow Technische Universität Berlin, Institute of Landscape Architecture and Environmental Planning, Department of Geoinformation Processing for Landscape and Environmental Planning |
| 2003-2008 | Research Scientist Technische Universität Berlin, Institute of Landscape Architecture and Environmental Planning, Department of Geoinformation Processing for Landscape and Environmental Planning |
| 2001-2003 | Consultant and GIS-Coordinator Environmental Consulting and Planning Agency - Froelich & Sporbeck, Potsdam, Germany |
| 1999-2001 | Research Associate Geo-Forschungs-Zentrum (GFZ) Potsdam, Section 1.4 (Remote Sensing) |
| 1998-1999 | Exchange Student (ERASMUS) University of Southampton, UK |
| 1996-2003 | Studies of Geoecology Universität Potsdam, Germany |
| 2003 | Diploma, University of Potsdam Grade: 1,1 (on a scale from 1 to 6, where 1 is highest) |
| 2009 | Doctorate, Technische Universität Berlin, summa cum laude |
Research Topics
- Development of methods to analyse the dynamics of ecosystems from time-series (optical and SAR), especially for degradation processes or abrupt damages (e.g. caused by fire or storms)
- Relation of temporal and spectral signals to plant traits and biophysical variables (xantophyll, nitrogen, chlorophyll and fluorescence)
- Derivation of operationalizable and comprehensive environmental indicators that are needed for the effective implementation of management measures (e.g. within the framework of the European NATURA 2000 requirements) or for a better understanding of ecosystems
- Interaction of vegetation structure, which can be measured with LiDAR or SAR, with spectral information for the evaluation of forest properties
- Combining spatially very high resolution data (drones) with satellite data to understand ecohydrological processes and especially to derive hydrological variables such as soil moisture content or interception
Other Publications
| Citation key | Fenske2020 |
|---|---|
| Author | Fenske, K. and Feilhauer, H. and Förster, M. and Stellmes, M. and Waske, B. |
| Pages | 1-13 |
| Year | 2020 |
| ISSN | 0303-2434 |
| DOI | https://doi.org/10.1016/j.jag.2019.102036 |
| Journal | International Journal of Applied Earth Observation and Geoinformation |
| Volume | 87 |
| Abstract | Mapping heathland habitats is generally challenging due to fine-scale habitats as well as spectral ambiguities between different classes. A multi-seasonal time-series of multispectral RapidEye data from several phenological stages was analysed towards the classification of different vegetation communities. A 3-level hierarchical dependent classification using Import Vector Machines was tested, based on the assumption that a probabilistic output per class would help the mapping. The first level of the hierarchical classification was related to the moisture gradient, which was derived from Ellenberg’s moisture indicative value. The second level aimed to separate plant alliances; the third level differentiated individual plant associations. For the final integration of the three classification levels, two approaches were implemented: (i) the F1-score and (ii) the maximum classification probability. The overall classification accuracies of both methods were found to be similar, around 0.7. Nevertheless, based on our expert knowledge we found the probabilistic approach to provide a more realistic picture and to be more practical compared to the result using the F1-score from the management point of view. In addition, the overall performance of the maximum probabilistic approach is better in the sense that the same accuracy of 0.7 was achieved with a differentiation of 33 classes instead of only 13 classes for the F1-score, meaning that the method is able to separate more spectral classes at a more detailed level providing the same accuracy. |