Remote Sensing of Bush Encroachment on Commercial Cattle Farms in Semi-Arid Rangelands in Namibia

Remote Sensing of Bush Encroachment on Commercial Cattle Farms in Semi-Arid Rangelands in Namibia

Matthias Schröter, Oliver Jakoby, Roland Olbrich, Marcus Eichhorn, Stefan Baumgärtner
DOI: 10.4018/978-1-60960-156-0.ch016
(Individual Chapters)
No Current Special Offers


Bush encroachment is one of the most extensive changes in land cover in semi-arid rangelands and an urgent problem for cattle farming, rapidly reducing the productivity of the rangeland. Despite the severity of these consequences, a complete and accurate assessment of bush encroached areas is still missing at large. This study aims at assessing bush encroachment on commercial cattle farms in central Namibia by employing remote sensing methods to distinguish between areas covered by bush and open rangeland. The authors use different classification techniques and vegetation indices to characterize the nature of vegetation cover. Their analysis shows that results are sensitive to specific classifications of indices. As an accuracy assessment could not be run on these results the authors could not analyze which classification approximates real bush encroachment best. Hence, this study highlights the need for further analysis. Ground truth data, in the form of field mappings, high resolution aerial photographs or local expert knowledge are needed to gain further insights and produce reliable results.
Chapter Preview


Remote sensing can be defined as “the science of collecting information about objects without coming into physical contact with them” (Hill, 2000). This definition applies to the recording of electromagnetic radiation by aircraft or satellite-born sensors (Richards & Jia 1999; de Lange, 2006; Albertz, 2007). Distinguishing between different objects on the images relies on the difference in their spectral reflectance behavior (Albertz, 2007). Sensors record electromagnetic radiation being reflected by objects on the earth’s surface, e.g. plants, buildings, water bodies, or in the atmosphere, e.g. clouds. These objects show characteristic patterns of reflectance across the wavelength spectrum allowing for the determination of specific types of objects (de Lange, 2006). These patterns, referred to as spectral signatures or “spectral fingerprints” (de Lange, 2006), differ among specific types of objects, e.g. vegetation, water and soil, but also between similar objects, e.g. different kinds of vegetation and soils (de Lange, 2006).

Complete Chapter List

Search this Book: