A comparison of change detection methods using MODIS images of Ruma National Park and surroundings (Kenya)
Katharina Müller (07/2017-01/2018)
Support: Christina Bogner, Valeska Scharsich, Britta Aufgebauer
National parks and protected areas in general play an important role in maintaining biodiversity, preserving endangered species and providing ecosystem services, such as water filtration or air cleaning. In many parts of the world, detailed information on land use and land cover (LULC) changes in national parks and surroundings is hardly available.
To decide whether the state of protection of the national park or management of surrounding areas have a quantifiable effect on LULC, remote sensing provides a valuable data source. We are interested in the changes in the Ruma National Park and its surroundings in Kenya. We use change vector analysis to detect changes of LULC using Landsat imagery. However, several methodological problems occurred in the process of finding a suitable threshold for changed areas as required in the change vector analysis. The goal of this work is to complement the analysis of Landsat images (temporal snapshots) by the break point analysis (BFAST) to derive objective thresholds for change detection.