Assessing post-socialist urban change with Landsat data; Case study Berlin, Germany

Germany´s capital Berlin has undergone tremendous urban structural changes in the past decades since the German re-unification. In addition to processes of urban growth and densification, processes of urban shrinkage occur in Berlin. As these processes have great impact on urban ecosystems and therefore on human well-being, monitoring of respective changes is required to support sustainable urban development. This study therefore aims to detect and quantify the urban structural changes in Berlin and its outskirts since the re-unification in 1989. Hence, two Landsat TM5 data sets covering the city of Berlin and adjacent rural areas were obtained, geometrically corrected and layer stacked. Subsequently, the layer stack was subject to a multi-temporal principal component analysis. Thereafter, an unsupervised classification of selected principal components (PC3, PC4, and PC6) was carried out. The change detection was conducted based on the classification results and yielded reliable results of approximately 90% accuracy. The change analysis showed that besides alternations in agricultural crops major changes occurred on the outskirts of Berlin, mostly in terms of large-area construction activity in the former eastern part. Furthermore, contrary processes could be identified in the former border region of the divided city. Whilst there was building activity in the more central parts, succession and renaturation were detected in most other parts of the Berlin Wall. The study showed that remote sensing has proven to be an appropriate tool to assess contrary post-socialist urban structural changes of Berlin, Germany.