Implementation und Test einer automatischen Verarbeitungskette zur Klassifizierung von Böden in Fernerkundungsdaten zur Verbesserung vorhandener Bodendaten für den Teststandort DEMMIN (Mecklenburg-Vorpommern)

Abstract

Within the scope of this master thesis, an automatic processing chain is developed in the programming language R, with which a soil map for the test field DEMMIN of DLR Neustrelitz is produced from remote sensing data. This soil map is used to improve existing soil data. For this purpose, relevant theoretical basics regarding the reflective properties of soils and vegetation as well as remote sensing-based soil mapping are discussed. A classification model is presented that combines a multi-criteria analysis with a classification algorithm. The data basis for this is data from the Sentinel-2 mission, digital elevation models and the available soil data for this area. Finally, the processing chain is tested using two test areas and compared with the soil data.