Time Series Analysis of Reef Arches and Coastal Dynamics at the East Frisian Islands in the North Sea Using Remote Sensing Data

Kurzfassung

This master's thesis examines the spatial-temporal dynamics of reef arcs along the East Frisian Islands in the German North Sea for the period 2016–2024 and shows how the location and area extent of these morphological structures can be derived from remote sensing data. The data basis consists of high-resolution PlanetScope scenes (3 m) and Sentinel-2 L2A. For radiometric harmonization of the multi-sensor time series, histogram matching according to Gorelick (2021) was implemented in Google Earth Engine and used as the structural basis for the script workflow. Several spectral indices (e.g., NDVI, NDWI, NDTI) were calculated from the harmonized mosaics and tested as inputs for an annual, unsupervised K-means classification (WEKA). The NDTI provided the best visual separation of reef arcs, which is why only this index was used for clustering. The resulting grids were further processed in QGIS to isolate relevant reef structures, generalize them, and vectorize them as centerlines. Using orthogonal transects, annual shifts relative to 2017 were determined as the median of the hub distances (m·a⁻¹). In addition, the annual areas (km²) per region of interest (ROI 1–5) were calculated and interpreted and validated in their morphological context using BSH bathymetry. The results show a pronounced west-east gradient: ROI 1 and ROI 2 are the most dynamic, while ROI 4 and ROI 5 are comparatively stable. ROI 3 plays a transitional role. A linear spatial trend analysis of the mean median values confirms a systematic decrease in dynamics from ROI 1 to ROI 5, consistent with sediment dynamics towards the east. Possible driving factors include changes in tidal/wave dynamics, sediment budget, and local interventions. The combined GEE/QGIS workflow with histogram matching and NDTI-driven classification thus proves to be a robust, reproducible method for monitoring tasks. Limitations include atmospheric disturbances, tidal differences, and manual rework. Overall, the spatial context dominates the extent of reef arc displacement over temporal dynamics. At the same time, the results provide reliable indicators for targeted, location-specific coastal monitoring and management.