Arctic Tundra Land Cover Classification on the Beaufort Coast using the Kennaugh Element Framework on Dual-Polarimetric TerraSAR-X Imagery

Kurzfassung

Arctic tundra landscapes are highly complex and are rapidly changing due to the warming climate. Datasets which document the spatial and temporal variability of the landscape are needed to monitor the rapid changes. Synthetic Aperture Radar (SAR) imagery is specifically suitable for monitoring the Arctic, as SAR, unlike optical remote sensing, can provide time series regardless of weather and illumination conditions. This study examines the seasonal backscatter mechanisms in Arctic tundra environments and their potential for land cover classification purposes using a time series of HH/HV TerraSAR-X imagery. A Random Forest classification was applied on multi-temporal backscatter intensity and Kennaugh matrix element data. The backscatter analysis revealed clear differences in the polarimetric response of water, soil and vegetation, while backscatter signal variations within different vegetation classes were more nuanced. The RF models show that the land cover classes can be distinguished with 92.4% accuracy using the Kennaugh element data, compared to 57.7% accuracy for backscatter intensity data. The accuracy was improved by adding texture measures to the predictor datasets, but the spatial resolution was reduced. TerraSAR-X acquisitions from the summer as well as from the autumn and winter seasons were important for the classification. The results of this study demonstrate that the Kennaugh elements derived from dual-polarized X-band imagery are a powerful tool for Arctic tundra land cover mapping.