There are still many unknowns in the understanding of spatiotemporal dynamics of Earth surface processes. A systematic monitoring of these processes can reveal new details on the process and is an important source for further process analysis or modelling.
Remote sensing imagery have always been a magnificent source for monitoring Earth surface processes. Recently available technologies, such as airborne and terrestrial laser scanning, and photogrammetry with structure from motion and multi-view stereo 3D reconstruction using images taken terrestrially or from unmanned aerial vehicles (UAVs), have allowed for greater flexibility in the design of Earth surface processes monitoring with unique spatial and temporal resolution. With these new technologies, detailed data on the process extent and dynamics, e.g. of rock glaciers, landslides, snow cover, vegetation types, leaf area index, or river dynamics can be obtained with high accuracy. These data are of high scientific interest for advancing our understanding of spatiotemporal relationships and dynamics of Earth surface processes. Additionally, important input variables for modeling future dynamics of surface processes can be obtained which may aid in increasing the quality of predictive modeling results. This is expected as every modeling result is highly influenced by the quality of the input variables.
Within the Earth Surface Processes Monitoring group at the Chair of GIScience, the potential and limitations of these new monitoring technologies are evaluated. Also, methods for the (automatic) analysis and delineation of spatiotemporal dynamics of the monitored earth surface processes from remote sensing imagery are developed. For these purposes commercial software as well as an in-house image analysis tool for segmentation and object oriented classification of high resolution remote sensing data has been developed.