Research Project
Geomorph4D – Characterising multi-process geomorphic change through high spatial- and high temporal-resolution monitoring
News
Research news can be found in our GIScience News Blog and on Twitter: #GEOMORPH4D.
New paper! Find all details on multi-directional change detection between point clouds in the full paper:
- Williams, J.G., Anders, K., Winiwarter, L., Zahs, V., Höfle, B. (2021): Multi-directional change detection between point clouds. ISPRS Journal of Photogrammetry and Remote Sensing. Vol. 172, pp. 95-113. https://doi.org/10.1016/j.isprsjprs.2020.12.002.
Objective
The objective of this project is to develop an approach capable of measuring and characterising multiple types of surface change across a point cloud scene, with a view to automatically extracting their contribution to the overall sediment budget (i.e. the supply, transport, and storage of sediment within the area of interest). We aim to use this information to explore the importance of rockfalls and talus accumulation at the headwall of rock glaciers, and to examine this within the context of changing climate conditions, and therefore stresses, within paraglacial environments.
Background
Near continuous, short-interval LiDAR monitoring has begun to provide new insights into a range of geomorphic processes.
Despite this potential, the signature of topographic change at a point in space rarely results from a single process. There remains a need to disaggregate multiple processes or mechanisms that may occur in unison to produce observed surface changes. On actively failing rock slopes, for example, such processes include the overall displacement of the slope due to deformation at the shear plane, as well as localised movements that may or may not be independent of this displacement, such as rockfalls. Disaggregating these processes and understanding their interaction is essential for understanding how patterns of movement at the surface reflect overall movement at depth, and for understand the cumulative set of mechanisms that contribute to failure.
In the case of rock glaciers, a dynamic equilibrium must exist between ice growth in the initiation zone and talus supply from the headwalls for continued to growth. Without enough talus, the shear stress of ice-saturated deposits will be insufficient to overcome the yield strength of the substrate, and movement will cease. The ability to characterise and distinguish ice thinning/thickening, rockfalls, individual boulder movements, and overall movement of the rock glacier is therefore critical for understanding this equilibrium.
At present, the techniques that we apply to analyse point cloud data, in particular from short-interval monitoring, are tailored to the individual style of movement that we expect or that is readily observable in the resulting change detection. Manual interpretation and/or measurement of superimposed processes within change detection data is complex, particularly in light of the considerable number of datasets (e.g. 103-104) generated from short-interval monitoring. The development of methods that allow for the disaggregation of multiple processes from 4D monitoring is therefore integral to an improved understanding of sediment movement through a point cloud scene.
Initial study site
One study site is the Äußeres Hochebenkar rock glacier, Obergurgl, Austria . During the summer period, where the ground is free of snow cover, the entire rock glacier and its source headwall will be monitored to constrain the various mechanisms of sediment production, transport, and removal from the system. Both the headwall and rock glacier are actively changing, and the site exhibits a range of surface processes.
Previous works and references
The importance of 4D monitoring for the improved understanding of rockfall mechanisms has been shown by Kromer et al. (2017) and Williams et al. (2018). In the latter, a series of algorithms was presented as a necessary tool for extracting discrete rockfall events from large numbers of point clouds, collected at hourly intervals. Critically, these techniques differ from the processing of datasets spaced at more conventional (e.g. monthly) periods but enable the detection of small events, which would otherwise be censored through superimposition or coalescence of neighbouring events. The ability to characterise a full range of event volumes was shown to provide unique insights into the mechanisms of rockfalls and their controls.
Multitemporal and multisource LiDAR data have been acquired over the Äußeres Hochebenkar rock glacier since 2006. The value of sub-annual monitoring in characterising the various mechanisms of rock glacier deformation, and in interpreting the mechanisms behind this has been demonstrated by Zahs et al. (2019) and Ulrich et al. (2021). Given the variety of processes that act to shape the rock glacier surface at any given point, improvements in the temporal resolution of monitoring have enabled the detection of processes that would superimpose one another over a year. While individual boulders are often tracked in order to characterise overall movement of the rock glacier, the analysis of LiDAR data in full 3D has provided useful insights regarding the extent to which these boulders move independently of the rock glacier as a whole.
Related Projects
- AHK-4D: High-resolution and high-frequency monitoring of the rock glacier Äußeres Hochebenkar (AHK) in Austria.
Publications
- Williams, J.G., Anders, K., Winiwarter, L., Zahs, V., Höfle, B. (2021): Multi-directional change detection between point clouds. ISPRS Journal of Photogrammetry and Remote Sensing. Vol. 172, pp. 95-113. https://doi.org/10.1016/j.isprsjprs.2020.12.002.
