Author Archives: andreas mayr

Drone-based laser scanning for erosion monitoring

Fig. 1: ULS system set up and flight at a 48-ha test site in the Dolomites (Italy).

Within the ERODYN project, we are using a drone (the RIEGL RiCOPTER system) equipped with the VUX-1LR laser scanner to produce high resolution 3D point clouds and derive detailed digital terrain models for automated geomorphological analyses. The aim is to detect and quantify changes at eroded areas by repeat scans. Such unmanned aerial vehicle laser scanning systems (ULS) are a relatively new technology, and tests in mountain areas are scarce. Thus, questions arise regarding ULS applicability in mountainous settings, the achievable accuracy of the data, and its potential for shallow erosion activity mapping. Recent investigations of the ERODYN project show promising results and, despite some limitations, changes at (sub-)decimetre level can be reliably observed.

Read more:

Mayr, A.; Bremer, M.; Rutzinger, M. (2020): 3D point errors and change detection accuracy of unmanned aerial vehicle laser scanning data. ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, V-2-2020, 765 – 772.

Mayr, A.; Bremer, M.; Rutzinger, M.; Geitner, C. (2019): Unmanned aerial vehicle laser scanning for erosion monitoring in Alpine grassland. ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, IV-2/W5, 405 – 412.

Fig. 2: Subset of the test site with surface deformation (3D point cloud distance) due to secondary erosion and deposition of eroded material. (a) Planimetric view of the point cloud coloured by deformation exceeding the LOD95 (level of detection at the 95%-confidence interval), with RGB coloured point cloud, shaded relief, and 5-m contours as background. (b) Oblique view of the RGB-coloured point cloud subset. (c) Oblique view of the point cloud subset coloured by deformation exceeding the LOD95.

Fig. 3: The level of detection (at the 95%-confidence interval; LOD95) indicates the magnitude of observable changes using multitemporal point clouds. The LOD95 is partly related to the registration error and the surface roughness. At some points, additional uncertainties related to laser footprint effects (modelled as a function of range, incidence angle, and beam divergence) affect the LOD95 negatively. Removing such erroneous points with a point error threshold improves the LOD95; though this comes at the cost of increasingly incomplete point clouds (depending on the threshold).

Our research groups at the University of Innsbruck:

Remote Sensing & Topographic LiDAR Research Group, https://www.uibk.ac.at/geographie/lidar/
Soil and Landscape Ecology Research Group, https://www.uibk.ac.at/geographie/bola/

Some hardware and software used:

Riegl RiCOPTER: http://www.riegl.com/products/unmanned-scanning/
Software used for flight planning: UgCS, https://www.ugcs.com/
Software for 3D point cloud processing: SAGA LIS, https://www.laserdata.at/software.html

How does vegetation influence the susceptibility of grassland slopes towards shallow erosion?

Vegetation plays (amongst other factors) an important role for the stability of slopes near their surface. Not only woody vegetation (i.e. shrubs and trees) but also herbaceous vegetation (dominating grasslands) can enhance the resistance of soil against shallow erosion. Being to some extent controllable via land management (in contrast to other factors, such as topography or the geological setting), vegetation and its impact on slope stability are of high interest for erosion prevention.

In a recent literature review in the journal Earth-Science Reviews, members of the ERODYN team summarize the current knowledge on this topic, and critically discuss the slope stabilization potential of herbaceous vegetation as compared to woody vegetation.

A second paper, just published in the Journal of Environmental Management, seeks to provide insights into the (surface-parallel) tensile strength of the topsoil in subalpine grasslands (< 10 cm depth). In an empirical study, tensile strength was measured in the field, and analysed with regard to potential impacts of soil and vegetation parameters. It turned out that densely interwoven roots and clonal structures often form a surprisingly strong “surface mat” as a small-scale reinforcement of the topsoil, possibly being important for the redistribution of stress. This surface-mat effect depends on the vegetation composition, and certain species provide stronger reinforcement than others. Species with well-developed root systems and a high capacity for clonal growth seem to be most advantageous, but also a balanced nitrogen supply and the plant and structural diversity appear to play a role.

Read more:

Löbmann, M.T.; Geitner, C.; Wellstein, C.; Zerbe, S. (2020): The influence of herbaceous vegetation on slope stability – A review. Earth-Science Reviews, 209, 103328.

Löbmann, M.T.; Tonin, R.; Stegemann, J.; Zerbe, S.; Geitner, C.; Mayr, A.; Wellstein, C. (2020): Towards a better understanding of shallow erosion resistance of subalpine grasslands. Journal of Environmental Management, 276, 111267.

For details on the measurement of the surface-mat effect, see also a previous study performed in montane grasslands: Löbmann et al. 2020, Catena.

A new method for measuring the tensile strength of grassland topsoil

Members of the ERODYN team at the Free University of Bozen-Bolzano have developed a novel in-situ method for measuring the surface-parallel tensile strength of the topsoil. The densely interwoven roots of herbaceous vegetation can form a strong reinforcement in the top soil (“surface-mat”), and this may be an important factor for resistance against small-scale shallow erosion, e.g. by lateral redistribution of stress and prevention of tension cracks in the turf. The method has been applied at montane grassland test sites to investigate potential effects of soil physical and biological factors. The results suggest that suitable vegetation management can enhance the topsoil stability through increasing the surface-mat effect.

More details are published here:

 

New publication: Object-based point cloud analysis for landslide and erosion monitoring

A new paper in the journal Photogrammetric Engineering & Remote Sensing presents methods for landslide and erosion monitoring based on terrestrial laser scanning point clouds. The 3D approach combines deformation detection and a geomorphological object classification automatically. This enhances the analysis and interpretability of a point cloud time series for erosion and landslide assessment.

Mayr, A.; Rutzinger, M.; Geitner, C. (2019): Object-based point cloud analysis for landslide and erosion monitoring. Photogrammetric Engineering & Remote Sensing, 85 (6), 455 – 462.

The following figure shows some example results obtained with the presented approach:

Left: Cumulative mean 3D deformation per class for a shallow landslide (calculated with a constant registration error parameter reg = 0.1 m), pointing to two time steps with pronounced changes at the landslide scarp. Right: Having a closer look at the first time step shows scarp erosion and the deposition of some dislocated scarp fragments), with (a) a 3D view of the second point cloud epoch coloured by mean deformation per object, (b) transect coloured by mean deformation per object, (c) transect overlaying the points of both epochs coloured by their class labels.

EGU conference contribution: UAV-based hyperspectral techniques for monitoring shallow erosion

Preliminary results from an ERODYN experiment will be presented with a poster at this years EGU (European Geosciences Union General Assembly) in Vienna:

Lloyd, J., Mejia-Aguilar, A., Tonin, R., Loebmann, M., Gild, Ch., Mayr, A., and Sonnenschein, R.:
UAV-based hyperspectral techniques for monitoring shallow erosion in alpine pastures

Wednesday 10 April, 08:30–10:15, Hall A

Traineeship / MSc thesis in hyperspectral sensing of vegetation

Eurac Research (Bozen, Italy) is looking for a MSc student / trainee within the ERODYN project. The traineeship will focus on hyperspectral data acquisition and processing to characterize grassland vegetation in the context of erosion susceptibility. There will likely be the possibility to write a MSc thesis on this topic after the traineeship.

More info here: EURAC_MScThesis_HyperspectralApps_RS.

Field work for vegetation assessment

This summer the team from the University of Bozen has sampled vegetation at test sites in the Nature Park Puez – Geisler, aiming at a spatial analysis and an assessment of different types of grassland vegetation with respect to landuse and erosion susceptibility.

At the same site the EURAC team was flying with an unmanned aerial vehicle (UAV), equipped with multi- and hyperspectral cameras, while field spectrometer measurements on the ground were collected for calibration of the UAV data. The aim of this experiment is to assess how multi-/hyperspectral data can help to characterize relevant grassland vegetation parameters in the context of erosion studies.

Terrestrial laser scanning of eroded areas

A grassland slope has been surveyed with a terrestrial laser scanner (TLS) to map eroded areas and their morphology. The detailed terrain data from TLS will also be integrated with multi-/hyperspectral data from an unmanned aerial vehicle (UAV) and botanical surveys from field plots to map and analyse vegetation types.

Conference paper online (ISPRS Archives)

Our contribution to the ISPRS Symposium in Riva del Garda is now online in the ISPRS Archives, the series of peer-reviewed proceedings published by the International Society of Photogrammetry and Remote Sensing (open access):

Mayr, A., Rutzinger, M. & Geitner, C. (2018): Multitemporal analysis of objects in 3D point clouds for landslide monitoring. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLII-2, 691–697.
[PDF]

Methods for landslide and erosion monitoring at ISPRS Symposium

Last week the ISPRS Technical Commission II Symposium 2018 took place in Riva del Garda, Italy, with participants from 37 countries. Researchers, practioners and companies presented and discussed the state-of-the-art in photogrammetry, computer vision, and remote sensing.

Members of the ERODYN team contributed to the session “3D Mapping for Environmental & Infrastructure Monitoring” with a talk about the latest methodological developments in object-based landslide and erosion monitoring using terrestrial laser scanning point clouds.