Landslides are one of the major weather-related geohazards. To assess their potential impact and design mitigation solutions, a detailed understanding of the slope processes is required. Landslide modelling is typically based on data-rich geomechanical models. Recently, machine learning has shown promising results in modelling a variety of processes. Furthermore, slope conditions are now also monitored from space, in wide-area repeat surveys from satellites. In the present study we tested if use of machine learning, combined with readily available remote sensing data, allows us to build a deformation nowcasting model. A successful landslide deformation nowcast, based on remote sensing data and machine learning, would demonstrate effective understanding of the slope processes, even in the absence of physical modelling. We tested our methodology on the Vögelsberg, a deep-seated landslide near Innsbruck, Austria. Our results show that the formulation of such a machine learning system is not as straightforward as often hoped for. The primary issue is the freedom of the model compared to the number of acceleration events in the time series available for training, as well as inherent limitations of the standard quality metrics such as the mean squared error. Satellite remote sensing has the potential to provide longer time series, over wide areas. However, although longer time series of deformation and slope conditions are clearly beneficial for machine-learning-based analyses, the present study shows the importance of the training data quality but also that this technique is mostly applicable to the well-monitored, more dynamic deforming landslides.@en

Continuous and slow-moving deep-seated landslides entail challenges for the effective planning of mitigation strategies aiming at the reduction of landslide movements. Given that the activity of most of these landslides is governed by pore pressure variations within the shear zone, profound knowledge about their hydrogeological control is required. In this context, the present study presents a new approach for the spatial assessment of probable recharge areas to better understand a slope's hydrogeological system. The highly automated geo-statistical approach derives recharge probability maps of groundwater based on stable isotope monitoring and a digital elevation model (DEM). By monitoring stable isotopes in both groundwater and precipitation, mean elevations of recharge areas can be determined and further constrained in space with the help of the DEM. The approach was applied to the Vögelsberg landslide, an active slab of a deep-seated gravitational slope deformation (DSGSD) in the Watten valley (Tyrol, Austria). Resulting recharge probability maps indicate that shallow groundwater emerging at springs on the landslide recharges between 1000 and 1650 a.s.l. In contrast, groundwater encountered in wells up to 49ĝ€¯m below the landslide's surface indicates a mean recharge elevation of up to 2200 a.s.l. matching the highest parts of the catchment. Further inferred proxies, including flow path length, estimated recharge area sizes, and mean transit times of groundwater, resulted in a profound understanding of the hydrogeological driver of the landslide. It is shown that the new approach can provide valuable insights into the spatial pattern of probable recharge areas where mitigation measures aiming at reducing groundwater recharge could be most effective.

@en

Continuous landslide monitoring is a crucial task for the management of natural hazards for identifying suitable mitigation measures, including nature-based solutions. In the present study, three monitoring techniques including (i) an automated tracking total station (ATTS), (ii) multi-temporal terrestrial laser scanning (TLS) and (iii) space-borne interferometric synthetic aperture radar (InSAR) are applied to monitor the spatio-temporal displacement patterns of the Vögelsberg landslide (Tyrol, Austria) between 2016/05 and 2020/06. The landslide shows spatially and temporally varying displacement rates with up to 12 cm/a and a mean annual displacement of 4 cm/a. The results show that only the ATTS provides sufficient temporal resolution and spatial accuracy for assessing the temporal behaviour of the landslide's movement. However, ATTS measurements are only available at the installed 53 retro-reflecting prisms. Multi-temporal TLS can provide additional insight into the spatial displacement pattern at various man-made and natural objects such as walls, fences, poles and tree stems. But the respective accuracy and data acquisition intervals do not allow to draw conclusions about the temporal dynamics of the landslide's movement. Results of the InSAR technique based on Sentinel-1 imagery show good agreement with ATTS measurements, but cannot provide real-time information on the landslide's acceleration and deceleration phases. However, in combination, the measurement techniques provide vital information in both the spatial and temporal domain.@en