Deformation characteristics and multi-slab formation of a deep-seated rock slide in a high alpine environment (Bliggspitze, Austria)
- 479 Downloads
This study presents the results of a more than 10-year-long field investigation and remote sensing monitoring campaign of a highly active deep-seated rock slide located in a glacial to periglacial environment (Bliggspitze, Tyrol, Austria). Data concerning (i) the terrain surface displacements based on imagery (webcam time-lapse, ortho-images) and both terrestrial and airborne laser scanning, and (ii) the geological-structural and geomorphological situation were analysed to develop a geological-geometrical model of the rock slide and to study the temporally variable activity behaviour and the formation of individual rock slide slabs. Results clearly show that at least seven rock slide slabs were formed at different times and under different slope stability conditions. Some of these rock slide slabs were displaced at slow to moderate velocities and reached scarp offsets of several tens of metres, whereas other, shallower slabs collapsed and formed extremely rapid rock falls and avalanches. Generally, the rock slide is affected by rock mass cataclasis, fracturing and loosening, which in turn cause extensive mass loss accompanied by debris accumulation at lower parts of the slope. The cause for the development of the Bliggspitze rock slide is poorly understood. However, there are clear indications that permafrost degradation and/or glacial retreat, particularly at the foot of the slide, during the recent decades may have adversely affected the slope stability situation.
KeywordsDeep-seated rock slide Metamorphic rock Rock slide slab formation High alpine environment Glacial changes Remote sensing Monitoring Airborne and terrestrial laser scanning
Deep-seated rock slides, located in well-foliated, mica-rich metamorphic rock masses, are usually compound slides characterised by significant internal fracturing and distortion of the displaced mass (Hungr et al. 2014). Typical geomorphological and structural features are stepped slope profiles, horst and graben structures, and/or secondary shear zones representing distinct boundaries of rock slide slabs/blocks of individual deformation and activity characteristics. As opposed to the formation of “single block” rock slides, the deformation behaviour of compound rock slide systems composed of individual slabs is different and complex (Zangerl et al. 2015). Herein, “rock slide slabs” are defined as individual parts of rock/soil masses which are delimited to other parts of the slide by shear zones composed of loose and non-cemented fault breccias and gouges. Deformation accumulates predominately along these shear zones and slope activity may vary from slab to slab. Generally, the development of individual slabs represents the slope stability evolution over time and leads to a considerable increase in complexity with regard to the kinematical, geomechanical and hydrogeological properties of a rock slide (Strauhal et al. 2017). Interaction with adjacent slope sections, displacement of rock slabs and soil deposits (e.g. new talus accumulation) cause ongoing slope processes which are generally difficult to characterise and quantify. It is even more challenging to numerically simulate such phenomena for process analyses, hazard assessments and mitigation measures. For numerical modelling, the complex mechanical interaction of different active slabs arranged on top of and/or laterally to each other have to be considered and often modelled in combination with large discrete shear displacements. Conceptually, this complex geomechanical behaviour has a considerable impact on the overall slope stability, and an understanding of such rock slide processes is crucial for risk and hazard assessment. Key issues concerning the failure and deformation behaviour of such rock slides require comprehensive investigation and prediction methods. Hence, there is a difference between a large rock slide that fails extremely rapidly as a whole in a single high-magnitude event, potentially developing a long runout path, and one that subdivides into smaller individual slabs that interact but do not fail synchronously as an entire mass. In the latter case, the failure of individual rock slide slabs is temporally different, i.e. not characterised by the collapse of the entire slope, but by sequential mobilisation of much smaller individual volumes (slabs). As such, even if such slabs fail rapidly as rock avalanches, but successively one slab after the other, a collapse is characterised by much shorter runout lengths (Willenberg et al. 2008). Such failure and runout behaviour may be expected and predicted based on the empirical correlation between failure volume and runout length (e.g. Heim 1932; Scheidegger 1973; Körner 1976; Okura et al. 2000; Legros 2002; Prager 2010). According to Hungr et al. 2014, the term “rock avalanche” is applied herein to failure events, independent of the landslide volume, characterised by extremely rapid flow-like motion of fragmented rocks featuring strong mechanical interaction between the displaced rock fragments. This is in contrast to rock fall events occurring singly or in clusters and interacting mainly with the substrate but having little dynamic interaction between the moving fragments. The increased internal fracturing, fragmentation and loosening of the rock mass during slope deformation and slab formation can produce considerable amounts of loose soil that freshly accumulate as talus at the foot of the slope (scree, rock fall and rock avalanche deposits). Conceptually, this material erosion and accumulation changes the slope geometry and the in situ stress conditions and thus modifies the overall slope stability.
Additional complexity may arise when deep-seated rock slides are located in permafrost regions characterised by ice-filled fractures and rock mass temperatures below the freezing point. Furthermore, the stability condition of a rock slope can be affected by the geomechanical and hydrogeological interaction with glaciers at the lower part or foot area of a potentially unstable slope. Glacial retreat and thawing permafrost influence the strength properties, the in situ stress conditions by unloading and the hydrological setting of a fractured rock mass (e.g. Davies et al. 2001; Deline et al. 2015; Krautblatter et al. 2013). Consequently, an increase in the instability of rock slopes in permafrost (e.g. Fischer et al. 2012; Huggel 2009; Paranunzio et al. 2016; Ravanel and Deline 2011) and recently de-glaciated areas (e.g. Cossart et al. 2008; McColl 2012; Deline et al. 2015) is observed. However, a full understanding of the processes and triggers leading to the formation and deformation of deep-seated rock slides in a glacial/periglacial environment is still lacking, and further research is needed (Ballantyne 2002; Evans and Clague 1994; McColl 2012; McColl and Davies 2013; Riva et al. 2018).
In order to assess a rock slide and monitor the slope evolution and deformation behaviour (i.e. spatial extent, 3D geometry, kinematics, location and shape of shear zones, slab formation, rock mass fragmentation and debris accumulation), mapping, rock mass characterisation, groundwater and multi-temporal 2D/3D deformation data are essential. These data can be obtained from different in situ methods, such as geomorphological and geological mapping, geophysical surveys, geodetic monitoring, and borehole drilling with geophysical logging, hydraulic packer testing, and inclinometer as well as piezometer measurements. Exemplarily, comprehensive rock slope investigations were performed at the Randa landslide in Switzerland (Willenberg et al. 2008; Gischig 2011), La Saxe in Italy (Crosta et al. 2014), La Clapiere in France (Palis et al. 2017), and in the Kaunertal valley in Austria (Strauhal et al. 2017; Zangerl et al. 2010, 2015). Although extremely useful and often a requirement for a comprehensive rock slide investigation campaign, these direct methods are costly and require direct access to the slide. In high mountain areas with harsh conditions, poor and dangerous accessibility, direct site investigation and monitoring is often hardly feasible or can be done only under enormous expenses. Therefore, the assessments of the Bliggspitze site presented herein have been based on geological field surveys and extensive remote sensing. In general, remote sensing techniques provide contactless tools for investigating both geomorphological and geological features (e.g. Abellán et al. 2014; Dewez et al. 2016; Wichmann et al. 2018) and to monitor surface displacements in 2D and 3D (e.g. Delacourt et al. 2007; Scaioni et al. 2014). In contrast to other techniques, a major advantage of remote sensing techniques is that site assessments and monitoring can be performed even at inaccessible areas and/or hazard zones (Jaboyedoff et al. 2009; Oppikofer et al. 2008; Oppikofer et al. 2011; Sturzenegger and Stead 2009).
After the first site visits since summer 2007 (Zangerl et al. 2010), a geological-geomorphological map of the study area at a survey scale of 1:4000 was produced by field mapping campaigns in summer 2011 and 2012 (Dejean de la Bâtie 2016). Shaded relief images derived from airborne laser scanning (ALS), digital elevation models (DEMs) and ortho-images were used to map lithological units, geomorphological features including primary and secondary scarps, up-hill facing scarps, extensional fractures, and geological structures comprising foliation, meso-scale fractures and brittle faults. In addition, hydrogeological data including the location and physical parameters (i.e. temperature, flow rate, electrical conductivity) of springs and streams were gathered. A direct mapping of the rock slide mass was only possible at the higher slope section near the head scarp. There, geological and geomorphological structures and permafrost indicators (i.e. ice-filled fractures) are well exposed and accessible. Loose and uncemented fault gouge and breccia samples of a brittle fault zone were taken for laboratory analyses to obatin the grain size distribution, the mineralogy (powder X-ray diffraction) and the shear strength properties (ring shear tests) (Strauhal 2015). Beside visual inspections, the occurrence of permafrost was investigated by measuring the bottom temperature of the snow cover (BTS) (Haeberli 1973; Hoelzle 1992; Keuschnig 2016) by using 23 temperature loggers placed in open fractures around the head scarp. As long as these loggers are covered by snow and thus are insulated against air temperature and solar radiation, they provide temperature data as a proxy for the occurrence of permafrost (Keuschnig 2016). The loggers recorded the temperature every 2 h between September 2013 and September 2016.
Remote sensing data used for the Bliggspitze rock slide analyses
Date of release
Information (relevance for the Bliggspitze area)
Historical topographic maps (Alpine Clubs, DAV, ÖAV)
1893, 1946, 1969, 1997, 2006
- Reconstruction of Bliggferner ice retreat and assessment of glacier volume changes
(Federal State of Tyrol)
2003, 2007, 2009, 2010
- Assessment and mapping of terrain units (e.g. bedrock, soils, structural and morphological, glacial ice)
- 2D measurement of terrain surface changes
- Visual assessment, interpretation of rock slide processes
(Federal State of Tyrol/TIWAG)
2006, 2007, 2008, 2009, 2010, 2012
- Identification and quantification of 2D/3D surface displacements, generation of displacement and distance change maps
- Identification and quantification of mass relocations (mass loss and accumulation)
- Analyses of geomorphological and structural features
(Fey et al. 2017a)
2012, 2014, 2015, 2016
(Federal State of Tyrol)
- Visual analyses and interpretation of slope deformation processes (based on time-lapse image sequences)
Terrestrial images from field investigation
2003, 2007, 2009, 2011
- Visual analyses and interpretation of slope deformation processes
In addition, images taken by two Mobotix M22 webcams (Fig. 1) provided valuable information concerning the temporal evolution of the rock slide. The two webcams were installed by Federal State of Tyrol in September 2007 at the foot of the slope and programmed to take three images per day, which were transmitted by a modem and thus quickly available online. The cameras were installed and configured in such a way that one camera captures the total overview of the rock slide and the other one captures a detailed view of the upper slope section.
Temporal and spatial changes of the Bliggferner glacier were assessed by analyses of historical topographical maps of the Austrian and German Alpine Clubs (maps dated 1893, 1946, 1969 and 1997, see Table 1).
Before deformation analyses were performed, the ALS and TLS raw data were pre-processed and transformed into a common coordinate system by using a relative registration and geo-referencing approach (Fey 2018). The ALS point cloud from 2012 available in the global coordinate system (ETRS 89 UTM 32 N) was defined as a reference point cloud, and all subsequent point clouds were matched onto this reference point cloud by the use of an iterative closest point algorithm (ICP). Registration of the TLS scans was done using an ICP algorithm implemented in RiSCAN Pro (Riegl 2013). The individual TLS scans were matched onto stable areas (i.e. areas which have not been modified by geomorphological and geological surface processes) of the reference point cloud from 2012. For slope deformation analyses, the precise relative alignment of two data sets is more important than the precise absolute geo-referencing. Therefore, the geo-referencing of the ALS point clouds was revised with the aim to increase the relative accuracy of the data sets (Fey 2018; Fey et al. 2015). A transformation matrix was derived by matching the stable areas of the respective ALS point cloud onto the reference point cloud (ALS scan from 2012) using an ICP algorithm implemented into OPALS (Pfeifer et al. 2014; Glira et al. 2015). Then, the transformation matrix was applied to the entire ALS point cloud data sets.
After the pre-processing step, ALS and TLS data were analysed focusing on both types of surface deformation analyses, i.e. slope displacement and distance change analyses, respectively. These deformation analyses were performed by applying the software SAGA GIS and the SAGA GIS add-on LIS Tools (www.laserdata.at, 2018). For displacement analyses, the distance between homologous parts detected in two data sets is measured by the identification of corresponding surface elements (e.g. terrain ridges or rock slide slabs, Fey 2018). Displacement analyses are used to quantify the amount of the 2D/3D displacement and the displacement direction of rock slide slabs. The displacement analyses are based on two techniques, i.e. the break line tracking and the image correlation method (Fey et al. 2015). For both methods, the point cloud data set was triangulated to a digital elevation model (DEM). Break-line tracking was used to analyse surface displacement processes accompanied by strong structural terrain changes. Break-line tracking is based on the identification of displacements of terrain ridges and depression structures which can be located on the two images. The ridge and depression structures are derived from curvature images calculated from ALS DEMs and then automatically skeletonised to break lines. Junction points of break lines were used to track manually the displacement of ridge and depression structures (Fey et al. 2015).
Level of detection (LoD) of displacement and distance change analyses (Fey 2018)
3D Distance change
ALS - ALS
ALS - ALS
ALS - ALS
ALS - ALS
ALS - ALS
ALS - TLS
TLS - TLS
TLS - TLS
As opposed to displacement analyses, distance change analyses quantify the distance between the pre- and post-failure terrain surface or between different post-failure terrain surfaces. The approach used for distance measurements (Fey and Wichmann 2017) is point cloud based, fully 3D and robust against surface roughness and low point densities. In contrast to elevation differences calculated from raster DEMs, the 3D distance measurement approach (Fey and Wichmann 2017) measures distances in 3D in the direction of a normal of a tangent plane which is fitted to the surface. Thus, the 3D distance measurement approach does not necessarily calculate the distance along an axis in the vertical direction. The LoD used for distance change analyses (Table 2) is based on a confidence interval considering spatially variable positional uncertainties, local surface roughness and registration errors (Fey and Wichmann 2017). The distance change maps were used to identify slope areas affected by (i) extremely rapid failure processes such as rock falls and rock avalanches, (ii) accumulation of soil deposits (i.e. debris generated from slope processes), and (iii) terrain surface lowering and uplift due to slope deformation processes.
In addition to laser scan data, multi-temporal ortho-images with spatial resolutions of 0.25 and 0.2 m, respectively, were analysed (airborne surveys, dated 2003, 2007, 2009 and 2010, see Table 1). The visual analyses and interpretation of ortho-images enabled the identification and documentation of slope changes, comprising the formation of new slabs and scarps, glacier cover and its changes, spatial distribution of rock outcrops and accumulation of debris at the time of image acquisition.
Selected webcam recordings were used to generate time-lapse image series for selected periods of interest, i.e. phases of intensified slope activity. These time-lapse analyses enabled a comprehensive reconstruction and visualisation of deformation processes such as the formation of new scarps, opening of extensional cracks, formation of new sliding slabs, rock/debris falls and slides. It should be noted, however, that these time-lapse analyses were sensitive to and limited by bad weather conditions (strong wind, snow fall, fog).
In addition, photographs taken during the numerous field campaigns since 2007 were analysed for the reconstruction and analyses of slope processes.
The reconstruction of the glacial retreat is based on historical topographical maps. In a first step, the historical maps were scanned, rectified and georeferenced, and the elevation lines of these maps were digitised and interpolated. In a second step, DEMs with a resolution of 20 m for a certain survey time were produced and used to calculate glacier change maps between two periods. In a third step, glacier volume changes were assessed by calculating DEM differences and multiplying the summarised raster cells of the DEM differences with the raster cell size.
Geological and (peri-)glacial characterisation
Concerning the brittle tectonic overprint of the area, some distinct brittle faults striking E–W to NE–SW and dipping steeply to moderately N and NW, respectively, were mapped (Fig. 2c). Faults were mapped in situ by field surveys and by GIS-based lineament mapping on ortho-images and ALS-shaded relief images. Brittle fault zone cores show either mm- to dm-thick zones of cemented cataclasites or loose uncemented fault gouges and breccias, with adjacent rims of increased fracture densities (i.e. fault damage zones, e.g. Strauhal et al. 2017). On several brittle fault planes, tectonically formed slickenside striations were observed (Fig. 2c). The head scarp of the Bliggspitze rock slide is formed along a prominent brittle fault plane dipping moderately towards N (i.e. mean fault dip direction/dip angle of 348/44). Slickenside striations on this fault plane are dipping with a mean trend/plunge of 309/36 oblique to the fault dip, and thus could not be related to the rock slide formation process. One prominent E–W-striking steeply dipping brittle fault zone extends from the stable bedrock outside of the slide into the rock slide and separates the rock slide into two parts, later defined as the upper rock slide “slab A” and lower “slab B” (Fig. 3). In order to characterise such brittle faults, a sample of fault zone material was taken near the main scarp at an altitude 3170 m a.s.l and investigated with regard to the grain size distribution, mineral content and shear strength properties. Mineral analyses based on X-ray powder diffraction (XRPD) and the standard Rietveld method show compositions of 28% quartz, 27% plagioclase, 3% alkali-feldspar, 25% mica (including both biotite and muscovite) and 17% chlorite (RVS 11.062 1987). This mineralogical composition is similar to that generally observed in the paragneissic host rock, and it was found that here, clay minerals were usually not formed from the host rock due to weathering or alteration processes (Hammer 1922; Purtscheller 1987).
Rock slide formation and deformation characteristics
Period 2006 to 2007: Beginning of slope activity and formation of rock slide slabs A, B, C and E
In north direction, the extension of slab A is limited by a E–W-striking, steeply inclined fault crossing the slope at a height of about 3120 m a.s.l. As already mentioned in section 4, it is suggested that this persistent fault plane subdivides the rock slide body into two slabs: slab A, located upslope of this fault, and slab B downslope (Fig. 3). Due to the ice cover, the scarp of slab B is for the most part not directly exposed.
In addition to slabs A, B and C, a further displaced rock mass evolved at the NW slope around 3050–2930 m a.s.l. (Figs. 7a and 8). There, a new scarp represents the source of slab E which was formed shortly after the initiation of the rock slide during summer 2007. The spatial relationship between the location of the scarp and the area of debris accumulation directly below suggest that a rapid slope process, i.e. a rock avalanche, had occurred. Based on the elevation of the upper rim of the scarp (3050 m a.s.l), the distal-most deposits (2650 m a.s.l.) and the runout distance of 550 m, a runout travel angle of 36° was determined. In the scarp area of slab E, a rock mass height loss of up to 22 m was obtained from distance change analyses, coinciding with a similar amount of debris accumulation at lower slope areas (Fig. 7a). During this time period, ongoing rock/ice fall and debris flow events released from the fractured and disintegrated rock slide slabs A, B, and C, which led to considerable talus depositions of several tens of metres at the foot of the N and NW slope (2740 to 2930 m a.s.l.).
Period 2007 to 2008: Ongoing slope displacements
The period between 2007 and 2008 is characterised by ongoing displacements of the slabs A, B and C (Figs. 7b and 9). The activity of slab A decreased significantly, reaching displacements of 40 m in summer 2007 to 4 m between 2007 and 2008. For slab B, a displacement of about 4 m was determined. At the toe of slab C, displacements of up to 43 m were measured. The N-directed slope displacements of slab C caused a lowering of the terrain surface by more than 20 m. From here towards N, the subsurface is covered by the glacier and affected by terrain uplift of up to 15 m (Fig. 7b).
Period 2008 to 2009: Ongoing displacements, formation of slab D
Generally, during the period 2008–2009, the Bliggspitze rock slide was characterised by decreasing slope displacements of the pre-existing slabs (Fig. 9), but also by the formation of a new slab, termed slab D (Figs. 7c and 8). Field observations, webcam time-lapse images, ALS distance maps and displacement analyses based on image correlation and break line tracking indicate ongoing slope deformations and terrain surface lowering of all three slabs A, B and C (Fig. 7c). A terrain surface lowering of less than 10 m was observed between the head scarp of slab A and the scarp of slab C. Given that most of this area was covered by the glacier, it remained unclear if the terrain surface lowering was caused by rock slide deformations only or additionally by the degradation of the ice cover. During the period 2008–2009, slab C is characterised by N-directed displacements of about 35 m, inducing a terrain surface lowering of more than 10 m (Fig. 7c).
Period 2010 to 2012: Ongoing displacements, initial formation of the soil slide F
Period 2010–2012 was characterised by ongoing slope deformations, primarily by large displacements of slab C with around 31 m at the slope toe (Fig. 9). At some locations, slope displacements caused terrain surface lowering of up to 15 m and terrain uplift of about 10 m (Figs. 7d and 8). Slab A and slab B were further affected by N-directed displacements, but at a much lower activity, i.e. displacement rates are between 0.2 and 0.4 m/a. Little activity of 0.15 m/a was also determined by ALS image correlation for slab D. Since the initial development of the Bliggspitze rock slide and during this period, material from other sections of the NW slope and scarp area of slab D were released and accumulated as talus deposits further down. The ongoing debris deposition led to a progressive steepening of the slope until limit-equilibrium condition was reached, and a new soil slide was formed with a scarp at and below 3000 m a.s.l. (soil slide F, Figs. 7d and 11b). From the scarp to the toe, the soil slide is characterised by a length of approx. 450 m and a maximum width of 130 m, respectively. ALS-based distance change analyses indicate that the scrap area was affected by a terrain surface lowering of about 3 m. At the toe, material accumulation of 3 m represented by terrain surface uplift was measured (Fig. 7d). The average slope angles are approx. 38° in the scarp and 35° in the toe area. Between 2010 and 2012, displacements of about 20 m were measured at the central part of the soil slide.
Period 2014 and 2016: Ongoing displacements, formation of slab G
The deformation characteristics of this period were obtained from TLS measurements performed at the foot of the slope. Displacement analyses, i.e. image correlation and break-line tracking on terrain surfaces outside of the scan shadow, show a minor slope activity of slab A and B of about 0.2 m/a. At the left flank of slab C, a new detachment area was detected, characterised by the shape of a scarp and a terrain surface lowering of up to 10 m (Figs. 7e and 8). This scarp indicates the source of a new event (i.e. slab G), and is approx. 50 m high and 70 m wide, respectively. The scarp area of slab G is located directly above a distinct accumulation area, which in turn is approx. 110 m long and 80 m wide and is showing up to 3 m of debris accumulation (Fig. 7e). The spatial relationship between the location of the scarp and the deposition area on the glacier below indicates the evolution of slab G, which represents a rock avalanche. The most conspicuous area of slope activity during the period 2014–2016 is related to the soil slide F, which is still moving towards NW (Fig. 7e). The scarp area further extended and the toe of the slide moved more than 100 m towards NW. These large displacements led to a reduction of the average slope angle at the toe from 35° at the previous period to approx. 29°. At the toe, the thickness of the soil slide increased to a maximum of 8 m. The shape of the soil slide surface is lobate and tongue-like, where an enlargement of the mass from 120 m to 150 m along the displacement path has occurred. The soil slide shows an average displacement rate of about 20 m/a.
Geological-geometrical rock slide model
The Bliggspitze rock slide comprises a maximum thickness of approx. 65 m, a length in the N–S direction of more than 530 m and a width of approx. 300 m (Fig. 3). A volume determination based on the DEM from 2006 and the reconstruction of the topography without the rock slide mass (i.e. topography of the basal shear zone) show a total volume in the range between 3.9 to 4.3 million m3. The large-scale slabs, i.e. A, B, C and D, comprise volumes in the range from several hundreds of thousands to several millions of m3. The extremely rapid rock avalanches originating from the scarps E and G are much smaller and show volumes of approx. 70,000 and 15,000 m3, respectively. The slowly moving soil slide F comprises a volume of approx. 100,000 m3. The intensive fracturing of the rock slide mass during the sliding process, and thus the associated long-term activity of rock falls, rock avalanches, debris flows and rock slides, caused considerable mass waste characterised by the deposition of loose soil at lower elevations. A comparison of DEMs measured before and after the development of the Bliggspitze rock slide reveals a mass loss of more than 0.9 million m3 from the rock slide area. In contrast, the deposition area outside of the slide boundary, i.e. mostly below the basal shear zone, comprises a volume increase by approx. 1.3 million m3. Given that the rock slide is composed of both fractured rock mass and glacial ice, the DEM-based GIS calculations cannot distinguish between those two features.
For prognoses of the slope deformation behaviour and hazard assessments, it is crucial to assess, if a fully persistent basal shear zone has been formed and which geometry and volume is resulting from the initial rock slide failure process. The volume of a rock slide, when collapsing rapidly, affects the runout length, as shown by several compiled data and empirical relationships obtained thereof (e.g. Heim 1932; Scheidegger 1973; Körner 1976; Okura et al. 2000; Legros 2002; Prager 2010).
In contrast to major single events, the Bliggspitze rock slide shows a spatially and temporally complex failure and deformation process characterised by the sequential development and movement of individual rock slide slabs. The rock slide slabs were formed laterally or vertically (on top of each other) adjacent, and when active were displaced at different velocities (Fig. 8). For most of the rock slide slabs, especially for the larger one, displacements of several tens of metres were measured, but for these slabs, no accelerations with total collapse were observed. However, a few smaller rock slide slabs featuring volumes of up to several 10,000 m3 failed extremely rapid and formed rock/ice avalanches and rock falls.
The impact of permafrost degradation on the evolution of the Bliggspitze rock slide is only poorly understood. Nevertheless, clear indications for the existence of permafrost were found either by visual inspection of ice-filled fractures at the head scarp or by applying the BTS method. From literature, it is known that the presence of ice-bonded discontinuities generally increases slope stability. For example, Davies et al. (2001) showed by direct shear tests in laboratory that the shear strength of ice is usually higher at lower temperatures and decreases under warming conditions. Thus, ice-filled fractures would increase the overall rock mass strength and would reduce sub-critical crack growth (Atkinson 1989), i.e. a process which is required for coalescence (linkage) of the fracture network and the generation of a continuous basal shear zone. Furthermore, Krautblatter et al. (2013) proposed a time-dependent rock-ice mechanical model for ice-filled fractures, which considers (i) secondary creep of ice, (ii) failure of rock–ice contacts, (iii) friction of rock–rock contacts of discontinuities, and (iv) fracturing of intact rock bridges. Concerning this model, ice-fillings of fractures can act as a nonlinear viscous material, whereby the shear strain rate increases by an order of a magnitude when the temperature rises from −10 to 0 °C. Furthermore, laboratory tests indicate that failure of rock–ice contacts is favoured by warming, i.e. temperature around 0 °C and above, rather than low freezing temperatures. For warming permafrost rocks close to thawing, Krautblatter et al. (2013) suppose that (i) ice-mechanical processes are dominating rock slope failures at shallow depths (less than 20 m), and (ii) rock-mechanical processes are more relevant for deep-seated rock slope failures. Additional laboratory tests focusing on water-saturated intact rock samples showed a reduction of the compressive and tensile strength by rock temperatures increasing from low freezing temperatures up to 0 °C (Kodama et al. 2013). Further mechanisms related to permafrost such as ice segregation may cause growth of continuous ice-filled cracks which in turn may trigger fracture widening and thus progressive rock mass loosening. Subsequent thawing conditions (e.g. freeze-thaw cycles in warming permafrost and periglacial environments) can substantially affect the rock mass strength. Taking these factors into account, an impact of permafrost on the Bliggspitze rock slide is likely. According to literature data above, we assume that permafrost degradation is more relevant for near-surface slope destabilisation processes (i.e. active layer) than at greater depths of some decametres where the basal rockslide shear zone is located.
Generally, it is widely accepted that glaciation and deglaciation have an important impact on slope stability of large valley flanks in the Alps. However, it is controversially discussed which processes or combinations/interactions of processes lead to slope failure and particularly which time-dependent rock strength weakening mechanisms are relevant. On the one hand, after LGM, long lag times of thousands of years are observed between deglaciation and large-scale slope instabilities (e.g. Prager et al. 2008; Ballantyne et al. 2014). Such large time delays between glacial retreat and slope failures are driven by progressive rock mass failure processes and thus cannot be modelled by simple time-independent geomechanical models (Riva et al. 2018). On the other hand, numerous case studies clearly show a temporal relationship with much shorter lag times (much less than decades) between glacial retreat and the evolution of rock slides (McColl and Davies 2013; Kos et al. 2016; Fey et al. 2017b). Although a much shorter time delay exists, progressive failure processes accompanied by strain localisation are also here essential to form a fully persistent failure surface. Glacial retreat induces in rock slopes complex coupled mechanical, hydrogeological and thermal processes, including stress changes and weight relief, loss of rock mass strength, rock mass fracturing and weathering, temperature changes and permafrost degradation, and modifications of the groundwater flow (Grämiger et al. 2017).
At the Bliggspitze rock slide, we assume a major impact on slope stability from glacial retreat during recent decades. As a mechanism, we infer that glacial unloading due to ice weight reduction rather than debuttressing, which typically occurs on slopes interacting with a large valley glacier, are relevant (Ballantyne 2002). Glacial unloading taking place predominately at lower parts of the slide may trigger force imbalances, which in turn cause slope deformations and the growth and linkage of fractures until a fully persistent shear zone is formed. GIS-based volume analyses indicate a total loss of glacial ice atop the rock slide of more than 1.5 million m3 between 1946 and 2006. Depending on this ice volume, more than 1 million m3 melted at the lower part and near the foot of the rock slide. Considering an average glacier ice density of 900 kg/m3, an ice mass reduction of 900 million kg resulted (Huss 2013). If the ice mass reduction is related to the total rock slide volume ranging from 3.9 to 4.3 million m3 and the rock mass density of 2300 kg/m3 (due to rock mass fracturing and loosening 10% porosity was assumed), a total mass reduction at the foot of the slide between 9 and 10% was obtained. The unloading of the rock slide due to the glacial retreat is remarkable and may have influenced slope stability, especially because the basal shear zone is curved and a rotational sliding mechanism is determined. Although not quantified so far by comprehensive numerical modelling in this study, it is presumed that the glacial retreat has a major impact on the overall in situ stress state of the slope, and thus may have influenced the failure and formation process of the Bliggspitze rock slide. In order to study and quantify these factors in greater detail, and additional potential trigger and stabilisation factors, coupled 2D/3D numerical modelling will be applied in the framework of subsequent research. On the basis of the findings and models from this study, numerical modelling will be focussing on the impact of (i) glacier unloading during recent decades, (ii) mass waste due to slope processes, and (iii) permafrost degradation on the initial failure process and the future stability evolution.
The Bliggspitze rock slide acted as a challenging site to test and validate different remote sensing techniques, i.e. ALS, TLS, ortho-images and time-lapse of webcam images in remote and inaccessible terrain influenced by permafrost and glacial retreat. Experience has demonstrated that ALS/TLS data from different campaigns are useful to obtain spatially and temporally variable slope deformation data. Such surface deformation maps, analysed and assessed by geological field surveys, enabled the detection and monitoring of different rock/soil slide slabs and the development of a geometrical-kinematical rock slide model. During the initial development stage and the long-term deformation history of the rock slide, three different slab types characterised by different volumes and deformation characteristics were determined. Firstly, large-scale slabs composed of fractured rock mass which are moving at low to moderate velocities were observed (i.e. slabs A, B, C and D). These slabs did not accelerate exponentially or even fail extremely rapidly, but rather reduced the displacement rates after some time (rock slab stabilisation mechanism). Secondly, small-scale slabs composed of fractured rock mass were identified showing continuous acceleration and slope failure as rock falls and rock avalanches (slabs E and G). Thirdly, a slowly-moving soil slide originating from a newly produced talus accumulation area at the foot of the slope (soil slide F) was recognised. Furthermore, long-term observations at the Bliggspitze rock slide demonstrate a complex and still ongoing loss of rock/soil of the slide mass through rock falls, rock avalanches, rock slides and debris flows, and the accumulation of the material at lower heights.
Open access funding provided by the University of Natural Resources and Life Sciences Vienna (BOKU). The authors wish to acknowledge and thank the TIWAG and the Federal State of Tyrol for providing remote sensing and monitoring data and supporting this research project. We would also like to thank Ass. Prof. Karl Krainer for supporting us with the BTS measurements in the field. This study was part of several alpS landslide projects, which were supported by TIWAG, ILF Consulting Engineers, geo.zt, and the Austrian Research Promotion Agency. The alpS K1-Centre was supported by the Federal Ministries BMVIT and BMWFW and by the States of Tyrol and Vorarlberg in the framework of “COMET - Competence Centres for Excellent Technologies”. COMET was funded and administrated by FFG.
- Atkinson BK (1989) Fracture mechanics of rock. Elsevier Science, Burlington, p 548Google Scholar
- Bortenschlager S (1984) Beitrage zur Vegetationsgeschichte Tirols I. Inneres Ötztal und unteres Inntal, Berichte Naturwiss-Mediz Vereins Innsbruck 71:19–56Google Scholar
- Cossart E, Braucher R, Fort M, Bourlès DL, Carcaillet J (2008) Slope instability in relation to glacial debuttressing in alpine areas (upper durance catchment, southeastern France): evidence from field data and 10Be cosmic ray exposure ages. Geomorphology 95:3–26. https://doi.org/10.1016/j.geomorph.2006.12.022 CrossRefGoogle Scholar
- Dejean de la Bâtie J (2016) Ingenieurgeologische und hydrogeologische Untersuchung der Massenbewegung im Bereich der Bliggspitze im Ötztal-Stubai-Kristallin (Kaunertal, Tirol). University of Innsbruck, Faculty of Geo- and Atmospheric Sciences, Innsbruck, Master Thesis 159 ppGoogle Scholar
- Deline P, Gruber S, Delaloye R, Fischer L, Geertsema M, Giardino M, Hasler A, Kirkbride M, Krautblatter M, Magnin F, McColl S, Ravanel L, Schoeneich P (2015) Ice loss and slope stability in High-Mountain regions. In: Snow and ice-related hazards, risks and disasters. Elsevier, pp 521–561Google Scholar
- Dewez TBJ, Girardeau-Montaut D, Allanic C, Rohmer J (2016) Facets: a cloud compare plugin to extract geological planes from unstructured 3d point clouds. Int Arch Photogramm Remote Sens Spatial Inf Sci XLI-B5:799–804. https://doi.org/10.5194/isprsarchives-XLI-B5-799-2016 CrossRefGoogle Scholar
- Fey C (2018) Remote sensing applications for deformation monitoring of landslides in alpine environment Doctoral thesis, 152 pp.Google Scholar
- Fey C, Wichmann V, Rett H, Perzlmaier S (2017a) Multi-temporal long-range laser scanning for spatial deformation monitoring of alpine slopes. In: Hanke K, Weinold T (eds) 19. Internationale geodätische Woche Obergurgl 2017, 1.Neuerscheinung. Wichmann, H; Wichmann Verlag, Berlin, pp. 247–252Google Scholar
- Gischig VS (2011) Kinematics and failure mechanisms of the Randa rock slope instability (Switzerland). Doctoral thesis, Nr. 19730, ETH, Zürich, 204 pp.Google Scholar
- Haeberli W (1973) Die Basis-Temperatur der winterlichen Schneedecke als möglicher Indikator für die Verbreitung von Permafrost in den Alpen. Z Gletscherk Glazialgeol 9:221–227Google Scholar
- Hammer W (1922) Geologischer Führer durch die Westtiroler Zentralalpen, Sam. geol. Führer XXII. Gebrüder Bornträger, Wien, 168 pp.Google Scholar
- Heim A (1932) Bergsturz und Menschenleben. Beiblatt zur Vierteljahrsschrift der Naturforschenden Gesellschaft in Zürich. Fretz & Wasmuth, 218 pp.Google Scholar
- Kerschner H (1979) Spätglaziale Gletscherstände im inneren Kaunertal (Ötztaler Alpen). Innsbrucker Geograph Studien (Leidlmair Festschrift II) 6:235–247Google Scholar
- Kerschner H, Nicolussi K (2014) Lateglacial and Holocene advance record of the Gepatschferner, Kaunertal, Tyrol. In: Kerschner H, Krainer K, Spötl C (eds) From the foreland to the Central Alps: field trips to selected sites of quaternary research in the Tyrolean and Bavarian Alps. Geozon, Berlin, pp 144–151CrossRefGoogle Scholar
- Keuschnig M (2016) Long term monitoring of permafrost affected rock walls. Technische Universität München, Ingenieurfakultät Bau Geo Umwelt, München, 117 pp.Google Scholar
- Nicolussi K, Patzelt G (2000) Untersuchungen zur Holozänen Gletscherentwicklung von Pasterze und Gepatschferner (Ostalpen). Z Gletscherk Glazialgeol 36:1–87Google Scholar
- Paranunzio R, Laio F, Chiarle M, Nigrelli G, Guzzetti F (2016) Climate anomalies associated to the occurrence of rockfalls at high-elevation in the Italian Alps. Nat Hazards Earth Syst Sci Discuss:1–24. https://doi.org/10.5194/nhess-2016-100
- Patzelt G (1972) Die spätglazialen Stadien und postglazialen Schwankungen von Ostalpengletschern. Ber D Dt Bot Ges 85:47–57Google Scholar
- Patzelt G (1977) Der zeitliche Ablauf und das Ausmaß postglazialer Klimaschwankungen in den Alpen, In: Dendrochronologie und postglaziale Klimaschwankungen in Europa Frenzel, B., Steiner (eds), Wiesbaden, 248–259Google Scholar
- Piccolruaz C (2004) Zur Quartärgeologie des Kaunertals im Bereich östlich des Gepatsch Stausees in den westlichen Ötztaler Alpen (Tirol) unter besonderer Berücksichtigung der Blockgletscher. Diplomarbeit, University of Innsbruck, Faculty of Geo- and Atmospheric Sciences, InnsbruckGoogle Scholar
- Prager, C. (2010) Geologie, Alter und Struktur des Fernpass Bergsturzes und tiefgründiger Massenbewegungen in seiner Umgebung (Tirol, Österreich). Dissertation, Universität Innsbruck. 309 pp.Google Scholar
- Purtscheller F (1987) Ötztaler und Stubaier Alpen, Sammlung geologischer Führer, 53. Gebr. Bornträger, Stuttgart, 128 pp.Google Scholar
- Reitner JM, Linner M (2009) Formation and preservation of large scale toppling related to alpine tectonic structures eastern Alps. Aust J Earth Sci 102:69–80Google Scholar
- Riegl 2013. Operating and Processing Software RiScan Pro. RIEGL LMSGoogle Scholar
- RVS 11.062 (1987) Blatt 10 & 11 - Prüfverfahren - Steinmaterial, Semiquantitative röntgenographische Mineralanalyse. RDA, Vienna, p 4Google Scholar
- Strauhal T (2015) Geological, hydrogeological and geotechnical characterisation of deep-seated rockslide in metamorphic rocks. Doctoral thesis, Institute of Geology, University of Innsbruck, 237 pp.Google Scholar
- Strauhal T, Zangerl C, Fellin W, Holzmann M, Engl DA, Brandner R, Tropper P, Tessadri R (2017) Structure, mineralogy and Geomechanical properties of shear zones of deep-seated rockslides in metamorphic rocks (Tyrol, Austria). Rock Mech Rock Eng 50:419–438. https://doi.org/10.1007/s00603-016-1113-y CrossRefGoogle Scholar
- van Husen D (1987) Die Ostalpen in den Einszeiten. – Die Ostalpen und ihr Vorland in der Letzten Eiszeit (Würm). Geol. Bundesanstalt, WienGoogle Scholar
- Weirich J, Bortenschlager S (1980) Beitrage zur Vegetationsgeschichte Tirols III: Stubaier Alpen - Zillertaler Alpen, Berichte naturwiss.-mediz. Vereins Innsbruck, vol 67, pp 7–30Google Scholar
- Willenberg H, Loew S, Eberhardt E, Evans KF, Spillmann T, Heincke B, Maurer H, Green AG (2008) Internal structure and deformation of an unstable crystalline rock mass above Randa (Switzerland): part I - internal structure from integrated geological and geophysical investigations. Eng Geol 101:1–14CrossRefGoogle Scholar
- Zangerl C, Prager C (2008) Influence of geological structures on failure initiation, internal deformation and kinematics of rockslides-ARMA 08–063. Proceedings 42nd U.S. / 2nd U.S.-Canada Rock Mechanics Symposium, San Francisco. 13 pp.Google Scholar
- Zangerl C, Prager C, Engl D (2010) Self-stabilisation mechanisms of slow rock slides in crystalline bedrock (Tyrol, Austria). In: Williams AL (ed) Geologically active: proceedings of the 11th IAEG congress, Auckland, New Zealand, 5–10 September 2010. CRC Press, Boca Raton, pp 903–910Google Scholar
- Zischinsky U (1966) On the deformation of high slopes. Proceedings of the first congress of the Int. Soc Rock Mech 6:179–185Google Scholar
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.