Cycles of glaciation impose mechanical stresses on underlying bedrock as glaciers advance, erode, and retreat. Fracture initiation and propagation constitute rock mass damage and act as preparatory factors for slope failures; however, the mechanics of paraglacial rock slope damage remain poorly characterized. Using conceptual numerical models closely based on the Aletsch Glacier region of Switzerland, we explore how in situ stress changes associated with fluctuating ice thickness can drive progressive rock mass failure preparing future slope instabilities. Our simulations reveal that glacial cycles as purely mechanical loading and unloading phenomena produce relatively limited new damage. However, ice fluctuations can increase the criticality of fractures in adjacent slopes, which may in turn increase the efficacy of fatigue processes. Bedrock erosion during glaciation promotes significant new damage during first deglaciation. An already weakened rock slope is more susceptible to damage from glacier loading and unloading and may fail completely. We find that damage kinematics are controlled by discontinuity geometry and the relative position of the glacier; ice advance and retreat both generate damage. We correlate model results with mapped landslides around the Great Aletsch Glacier. Our result that most damage occurs during first deglaciation agrees with the relative age of the majority of identified landslides. The kinematics and dimensions of a slope failure produced in our models are also in good agreement with characteristics of instabilities observed in the field. Our results extend simplified assumptions of glacial debuttressing, demonstrating in detail how cycles of ice loading, erosion, and unloading drive paraglacial rock slope damage.

Key Points

Simply adding then removing glacier ice from an alpine valley has little net effect on rock wall damage
Glacial erosion, i.e., rock debuttressing, creates significant new rock slope damage during first deglaciation
Damage kinematics vary during a glacial cycle: ice advance favors toppling, while retreat promotes sliding

1 Introduction

Glacier advance and retreat imposes mechanical stress cycles on underlying bedrock and alters stress trajectories in adjacent valley slopes. The rock slope response to these stress changes will vary with in situ stress conditions, rock mass properties, valley geometry, and local environment [Augustinus, 1995; Ballantyne, 2002; McColl, 2012]; however, induced mechanical stress changes can generate both elastic (e.g., glacioisostatic rebound) and inelastic (i.e., irreversible) deformations. Creation of new fractures, propagation of slip along existing joints, and failure of intact rock bridges connecting nonpersistent discontinuities constitutes rock slope damage and rock mass strength degradation, conditioning slope instability and preparing paraglacial valley walls for failure [Terzaghi, 1962; Eberhardt et al., 2004; McColl, 2012].

In most previous studies, the mechanical reasoning explaining the temporal and spatial distributions of paraglacial (in the sense of Slaymaker [2009]) slope failures remains vague. Some studies relate slope failure activity with confinement loss due to glacier retreat [Cossart et al., 2008; Deline, 2009], while others emphasize the role of stress redistribution induced by valley erosion [Augustinus, 1995; Leith et al., 2014a] or glacial rebound and uplift [Cossart et al., 2014; Ballantyne et al., 2014a, 2014b]. Slope debuttressing associated with deglaciation (i.e., removal of an ice buttress) is often assumed to be the predominant cause of postglacial alpine slope failures [e.g., Bovis, 1990; Cossart et al., 2008; Jaboyedoff et al., 2012], and while spatial correlation of landslides with glacial debuttressing patterns could be identified in some studies [Holm et al., 2004; Cossart et al., 2008], it was not evident in others [Cossart et al., 2014]. McColl et al. [2010] questioned the mechanical reasoning behind glacial debuttressing, pointing out that at long time scales (more than tens of years) and small strain rates (<10⁻³ s⁻¹ [Schulson, 1990]), ice behaves in a ductile manner, thereby loading underlying bedrock by its weight alone and not providing significant rigid lateral support to adjacent rock slopes. Field evidence of squeezed glaciers adjacent to active slope failures supports this hypothesis [McColl and Davies, 2013]. Meanwhile, frequently observed large lag times between deglaciation and large-scale slope instability [e.g., Prager et al., 2008; Ivy-Ochs et al., 2009a; Ballantyne et al., 2014a, 2014b; Zerathe et al., 2014] cast doubt on the importance of glacial debuttressing as a direct failure trigger and point to the need to further understand time-dependent effects. These results underscore the importance of additional research into the mechanics of the paraglacial rock slope response and specifically how stress cycles imposed by the changing weight of glaciers generate rock mass damage as a first-order control on slope failure processes.

Unique measurements seek to unravel the forces acting at the rock-ice interface beneath a glacier over time. Ongoing elastic rebound of bedrock on the margins of an actively retreating ice sheet is currently monitored in Greenland by using GPS [Khan et al., 2010; Bevis et al., 2012], showing nominal uplift rates in the range of a few mm yr⁻¹. In situ stress measurements at an ice/bedrock contact are rare, and only a few measurements of ice pressure on bedrock exist from subglacial laboratories showing a wide range of measured normal stresses, which is attributed to disturbances of the local ice flow, with values commonly close to the hydrostatic pressure of the ice overburden [Hagen et al., 1993; Cohen et al., 2000; Cohen et al., 2005; Cohen et al., 2006]. Measured basal shear stress reached up to 350 kPa with 210 m ice overburden [Cohen et al., 2000]. Till deformation beneath glaciers has been studied in detail [Alley et al., 1986; Iverson et al., 1999; Iverson et al., 2007] but rarely reported are deformation measurements in subglacial bedrock. In one study, strainmeters in tunnels reaching within 10 m of the ice/rock interface below an alpine glacier detected small elastic strain excursions perpendicular to glacier flow, which were believed to originate from changes in surface traction between the glacier and its bed due to frozen patches at the bedrock interface [Goodman et al., 1979].

Increasingly detailed attempts to quantify the bedrock response to glacial cycles have been accomplished through numerical modeling studies. Ustaszewksi et al. [2008] studied fault slip induced by postglacial rebound in a Swiss alpine valley and connected the formation of uphill-facing scarps with postglacial unloading. In the context of hazard assessment for a deep repository, the role of coupled thermo- and hydromechanical bedrock responses to a glacial cycle (ice sheet) has been previously studied [Boulton et al., 2004; Chan et al., 2005; Vidstrand et al., 2008; Selvadurai et al., 2015]. However, such studies lack holistic treatment coupling an alpine valley glacier with adjacent rock slopes. Several site-specific modeling studies have been conducted by investigating the evolution of rock slopes undergoing glacier retreat [e.g., Eberhardt et al., 2004; Fischer et al., 2010; Jaboyedoff et al., 2012; Agliardi and Crosta, 2014], using both continuum and discontinuum approaches. In all of these studies, however, glacier ice was implemented as a rigid, elastic material, which is most likely inappropriate when considering long time scales (more than tens of years) and small strain rates (<10⁻³ s⁻¹) [McColl et al., 2010; McColl and Davies, 2013]. Furthermore, these studies simply model multiple stages of glacier retreat, ignoring past glacial cycles and inherited rock mass damage. Leith et al. [2014a], on the other hand, examined the path dependency of subglacial fracturing in a detailed manner by using a continuum modeling approach, simulating alpine valley evolution over Pleistocene time scales.

In this study, we seek to clarify the mechanics of how cyclic stress changes associated with glacial cycles create rock mass damage in adjacent valley slopes, helping prepare these slopes for future failure. Our investigation spanning glacial time scales necessitates a modeling study parameterized and tested against conclusions from present-day observations. We present a new numerical modeling framework, founded in extensive field mapping and characterization at our Aletsch Valley study site, Switzerland, and implemented in a distinct element code. Coupled thermo- and hydromechanical effects are detailed in the following companion studies. Here we show how simple stress changes associated with repeat glacier advance and retreat cycles may propagate fractures, enhance slip along discontinuities, and lead to failure of intact rock bridges; all mechanisms result in time-dependent rock mass strength degradation and preparation of rock slope failures. We describe spatial and temporal damage patterns, stress redistribution, and displacement associated with Late Pleistocene and Holocene glacial cycles and compare numerical predictions with spatial and temporal landslide patterns in the Aletsch area (Figure 1). Our results help quantify the mechanical role of glacier advance, retreat, and erosion cycles as a preparatory factor for rock slope instabilities and further improve the mechanical understanding underlying development of paraglacial slope failures.

Details are in the caption following the image — **Figure 1**
Open in figure viewer PowerPoint

Concentration of rock slope instabilities around the glacier terminus and near LIA extents. The lower case labels correspond to mapped extents in Figure 2. (a) Tälli instability with Driest instability at the western valley flank at the junction of Upper Aletsch Glacier and Great Aletsch Glacier. (b) Historic photos of Tälli from 1967 and 1975 reproduced from *Kasser et al.* [1982]. (c) Moosfluh instability and Silbersand instability on the eastern valley flank. (d) Zoom of the Silbersand instability. (e) Small rock slope instability at the crest with debris cones truncated at LIA extents and a possible relict landslide being revealed by present-day ice retreat. (f) Head scarp of the Moosfluh instability. (g) Uphill facing scarps with offset Egesen moraine at Galkina.

2 Paraglacial Setting of the Aletsch Region

2.1 Study Site

Our study area encompasses the Aletsch Glacier region of Switzerland, including the Upper-, Middle-, and Great Aletsch Glaciers (Figure 2), and focuses primarily on the Great Aletsch Glacier and surrounding valley rock slopes. The Great Aletsch Glacier is the largest and longest glacier in Europe, extending nearly 22 km through the heart of the Central Alps. It is thickest at Concordia, where the depth to bedrock has been measured at over 900 m [Hock et al., 1999]. This study area was selected based upon a number of factors, including large ice volume changes over time affecting well-mapped and relatively homogenous bedrock, well-established spatial and temporal extents of the glacial record, and steep valley rock slopes prone to instability (Figure 1), some of which have been previously monitored in detail [e.g., Strozzi et al., 2010; Kos et al., 2016; Loew et al., 2017]. The study area is situated in the Aar Massif, the largest external crystalline massif in the Central Alps. Lithologies around the Great Aletsch Glacier consist primarily of gneisses of the metamorphic Altkristallin near the glacier terminus and the Central Aare granites in the upper parts of the study area [Steck, 2011].

2.2 Late Pleistocene and Holocene Glaciation

Detailed information describing Late Glacial and Holocene glacier extents exists for the Aletsch region. Late Glacial moraines and Little Ice Age (LIA) maxima are well preserved and documented, providing information on spatial extents, while absolute ages of different glacial features in the study area constrain the timing of several stadia. We compiled and combined glacial extents from available literature [Landestopographie and VAW, 1962; Holzhauser, 1995; Kelly et al., 2004a] with our own mapping of moraines and trimlines by using available aerial photography and LiDAR digital elevation model (DEM) data and confirmed findings through field inspection. Figure 2 displays a synopsis of the Late Glacial and Holocene glacier extents in the Aletsch region, while Figure 3 illustrates the change in glacier length over time.

Previous studies have suggested that the deep trough form of major Alpine valleys was initially carved around the Mid-Pleistocene revolution (~0.9 Ma), during the onset of the first major Pleistocene glaciation in the Alps [Muttoni et al., 2003; Haeuselmann et al., 2007; Leith et al., 2014a]. Several glacial/interglacial cycles since that time helped revitalize and maintain these characteristic glacial trough valleys. The Aletsch region was most likely ice-free during the penultimate Eemian interglacial period (~130 to ~115 kyr [Dahl-Jensen et al., 2013]) (Figure 3). Eemian climate was likely warmer than the Holocene [Dahl-Jensen et al., 2013], and therefore, we assume that ice abandoned the Aletsch area completely. The last glacial period (Würmian) peaked at the Last Glacial Maximum (LGM) [Ivy-Ochs et al., 2008]. The LGM in the Alps is dated at ~28 to 18 kyr [Ivy-Ochs et al., 2008; Ivy-Ochs, 2015] (Figure 3), and ice extents have been mapped by Kelly et al. [2004b]. The LGM glacier system had extensively retreated by ~19 to 18 kyr [Ivy-Ochs, 2015]. Between ~17 and ~11 kyr, a series of successive Late Glacial readvances occurred, termed (from oldest to youngest) Gschnitz, Clavadel, Daun, and Egesen stadia (Figure 3) [Maisch et al., 1999; Ivy-Ochs et al., 2008; Darnault et al., 2012]. In the Aletsch region, the elevation of LGM ice is visible through trimlines (Figure 2). During the Oldest Dryas (Gschnitz, Clavadel, and Daun stadia), ice elevations dropped steadily with several readvances but the Aletsch Glacier still flowed over the ridge above Bettmeralp toward the Rhone Valley, covering rock slopes in our study area (Figure 2). Ice retreated significantly during the Bølling/Allerød interstadial [Ivy-Ochs et al., 2008] (Figure 3).

The onset of the Younger Dryas (YD) caused dramatic ice readvance in high Alpine valleys, sending the Aletsch Glacier several km downstream to the Rhone Valley at Brig [Kelly et al., 2004a and references therein] (Figure 3). Nested moraines of the Egesen stadial are well preserved today (Figure 2) and age constraints provided by surface exposure dating (¹⁰Be) [Kelly et al., 2004a; Schindelwig et al., 2012]. We recalculated published exposure ages by using the northeastern North America (NENA) production rate for cosmogenic ¹⁰Be of 3.88 ± 0.19 at. g⁻¹ yr⁻¹ [Balco et al., 2009] and a time-dependent spallation production model [Lal, 1991; Stone, 2000]. Recalculated ages of boulders from the left-lateral moraine (AG-1, 2, 4-WM, and 5 in Kelly et al. [2004a]) resulted in a mean age of 12.3 ± 0.9 kyr (Figure 3), while the recalculated mean age of glacially striated bedrock at the right-lateral moraine (VBA-8, 9, and 10 in Schindelwig et al. [2012]) was 13.7 ± 1.0 kyr (Figure 3). These ages match global timing for the YD of 12.8–11.5 kyr B.P. [Alley et al., 1993]. Boulder ages of the moraine system at the Unnerbäch cirque at Belalp (Figure 2) dated by Schindelwig et al. [2012] were also recalculated: the outer moraine (VBA-1 to 6, 11 to 16, and 22 to 26) has a mean age of 12.1 ± 0.9 kyr, while the age of the inner moraine (VBA-17 to 20) is 10.6 ± 0.8 kyr (Figure 3). Recalculated exposure ages show good fit with the YD for the outer moraine and may relate the inner moraine to the pre-Boreal Oscillation [Moran et al., 2016]. A single boulder (VBA-21) beyond these moraines at Belalp has a recalculated age of 14.2 ± 1.0 kyr (Figure 3), which may be related to the latest Late Glacial readvance (Daun: >14.7 kyr [Ivy-Ochs et al., 2008]). Local dated peat bog profiles by Welten [1982] provide complementary evidence of the timing of Egesen deglaciation (Figure 3).

Retreat of Egesen glaciers following the YD marked the onset of the Holocene, which saw a number of minor glacier fluctuations culminating in the Little Ice Age (LIA) around 1850 [Joerin et al., 2006; Ivy-Ochs et al., 2009b]. Early Holocene readvances in the Aletsch region are not well constrained, but Late Holocene (past ~3500 years) glacier fluctuations are revealed through radiocarbon-dated fossil tree trunks, which were overrun by advancing ice and exposed during later retreat [Holzhauser et al., 2005] (Figures 2 and 3). The LIA marks the Holocene glacial maximum; however, this extent may have been reached several times, creating compound moraines [Röthlisberger and Schneebeli, 1979; Schimmelpfennig et al., 2012] (Figures 2 and 3). Complementary studies at nearby sites [Hormes et al., 2001; Goehring et al., 2011; Luetscher et al., 2011], together with tree line variability in the Kauner Valley, Austria [Nicolussi and Patzelt, 2000], suggest that Alpine glacier extents during the Mid-Holocene were mostly smaller than today but interrupted by a few readvances not exceeding the LIA (e.g., 8.2 kyr event [Nicolussi and Schlüchter, 2012]). Holzhauser et al. [2005] postulated that during the Bronze Age Optimum (3350–3250 years ago) the Great Aletsch Glacier was approximately 1 km shorter than today. Combining this assumption with three-dimensional retreat models of the Great Aletsch Glacier [Jouvet et al., 2011] allows us to estimate a plausible Holocene minimum extent (Figures 2 and 3). However, we cannot exclude that glacier retreat during the Holocene Climatic Optimum (Figure 3) exceeded this minimum. These results suggest that bedrock above and outside of the LIA extents most likely experienced only a single glacier readvance (Egesen stadial) following LGM ice retreat, whereas rock slopes within and below the LIA extent were affected by five (or more) glacier cycles.

2.3 Spatial and Temporal Distributions of Paraglacial Rock Slope Instabilities

We mapped the distribution of landslides in the Aletsch region by using available aerial photography and LiDAR DEM data, as well as relevant literature, and complemented these findings with field inspection. To better understand links between deglaciation and initiation of landsliding, we also attempted to estimate or constrain initial failure ages of the mapped slope instabilities. Observations of glacially striated bedrock and/or offset moraines allowed relative age constraints for several mapped slope failures (e.g., post-LIA, pre-LIA/post-Egesen, and pre-Egesen/post-LGM), while absolute cosmogenic surface exposure dating using was applied at the Driest instability (Figure 2). The summarized spatial and temporal distributions of paraglacial rock slope instabilities, superimposed with the Late Glacial and Holocene ice extents, are displayed in Figure 2.

The largest landslide in our study area is the deep-seated gravitational slope deformation (DSGSD) along the western flank of the Rhone Valley, extending from Riederalp to Fiescheralp. This so-called Riederalp-Bettmeralp-Fiescheralp DSGSD (Figure 2a) is evident from morphological surface features [Crosta et al., 2013] but does not show signs of recent activity. Unusually high sedimentation rates in a peat bog within the landslide investigated by Welten [1982] (P5 in Figure 2) may indicate activity during the Mid-Holocene. This landslide does not directly affect the Aletsch Glacier system. Meanwhile, the Nessel rock slope instability (Figure 2b) is located at the lower end of the Aletsch Valley on the western valley flank. Unmodified landslide deposits within the Egesen extents indicate a post-Egesen failure age. The Belalp DSGSD (Figure 2c) is located on the same bench. Large graben structures indicate displacement, and the prominent, well-preserved Egesen moraine described by Schindelwig et al. [2012] is offset indicating post-Egesen activity. Geodetic measurements at Belalp and Nessel show recent displacement rates of a few mm yr⁻¹ at both sites [Glaus, 1992].

Higher in the Aletsch Valley, we mapped a concentration of large landslides around the present-day glacier terminus. The Moosfluh instability on the southeastern valley flank (Figures 1c and 1f and 2d) is an active toppling-mode landslide, described and monitored by Strozzi et al. [2010], Kos et al. [2016], and Loew et al. [2017]. Remote sensing reveals accelerated displacements over the past 20 years from ~4 cm yr⁻¹ to more than 30 cm yr⁻¹ [Strozzi et al., 2010; Kos et al., 2016]. Historic maps and orthoimages confirm the existence of the head scarp (Figure 1f) prior to this recent acceleration. We postulate a post-Egesen initialization age due to the discontinuity and deformation of the Egesen moraine in the landslide area (Figure 2d). The nearby Silbersand instability (Figures 1c and 1d and 2e) has a relict appearance: large displacement at the head scarp with a missing landslide deposit and glacial erosion features along the scarp indicate post-Egesen/pre-LIA age, although offset moraines from 1926/1927 on the landslide body reveal recent movement [Crisinel, 1978]. The Tälli instability is located at the intersection of the Upper Aletsch Glacier and the Great Aletsch Glacier (Figures 1a and 2f). Historic photographs (Figure 1b) and orthoimages (Swisstopo) limit the initial failure timing to 1965 or 1966, when the Great Aletsch Glacier was just retreating from the toe of the developing unstable slope [Kasser et al., 1982]. During the following ~5 years, the ice-free slope developed into a disaggregated body. The adjacent Driest instability (Figures 1a and 2g) is a compound rock slide that shows recent movement, visible by the freshly exposed band at the bottom of the head scarp and evaluated through remote monitoring [Kääb, 2002; Vogler, 2015; Kos et al., 2016]. The steep head scarp lies around 50 m below the Egesen moraine. A LIA moraine is clearly visible within the disaggregated landslide body, around 100 m below the head scarp. A secondary failure occurred after LIA retreat. To constrain the initial failure age of the Driest landslide, we extracted five bedrock samples along a vertical transect down the scarp for ¹⁰Be cosmogenic nuclide surface exposure dating (see Appendix A1). Exposure dates reveal an initiation age of 7.4 ± 0.7 kyr (Figure 4), not directly following LGM or Egesen ice retreat (lag time of several thousand years). Figures 3a and 3b illustrate the presumed interaction of major landslide activity in our study area with Late Glacial/Holocene glacier fluctuations.

Smaller rock slope instabilities are located at Hohbalm (Figures 1e and 2h), just above the Egesen moraine, resulting in debris cones cross-cutting the Egesen but not the LIA extents. Therefore, these instabilities must be post-Egesen/pre-LIA, although recent rockfall suggests ongoing activity. Areas above LGM trimlines are strongly weathered with a broken rock mass structure in the near surface (especially the region between Bettmerhorn and Eggishorn; Figure 2), being exposed to weathering processes over the entire last glacial period. Rockfall from some of these regions is the source for several rock glaciers in the study area. We also observed the head scarp of a possible relict landslide being revealed by recent glacier retreat (Figures 1e and 2i), suggesting that relict landslides from previous retreat episodes may be hidden beneath the present glacier. Prominent uphill-facing scarps at Galkina (Figures 1g and 2k) are described by Eckardt et al. [1983]. Our on-site investigations show relative scarp displacements up to 4 m, and tilted Egesen moraines constrain this activity to post-Egesen/pre-LIA. Despite observations and structural measurements in the field, it remains unclear whether the uphill-facing scarps are part of a localized instability with toppling style kinematics or whether they represent regional large-scale differential uplift [Ustaszewksi et al., 2008].

3 Rock Mass Characterization and Structural Analysis

In order to create an adequate geological model representing the Aletsch Valley, we performed field-based rock mass characterization. The following rock mass parameters were collected on a regional scale: Geological Strength Index (GSI), Schmidt hammer hardness, and joint properties including dip/dip direction, spacing, persistence, and Joint Roughness Coefficient (JRC). Orthoimages and a LiDAR DTM were used to map large-scale lineaments (i.e., faults) across the study area (Figure 5), and orientations were determined by fitting a plane through the trace data by using linear regression. The results of our structural analysis are summarized in Figure 5. Three main joint sets were observed with the following average properties across all measurements:

F1 follows foliation, dipping steep to subvertical toward SE (dip direction/dip: 122°/76°). Joints are persistent (trace length of 3–10 m), with spacing in the range of cm. This is the most abundant joint set, correlating with the alpine foliation in the Aar Massif [Steck, 2011]. On occasion, foliation is divided into two subsets: F1a (138°/77°) and F1b (101°/78°), whereas F1a is more abundant.
F2 dips very steep to subvertical toward S to SW (198°/83°). Joints have minor persistence (1–3 m) and medium spacing (0.5–2 m).
F3 dips gently SW (240°/20°), although the orientation can vary. Joints are moderately persistent (3–5 m) and exhibit larger spacing (1–5 m).
F4 includes large-scale lineaments (i.e., faults), which in general follow foliation (F1). Minor lineaments may be oriented W-E or N-S, following the subsets F1a or F1b. Field observation correlates these lineaments with fault zones, which typically exhibit strong internal foliation and fracturing due to tectonic shearing and rarely contain gouge.

GSI [Hoek and Brown, 1997] was measured across the study area and found to be relatively uniform between lithologies. The rock mass is generally blocky, undisturbed with good surface conditions (GSI 65–80). We evaluated detailed joint properties at several outcrops within the Augen gneiss at Bettmerhorn and Chatzulecher (Figures 5k and 5i). Mean measurement values are given together with one standard deviation and the number of measurements (#):

Spacing was measured at Bettmerhorn and Chatzulecher: F1 = 0.4 ± 0.3 m (#68), F2 = 0.5 ± 0.3 m (#52), and F3 = 0.6 ± 0.4 m (#32). Since the upper boundary of the spacing was often not captured due to the scale of the outcrop, these measurements likely represent lower limits. Additional field observations were therefore used for global description.
Schmidt hammer rebound hardness was evaluated by using an L-type instrument and processed according to Aydin [2015] at Bettmerhorn for joint sets: F1 = 48 ± 10 (#53), F2 = 52 ± 11 (#41), and F3 = 51 ± 5 (#31). The estimated joint wall compressive strengths from those rebound values are F1 = 90–200 MPa, F2 = 90–270 MPa, and F3 = 130–200 MPa.
JRC was evaluated at Bettmerhorn and resulted in relatively smooth values: F1 = 6.1 ± 3.1 (#10), F2 = 6.1 ± 2.8 (#7), and F3 = 6.9 ± 3.5 (#11).

Table 1 summarizes discontinuity characterization along the cross section between Bettmerhorn and Geisshorn. In the more massive Aar granite, joint set F3 is accompanied by exfoliation joints, e.g., in Figure 5g.

Table 1. Summary of Discontinuity Characterization Along the Cross Section M in Figure 5

Set	Type	Dip Direction/Dip	Spacing	Persistence	JRC	Schmidt Hammer Rebound Hardness	Description
F1	Foliation	124°/75°	0.1–0.5 m	Very persistent (3–10 m)	6 ± 3	48 ± 10	No infilling, no aperture, unweathered to slightly weathered
F2	Fractures	198°/84°	0.5–2 (>5) m	Minor persistent (1–3 m)	6 ± 3	52 ± 11	No infilling, 1–5 mm aperture, unweathered to slightly weathered
F3	Fractures	244°/18°	1–5 (>5) m	Moderately persistent (3–5 m)	7 ± 4	51 ± 5	No infilling, 1–5 mm aperture, unweathered to slightly weathered
F4	Fault zones	141°/61°	25–50 m	Very persistent (>20 m)			Strongly foliated, 1–5 mm aperture, moderately to highly weathered

4 Numerical Study of Paraglacial Rock Slope Damage and Displacement

4.1 Model Approach and Inputs

Paraglacial rock slope damage and displacements were computed by using the 2-D distinct-element code UDEC [Cundall and Hart, 1992; Itasca, 2014], which is well-suited for analyzing the behavior of a discontinuous rock mass, i.e., a large number of discontinuities separating intact blocks. Deformable blocks are connected through contacts representing discontinuities. The contacts mimic joint stiffness and are assigned a strength criterion (e.g., Mohr-Coulomb). Shear slip or joint opening can promote irreversible displacement once the failure criterion is reached.

Our model geometry and initialization steps are shown in Figure 6. To accurately represent the stress field in the Aletsch Valley and to minimize boundary effects, our area of interest was embedded into a 10 km wide large-scale model including the neighboring Rhone Valley. Model cross sections are based on a DTM (swissALTI3D by Swisstopo) complemented with ice penetrating radar data [Farinotti et al., 2009] below the present-day glacier. The area of interest (Figure 6) contains three rock mass elements: (1) intact rock, (2) discontinuities (joints), and (3) brittle-ductile fault zones. Intact rock is represented by randomly oriented discontinuities, so-called Voronoi polygons [Lorig and Cundall, 1989], which allow failure along new potential pathways simulating failure of rock bridges. The orientation, spacing, persistence, and strength parameters for each rock mass constituent were evaluated through field assessment and data collection at outcrops along the representative cross-section M (Figure 5). Blocks between discontinuities were assigned elastic properties, while Voronoi contacts, discontinuities, and faults were assigned a Mohr-Coulomb constitutive law including slip-weakening of friction, cohesion, and tensile strength properties (i.e., a drop from peak to residual strength at failure).

Based on reported literature values for intact rock (Table 2) and an assumed Geological Strength Index (GSI) of 75 obtained from field observation, we calculated rock mass strength properties by using the GSI approach [Hoek et al., 2002]. This resulted in estimated values of cohesion, c = 7.8 MPa; friction angle, φ = 50.6°; tensile strength, t = −0.6 MPa; and elastic modulus, E = 31.8 GPa. Rounded values were implemented in UDEC for blocks and Voronoi contacts representing the rock mass (Tables 2 and 3). These bulk rock mass properties match well with previous studies in similar lithologies [e.g., V. Gischig et al., 2011; Kinakin and Stead, 2005]. Bulk modulus (K) and shear modulus (G) were calculated by assuming homogeneous isotropic materials [Hudson and Harrison, 1997].

Table 2. Intact Rock and Rock Mass Properties Implemented in UDEC


Intact Rock
Density ρ	(kg m⁻³)	2700	Wegmann [1998]
Poisson's ratio	()	0.2
Stress ratio k	()	1.0	Kastrup et al. [2004]
UCS	(MPa)	110	Steiner et al. [1996]
Young's modulus	(GPa)	39	Wegmann [1998]
Rock Mass (GSI = 75)
Young's modulus	(GPa)	30
Friction angle φ	(°)	50
Cohesion c	(MPa)	8
Tensile strength t	(MPa)	1

Table 3. Discontinuity Properties for the Mohr-Coulomb Constitutive Law Including Slip-Weakening Implemented in UDEC

Discontinuity Parameters	Unit	Intact Rock (Voronoi)	F1 Foliation	F3	F4 Faults
Peak friction angle φ	(°)	50	33.7	37.2	27
Peak cohesion c	(MPa)	8	1.8	3.5	0.03
Peak tensile strength t	(MPa)	1	0.4	0.8	0
Residual friction angle φ_R	(°)	27	27	27	27
Residual cohesion c_R	(MPa)	0.03	0.03	0.03	0.03
Residual tensile strength t_R	(MPa)	0	0	0	0
Dilation angle	(°)	5	5	5	5
Dip angle	(°)	-	75	6	75
Normal stiffness	(GPa m⁻¹)	20	10	10	1
Shear stiffness	(GPa m⁻¹)	10	5	5	0.5

For discontinuities (peak strength without intact rock bridges), we assumed a peak friction angle of 30° and a peak cohesion of 0.1 MPa, with no tensile strength [V. Gischig et al., 2011]. Residual friction was set to 27° and residual cohesion to 0.03 MPa (Table 3) [V. Gischig et al., 2011]. Jennings's approach [Jennings, 1970] was used to determine peak composite Mohr-Coulomb strength properties for discontinuities including effects of intact rock bridges; we assumed that F1 contains 10% rock bridges and F3 20% rock bridges (see persistence in Table 1). The resulting discontinuity friction angles (Table 3) are within the range of past measurements in the same lithology [e.g., Steiner et al., 1996]. Peak strength values of faults (F4) were set to residual values.

The orientation and spacing of discontinuities and faults in our model are taken from field observations along cross-section M (Table 1). For this profile, the apparent dip of foliation (F1) and faults (F4) does not change. However, the apparent dip of set F2 reduces to 77° and falls together with F1. The flat dip angle of F3 is reduced to 6° apparent dip along the cross section. Input spacing for discontinuities in UDEC is based on field observations (Table 1); however, a single joint in UDEC represents several joints in reality. The spacing of F1 discontinuities was thus set to 40 m, F3 to 80 m, and faults F4 to 200 m. Joint spacing in the model increases with depth (>300 m) by a factor of 2 accounting for a near-surface fractured zone in alpine rock slopes [Masset and Loew, 2010; Zangerl et al., 2008a, 2008b]. The Voronoi contacts have a maximum length of 40 m. Blocks generated by intersecting joints are meshed with a mesh size of 15 m in the upper 300 m, while the mesh size increased stepwise with depth and beyond the area of interest. We assumed a linear joint normal stiffness of 10 GPa m⁻¹ and a joint shear stiffness of 5 GPa m⁻¹ (Table 3), as used in past similar studies [V. Gischig et al., 2011]. The stiffness of Voronoi contacts was twice as high. For faults, we assigned a normal stiffness of 1 GPa m⁻¹ and a shear stiffness of 0.5 GPa m⁻¹ (Table 3) [Zangerl et al., 2008b, 2008c].

Selecting a physically meaningful modeling approach to represent glacier ice and its influence on surrounding rock slopes is crucial for analyzing paraglacial rock slope mechanics. Previous studies [e.g., Eberhardt et al., 2004; Jaboyedoff et al., 2012] have modeled glacier ice as either an elastic or plastic material. Recent investigations [McColl et al., 2010; McColl and Davies, 2013] emphasize that ice will undergo ductile flow under small strain rates and is not capable of providing shear resistance to adjacent rock slopes, instead loading underlying bedrock by its weight alone. Furthermore, ice will relax through plastic deformation (creep) at stresses above a yield shear stress of ~100 kPa [Schulson, 1999; Cuffey and Paterson, 2010]. We therefore model glacier ice as a hydrostatic stress boundary condition, rather than an elastic material, as similarly applied by Leith et al. [2014a, 2014b]. A comparison between modeling ice as a time-varying stress boundary condition versus an elastic material is presented in Appendix A2. Out-of-plane stresses at the glacier bed (i.e., basal shear stress due to ice flow) are limited by the yield shear stress of ice and the presence of water at the interface. Basal shear stresses are typically in a range from 50 to 150 kPa [Cuffey and Paterson, 2010]), an order of magnitude smaller than overburden stress during Late Glacial and Holocene glacier cycles.

We also perform large-scale transient models in UDEC to model ice loading cycles, once the initial conditions have been established. These models are computationally intensive and require a reduced geometry (Figure 6). The area of interest is restricted to a smaller window around the valley and surrounded by a buffer zone with fewer model elements. Voronoi contacts, allowing creation of new joints through intact rock, are restricted to the uppermost 300 m. Joint spacing is increased by a factor of 3 within the buffer zone, and to account for the change in compliance, joints within this zone are assigned a normal stiffness which is 3 times lower [cf. Zangerl et al., 2003]. The mesh size in the uppermost 300 m is 35 m, increasing stepwise toward the boundary of the area of interest.

4.2 Initialization: Critically Stressed Alpine Valley Before and During the LGM

We argue for ice-free initial conditions that represent the Aletsch Valley prior to the LGM. Otherwise initial damage (i.e., failure occurring during initialization, representing inherited damage resulting from all prior processes) is strongly underestimated due to relaxation under LGM ice occupation. Initialization of our model is thus undertaken in two primary steps (Figure 7): (1a) First, initial stresses are calculated under elastic, ice-free conditions (Figures 7a–7d). The initial far-field stresses applied represent exhumation-induced and tectonic stresses in a simplified (linear hillslope) paleoalpine valley (i.e., σ_yy = −ρgΔz, where g is gravitational acceleration and Δz is vertical distance to paleoalpine valley). A strike-slip stress regime prevails in our study area [Kastrup et al., 2004]; we therefore use a stress ratio (σ_h/σ_v) of k = 1 (i.e., σ_xx = kσ_yy), as values of k ≠ 1 would represent normal or thrust faulting in a 2-D model. Out-of-plane stresses (σ_zz) are calculated by assuming plane-strain conditions. (1b) Unrealistically high stresses are avoided during initialization by using an elastoplastic equilibration phase with a simple Mohr-Coulomb failure criterion for blocks (φ = 50°, c = 8 MPa, t = 1 MPa). (1c) The elastoplastic failure criterion for discontinuities is assigned (strength properties listed in Table 3), allowing joints to fail. Thereby, initial damage was simulated during the ice-free pre-LGM interglacial (Eemian) period (Figures 7e and 7f). (2) In the second step, we add ice loading to the mapped LGM elevation (2800 m above sea level (asl)). Stress redistribution is calculated under elastic conditions before again allowing joint failure (Figures 7g and 7h). This represents the starting point for subsequent transient models investigating Late Glacial and Holocene glacier cycles.

In Figure 7, we display the stress state along cross-section M (see Figure 2) during our initialization procedure. Under initial ice-free conditions, in-plane major principle stresses (σ₁) reach maximum values at the valley bottom (up to ~50 MPa) and decrease toward the flanks (Figure 7a). Stresses are around 10–20 MPa at the valley shoulders and stress orientations generally parallel to topography. In-plane minor principle stresses (σ₃) are defined by overburden and oriented orthogonal to topography (Figure 7b). Out-of-plane stresses (σ_zz) are in a similar range as σ₁ (Figure 7c). Differential stresses in the valley bottom exceed 25 MPa within ~500 m of the axis (Figure 7d). Allowing blocks then joints to fail, these differential stresses lead to a ~300 m deep damage zone focused mostly around the lower valley flanks (Figures 7e and 7f). All failed joint segments are critically stressed (here defined as within 1 MPa of the Mohr-Coulomb failure criterion), while steeply dipping joints on the western flank are critically stressed in tension at the surface and faults are critically stressed in shear on both sides (Figure 7f). The ice-free Aletsch Valley is at critical conditions. Adding LGM ice in the second initialization step changes the differential stress state (assuming zero glacial erosion during LGM). Differential stresses decrease in the valley bottom by up to 18 MPa but increase slightly at the surface on the western flank and even more at the toe of the slope on both sides (Figure 7g). LGM ice occupation thus alleviates critical stresses within the valley; only a few failed joints are still critical under tension (Figure 7h). Furthermore, stress redistribution under LGM ice leads to a change in failure mode from a mixed shear/tensile regime toward a shearing dominated regime. The glacier ice load prevents most tensile failure. This illustrates the importance of selecting a model starting point prior to LGM ice occupation; otherwise the unstressed rock slope experiences unrealistically high stresses and damage accumulation during deglaciation.

We simulated initial damage for different cross sections along the Aletsch Valley. The same modeling procedure as described above was applied to the extended model geometry (see Figure 6) for different profiles (L–P, Figure 2). The thickness of LGM ice was adapted for each cross section based on field data, and the same model geometry and rock structure were used for all profiles. Figure 8 shows the resulting initial damage distribution and damage propagation during LGM ice occupation. The total failed joint length ranges between 8 and 21 × 10³ m (summed length of discontinuities at residual strength). Additional damage propagation during LGM ice occupation varies between 0.5% and 2.5%. Profiles M, N, and O show small variability in the damage field but differ from damage in profile L with its distinct U-shaped form. Additional damage occurs in all profiles close to the valley bottom during LGM conditions, even though general rock slope conditions are less stressed. We conclude that topographic effects on the initial damage field and subsequent damage during LGM ice occupation exist, but variations are small and the general patterns of damage are comparable throughout the Aletsch Valley.

A fluctuating glacier affects different areas of an alpine valley over time. Since we seek to investigate the isolated influence of glacier cycles on an adjacent rock slope, and not focus on variable topographic effects, we model different glacier histories along the same cross section, conceptually representing different sections of the valley. The Aletsch Valley is situated in a relatively homogeneous crystalline massif and aligned parallel to the main foliation. In the preceding Figure 8, we showed that the spatial damage distribution for different topographic profiles along the valley does not vary markedly, justifying the following analysis along a single cross section for different glacier histories.

4.3 Transient Rock Slope Damage

Fracture development in an alpine valley subject to Late Pleistocene glaciation is strongly path-dependent [Leith et al., 2014a, 2014b]. Transient numerical models are therefore crucial to investigate the influence of repeat glacier cycles on progressive damage. Here we use a quasi steady state approach to simulate cycles of glacier retreat and advance. Changes in stress boundary conditions over glacial time scales are relatively slow, so the system has sufficient time to maintain mechanical equilibrium. Discretization into small steps, with subsequent equilibration to steady state after each step, is therefore a valid approach for our simulations. We conducted a sensitivity analysis for the model step size to justify that the value we use to discretize glacier change is small enough so that model behavior is stage-dependent and not rate-dependent; i.e., the model is always in steady state. We emphasize that due to the quasi steady state approach, there is no real time in these models. Slope response is an instantaneous reaction to the glacier load at each model step.

Transient models are performed by using the reduced model geometry (Figure 6). Cross-section M was selected to represent the Aletsch Valley and subjected to different glacier histories related to other cross sections (M, N, O, and P in Figure 2). We performed a sensitivity study to evaluate the impact of reduced model complexity on initial and subsequent damage. We compared damage accumulation during our initialization procedure for different model configurations: (a) Voronoi contacts throughout the area of interest, (b) Voronoi contact restricted to the uppermost 300 m, and (c) finer model geometry with joint spacing reduced by a factor of 2. A reduction in joint density results in smaller absolute damage, but the amount of failed joints relative to total available joints remains similar. Initial damage patterns, as well as the location and amount of subsequent damage with added LGM ice, remained similar. While the reduced model geometry may slightly underestimate progressive damage, results do not vary significantly. Therefore, we use the more computationally efficient reduced geometry in the following transient models.

Based on mapped Late Glacial and Holocene ice extents along profiles M, N, O, and P (Figure 2), we generated different conceptual glacier scenarios for cross-section M (Figure 9a). All scenarios begin with LGM ice and the initialization procedure described previously. In the first 300 model steps, LGM ice is lowered in the Aletsch and Rhone Valleys simultaneously, until the glacier elevation reaches the crest that divides the valleys. LGM deglaciation continues in the Rhone Valley until 1000 steps. All four glacier scenarios undergo Egesen readvance after LGM deglaciation. Subsequent Holocene fluctuations are represented by three repeat cycles. Depending on the position of the profile, the Egesen readvance and Holocene fluctuations vary in amplitude.

The temporal evolution of damage under different glacier scenarios is presented in Figure 9b. Damage within the rock slope is quantified as the summed length of failed discontinuities (i.e., joints, faults, and Voronoi contacts) at each model step. The final additional damage accumulated in all models is relatively small (3–4% of initial damage). We find that most damage occurs when ice elevations lower from 2200 to 2000 m asl during first glacier retreat. Holocene fluctuations only result in subsequent damage for scenario N (0.5% of initial). Damage propagation is greater during Holocene glacier advance than during subsequent retreat. Model scenario M, which never reaches ice-free conditions, shows the least total damage. We conclude that the amplitude of glacial cycles must be large (>300 m in this example) and the Holocene minimum must reach close to the valley bottom causing the slope to become most critically stressed, to be effective in damage propagation, and even then only the first cycle appears to produce additional damage.

Figure 9c presents the amount of failed and critically stressed discontinuities over time for scenario N containing pronounced Holocene glacier fluctuations. While Holocene cycles produce only minor additional damage, the amount of critically stressed joints varies strongly. Deglaciation is accompanied by an increase in critically stressed joints as normal stresses are alleviated. During glacier advance, the number of critically stressed joints decreases and discontinuities preferring tensile failure shift to shearing mode, while the reverse is encountered with glacier retreat.

4.4 Spatial Damage Patterns

Depending on the site-specific glacial history, adjacent rock slopes are affected differently by the ice load, leading to a spatially variable damage patterns. Figure 10 compares spatial damage predictions for the different glacier model scenarios introduced previously. Spatial damage over time for model scenario N is displayed in Figures 10a and 10b. New damage occurs mainly on the eastern flank in the form of fracture propagation of initially failed discontinuities, or a fault zone at its yield limit in the midportion of the slope. Comparison of damage for alternative model scenarios (Figure 10b) shows that similar spatial patterns occur with or without Holocene glacier fluctuations. Most new rock slope damage accumulates in the midportion of the valley at the time when the glacier surface first lowers past this area. By comparison, in Figure 7 we showed that this midslope region was critically stressed, but not yet failed, at initialization. Through stress redistribution, i.e., due to additional damage during LGM occupation and deglaciation, joints in the midslope area may now reach their failure limit. The location of critically stressed areas prone to failure is thus more important in controlling the pattern of spatial damage than the elevation of Holocene ice extents.

4.5 Stress Redistribution

Stress redistribution as a result of incremental failure drives further damage within the slopes during our simulations. Figure 11 displays stress redistribution in the valley slopes for various glacier cycles. Limited by failure within the rock slope, differential stresses increase during LGM deglaciation and attain a maximum value of ~10 MPa at the valley bottom (Figure 11a). In contrast, the valley flanks experience a reduction in differential stress of similar magnitude. LIA deglaciation shows a similar pattern but with approximately half the magnitude (Figure 11b). The missing weight of the valley glacier allows higher stresses to concentrate on the valley bottom, while the slope flanks experience concomitant stress reduction. Deglaciation, independent of the model scenario, results in stress redistribution, which reallocates high stresses from the midslope area to the valley bottom. Internal failure limits the magnitude of stress increase and promotes further stress redistribution. The full extent of stress redistribution during a complete glacial cycle is shown in Figure 11c; at the tips of failed joint segments, stress changes are in the range of ±2.5 MPa and affect depths up to 200 m on the slope.

An example stress path is shown in Figure 12a for a failing midslope joint, highlighting localized stress changes during glacier fluctuations. At initial stress conditions, this discontinuity is close to failure. The additional load of LGM ice makes the joint less critically stressed (i.e., stresses move away from the failure envelope). Nearby joint failures and accompanying stress redistribution then lead to an irreversible change in the stress path during first deglaciation (LGM/Egesen ice retreat). The joint segment is more critically stressed during deglaciation than during initial ice-free conditions (shift from 1 to 3 in Figure 12a). Stresses exceed the failure criterion, the joint fails, and stress redistribution results as the joint assumes residual strength. This particular discontinuity then remains critically stressed (at residual strength) throughout the Holocene, although glacier advance leads to slight stress relaxation. The stress path during repeat Holocene cycles describes a closed loop following the residual strength envelope. Shear displacements attributed to the stress path are illustrated in Figure 12b. Irreversible joint slip occurs during first deglaciation as the joint fails. Further minor irreversible slip occurs during the first Holocene advance and later during repeat glacier retreat.

4.6 Slope Displacement

The modeled Aletsch Valley rock slopes experience not only distinct, localized failure along discontinuities but also bulk rock mass displacement as a result of elastic rock mass deformation and accumulated joint displacement (Figure 13). Estimated uplift magnitudes for LGM deglaciation range from 300 to 600 mm, whereas LIA deglaciation results in 60–160 mm of simulated uplift (Figures 13a and 13b). The displacement pattern for LGM deglaciation is asymmetric due to greater ice loss in the Rhone Valley (see Figure 6). We observe left-handed block displacement along steeply dipping joints on the western slope and right-handed shearing on the eastern slope, resulting in uphill-facing counterscarps (inset in Figure 13a). These calculated uplift values are within the same order of magnitude as determined in previous studies; e.g., Memin et al. [2009] estimated 5–9 mm uplift for 30–50 m ice loss in the Mont Blanc region. Our predicted post-LIA uplift at Aletsch corresponds to a mean uplift rate of 0.4–1.1 mm yr⁻¹, which is in good agreement with a rebound modeling study at Aletsch estimating an uplift rate of up to 1.5 mm yr⁻¹ due to recent glacier retreat [Melini et al., 2015].

Figure 13c shows that the eastern midslope region and the toe area in the west are most strongly affected by a complete modeled glacial cycle. The differences between valley slopes arise mainly from the prevailing joint patterns. Although modeled glacier cycles during the Holocene do not produce a large amount of additional rock slope damage (Figure 9), varying ice extents do influence slope displacements. The displacement patterns in scenarios M and N are similar, but the magnitudes increase by a factor of four with larger amplitude glacier fluctuations (Figures 13d and 13e). Holocene displacements for scenario O, with only minor Holocene ice fluctuation, are <0.1 mm.

4.7 Influence of LGM Valley Erosion

Current rates of glacial erosion measured in the Aletsch Valley are around 1 mm yr⁻¹ [Hallet et al., 1996 and references therein], but these values may decrease by 1 or 2 orders of magnitude when integrated over glacial cycles [Koppes and Montgomery, 2009]. The influence of topographic change due to glacial erosion during the last glacial period (~100 kyr) may be small compared to landscape modifications during the Mid-Pleistocene revolution [Haeuselmann et al., 2007] but plays an important role in our mechanical model. We investigated rock slope damage under different erosion scenarios, assuming maximum glacial erosion rates of 0.1, 0.2, and 0.4 mm yr⁻¹ for the modeled cross section (maximum values occur in the Rhone Valley). For simplicity, we let the amount of total erosion decrease linearly with decreasing ice overburden. This resulted in maximum erosion within the Aletsch Valley of 8, 15, and 30 m, respectively, thinning toward adjacent valley flanks (Figure 14a). Less than 30 m of total erosion through abrasion during the last glacial cycle is likely a reasonable assumption [see Leith et al., 2014a and references therein]. In all modeled erosion scenarios, the ice-free valley was initialized with plastic deformation including the additional, uneroded rock overburden, which was then instantaneously removed during LGM ice occupation.

The temporal evolution of damage for model scenario N including erosion is shown in Figure 14b, revealing a strong increase in damage. A major damage event occurs when the ice elevation first reaches the midslope region and continues until ice-free conditions are reached. Greater uneroded rock overburden included during initialization leads to a reduction in initial damage by loading the toe of the slope. New damage during deglaciation varies between 27% and 35% of initial damage, increasing with greater glacial erosion. The increased successive damage does not compensate the lower initial damage that comes with these erosion scenarios. Therefore, the smallest erosion scenario creates the greatest amount of final damage. However, the period of adjustment to LGM erosion is longer for greater erosion rates, since stress conditions are shifted farther from the initial conditions. Therefore, Holocene cycles can create more new damage. Figures 14c and 14d display the spatial damage distribution for various erosion scenarios. New damage appears mostly in the midslope and lower regions of the eastern flank. We conclude that glacial erosion (i.e., rock debuttressing) during the last glacial period has a strong influence on damage accumulation during first deglaciation and during subsequent glacial cycles. Buttressing by uneroded rock during the preceding ice-free interglacial (Eemian) prevents the development of initial damage.

4.8 Weakened Rock Slope Response

Here we investigate the efficacy of glacial loading cycles in creating damage within an already weakened rock slope, performing a series of simulations by assuming reduced initial rock mass strength. The initial stress state remained the same for all scenarios, but we reduced the peak friction angle, peak cohesion, and peak tensile strength for all rock mass elements (except fault zones, which were already at residual strength). Reduced strengths were varied between the previously applied peak and residual values (see Table 3), scaled linearly by a factor (α); e.g., c_red = c_R + α(c − c_R) (i.e., α = 100% represents peak strength properties as in previous models, while α = 0% represents residual strength).

Figure 15 displays the results for strength reduction factors α = 70%, 50%, 30%, and 10% for model scenario N. Reduced rock mass strength results in increased initial damage (Figure 15b). However, new damage accumulation during Holocene cycles is greatest for α = 50% and 30%. In a weaker rock slope close to residual strength (e.g., α = 10%), all critically stressed discontinuities have already failed during initialization and the magnitude of stress redistribution by slip-weakening is small. Major damage propagation for α = 50% and 30% occurs during Egesen ice retreat, while minor damage events follow subsequent Holocene advance and retreat cycles. Compared with models assuming higher rock mass strength (Figure 10a), weaker slopes have increased initial damage with more intact rock failure near the valley bottom (Figure 15e). New damage during glacial cycles occurs mostly on the eastern valley flank, in the form of fracture propagation along pre-existing, steeply dipping discontinuities. Failure of intact rock bridges (i.e., Voronoi contacts) parallel to topography connects steeply dipping joints and generates a shear failure surface. Maximum displacements in Figure 15f reach 0.5 m, revealing the extent of a slope instability with toppling kinematics and a graben structure as back scarp (see magnified displacement in Figure 15f). The kinematics and dimensions of the unstable rock slope produced in this simulation (Figures 15e and 15f) generally resemble field observations at the Moosfluh instability (Figures 1c and 1f) as well as the instability at Hohbalm (Figure 15i). No substantial sliding was produced on the western valley slope as is observed in the field (e.g., at the Driest instability, Figure 1a).

The evolution of slope instability is illustrated in the temporal damage and point displacement plot in Figures 15c and 15d. Damage during Egesen ice retreat leads to major irreversible displacement (Figure 15d), initiating slope failure. Interestingly, further displacement accumulates during the first and second Holocene ice advances, while later glacier retreat does not promote significant movement. The reason for enhanced displacement during Holocene advances is revealed by the elastic rock slope response shown in Figure 16: Glacier advance pushes the lower slope away from the valley axis, rotating the upper slope inward toward the valley, and vice versa during glacier retreat (Figures 16c and 16d). Inward rotation of the upper slope during glacier advance helps drive irreversible slip and damage propagation in the rock slope with joint orientations favoring flexural toppling.

Figure 17 provides further insight into the mechanics of a weakened rock slope (α = 30%) and illustrates how glacier cycles promote additional damage and shear displacement; different glacial stages applied in our models are queried in detail for shear displacement and additional failed discontinuities. LGM and Egesen deglaciation leads to large shearing near the valley bottom and along steeply dipping discontinuities (F1) on the eastern flank near the slope crest (Figure 17a; 1). Most fault zones are affected by minor shear displacement of a few millimeters. New damage is distributed over the entire rock slope but concentrates in the upper part of the eastern slope in the form of fracture propagation extending pre-existing joints. During the first Holocene advance (Figure 17a; 2), right-handed shearing dominates the steeply dipping joints on the eastern flank (i.e., toppling), induced by inward rotation of the upper slope (see Figure 16). Prevailing joint patterns promote toppling kinematics. Although the glacier surface remains near the toe of the slope, the entire flank is affected. Later Holocene ice advance (Figure 17a; 3) reduces the affected region of the rock slope but enhances shearing along faults close to the valley axis. Glacier fluctuations around the toe of an unstable rock slope are more effective at driving damage than glacier fluctuations affecting the upper part of the instability. Holocene ice retreat generates only minor shearing and damage in the lower slope region, mostly as the glacier reaches the toe of the slope (Figure 17a; 4 and 5). The same mechanical interaction between the glacier and rock slope appears during the second and third cycles, although the magnitude of shearing and the amount of new damage decreases with each cycle (Figure 17a; 6 to 9).

Shear displacements of four example points are shown in Figure 17c with location and corresponding joint orientation in Figure 17b. Instances of irreversible slip along these joints are highlighted. Planar sliding along gently dipping joints (points 4 and 5 in Figure 17c) occurs only during glacier retreat. On the other hand, shearing along steeply dipping toppling joints (points 6 and 7 in Figure 17c) happens during Egesen retreat and again during Holocene ice advances. This demonstrates that different joint patterns are affected differently by glacial loading and unloading cycles.

5 Implications for Paraglacial Rock Slope Instabilities

Our simulations provide new insights into the mechanical development of damage produced by glacial cycles as a preparatory factor for paraglacial rock slope instabilities. At the Aletsch Glacier, we observe a concentration of large landslides around the present-day terminus (Figure 2). The landslide density decreases in the lower part of the valley affected only by the LGM and subsequent Egesen stadia, and we observe fewer slope failures in the upper Aletsch Valley, which was affected by fewer Holocene glacier fluctuations (Figure 2). There exist numerous other examples of rock slope failures located around present-day glacier termini [Bovis, 1990; Oppikofer et al., 2008; Clayton et al., 2013; McColl and Davies, 2013], the area where ice fluctuated most during the Holocene. In our study area, the majority of identified instabilities have post-Egesen/pre-LIA relative initialization age, while post-LIA landslides are less frequent (Figure 2). Post-LIA initiation of the Tälli instability coincided with ice retreat from the toe of the unstable slope. Exposure age dating constraining initiation of the Driest instability (7.4 ± 0.7 kyr) shows a large lag time between LGM or Egesen ice retreat and initial displacement, although initiation may be related to a minor Holocene readvance around 8.2 kyr reaching an extent similar to today [Nicolussi and Schlüchter, 2012] affecting the toe of the landslide. Other past studies have suggested that paraglacial rock slope instabilities were more frequent during the early Holocene after deglaciation but often with large lag times [e.g., Prager et al., 2008; Ivy-Ochs et al., 2009a; Ballantyne et al., 2014a, 2014b]. Rock slope failures associated with LIA retreat are generally less common [e.g., Evans and Clague, 1994; Jaboyedoff et al., 2012].

In our simulated rock slope cross section assuming realistic, moderate rock mass strength properties, the effects of glacier cycles as purely mechanical loading and unloading phenomena resulted in only minor damage (Figure 10). Glacier ice loading does not significantly buttress the adjacent slope but does alter the in situ stress field (Figure 11) potentially driving damage propagation. Most observed damage in our model was inherited and occurred during initialization under Eemian ice-free conditions (Figure 10). However, we demonstrated that fluctuating glacier elevations strongly influence the criticality of adjacent rock slopes, which may in turn control the efficacy of other fatigue processes not explicitly considered in our models. Glacial erosion during the last glacial period is an effective way to change the in situ stress conditions, stimulating new damage during LGM deglaciation and potentially acting as a significant contributor in preparing future rock slope instabilities. Meanwhile, reduced initial rock mass strength similarly enhances damage accumulation during subsequent glacial cycles. Therefore, acting in concert with a change in boundary conditions or material properties, fluctuating glacier ice can represent a significant preparatory factor for paraglacial rock slope failures.

Spatial patterns of landslides at Aletsch correlate with results of our numerical modeling. The kinematics and dimensions of the instability produced in our weakened slope closely resemble characteristics of the Moosfluh instability [Strozzi et al., 2010; Kos et al., 2016]; however, our numerical analysis was unable to reproduce substantial displacement on the western valley slope as observed at Driest [Kääb, 2002; Kos et al., 2016]. Using moderate, presumed realistic, rock mass strength conditions, we were unable to generate large-displacement slope instability through glacial cycling. However, we did observe greater slope displacements in the presence of larger Holocene ice fluctuations, representing a slope profile around the present-day glacier terminus, as compared to smaller glacial cycles representing a higher valley profile (Figure 13). Furthermore, only large amplitude Holocene cycles, whose minima reached near the valley bottom, produced additional damage. These results indicate the potential influence of Holocene glacier fluctuations on the preparation of rock slope instabilities and agree with the mapped landslide concentration around the present glacier tongue, as well as with field observations of fewer landslides in the upper Aletsch Valley, where ice has likely remained throughout the Holocene. However, we note that the rock slope reaction to glacier activity also depends strongly on site-specific rock mass conditions [cf. McColl, 2012], which are in turn controlled by geological predisposition (i.e., rock strength and structure) as well as the damage history [Terzaghi, 1962; Augustinus, 1995; Stead and Wolter, 2015].

Correlating the temporal distribution of landslides assessed from field evidence with damage propagation in our numerical simulations reveals several similarities. In our models, most damage occurs during first deglaciation, when the glacier elevation drops for the first time below the critically stressed midslope region. For a weakened valley flank, this damage event may even initiate slope failure. The timing of large damage events during first deglaciation correlates well with the majority of post-Egesen/pre-LIA landslide ages assessed at Aletsch. Subsequent fracture propagation and slope displacement in our models accumulate during each Holocene cycle, especially at times when the glacier reaches the toe of the slope. Possible correlation between initiation of the Driest instability and timing of the minor 8.2 kyr Holocene advance, as well as the coincidence of the Tälli landslide with recent glacier retreat, supports the role of Holocene fluctuations in creating new slope damage. High sensitivity to glaciers in the slope toe region matches field observations indicating highly active and accelerating displacements at the Moosfluh instability as the present-day Aletsch Glacier retreats from its toe [Strozzi et al., 2010; Kos et al., 2016; Loew et al., 2017]. Rock slopes higher in the Aletsch Valley, which have likely remained ice-covered since the LGM, might similarly be more prone to damage as they undergo first-time glacier retreat, potentially resulting in increased instances of slope failure.

Detailed temporal correlations between field observations and our numerical results remain challenging to assess, since damage propagation in our models is the immediate response to glacier change. Aside from glacier cycles, no other time-dependent processes act in our model and equilibrium is reached after each step; therefore, it is not possible to simulate lag times between deglaciation and slope failure. Furthermore, it is important to point out that initial gross displacement of a landslide body does not necessarily correlate with the timing of the genesis of that landslide. Slope displacement can appear long after the amount of internal rock mass damage has reached a critical level when an ultimate trigger finally initiates movement. We investigated the mechanics and evolution of rock mass damage as a preparatory factor for paraglacial slope instabilities, indicating times when new damage accumulation may be most prominent. The temporal evolution of a fully developed slope instability and its interplay with retreating or advancing ice are beyond the focus of this research [see McColl and Davies, 2013]. We demonstrated that even when neglecting ice buttressing effects, most damage occurs during first deglaciation, bringing the slope closer to potential failure. In nature, other environmental processes following deglaciation may additionally reduce slope stability over time until failure occurs [Eberhardt et al., 2004; Prager et al., 2008; McColl, 2012].

In our models, we include explicit mechanical reasoning explaining the development and accumulation of rock slope damage associated with cyclic ice loading. Driving mechanisms for damage include stress changes during glacier cycles, stress redistribution by slip-weakening following incremental failure, or changes in rock slope boundary conditions. However, additional driving mechanisms may be important for preparing paraglacial rock slope instabilities. We demonstrated that glacial cycles strongly affect the amount of critically stressed joints within an alpine valley (Figures 7f and 7h) and each phase of glacier retreat places adjacent rock slopes into a more critically stressed condition (Figure 9c). Other environmental processes can act on the critically stressed slopes contributing to additional damage and promoting time-dependent failure, e.g., chemical weathering within joints [Jaboyedoff et al., 2004], stress corrosion at fracture tips [Faillettaz et al., 2010], ice segregation [Wegmann et al., 1998; Hales and Roering, 2007; Sanders et al., 2012; Krautblatter et al., 2013], changes in joint water pressure [Hansmann et al., 2012; Preisig et al., 2016], thermal stresses [Wegmann and Gudmundsson, 1999; Gischig et al., 2011a, 2011b; Baroni et al., 2014], or seismic fatigue [Gischig et al., 2015]. Each of these processes can contribute to further rock slope damage, especially at times when ice loading conditions increase the criticality of the slope. Over time, reduced rock mass strength may then favor increasing slope sensitivity to glacial mechanical loading and unloading cycles as shown in Figure 15. Thermo- and hydromechanical effects acting in concert with glacier cycles likely also play an important additional role in preparing slopes for failure, which we investigate in detail in following companion studies.

6 Summary and Conclusions

We investigated the mechanical development of rock slope damage induced by loading and unloading of Late Pleistocene and Holocene glaciers. Using detailed, conceptual numerical models, based on realistic site conditions at the Aletsch Glacier in Switzerland, we explored how stress changes associated with repeat glacial cycles generate progressive rock mass failure and prepare paraglacial slope instabilities. We first evaluated the transition from an ice-free valley during the Eemian interglacial toward LGM conditions for several cross sections along the Aletsch Valley. Conceptualized glacier scenarios based on site-specific field mapping were then applied to investigate transient rock slope damage, in situ stress changes, and displacement. We further highlighted the role of LGM valley erosion on damage propagation during deglaciation and showed how a weaker rock mass is more sensitive to damage accumulation. Stress changes and resulting damage distributions were compared with mapped spatial and temporal landslide distributions at our Aletsch Valley study sites. Key outcomes of our study are as follows:

Following initialization under ice-free conditions and reoccupation by LGM glaciers, damage accumulation during subsequent deglaciation and Late Glacial/Holocene cycles was minor and originated in our models from stress changes during glacier cycles and stress redistribution as a result of subglacial fracturing. The purely mechanical response to simple glacial loading and unloading thus represents a comparatively minor preparatory factor for paraglacial rock slope instabilities under moderate strength conditions; glacial debuttressing alone has a limited effect. However, including even minor amounts of bedrock erosion, i.e., rock debuttressing, during the LGM promoted significant new damage accumulation during first deglaciation. Major damage occurs during first ice retreat in our models not due to the removal of an ice buttress but rather because stress conditions within the slope reach a critical state for the first time.
The mechanical interaction between a rock slope and glacier varies over time: the location of damage changes in conjunction with changing ice thickness. The first deglaciation (LGM and Egesen retreat) and subsequent Holocene ice advances are more effective in creating damage than glacier retreat in general. Ice advance generates shear dislocation and damage along toppling discontinuities in our models, while ice retreat promotes planar sliding. Weaker rock slopes showed increased sensitivity to glacial loading cycles, accumulating greater damage and displacements, which in some cases led to full development of an instability.
Temporal and spatial landslide patterns assessed in the field support conclusions from our numerical study. The kinematics and dimensions of a modeled instability on the eastern slope match characteristics of landslides at Aletsch (Moosfluh and Hohbalm instabilities). Major damage during first deglaciation in our models correlates with the postulated post-Egesen initiation ages for these failures. Our modeled rock slope is most sensitive to ice loss in the toe region, which is confirmed by recent landslide monitoring at Aletsch. The eastern flank (toppling) in our numerical study showed enhanced slope displacement and subsequent damage propagation with repeat high-amplitude ice elevation changes during Holocene advances. Local predisposition, in combination with large amplitude ice fluctuations, may explain the observed concentration of landslides around the current tongue of the Great Aletsch Glacier.
Fluctuating ice in an alpine valley has a strong influence on the criticality of rock slopes. Retreating ice places adjacent slopes into a more critical state (reducing normal stress and increasing shear stress on joints). Critically stressed joints may be more susceptible to fatigue weathering processes resulting in time-dependent damage. Coupled processes acting in parallel with glacial cycles, e.g., changing ground temperatures or hillslope hydrology, should be considered in order to more broadly evaluate the efficacy of glacial cycles as a preparatory factor for paraglacial rock slope instabilities. Such processes likely play a significant role in creating new rock mass damage during deglaciation.
We demonstrate the importance of exploring paraglacial rock mechanics beyond simple glacial debuttressing through physically and geologically meaningful numerical models. We highlight proper modeling assumptions essential for implementing glacial ice into mechanical models over long time-scales, arguing for the use of stress boundary conditions rather than simulating ice as an elastic material. Furthermore, we showed that initial conditions are crucial for obtaining valid model results. We argue for using an ice-free valley during the Eemian interglacial as initial conditions, rather than peak LGM ice occupation, in order to avoid overestimating new damage during deglaciation. An entire transient glacial cycle should be considered to evaluate rock slope damage, since both glacier retreat and advance affect the valley's in situ stress conditions.

Acknowledgments

This project was funded by the Swiss National Science Foundation (projects 135184 and 146593). The data used for this paper are properly cited and referred to in the reference list. The data output is included in the tables and figures. Raw data are available on request from L.G. (e-mail: [email protected]). We thank Marcus Christl and the Ion Beam Physics group at ETH Zurich for the ¹⁰Be accelerator mass spectrometry measurements. Special thanks to Johnny Sanders for field assistance and initial data collection and Christian Wirsig for help in the lab. Thanks to Kerry Leith and Florian Amann for fruitful discussions and to Martin Funk and Martin Lüthi for input on the behavior of ice. Andreas Bauder provided the ice thickness distribution data of the Great Aletsch Glacier. Constructive comments from Sam McColl and Stuart Dunning, as well as the associate editor, are greatly appreciated and helped improve this manuscript.

Appendix A

A1 Cosmogenic Nuclide Exposure Dating of the Driest Instability

Exposure dating, exploiting the concentration of in situ cosmogenic nuclides produced by cosmic rays [Ivy-Ochs and Kober, 2008], can be used to determine the initiation ages and paleoslip rates of landslides [e.g., Hermanns et al., 2013; Zerathe et al., 2014]. The Driest instability is a promising site for exposure dating in the Aletsch area since it has a clear head scarp that was not covered by the LIA glacier. We collected five samples along a transect down the scarp to constrain the initial age of the Driest instability by using cosmogenic ¹⁰Be (Figures 2 and 4). DRIEST 01 was taken from glacially polished bedrock above the LIA trimline but 50 m below the Egesen moraine. DRIEST 02 was located on a 3 m high wall behind the main back-scarp. DRIEST 03–05 were located on the 50–70 m high main backscarp. Sampling locations showing evidence of recent slabbing were avoided.

Sample preparation, quartz separation, and Be extraction were undertaken according to procedures described by Ivy-Ochs [1996]. Total Be and ¹⁰Be measurements were carried out on the 600 kV TANDY system [Christl et al., 2013] at the accelerator mass spectrometry facility of the Laboratory of Ion Beam Physics, ETH Zurich (Table A1). The ETH internal standard S2007 N, calibrated against the primary 07KNSTD standard, was used to normalize the ¹⁰Be/⁹Be ratios of the samples [Christl et al., 2013]. Measurements were corrected by subtracting full process chemistry blanks with a ¹⁰Be/⁹Be ratio of (3.6 ± 2.6) × 10⁻¹⁵. Exposure ages were calculated with the CRONUS-Earth online calculator [Balco et al., 2008] by using local production rates derived from the NENA calibration data set [Balco et al., 2009] and a time-dependent spallation production model [Lal, 1991; Stone, 2000]. Corrections for topographic shielding were calculated with the CRONUS-Earth online calculator [Balco et al., 2008]. A surface erosion rate of 1 mm kyr⁻¹ was assumed for all samples, while corrections for snow cover were not included. DRIEST 04 was lost during sample processing. Calculated exposure ages are shown in Figure 4 and Table A1.

Table A1. Sample Names, Measured ¹⁰Be Concentrations, and Calculated Exposure Ages of the Driest Instability

Sample Name	Latitude	Longitude	Elevation	Thickness	Density	Shielding Correction Factor	¹⁰Be	¹⁰Be Error	Exposure Age	External Uncertainty
Units	(DD.DDDD)	(DD.DDDD)	(m asl)	(cm)	(g cm⁻³)	()	(10⁴ at. g⁻¹)	(10⁴ at. g⁻¹)	kyr	kyr
DRIEST 1	46.4057	8.0245	2163	3.5	2.65	0.989	22.66	1.22	10.42	0.76
DRIEST 2	46.4056	8.0245	2157	3.0	2.65	0.752	14.62	2.01	8.77	1.29
DRIEST 3	46.4054	8.0244	2148	2.0	2.65	0.789	13.12	0.70	7.46	0.54
DRIEST 4	46.4054	8.0243	2143	4.0	2.65	0.701
DRIEST 5	46.4053	8.0242	2129	2.0	2.65	0.752	12.18	1.20	7.37	0.82

DRIEST 01 was sampled from glacially polished bedrock exposed by the retreating Egesen glacier. Therefore, the expected exposure age is close to Egesen. Recalculation (using the NENA production rate [Balco et al., 2009]) of nearby bedrock exposure ages within the Egesen extent from Schindelwig et al. [2012] resulted in a mean age of 13.7 ± 1.0 kyr (Figures 2 and 4). These ages, slightly older than Egesen (during YD dated at 12.8–11.5 kyr B.P. [Alley et al., 1993]), may indicate inherited nuclides from pre-exposure [Ivy-Ochs and Kober, 2008] during the preceding ice-free Bølling/Allerød. The calculated exposure age of DRIEST 01 (10.4 ± 0.8 kyr), however, is slightly younger than Egesen. DRIEST 03 and 05, with a mean exposure age of 7.4 ± 0.7 kyr, best represent the initiation age of sliding along the head scarp. Initial exposure of the head scarp thus seems to have occurred during the Holocene Climatic Optimum and not directly following LGM or Egesen ice retreat. DRIEST 02, not being located on the main head scarp but also not showing distinct marks of glacial erosion, has an intermediate exposure age (8.8 ± 1.3 kyr). The data are too sparse to calculate estimates of paleo slip rates, but nonetheless, more than 20 m (elevation difference between DRIEST 03 and 05) of the sliding surface was exposed in a relatively short period (within the uncertainty of the dating method).

A2 Modeling Approach for Glacial Ice

Changing glacial ice loading is the main factor driving rock slope damage in this study. The mechanical behavior of ice on different time scales and its effect on adjacent slopes is complex. An adequate modeling approach is therefore crucial to simulate glacial ice loading in a realistic manner. Here we present a comparison between modeling glacial ice as a hydrostatic stress boundary condition versus as an elastic material. The comparison emphasizes differences in the in situ stress field of underlying bedrock by using these two approaches and the resulting rock slope damage.

Our extended model geometry presented in Figure 6 was initialized elastically with a glacier level at 2200 m asl. We then subsequently allowed plastic deformation under the ice load. The glacier ice was removed completely under elastic conditions, and a new mechanical equilibrium was established, before plasticity was again allowed. The same procedure was calculated with hydrostatic stress boundary conditions (Figure A1a) and with an elastic ice body filling the valley (Figure A1b). A density of 917 kg m⁻³, Young's modulus (E) of 10 GPa, and a Poisson's ratio (ν) of 0.3 were assumed for the elastic properties of ice [Schulson, 1999], similar to other studies [e.g., Eberhardt et al., 2004].

The results of our comparison are shown in Figure A1 and reveal that modeling the glacier as a stress boundary produces more initial damage in the rock slope than treating ice as an elastic material. In the former case, the glacier provides less lateral support and the situation is closer to an ice-free valley. In the latter case, almost no damage appears initially underneath the ice due to the strong buttressing effect of the elastic ice body. Removing the ice leads to an extreme increase in rock slope damage for the elastic material case (+351% of initial damage) compared to simulating ice as a stress boundary (+85% of initial damage) (Figures A1a and A1b). Both models are similar in how the vertical glacier load affects underlying bedrock (Figure A1d). However, we observe higher horizontal stress in the valley bottom for the stress-boundary condition as compared to the elastic assumption (Figure A1c). Vice versa, higher horizontal stresses exist at the glacier surface elevation for the elastic assumption. Subsequent slope debuttressing causes a strong damage increase for the elastic material approach.

This brief comparison shows that modeling a valley glacier as an elastic material provides significantly more lateral confinement to adjacent rock slopes and reduces stresses in the valley bottom. Modeling ice as an elastic material will thus lead to significantly overestimated damage accompanying glacier retreat. Our comparison further shows that stress redistribution in the adjacent rock slope is also not comparable between reducing the thickness of a valley glacier by 100 m (ρ = 917 kg m⁻³) and eroding 30 m of rock at the valley bottom (ρ = 2700 kg m⁻³), since the latter provides strong lateral support, unlike plastic ice. A modeling approach for glacial ice assuming elastic material properties might be reasonable for short-term rapid loading [e.g., McColl et al., 2012] but not for long-term mechanical studies on glacial time scales.

References

Agliardi, F., and G. Crosta (2014), Long-and short-term controls on the Spriana rockslide (Central Alps, Italy), in Landslide Science for a Safer Geoenvironment, edited by K. Sassa, P. Canuti, and Y. Yin, pp. 243–249, Springer International Publishing, Cham, Switzerland, doi:10.1007/978-3-319-05050-8_39.
10.1007/978-3-319-05050-8_39
Google Scholar
Alley, R. B., D. D. Blankenship, C. R. Bentley, and S. T. Rooney (1986), Deformation of till beneath ice stream B, West Antarctica, Nature, 322(6074), 57–59.
10.1038/322057a0
ADSWeb of Science®Google Scholar
Alley, R. B., et al. (1993), Abrupt increase in Greenland snow accumulation at the end of the Younger Dryas event, Nature, 362(6420), 527–529.
10.1038/362527a0
ADSWeb of Science®Google Scholar
Augustinus, P. C. (1995), Glacial valley cross-profile development: The influence of in situ rock stress and rock mass strength, with examples from the Southern Alps, New Zealand, Geomorphology, 14(2), 87–97, doi:10.1016/0169-555x(95)00050-x.
10.1016/0169-555x(95)00050-x
ADSWeb of Science®Google Scholar
Aydin, A. (2015), ISRM suggested method for determination of the Schmidt Hammer Rebound Hardness: Revised version, in The ISRM Suggested Methods for Rock Characterization, Testing and Monitoring: 2007–2014, edited by R. Ulusay, pp. 25–33, Springer International, Cham, Switzerland, doi:10.1007/978-3-319-07713-0.
10.1007/978-3-319-07713-0
Google Scholar
Balco, G., J. O. Stone, N. A. Lifton, and T. J. Dunai (2008), A complete and easily accessible means of calculating surface exposure ages or erosion rates from 10Be and 26Al measurements, Quat. Geochronol., 3(3), 174–195, doi:10.1016/j.quageo.2007.12.001.
10.1016/j.quageo.2007.12.001
Web of Science®Google Scholar
Balco, G., J. Briner, R. C. Finkel, J. A. Rayburn, J. C. Ridge, and J. M. Schaefer (2009), Regional beryllium-10 production rate calibration for late-glacial northeastern North America, Quat. Geochronol., 4(2), 93–107, doi:10.1016/j.quageo.2008.09.001.
10.1016/j.quageo.2008.09.001
Web of Science®Google Scholar
Ballantyne, C. K. (2002), Paraglacial geomorphology, Quat. Sci. Rev., 21(18–19), 1935–2017, doi:10.1016/s0277-3791(02)00005-7.
10.1016/s0277-3791(02)00005-7
ADSWeb of Science®Google Scholar
Ballantyne, C. K., G. F. Sandeman, J. O. Stone, and P. Wilson (2014a), Rock-slope failure following Late Pleistocene deglaciation on tectonically stable mountainous terrain, Quat. Sci. Rev., 86, 144–157, doi:10.1016/j.quascirev.2013.12.021.
10.1016/j.quascirev.2013.12.021
ADSWeb of Science®Google Scholar
Ballantyne, C. K., P. Wilson, D. Gheorghiu, and À. Rodés (2014b), Enhanced rock-slope failure following ice-sheet deglaciation: Timing and causes, Earth Surf. Processes Landforms, 39(7), 900–913, doi:10.1002/esp.3495.
10.1002/esp.3495
ADSWeb of Science®Google Scholar
Baroni, C., S. Martino, M. C. Salvatore, G. Scarascia Mugnozza, and L. Schilirò (2014), Thermomechanical stress-strain numerical modelling of deglaciation since the Last Glacial Maximum in the Adamello Group (Rhaetian Alps, Italy), Geomorphology, 226, 278–299, doi:10.1016/j.geomorph.2014.08.013.
10.1016/j.geomorph.2014.08.013
ADSWeb of Science®Google Scholar
Bevis, M., et al. (2012), Bedrock displacements in Greenland manifest ice mass variations, climate cycles and climate change, Proc. Natl. Acad. Sci. U.S.A., 109(30), 11,944–11,948, doi:10.1073/pnas.1204664109.
10.1073/pnas.1204664109
CASADSWeb of Science®Google Scholar
Blunier, T., and E. J. Brook (2001), Timing of millennial-scale climate change in Antarctica and Greenland during the last glacial period, Science, 291(5501), 109–112, doi:10.1126/science.291.5501.109.
10.1126/science.291.5501.109
CASADSPubMedWeb of Science®Google Scholar
Boulton, G., T. Chan, R. Christiansson, L. O. Ericsson, J. Hartikainen, M. R. Jensen, F. W. Stanchell, and T. Wallroth (2004), Thermo-hydro-mechanical (T-H-M) impacts of glaciation and implications for deep geologic disposal of nuclear waste, in Elsevier Geo-Engineering Book Series, edited by S. Ove, pp. 299–304, Elsevier, Amsterdam, doi:10.1016/S1571-9960(04)80057-0.
10.1016/S1571-9960(04)80057-0
Google Scholar
Bovis, M. J. (1990), Rock-slope deformation at Affliction Creek, southern Coast Mountains, British Columbia, Can. J. Earth Sci., 27(2), 243–254, doi:10.1139/e90-024.
10.1139/e90-024
ADSWeb of Science®Google Scholar
Chan, T., R. Christiansson, G. S. Boulton, L. O. Ericsson, J. Hartikainen, M. R. Jensen, D. Mas Ivars, F. W. Stanchell, P. Vistrand, and T. Wallroth (2005), DECOVALEX III BMT3/BENCHPAR WP4: The thermo-hydro-mechanical responses to a glacial cycle and their potential implications for deep geological disposal of nuclear fuel waste in a fractured crystalline rock mass, Int. J. Rock Mech. Mining Sci., 42(5–6), 805–827, doi:10.1016/j.ijrmms.2005.03.017.
10.1016/j.ijrmms.2005.03.017
Web of Science®Google Scholar
Christl, M., C. Vockenhuber, P. W. Kubik, L. Wacker, J. Lachner, V. Alfimov, and H. A. Synal (2013), The ETH Zurich AMS facilities: Performance parameters and reference materials, Nucl. Instruments Methods Phys. Res. Section B, 294, 29–38, doi:10.1016/j.nimb.2012.03.004.
10.1016/j.nimb.2012.03.004
CASADSWeb of Science®Google Scholar
Clayton, M. A., D. Stead, and D. Kinakin (2013), The Mitchell Creek Landslide, B.C., Canada: Investigation using remote sensing and numerical modeling, 47th U.S. Rock Mechanics/Geomechanics Symposium, pp. 180-187, American Rock Mechanics Association, San Francisco, Calif.
Google Scholar
Cohen, D., R. L. Hooke, N. R. Iverson, and J. Kohler (2000), Sliding of ice past an obstacle at Engabreen, Norway, J. Glaciol., 46(155), 599–610.
10.3189/172756500781832747
ADSWeb of Science®Google Scholar
Cohen, D., N. R. Iverson, T. S. Hooyer, U. H. Fischer, M. Jackson, and P. L. Moore (2005), Debris-bed friction of hard-bedded glaciers, J. Geophys. Res., 110, F02007, doi:10.1029/2004JF000228.
10.1029/2004JF000228
CASADSWeb of Science®Google Scholar
Cohen, D., T. S. Hooyer, N. R. Iverson, J. F. Thomason, and M. Jackson (2006), Role of transient water pressure in quarrying: A subglacial experiment using acoustic emissions, J. Geophys. Res., 111, F03006, doi:10.1029/2005JF000439.
10.1029/2005JF000439
ADSPubMedWeb of Science®Google Scholar
Cossart, E., R. Braucher, M. Fort, D. L. Bourlès, and J. Carcaillet (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(1–2), 3–26, doi:10.1016/j.geomorph.2006.12.022.
10.1016/j.geomorph.2006.12.022
ADSWeb of Science®Google Scholar
Cossart, E., D. Mercier, A. Decaulne, T. Feuillet, H. P. Jónsson, and Þ. Sæmundsson (2014), Impacts of post-glacial rebound on landslide spatial distribution at a regional scale in northern Iceland (Skagafjörður), Earth Surf. Processes Landforms, 39(3), 336–350, doi:10.1002/esp.3450.
10.1002/esp.3450
ADSWeb of Science®Google Scholar
Crisinel, A. (1978), Geologie de la reserve naturelle de la forêt d'Aletsch (Valais-Suisse), Bull. Murithienne, 95, 45–58.
Google Scholar
Crosta, G. B., P. Frattini, and F. Agliardi (2013), Deep seated gravitational slope deformations in the European Alps, Tectonophysics, 605, 13–33, doi:10.1016/j.tecto.2013.04.028.
10.1016/j.tecto.2013.04.028
ADSWeb of Science®Google Scholar
Cuffey, K. M., and W. S. B. Paterson (2010), The Physics of Glaciers, Butterworth-Heinemann/Elsevier, Burlington.
Google Scholar
Cundall, P. A., and R. D. Hart (1992), Numerical modelling of discontinua, Eng. Comput., 9(2), 101–113, doi:10.1108/eb023851.
10.1108/eb023851
Google Scholar
Dahl-Jensen, D., et al. (2013), Eemian interglacial reconstructed from a Greenland folded ice core, Nature, 493(7433), 489–494, doi:10.1038/nature11789.
10.1038/nature11789
CASADSPubMedWeb of Science®Google Scholar
Darnault, R., Y. Rolland, R. Braucher, D. Bourlès, M. Revel, G. Sanchez, and S. Bouissou (2012), Timing of the last deglaciation revealed by receding glaciers at the Alpine-scale: Impact on mountain geomorphology, Quat. Sci. Rev., 31, 127–142, doi:10.1016/j.quascirev.2011.10.019.
10.1016/j.quascirev.2011.10.019
ADSWeb of Science®Google Scholar
Deline, P. (2009), Interactions between rock avalanches and glaciers in the Mont Blanc massif during the late Holocene, Quat. Sci. Rev., 28(11–12), 1070–1083, doi:10.1016/j.quascirev.2008.09.025.
10.1016/j.quascirev.2008.09.025
ADSWeb of Science®Google Scholar
Eberhardt, E., D. Stead, and J. S. Coggan (2004), Numerical analysis of initiation and progressive failure in natural rock slopes—The 1991 Randa rockslide, Int. J. Rock Mech. Mining Sci., 41(1), 69–87, doi:10.1016/s1365-1609(03)00076-5.
10.1016/s1365-1609(03)00076-5
Web of Science®Google Scholar
Eckardt, P., H. Funk, and T. Labhart (1983), Postglaziale Krustenbewegungen an der Rhein-Rhone-Linie, Mensuration, Photogramm., Génie rural, 2, 43–65.
Google Scholar
Evans, S. G., and J. J. Clague (1994), Recent climatic change and catastrophic geomorphic processes in mountain environments, Geomorphology, 10(1–4), 107–128, doi:10.1016/0169-555x(94)90011-6.
10.1016/0169-555X(94)90011-6
ADSWeb of Science®Google Scholar
Faillettaz, J., D. Sornette, and M. Funk (2010), Gravity-driven instabilities: Interplay between state- and velocity-dependent frictional sliding and stress corrosion damage cracking, J. Geophys. Res., 115, B03409, doi:10.1029/2009JB006512.
10.1029/2009JB006512
ADSPubMedWeb of Science®Google Scholar
Farinotti, D., M. Huss, A. Bauder, and M. Funk (2009), An estimate of the glacier ice volume in the Swiss Alps, Global Planet. Change, 68(3), 225–231, doi:10.1016/j.gloplacha.2009.05.004.
10.1016/j.gloplacha.2009.05.004
ADSWeb of Science®Google Scholar
Fischer, L., F. Amann, J. R. Moore, and C. Huggel (2010), Assessment of periglacial slope stability for the 1988 Tschierva rock avalanche (Piz Morteratsch, Switzerland), Eng. Geol., 116(1–2), 32–43, doi:10.1016/j.enggeo.2010.07.005.
10.1016/j.enggeo.2010.07.005
Web of Science®Google Scholar
Gischig, V., F. Amann, J. R. Moore, S. Loew, H. Eisenbeiss, and W. Stempfhuber (2011), Composite rock slope kinematics at the current Randa instability, Switzerland, based on remote sensing and numerical modeling, Eng. Geol., 118(1–2), 37–53, doi:10.1016/j.enggeo.2010.11.006.
10.1016/j.enggeo.2010.11.006
Web of Science®Google Scholar
Gischig, V., G. Preisig, and E. Eberhardt (2015), Numerical investigation of seismically induced rock mass fatigue as a mechanism contributing to the progressive failure of deep-seated landslides, Rock Mech. Rock Eng., 49, 2457–2478, doi:10.1007/s00603-015-0821-z.
10.1007/s00603-015-0821-z
ADSWeb of Science®Google Scholar
Gischig, V. S., J. R. Moore, K. F. Evans, F. Amann, and S. Loew (2011a), Thermomechanical forcing of deep rock slope deformation: 1. Conceptual study of a simplified slope, J. Geophys. Res., 116, F04010, doi:10.1029/2011JF002006.
10.1029/2011JF002006
Web of Science®Google Scholar
Gischig, V. S., J. R. Moore, K. F. Evans, F. Amann, and S. Loew (2011b), Thermomechanical forcing of deep rock slope deformation: 2. The Randa rock slope instability, J. Geophys. Res., 116, F04011, doi:10.1029/2011JF002007.
10.1029/2011JF002007
Google Scholar
Glaus, R. (1992), Rutschungsmessungen im Gebiet Blatten-Belalp. Diplomvermessungskurs 1991 Naters, Rep., Institute für Geodäsie und Photogrammetrie, ETH Zürich.
Google Scholar
Goehring, B. M., J. M. Schaefer, C. Schluechter, N. A. Lifton, R. C. Finkel, A. J. T. Jull, N. Akçar, and R. B. Alley (2011), The Rhone Glacier was smaller than today for most of the Holocene, Geology, doi:10.1130/g32145.1.
10.1130/g32145.1
ADSWeb of Science®Google Scholar
Goodman, D. J., G. C. P. King, D. H. M. Millar, and Q. Robin (1979), Pressure-melting effects in basal ice of temperate glaciers: Laboratory studies and field observations under glacier d'Argentière, J. Glaciol., 23(89), 259–271.
10.1017/S0022143000029889
ADSGoogle Scholar
Haeuselmann, P., D. E. Granger, P.-Y. Jeannin, and S.-E. Lauritzen (2007), Abrupt glacial valley incision at 0.8 Ma dated from cave deposits in Switzerland, Geology, 35(2), 143–146, doi:10.1130/g23094a.
10.1130/G23094A
ADSWeb of Science®Google Scholar
Hagen, J. O., O. Liestøl, J. L. Sollid, B. Wold, and G. Østrem (1993), Subglacial investigations at Bondhusbreen, Folgefonni, Norway, Norsk Geografisk Tidsskrift—Norwegian J. Geogr., 47(3), 117–162, doi:10.1080/00291959308621974.
10.1080/00291959308621974
Google Scholar
Hales, T. C., and J. J. Roering (2007), Climatic controls on frost cracking and implications for the evolution of bedrock landscapes, J. Geophys. Res., 112, F02033, doi:10.1029/2006JF000616.
10.1029/2006JF000616
ADSWeb of Science®Google Scholar
Hallet, B., L. Hunter, and J. Bogen (1996), Rates of erosion and sediment evacuation by glaciers: A review of field data and their implications, Global Planet. Change, 12(1–4), 213–235, doi:10.1016/0921-8181(95)00021-6.
10.1016/0921-8181(95)00021-6
ADSWeb of Science®Google Scholar
Hansmann, J., S. Loew, and K. Evans (2012), Reversible rock-slope deformations caused by cyclic water-table fluctuations in mountain slopes of the Central Alps, Switzerland, Hydrogeol. J., 20(1), 73–91, doi:10.1007/s10040-011-0801-7.
10.1007/s10040-011-0801-7
ADSWeb of Science®Google Scholar
Hermanns, R. L., T. Oppikofer, H. Dahle, and T. Eiken (2013), Understanding long-term slope deformation for stability assessment of rock slopes: The case of the Oppstadhornet rockslide, Norway, paper presented at International Conference Vajont 1963–2013. Thoughts and analyses after 50 years since the catastrophic landslide, Padua, Italy—8–10 October 2013.
Google Scholar
Hock, R., A. Iken, and A. Wangler (1999), Tracer experiments and borehole observations in the overdeepening of Aletschgletscher, Switzerland, Ann. Glaciol., 28(1), 253–260, doi:10.3189/172756499781821742.
10.3189/172756499781821742
ADSWeb of Science®Google Scholar
Hoek, E., and E. T. Brown (1997), Practical estimates of rock mass strength, Int. J. Rock Mech. Mining Sci., 34(8), 1165–1186.
10.1016/S1365-1609(97)80069-X
Web of Science®Google Scholar
Hoek, E., C. Carranza-Torres, and B. Corkum (2002), Hoek-Brown failure criterion—2002 edition, Proceedings NARMS-TAC Conference, pp. 267–273, Toronto.
Google Scholar
Holm, K., M. Bovis, and M. Jakob (2004), The landslide response of alpine basins to post-Little Ice Age glacial thinning and retreat in southwestern British Columbia, Geomorphology, 57(3–4), 201–216, doi:10.1016/s0169-555x(03)00103-x.
10.1016/s0169-555x(03)00103-x
ADSWeb of Science®Google Scholar
Holzhauser, H.. (1995), Gletscherschwankungen innerhalb der letzten 3200 Jahre am Beispiel des Grossen Aletsch- und des Gornergletschers. Neue Ergebnisse, vdf Hochschulverlag AG an der ETH Zürich, ETH Zürich.
Google Scholar
Holzhauser, H., M. Magny, and H. J. Zumbühl (2005), Glacier and lake-level variations in west-central Europe over the last 3500 years, Holocene, 15(6), 789–801, doi:10.1191/0959683605hl853ra.
10.1191/0959683605hl853ra
ADSWeb of Science®Google Scholar
Hormes, A., B. U. Müller, and C. Schlüchter (2001), The Alps with little ice: Evidence for eight Holocene phases of reduced glacier extent in the Central Swiss Alps, Holocene, 11(3), 255–265, doi:10.1191/095968301675275728.
10.1191/095968301675275728
ADSWeb of Science®Google Scholar
Hudson, A. J., and P. J. Harrison (1997), Engineering Rock Mechanics—An Introduction to the Principles, Elsevier, Amsterdam.
Google Scholar
Itasca (2014), UDEC—Universal Distinct Element Code, Version 6.0, User's Manual, Itasca Consulting Group, Inc., Minneapolis.
Google Scholar
Iverson, N. R., R. W. Baker, R. L. Hooke, H. Brian, and J. Peter (1999), Coupling between a glacier and a soft bed. I. A relation between effective pressure and local shear stress determined from till elasticity, J. Glaciol., 45(149), 31–40.
10.1017/S0022143000003014
ADSWeb of Science®Google Scholar
Iverson, N. R., T. S. Hooyer, U. H. Fischer, D. Cohen, P. L. Moore, M. Jackson, G. Lappegard, and J. Kohler (2007), Soft-bed experiments beneath Engabreen, Norway: Regelation infiltration, basal slip and bed deformation, J. Glaciol., 53(182), 323–340, doi:10.3189/002214307783258431.
10.3189/002214307783258431
ADSWeb of Science®Google Scholar
Ivy-Ochs, S. (1996), The dating of rock surfaces using in situ produced 10Be, 26Al and 36Cl, with examples from Antarctica and the Swiss Alps, Doctoral thesis, ETH Zurich, Zurich.
Google Scholar
Ivy-Ochs, S. (2015), Glacier variations in the European Alps at the end of the last glaciation, Cuad. Invest. Geogr., 41(2), 295–315, doi:10.18172/cig.2750.
10.18172/cig.2750
Web of Science®Google Scholar
Ivy-Ochs, S., and F. Kober (2008), Surface exposure dating with cosmogenic nuclides, Quat. Sci. J., 57(1–2), 179–209, doi:10.3285/eg.57.1-2.7.
10.3285/eg.57.1-2.7
Google Scholar
Ivy-Ochs, S., H. Kerschner, A. Reuther, F. Preusser, K. Heine, M. Maisch, P. W. Kubik, and C. Schlüchter (2008), Chronology of the last glacial cycle in the European Alps, J. Quat. Sci., 23(6-7), 559–573, doi:10.1002/jqs.1202.
10.1002/jqs.1202
ADSWeb of Science®Google Scholar
Ivy-Ochs, S., A. V. Poschinger, H. A. Synal, and M. Maisch (2009a), Surface exposure dating of the Flims landslide, Graubünden, Switzerland, Geomorphology, 103(1), 104–112, doi:10.1016/j.geomorph.2007.10.024.
10.1016/j.geomorph.2007.10.024
ADSWeb of Science®Google Scholar
Ivy-Ochs, S., H. Kerschner, M. Maisch, M. Christl, P. W. Kubik, and C. Schlüchter (2009b), Latest Pleistocene and Holocene glacier variations in the European Alps, Quat. Sci. Rev., 28(21–22), 2137–2149, doi:10.1016/j.quascirev.2009.03.009.
10.1016/j.quascirev.2009.03.009
ADSWeb of Science®Google Scholar
Jaboyedoff, M., F. Baillifard, E. Bardou, and F. Girod (2004), The effect of weathering on Alpine rock instability, Q. J. Eng. Geol. Hydrogeol., 37(2), 95–103, doi:10.1144/1470-9236/03-046.
10.1144/1470-9236/03-046
CASWeb of Science®Google Scholar
Jaboyedoff, M., M.-H. Derron, J. Jakubowski, T. Oppikofer, and A. Pedrazzini (2012), The 2006 Eiger Rockslide, European Alps, 1st ed., Cambridge Univ. Press, Cambridge.
10.1017/CBO9780511740367.024
Google Scholar
Jennings, J. E. (1970), A mathematical theory for the calculation of the stability of open cast mines, paper presented at Symposium on the Theoretical Background to the Planning of Open Pit Mines, Johannesburg.
Google Scholar
Joerin, U. E., T. F. Stocker, and C. Schlüchter (2006), Multicentury glacier fluctuations in the Swiss Alps during the Holocene, Holocene, 16(5), 697–704, doi:10.1191/0959683606hl964rp.
10.1191/0959683606hl964rp
ADSWeb of Science®Google Scholar
Jouvet, G., M. Huss, M. Funk, and H. Blatter (2011), Modelling the retreat of Grosser Aletschgletscher, Switzerland, in a changing climate, J. Glaciol., 57(206), 1033–1045, doi:10.3189/002214311798843359.
10.3189/002214311798843359
ADSWeb of Science®Google Scholar
Kääb, A. (2002), Monitoring high-mountain terrain deformation from repeated air- and spaceborne optical data: Examples using digital aerial imagery and ASTER data, ISPRS J. Photogram. Remote Sens., 57(1–2), 39–52, doi:10.1016/s0924-2716(02)00114-4.
10.1016/s0924-2716(02)00114-4
ADSWeb of Science®Google Scholar
Kasser, P., M. Aellen, and H. Siegenthaler (1982), Die Gletscher der Schweizer Alpen 1973/74 und 1974/75, Rep., 159 pp, Zürich.
Google Scholar
Kastrup, U., M. L. Zoback, N. Deichmann, K. F. Evans, D. Giardini, and A. J. Michael (2004), Stress field variations in the Swiss Alps and the northern Alpine foreland derived from inversion of fault plane solutions, J. Geophys. Res., 109, B01402, doi:10.1029/2003JB002550.
10.1029/2003JB002550
ADSWeb of Science®Google Scholar
Kelly, M. A., P. W. Kubik, F. Von Blanckenburg, and C. Schlüchter (2004a), Surface exposure dating of the Great Aletsch Glacier Egesen moraine system, western Swiss Alps, using the cosmogenic nuclide 10Be, J. Quat. Sci., 19(5), 431–441, doi:10.1002/jqs.854.
10.1002/jqs.854
ADSWeb of Science®Google Scholar
Kelly, M. A., J. F. Buoncristiani, and C. Schluchter (2004b), A reconstruction of the Last Glacial Maximum (LGM) ice-surface geometry in the western Swiss Alps and contiguous Alpine regions in Italy and France, Eclogae Geol. Helv., 97(1), 57–75, doi:10.1007/s00015-004-1109-6.
10.1007/s00015-004-1109-6
Web of Science®Google Scholar
Khan, S. A., L. Liu, J. Wahr, I. Howat, I. Joughin, T. van Dam, and K. Fleming (2010), GPS measurements of crustal uplift near Jakobshavn Isbræ due to glacial ice mass loss, J. Geophys. Res., 115, B09405, doi:10.1029/2010JB007490.
10.1029/2010JB007490
ADSPubMedWeb of Science®Google Scholar
Kinakin, D., and D. Stead (2005), Analysis of the distributions of stress in natural ridge forms: Implications for the deformation mechanisms of rock slopes and the formation of sackung, Geomorphology, 65(1–2), 85–100, doi:10.1016/j.geomorph.2004.08.002.
10.1016/j.geomorph.2004.08.002
ADSWeb of Science®Google Scholar
Koppes, M. N., and D. R. Montgomery (2009), The relative efficacy of fluvial and glacial erosion over modern to orogenic timescales, Nat. Geosci., 2(9), 644–647.
10.1038/ngeo616
CASADSWeb of Science®Google Scholar
Kos, A., F. Amann, T. Strozzi, R. Delaloye, J. von Ruette, and S. Springman (2016), Contemporary glacier retreat triggers a rapid landslide response, Great Aletsch Glacier, Switzerland, Geophys. Res. Lett., 43, 12,466–12,474, doi:10.1002/2016GL071708.
10.1002/2016GL071708
ADSWeb of Science®Google Scholar
Krautblatter, M., D. Funk, and F. K. Günzel (2013), Why permafrost rocks become unstable: A rock-ice-mechanical model in time and space, Earth Surf. Processes Landforms, 38(8), 876–887, doi:10.1002/esp.3374.
10.1002/esp.3374
ADSWeb of Science®Google Scholar
Lal, D. (1991), Cosmic ray labeling of erosion surfaces: In situ nuclide production rates and erosion models, Earth Planet. Sci. Lett., 104(2–4), 424–439, doi:10.1016/0012-821X(91)90220-C.
10.1016/0012-821X(91)90220-C
CASADSWeb of Science®Google Scholar
Landestopographie, and VAW (1962), Aletschgletscher Kartenmaterial—Gletscherausdehnung 1957, Wabern-Bern.
Google Scholar
Leith, K., J. R. Moore, F. Amann, and S. Loew (2014a), Subglacial extensional fracture development and implications for Alpine Valley evolution, J. Geophys. Res. Earth Surf., 119, 62–81, doi:10.1002/2012JF002691.
10.1002/2012JF002691
ADSWeb of Science®Google Scholar
Leith, K., J. R. Moore, F. Amann, and S. Loew (2014b), In situ stress control on microcrack generation and macroscopic extensional fracture in exhuming bedrock, J. Geophys. Res. Solid Earth, 119, 594–615, doi:10.1002/2012JB009801.
10.1002/2012JB009801
ADSWeb of Science®Google Scholar
Loew, S., V. S. Gischig, F. Glueer, R. Seifert, and J. R. Moore (2017), Multidisciplinary monitoring of progressive failure processes in brittle rock slopes, in Rock Mechanics and Engineering, edited by X.-T. Feng and J. A. Hudson, CRC Press/Balkema-Taylor and Francis, London.
Google Scholar
Lorig, L. J., and P. A. Cundall (1989), Modeling of reinforced concrete using the distinct element method, in Fracture of Concrete and Rock, edited by S. P. Shah, and S. E. Swartz, pp. 276–287, Springer, New York, doi:10.1007/978-1-4612-3578-1_28.
10.1007/978-1-4612-3578-1_28
Google Scholar
Luetscher, M., D. L. Hoffmann, S. Frisia, and C. Spötl (2011), Holocene glacier history from alpine speleothems, Milchbach cave, Switzerland, Earth Planet. Sci. Lett., 302(1–2), 95–106, doi:10.1016/j.epsl.2010.11.042.
10.1016/j.epsl.2010.11.042
CASADSWeb of Science®Google Scholar
Maisch, M., A. Wipf, B. Denneler, J. Battaglia, and C. Benz (1999), Die Gletscher der Schweizer Alpen: Gletscherhochstand 1850, Aktuelle Vergletscherung, Gletscherschwund-Szenarien. Schlussbericht NFP 31, vdf Hochschulverlag.
Google Scholar
Masset, O., and S. Loew (2010), Hydraulic conductivity distribution in crystalline rocks, derived from inflows to tunnels and galleries in the Central Alps, Switzerland, Hydrogeol. J., 18(4), 863–891, doi:10.1007/s10040-009-0569-1.
10.1007/s10040-009-0569-1
CASADSWeb of Science®Google Scholar
McColl, S. T. (2012), Paraglacial rock-slope stability, Geomorphology, 153–154, 1–16, doi:10.1016/j.geomorph.2012.02.015.
10.1016/j.geomorph.2012.02.015
ADSWeb of Science®Google Scholar
McColl, S. T., and T. R. H. Davies (2013), Large ice-contact slope movements: Glacial buttressing, deformation and erosion, Earth Surf. Processes Landforms, 38(10), 1102–1115, doi:10.1002/esp.3346.
10.1002/esp.3346
ADSWeb of Science®Google Scholar
McColl, S. T., T. R. H. Davies, and M. J. McSaveney (2010), Glacier retreat and rock-slope stability: Debunking debuttressing, Geologically. Active 11th Congress of the International Association for Engineering Geology and the Environment, pp. 467–474, Auckland, New Zealand.
Google Scholar
McColl, S. T., T. R. H. Davies, and M. J. McSaveney (2012), The effect of glaciation on the intensity of seismic ground motion, Earth Surf. Processes Landforms, 37(12), 1290–1301, doi:10.1002/esp.3251.
10.1002/esp.3251
ADSWeb of Science®Google Scholar
Melini, D., P. Gegout, M. King, B. Marzeion, and G. Spada (2015), On the rebound: Modeling Earth's ever-changing shape, Eos Trans. AGU, 96, doi:10.1029/2015EO033387.
10.1029/2015EO033387
Google Scholar
Memin, A., Y. Rogister, J. Hinderer, M. Llubes, E. Berthier, and J. P. Boy (2009), Ground deformation and gravity variations modelled from present-day ice thinning in the vicinity of glaciers, J. Geodyn., 48(3–5), 195–203, doi:10.1016/j.jog.2009.09.006.
10.1016/j.jog.2009.09.006
ADSWeb of Science®Google Scholar
Moran, A. P., S. Ivy-Ochs, M. Schuh, M. Christl, and H. Kerschner (2016), Evidence of central alpine glacier advances during the Younger Dryas-early Holocene transition period, Boreas, 45(3), 398–410, doi:10.1111/bor.12170.
10.1111/bor.12170
Web of Science®Google Scholar
Muttoni, G., C. Carcano, E. Garzanti, M. Ghielmi, A. Piccin, R. Pini, S. Rogledi, and D. Sciunnach (2003), Onset of major Pleistocene glaciations in the Alps, Geology, 31(11), 989–992, doi:10.1130/g19445.1.
10.1130/G19445.1
ADSWeb of Science®Google Scholar
Nicolussi, K., and C. Schlüchter (2012), The 8.2 ka event—Calendar-dated glacier response in the Alps, Geology, doi:10.1130/g32406.1.
10.1130/g32406.1
ADSWeb of Science®Google Scholar
Nicolussi, K., and G. Patzelt (2000), Discovery of early Holocene wood and peat on the forefield of the Pasterze Glacier, Eastern Alps, Austria, Holocene, 10(2), 191–199, doi:10.1191/095968300666855842.
10.1191/095968300666855842
ADSWeb of Science®Google Scholar
Nicolussi, K., M. Kaufmann, G. Patzelt, J. van der Plicht, and A. Thurner (2005), Holocene tree-line variability in the Kauner Valley, Central Eastern Alps, indicated by dendrochronological analysis of living trees and subfossil logs, Veg. Hist. Archaeobotany, 14, 221–234.
10.1007/s00334-005-0013-y
Web of Science®Google Scholar
Oppikofer, T., M. Jaboyedoff, and H.-R. Keusen (2008), Collapse at the eastern Eiger flank in the Swiss Alps, Nat. Geosci., 1(8), 531–535, doi:10.1038/ngeo258.
10.1038/ngeo258
CASADSWeb of Science®Google Scholar
Prager, C., C. Zangerl, G. Patzelt, and R. Brandner (2008), Age distribution of fossil landslides in the Tyrol (Austria) and its surrounding areas, Nat. Haz. Earth Syst. Sci., 8(2), 377–407, doi:10.5194/nhess-8-377-2008.
10.5194/nhess-8-377-2008
ADSWeb of Science®Google Scholar
Preisig, G., E. Eberhardt, M. Smithyman, A. Preh, and L. Bonzanigo (2016), Hydromechanical rock mass fatigue in deep-seated landslides accompanying seasonal variations in pore pressures, Rock Mech. Rock Eng., 49, 2333–2351, doi:10.1007/s00603-016-0912-5.
10.1007/s00603-016-0912-5
ADSWeb of Science®Google Scholar
Röthlisberger, F., and W. Schneebeli (1979), Genesis of lateral moraine complexes, demonstrated by fossil soils and trunks: Indicators of postglacial climatic fluctuations, in Moraines and Varves: Origin, Genesis, Classification, edited by C. Schlüchter, pp. 387–419, A.A. Balkema, Rotterdam, Netherlands.
Google Scholar
Sanders, J. W., K. M. Cuffey, J. R. Moore, K. R. MacGregor, and J. L. Kavanaugh (2012), Periglacial weathering and headwall erosion in cirque glacier bergschrunds, Geology, doi:10.1130/g33330.1.
10.1130/g33330.1
ADSWeb of Science®Google Scholar
Schimmelpfennig, I., J. M. Schaefer, N. Akçar, S. Ivy-Ochs, R. C. Finkel, and C. Schlüchter (2012), Holocene glacier culminations in the Western Alps and their hemispheric relevance, Geology, doi:10.1130/g33169.1.
10.1130/g33169.1
ADSWeb of Science®Google Scholar
Schindelwig, I., N. Akçar, P. W. Kubik, and C. Schlüchter (2012), Lateglacial and early Holocene dynamics of adjacent valley glaciers in the Western Swiss Alps, J. Quat. Sci., 27(1), 114–124, doi:10.1002/jqs.1523.
10.1002/jqs.1523
ADSWeb of Science®Google Scholar
Schulson, E. M. (1990), The brittle compressive fracture of ice, Acta Metallurgica et Materialia, 38(10), 1963–1976, doi:10.1016/0956-7151(90)90308-4.
10.1016/0956-7151(90)90308-4
Web of Science®Google Scholar
Schulson, E. M. (1999), The structure and mechanical behavior of ice, JOM, 51(2), 21–27, doi:10.1007/s11837-999-0206-4.
10.1007/s11837-999-0206-4
CASADSWeb of Science®Google Scholar
Selvadurai, A. P. S., A. P. Suvorov, and P. A. Selvadurai (2015), Thermo-hydro-mechanical processes in fractured rock formations during a glacial advance, Geosci. Model. Dev., 8(7), 2167–2185, doi:10.5194/gmd-8-2167-2015.
10.5194/gmd-8-2167-2015
CASADSWeb of Science®Google Scholar
Slaymaker, O. (2009), Proglacial, periglacial or paraglacial?, Geol. Soc. London, Spec. Publ., 320(1), 71–84, doi:10.1144/sp320.6.
10.1144/sp320.6
ADSWeb of Science®Google Scholar
Stead, D., and A. Wolter (2015), A critical review of rock slope failure mechanisms: The importance of structural geology, J. Structural Geol., 74, 1–23, doi:10.1016/j.jsg.2015.02.002.
10.1016/j.jsg.2015.02.002
ADSWeb of Science®Google Scholar
Steck, A. (2011), 1269 Aletschgletscher mit Teil von 1249 Finsteraarhorn, Geolog. Atlas der Schweiz 1:25000.
Google Scholar
Steiner, W., U. Graber, and H.-R. Keusen (1996), Construction in rock at 3550 meters elevation (Jungfraujoch, Switzerland), in EUROCK 96, pp. 543–550, International Society for Rock Mechanics, Torino, Italy.
Google Scholar
Stone, J. O. (2000), Air pressure and cosmogenic isotope production, J. Geophys. Res., 105, 23,753–23,759, doi:10.1029/2000JB900181.
10.1029/2000JB900181
CASADSWeb of Science®Google Scholar
Strozzi, T., R. Delaloye, A. Kääb, C. Ambrosi, E. Perruchoud, and U. Wegmüller (2010), Combined observations of rock mass movements using satellite SAR interferometry, differential GPS, airborne digital photogrammetry, and airborne photography interpretation, J. Geophys. Res., 115, F01014, doi:10.1029/2009JF001311.
10.1029/2009JF001311
ADSPubMedWeb of Science®Google Scholar
Terzaghi, K. (1962), Stability of steep slopes on hard unweathered rock, Geotechnique, 12(4), 251–270.
10.1680/geot.1962.12.4.251
Google Scholar
Tinner, W., and J.-P. Theurillat (2003), Uppermost limit, extent, and fluctuations of the timberline and treeline ecocline in the Swiss Central Alps during the past 11,500 years, Arct. Antarct. Alpine Res., 35(2), 158–169, doi:10.1657/1523-0430(2003)035[0158:uleafo]2.0.co;2.
10.1657/1523-0430(2003)035[0158:uleafo]2.0.co;2
Web of Science®Google Scholar
Ustaszewksi, M. E., A. Hampel, and O. A. Pfiffner (2008), Composite faults in the Swiss Alps formed by the interplay of tectonics, gravitation and postglacial rebound: An integrated field and modelling study, Swiss J. Geosci., 101(1), 223–236.
10.1007/s00015-008-1249-1
Web of Science®Google Scholar
Vidstrand, P., T. Wallroth, and L. Ericsson (2008), Coupled HM effects in a crystalline rock mass due to glaciation: Indicative results from groundwater flow regimes and stresses from an FEM study, Bull. Eng. Geol. Environ., 67(2), 187–197, doi:10.1007/s10064-008-0123-8.
10.1007/s10064-008-0123-8
CASWeb of Science®Google Scholar
Vinther, B. M., et al. (2009), Holocene thinning of the Greenland ice sheet, Nature, 461(7262), 385–388, doi:10.1038/nature08355.
10.1038/nature08355
CASADSPubMedWeb of Science®Google Scholar
Vogler, M. (2015), Hydro-mechanical interactions between the Great Aletsch Glacier and the Driest Landslide (Switzerland), MSc thesis, Department of Earth Sciences, ETH, Zurich.
Google Scholar
Wegmann, M. (1998), Frostdynamik in hochalpinen Felswänden—am Beispiel der Region Jungfraujoch-Aletsch, PhD thesis, ETH, Zurich.
Google Scholar
Wegmann, M., and G. H. Gudmundsson (1999), Thermally induced temporal strain variations in rock walls observed at subzero temperatures, in Advances in Cold-Region Thermal Engineering and Sciences, edited by K. Hutter, Y. Wang and H. Beer, pp. 511–518, Springer, Berlin Heidelberg, doi:10.1007/BFb0104208.
10.1007/BFb0104208
Web of Science®Google Scholar
Wegmann, M., G. H. Gudmundsson, and W. Haeberli (1998), Permafrost changes in rock walls and the retreat of alpine glaciers: A thermal modelling approach, Permafrost Periglacial Process., 9(1), 23–33, doi:10.1002/(sici)1099-1530(199801/03)9:1<23::aid-ppp274>3.0.co;2-y.
10.1002/(sici)1099-1530(199801/03)9:1<23::aid-ppp274>3.0.co;2-y
Web of Science®Google Scholar
Welten, M. (1982), Vegetationsgeschichtliche Untersuchungen in den westlichen Schweizer Alpen: Bern-Wallis, Denkschrift SNG, 95, 104.
Google Scholar
Zangerl, C., E. Eberhardt, and S. Loew (2003), Ground settlements above tunnels in fractured crystalline rock: Numerical analysis of coupled hydromechanical mechanisms, Hydrogeol. J., 11(1), 162–173, doi:10.1007/s10040-002-0234-4.
10.1007/s10040-002-0234-4
ADSWeb of Science®Google Scholar
Zangerl, C., K. F. Evans, E. Eberhardt, and S. Loew (2008a), Consolidation settlements above deep tunnels in fractured crystalline rock: Part 1—Investigations above the Gotthard highway tunnel, Int. J. Rock Mech. Mining Sci., 45(8), 1195–1210, doi:10.1016/j.ijrmms.2008.02.002.
10.1016/j.ijrmms.2008.02.002
Web of Science®Google Scholar
Zangerl, C., E. Eberhardt, K. F. Evans, and S. Loew (2008b), Consolidation settlements above deep tunnels in fractured crystalline rock: Part 2—Numerical analysis of the Gotthard highway tunnel case study, Int. J. Rock Mech. Mining Sci., 45(8), 1211–1225, doi:10.1016/j.ijrmms.2008.02.005.
10.1016/j.ijrmms.2008.02.005
Web of Science®Google Scholar
Zangerl, C., K. F. Evans, E. Eberhardt, and S. Loew (2008c), Normal stiffness of fractures in granitic rock: A compilation of laboratory and in-situ experiments, Int. J. Rock Mech. Mining Sci., 45(8), 1500–1507.
10.1016/j.ijrmms.2008.02.001
Web of Science®Google Scholar
Zerathe, S., T. Lebourg, R. Braucher, and D. Bourlès (2014), Mid-Holocene cluster of large-scale landslides revealed in the Southwestern Alps by 36Cl dating. Insight on an Alpine-scale landslide activity, Quat. Sci. Rev., 90, 106–127, doi:10.1016/j.quascirev.2014.02.015.
10.1016/j.quascirev.2014.02.015
ADSWeb of Science®Google Scholar

Citing Literature

Volume122, Issue4

April 2017

Pages 1004-1036

Beyond debuttressing: Mechanics of paraglacial rock slope damage during repeat glacial cycles

Abstract

Key Points

1 Introduction