Multiple rock-slope failures from Mannen in Romsdal Valley,western Norway,revealed from Quaternary geological mapping and 10 Be exposure dating

Abstract

Oversteepened valley walls in western Norway have high recurrences of Holocene rock-slope failure activity causing significant risk to communities and infrastructure. Deposits from six to nine catastrophic rock-slope failure (CRSF) events are preserved at the base of the Mannen rock-slope instability in the Romsdal Valley, western Norway. The timing of these CRSF events was determined by terrestrial cosmogenic nuclide dating and relative chronology due to mapping Quaternary deposits. The stratigraphical chronology indicates that three of the CRSF events occurred between 12 and 10 ka, during regional deglaciation. Congruent with previous investigations, these events are attributed to the debuttressing effect experienced by steep slopes following deglaciation, during a period of paraglacial relaxation. The remaining three to six CRSF events cluster at 4.9 ± 0.6 ka (based on 10 cosmogenic ¹⁰Be samples from boulders). CRSF events during this later period are ascribed to climatic changes at the end of the Holocene thermal optimum, including increased precipitation rates, high air temperatures and the associated degradation of permafrost in rock-slope faces. Geomorphological mapping and sedimentological analyses further permit the contextualisation of these deposits within the overall sequence of post-glacial fjord-valley infilling. In the light of contemporary climate change, the relationship between CRSF frequency, precipitation, air temperature and permafrost degradation may be of interest to others working or operating in comparable settings.

Keywords

climate change cluster debuttressing fjord-valley infill rockslides

Introduction

Formerly glaciated valleys often exhibit slope instabilities that lead to catastrophic rock-slope failures (CRSF; Hermanns and Longva, 2012). Although CRSF are often attributed to seismic activity (e.g. Agliardi et al., 2009; Hewitt et al., 2008; Korup, 2004; Moreiras et al., 2015; Penna et al., 2011), many are triggered by climatogenic destabilisations owing to post-glacial debuttressing, permafrost degradation or precipitation increases (Evans and Clague, 1994; Soldati et al., 2004; Trauth et al., 2000). Deglaciation and valley adjustment in the Lateglacial and Holocene provoke localised stress concentrations in steep fjord rock walls exceeding the crack initiation threshold; coincidently advancing and retreating permafrost significantly alters rock mass strength properties (Krautblatter and Leith, 2015; Leith et al., 2014). Both in situ stress evolution and rapidly changing material strength due to permafrost advance play a key role in controlling rock slope failure evolution. Changing permafrost conditions alter the strength of (1) intact rock, (2) ice infill and (3) rock ice interfaces by 20–80% and freezing causes high cryostatic stresses and irreversible rock fatigue (Jia et al., 2015, 2017; Krautblatter et al., 2013). Previous studies from Scotland and Norway indicate that CRSF commonly occur within the first 2 ka of deglaciation (Ballantyne and Stone, 2013; Cossart et al., 2008; Holm et al., 2004). Events within this time window are commonly attributed to ‘debuttressing’ – the unloading and stress release experienced in paraglacial environments following deglaciation, although it is difficult to preclude seismicity caused by post-glacial isostatic adjustment. CRSF occurring outside of this interval are often attributed to changing climate, in particular higher precipitation rates, air temperatures and, in some cases, permafrost degradation (Allen et al., 2009; Fischer et al., 2006; Krautblatter et al., 2013; Nagelisen et al., 2015). The temporal and spatial distribution of CRSF in Norway reflects the strong influence of debuttressing on slope stability. However, continued rock-slope activity throughout the Holocene has also been identified (Blikra et al., 2006; Hermanns et al., 2017). Few studies have investigated the relation between climatic variability and CRSF frequency in Norway (Blikra and Christiansen, 2014; Böhme et al., 2015).

A precise geochronology is necessary to establish cause and assess hazard of CRSF which may have recurrences exceeding millennia. Until the late 90s, most CRSF event chronologies relied on independently dated stratigraphically related sediments for limiting or contemporaneous age control (e.g. Clavero et al., 2002; Topping, 1993). For post-glacial events, radiocarbon dating was frequently used (Hermanns et al., 2000; Orwin et al., 2004; Ostermann et al., 2016). Recently, terrestrial cosmogenic nuclide (TCN) exposure dating is more frequently used to directly date CRSF deposits (Gosse and Phillips, 2001; Ivy-Ochs and Kober, 2008; Sturzenegger et al., 2015; Ostermann et al., 2016). Recent studies have utilised TCN methods to determine the timing of CRSF events and better relate them to otherwise hypothetical triggering mechanisms (e.g. Ballantyne et al., 1998; Dortch et al., 2009; Hermanns et al., 2004, 2015). However, in most mountain environments, the limited number of investigations precludes a comprehensive assessment of the conditioning variables that lead to failure. Exceptions exist for the European Alps, Scotland and Norway where many studies have been conducted (e.g. Ballantyne et al., 2014; Hermanns et al., 2017; Ivy-Ochs et al., 2017).

The rock-slope instability ‘Mannen’ in Romsdal Valley, Norway, is located in a key region for tourism and is situated above the farming community of Marstein. The risk of rock-slope failure at Mannen has been considered high owing to geological structural pre-conditioning and observed deformation and sliding rates of cliff blocks. In this region, the recurrence interval for a 0.15 Mm³ failure is estimated to be <100 years while the recurrence of a 2–4 Mm³ volume is 100–1000 years (Blikra et al., 2016). Extensive CRSF deposits with volumes between 0.05 and 1.95 Mm³ from previous events have accumulated in the valley bottom beneath Mannen. It is therefore important to decipher the frequency of previous failures in order to assess if historic CRSF events were randomly distributed or clustered in time and if they can be linked to a particular failure mechanism. Using techniques of Quaternary geological mapping, sedimentary stratigraphy, electrical resistivity tomography (ERT), ground-penetrating radar (GPR) and cosmogenic ¹⁰Be exposure dating, we (1) determine the timing of post-glacial CRSF events at Mannen, (2) contextualise the colluvium within the framework of a fjord-valley fill succession and (3) identify the antecedent climate and glacially generating conditions resulting in periods of heightened CRSF activity.

Setting

Mannen is situated along a north-facing slope of the glacially formed, U-shaped Romsdal Valley (62.46°N, 7.77°E; Møre og Romsdal, Norway). This lower reach of the valley is a fjord-valley, or sediment-filled palaeofjord (Corner, 2006) with an underfit stream. Within 10 km of the site, the valley walls reach between 800 and 1400 m above the valley floor. The rock-slope section at Mannen is 1295 m in height with an average slope gradient of 47°. At present, there are four active rock-slope instabilities along the southern and south-western slope of the lower Romsdal Valley (Saintot et al., 2012). Each site is situated above massive rock-avalanche and debris-flow deposits on the valley floor (Figure 1).

Figure 1.

Overview over the study area: (a) Quaternary geology of the lower and mid Romsdal Valley and (b) detail of the main study site and the locations of geophysical surveys and composite sediment logs. Note the stretched colour code.

Mannen is located in the Western Gneiss Region, which consists of Precambrian crystalline basement rocks of the Scandinavian Caledonides (Roberts, 2003). The Romsdal Valley cuts east-west through dioritic-granitic gneiss with local transitions to quartz-rich gneiss with sillimanite and kyanite and coarse granitic gneiss (Tveten et al., 1998). The Caledonide structural fabrics and mineralogical banding impart critical weaknesses inherent along the glacially oversteepened valley sides.

Here, post-Weichselian deglaciation began by thinning during the Bølling/Allerød interstadial (ca. 15–13 ka), as the outer coast of western Norway became ice-free (Longva et al., 2009). Ice marginal retreat up the Romsdal Valley has been completed between 12.8 and 11.7 ka, following the Younger Dryas cold period (Hermanns et al., 2017; Hughes et al., 2016; Stroeven et al., 2016). During deglaciation, the landscape was inundated by the sea, reaching a marine limit of 120 m above modern sea level in the Romsdal Valley (that is, 60 m above the Rauma River at Mannen; Høgaas et al., 2012). Subsequently, the relative sea level lowered approximately exponentially with glacioisostatic uplift (Svendsen and Mangerud, 1987).

The Quaternary valley infill is the product of processes connected to the Pleistocene glaciation, post-glacial sea-level fall and paraglacial colluvial activity (Figure 1). The geomorphology of the Mannen area in the lower Romsdal Valley is dominated by talus, debris-flow cones and CRSF deposits (Figure 1; Blikra et al., 2006; Saintot et al., 2012). More than 30 historical mass wasting events are documented along the valley (NVE, 2018), dominated by small rock-fall events with a volume <100 m³. Relative to the sea-level history, these deposits stratigraphically post-date ice marginal retreat with varying failure timing throughout the Holocene (Blikra et al., 2006).

The mean annual air temperature ranges from 4°C in the valley bottom to −1.5°C on the plateau above Mannen (1295 m a.s.l.). Annual precipitation during the reference period 1961–1990 lies between 1000 and 1500 mm in the Romsdal Valley. During the winter season, the snow precipitation accumulates an average of 37 cm of snow cover with monthly means of 12–25 cm in November and Mai and 50–56 cm in December and January. During the last 20 years the values have increased slightly, now usually being close to the upper limits of these ranges. Temperatures have increased as well, particularly since the year 2000 with 1°C in relation to the last reference period 1961–1990 (NVE, met.no and Kartverket, 2018).

Methods

Quaternary geological mapping

High-resolution (1 and 5 m) bare-earth LiDAR-derived digital elevation models (DEM) are the main sources for the recognition of landforms (e.g. Schleier et al., 2016). The digital relief analysis was complemented by field mapping in the summers 2016 and 2017 in the area around Mannen but included a ca. 40 km region extending from the contemporary fjord-head delta to the Skiri rock avalanche (Figure 1). Landform elevations were extracted from the DEM and plotted together with the local relative sea-level curve in order to establish the relationships between the marine transgression and regression and the onshore stratigraphy (cf. Eilertsen et al., 2015). Small excavations and field observations of the sediment characteristics facilitate the classification of different surface levels along the entire Romsdal Valley. The sedimentology characterisations and interpretations are based on field observations, geographic information system (GIS) analyses and previous studies about typical valley-fill stratigraphy in Norwegian fjord valley settings (cf. Corner, 2006; Eilertsen et al., 2006).

The volume of the CRSF deposits was estimated by reconstructing the pre-failure surface from the most recent high-resolution (1 m) LiDAR data. Based on interpolated and modified 5-m contour lines, we created new pre-failure DEM, which were then used to extract the elevation differences and subsequently calculate the volume on a pixel basis.

Two 2D ERT profiles, 900 and 700 m in length and parallel to the valley axis, were obtained on the northern and southern sides of the Rauma River (Figure 1b). ERT utilised an ABEM Terrameter LS and field testing employed the roll-along method using the Schlumberger protocol (e.g. Aizebeokhai, 2010) with four 100-m-long cables and 5 m electrode spacing. To interpret the subsurface geology, the inverse electrical conductivity was derived with RES2DINV (© M.H. Loke, 1995-2015). After the manual extermination of bad data points, the inversion was derived using the robust L₁-norm.

A single GPR profile was conducted, running along the foot of the valley parallel 20-m step in the relief in the north-west of the study site (Figure 1b). We used a snake antenna with a frequency of 100 MHz, which was towed behind the surveyor. The measurement frequency was 0.5 m. The post-processing was conducted with RadExplorer (©MALÅ Geoscience).

Geochronology

A total of 13 samples were collected and processed for surface exposure dating with cosmogenic ¹⁰Be in quartz. We sampled boulders representing at least four of the CRSF deposits below the Mannen rock-slope instability. The boulders are well distributed and their sampled surfaces lie more than 1 m above the surroundings and have an either flat or convex geometry. All samples were taken with hammer and chisel and are 1–6 cm thick (Table 1). Where the foliation was favourable the collected samples have a rather even thickness of 2–3 cm and 20–30 cm of diameter. Generally, the selected sample locations on the CRSF deposits are either on flat or convex boulder surfaces with at least 40 cm distance to the boulder edge to minimise the effect of neutron loss and local shielding. In two cases the samples are only 20 cm from the boulder edge (MANN-31 and MANN-38) and three samples are collected from boulder surfaces steeper than 30° (MANN-23, MANN-32 and MANN-35). Most sampled boulders were in continuous boulder fields with deep interstitial gaps, lacking the infill of a finer matrix (Figure 2b and c). Many boulders were covered by up to 5-cm-thick moss with a (dry) density of 0.05 g/cm³. An open birch forest with 5–20 cm trunk diameter and 2–12 m height covers the CRSF deposits today, with solitary pine trees in places.

Table 1.

Sample characteristics. The boulder height refers to the surrounding ground or boulders.

Sample name	Rock type	Sample thickness (cm (estimated average))	Orientation (dip direction/dip)	Boulder dimensions (a- and b-axis) (m)	Boulder (sample) height (m)	Shortest distance to edge (cm)	Moss cover (cm)
MANN-07	Medium/coarse grained granite	1.0	264/20	2 × 4	3.0	100	3.0
MANN-10	Medium/coarse grained gneiss	2.0–4.0 (2.5)	302/10	5 × 5	3.0	50	5.0
MANN-23	Medium/coarse porphyritic granite	5.0	074/30	1 × 2	1.0	Convex boulder	–
MANN-26	Fine grained micaceous gneiss	3.0	000/20	15 × 7	7.0	70	3.0
MANN-28	Fine grained felsic gneiss	3.0–6.0	234/10	4 × 3	2.5	50	4.0
MANN-31	Medium grained, strongly foliated gneiss	0.5–5.0 (3.5)	358/36	1.5 × 1.5	3.0	20	1.0
MANN-32	Fine grained gneiss	1.0–5.0 (4.0)	254/18; 149/49	12 × 3.5	2.0	Convex boulder	0.5
MANN-35	Medium/coarse grained granitic gneiss	2.5	181/32	7 × 6	2.5	300	–
MANN-36	Medium grained strongly foliated gneiss	2.5	315/12	4 × 3	2.5	150	4.0
MANN-37	Weakly foliated granitic gneiss	1.0–2.0 (1.5)	071/26	8 × 6	8.0	300	4.0
MANN-38	Fine grained gneiss with mid-strength foliation	1.0–6.0 (5.0)	084/20	5 × 2	1.5	20	–
MANN-39	Medium grained gneiss with mid-strength foliation	1.0–2.0 (2.0)	132/26	5 × 2	1.5	40	2.0
MANN-40	Fine grained and weakly foliated gneiss	1.0–5.0 (4.0)	305/15	2.5 × 2	1.0	50	5.0

Figure 2.

Images of (a) the sample location MANN-36 in a surrounding characterised by soil in contrast to (b) the chaotic boulder field of the south eastern CRSF deposits and (c) the sample location MANN-28 with a pulled back moss cover.

Selected samples were cleaned by brushing, crushed, ground and sieved, optimising the 250–355 µm fraction. We subsequently concentrated and purified the quartz at the Cosmic Ray Isotope Sciences at Dalhousie University (CRISDal) lab, Halifax, Canada, using magnetic separation, froth flotation, heavy liquid separation and chemical leaching. The abundance of selected cations including Be was measured with ICP-OES at CRISDal to ensure purity (<100 ppm Al and Ti). Following carrier addition (240 mg of Be) to 30 g of pure quartz for each sample, the samples were digested in a mixture of concentrated trace-metal grade perchloric, hydrofluoric, and aqua regia. Be-Carrier-B31 was produced at CRISDal on 28 September 2012 from a deeply sourced Ural Mountains phenacite with low levels of ¹⁰Be (averaging 150 atoms ¹⁰Be per mg ⁹Be over multiple years, for example, ¹⁰Be/⁹Be of the carrier averaging 10⁻¹⁷ and lower, usually with zero or one counts over 400 s with 20 µA current at Lawrence Livermore National Lab. The Be concentration of the carrier was determined by ICP-OES at CRISDal and by ICP-OES at PRIME Lab to be 282 ± 5 µg/mL with a density of 1.013 g/mL, and this 2% uncertainty is included in the total analytical error of each measurement). Following routine column chemistry with sulphination, pH-controlled precipitations with ammonia gas and calcination to BeO over a Bunsen burner flame, the BeO was pulverised in its low-boron quartz vial and mixed well with niobium powder (1:1.5 BeO:Nb by volume). The prepared targets were measured by accelerator mass spectrometry (AMS) at Lawrence Livermore National Laboratory, Livermore (USA) against standard 07KNSTD-3110 (¹⁰Be/⁹Be 2.85 × 10⁻¹²) and achieved 2–3% AMS precision on most samples. The process blank correction (5.59 × 10³ atoms, ¹⁰Be/⁹Be 3.3 × 10⁻¹⁶, which is very close to the average blank since 2016) resulted in the subtraction of <1% of the measured concentrations.

To estimate the topographic shielding, we measured with an inclinometer the azimuth and gradient to several skyline inflection points. As low-level clouds and vegetation affected some measurements, the shielding correction was verified with a high-resolution (5-m) DEM. For this, the elevation for each sample was corrected using the LiDAR data and the angle to the horizon derived for each azimuth. These values are on average 1.3° higher than the corresponding inclinometer measurements but have a much higher resolution and should thus overall be more accurate. Snow shielding was derived after Gosse and Phillips (2001) using historic and modern climate data to estimate the average seasonal snow cover (snow density ~0.3 g/cm³). This method only represents the snow cover for the last decades and does not include local effects such as vegetation and wind drift. An erosion rate of 1 mm/ka was used for the calculation (Zimmerman et al., 1994). For the calculation of the exposure ages, we used version 3 of the online exposure age calculator formerly known as the CRONUS-Earth online exposure age calculator written by G. Balco, 2017, and the LSDn scaling scheme. The reported 1σ uncertainty for an exposure age includes the internal and external errors (details in supplementary files, available online).

Results

Quaternary geology and geomorphology

Geomorphological mapping using high-resolution (1 m) LiDAR data

The DEM revealed well-defined 20-m high steps in the relief on both sides of the valley below the Mannen rock-slope instability (transition from green to yellow colouring in Figure 1b; an additional plain hillshade-map can be found in the supplementary files, available online, for comparison). The steep slopes with angles around 35° run almost parallel to the Rauma River and are connected to rather flat (<5°) elevated surfaces. Upstream, these elevated surfaces end abruptly, where the steep slopes turn towards the rock walls. These stepped landforms are overlain by at least five but possibly up to seven lobate CRSF deposits. While one of these mapped CRSF events, featuring a secondary failure scar, has overrun and modified the steep 20 m high slope, the latter draws through most of the other CRSF deposits. A lobate and hammocky landform on a low elevation basin-like section north-east of the high elevated landforms suggests up to two additional CRSF deposits. A third larger event, exceeding the Rauma River, has been mapped by an intensive GPR survey (Tønnesen, 2009). But since the deposits of this event are no longer visible at the surface, it was not possible to include them into the age determinations of this study.

Longitudinal valley profile and terrace mapping

The valley-fill deposits along the Romsdal Valley were mapped in relation to the longitudinal profile of the Rauma River (Figure 3). The terraced landforms below the Mannen rock-slope instability are the highest valley fill deposits along the Romsdal Valley, exceeding all other valley-fill sediments by more than 10 m. The next highest terraces have been investigated with small (~1 m deep) excavations along the entire valley (Figure 1a). They revealed coarse sand interbedded with fine sand with 1–50 cm large rounded to well-rounded clasts in the upper Romsdal Valley (Figure 3a and b) and patterns of altering sand layers of varying grain sizes without pebbles and cobbles close to the fjord head. The bedded structure and clast roundness imply fluvial deposition processes but because of their high elevations, these terraces are interpreted to be of glacio-fluvial origin, which is in accordance with the most recent regional quaternary geological map (NGU, 2018). High elevated sand deposits with nearly horizontal sandy layers are found on either side of the valley in bay-like settings (Figure 3c) and have previously been interpreted as beach deposits (NGU, 2018). The correlating elevation of these beach deposits and the sandy terrace segments covering large parts of the lower Romsdal Valley (Figure 1a) suggests a possible deposition of distal fine grained glacio-fluvial sediments in a deltaic environment and thus indicating the sea level at the time of deposition.

Figure 3.

Surface profiles of terrace segments along the Romsdal Valley characterised according to their relative position and (a–c) field observations. Sea-level elevations and their approximate timing are indicated as waved grey lines next to the glacio-fluvial terrace segments, which could be connected to a delta progradation. The vertical axis is exaggerated by 10 in relation to the horizontal axis.

Sedimentary stratigraphy and interpretation below the Mannen rock-slope instability

The upper ca. 40 m of the fjord-valley fill below the Mannen instability were investigated in three locations (A–C; Figure 1b). Four sedimentary facies associations (FA I-IV) are identified and likely relate to the progradation of a fjord-head delta system and its associated braided river, rock-slope failure and related processes, including debris flows on colluvial fans. A brief tabular summary of the following sections is provided in the supplementary files, available online.

FA I stratigraphically occupies the lowermost position at the locations A and B (Figure 4). Observed exposures were ca. 22 m in height. FA I consists of stratified sands and gravels, which dip ca. 5° in downstream direction (Figure 4a). Gravel clasts are typically subrounded to rounded. FA I is interpreted to be stratified drift. Similar units of stratified drift consisting of sands and gravels are frequently described in fjord-valleys in western Norway (cf. Corner, 2006). Considering the high relative elevation of these sediments together with their morphological appearance connected to a steep slope facing upstream and downstream dipping sediment layers, FA I could be interpreted to represent an ice-contact glaciomarine fan or delta (cf. Corner, 2006; Eilertsen et al., 2006; Lønne, 1995).

Figure 4.

Impressions from the locations of the composite sediment profiles and the sediment characteristics: (a) gravel pit with ca. 35 m stratified drift (FA I) and ca. 7 m interbedded fluvial sands in angular debris-flow gravel (FA II), person is 1.70 m tall; (b) ca. 15 m thick chaotic boulder deposits (FA III) on top of FA I, which is only visible in small outcrops; (c) silty sands with outsized cobbles on top of ca. 10 m thick rock-slope failure boulders.

FA II is observed overlying FA I at location A (Figure 1). The deposits are ca. 5 m thick and consist of flat-laying interbedded silty sands and gravels. Gravel clasts are angular to sub-angular. Variations in clast shape reflect changes in the transport distance and sediment source. Angular clasts likely entered the fluvial domain as debris flows from the proximal upstream reaches and hence were transported a short distance prior to deposition. Based on the sedimentary architecture and composition, FA II is interpreted to be fluvial with debris-flow deposits, deposited on the distal parts of colluvial fans (cf. Blikra and Nemec, 1998).

FA III is encountered at locations B and C (Figure 1) and varies from ca. 8 to >10 m in thickness. Deposits consist of boulders (up to 10 m in diameter) as chaotic block fields or suspended in a gravelly, sandy matrix covered with soil (Figure 4b). FA III is interpreted as CRSF deposits.

FA IV is observed in the upper 7 m at location C. The deposits consist of interlaminated sands and silts (Figure 4c). Isolated, sub-angular to angular, cobble-size fragments are incorporated within FA IV. FA IV is tentatively interpreted as overbank fluvial deposits (cf. Corner, 2006; Eilertsen et al., 2006) with outsized cobbles being debris originating from a steep, proximal slope (Blikra and Nemec, 1998). If so, FA IV is an indicator for a higher water level than today’s river level at time of deposition.

Direct current (DC) resistivity

The tomography of the ERT profiles (Figure 1) generally support our sedimentological and geomorphological observations. Based on the 2D distribution of the electrical resistivity of the profiles ERT (a) and (b), we defined five main electrical resistivity units (ERU): ERU 0, I, II, III and IV, with ERU I, II and III corresponding to FA I, II and III, respectively.

ERU 0 occupies the lower 40 m of both DC resistivity profiles (Figure 5a and b). The unit is characterised by resistivity values from <400 Ωm to 5 kΩm and a transition to high resistivity values of >14 kΩm at 50–60 m a.s.l. The elevation of this transition coincides with the lower limit of the mapped dry and coarse-grained stratified drift (FA I). However, we have no field information about sediment characteristics below this elevation. The similar resistivity patterns of the two profiles suggest that the bottom geologic characteristics are similar over the entire width of the Romsdal Valley. While the fjord-valley models would expect fine grained glaciomarine sediments at the valley bottom, the typical resistivity values for clays do not exceed 100 Ωm. We therefore suggest this lower unit to be either bedrock (Palacky, 1988) or inversion artefacts due to the high values above.

Figure 5.

Results of the geophysical surveys: Top: 2D DC resistivity pseudosections of the profiles ERT (a) and ERT (b) (NW-SE) as indicated in Figure 1b. The lowercased letters along the profiles represent prominent relief changes (Figure 1b) that are connected to a change in sediment characteristics in places. Steep and ca. 20 m high steps in the relief are marked by the sections b-c and d-e, respectively. (c) GPR survey over CRSF deposits. Note that the scales differ for visualisation. The y-axes are exaggerated by 1.5 in all profiles. The composite stratigraphy logs from Figure 3 are included in the approximate locations.

ERU I is characterised by resistivity values of 14 to >36 kΩm and occupies the largest parts of both sections ERT (a) and (b). This unit lies above ERU 0 between ca. 50 and 90 m a.s.l. The highest values in ERT (a) are most likely an artefact due to the proximity of the edge to the open gravel pit. ERU I is interpreted to represent FA I with glacio-fluvial sand and gravels (Palacky, 1988).

ERU II occupies the upper 5–7 m in section ERT (a). Resistivity values range from 350 Ωm in the uppermost 2 m to values above 14 kΩm. But the values are generally lower than in unit I. ERU II represents the sediments interbedded silty sand layers in debris-flow gravels of FA II. The variation of the resistivity values are interpreted to originate from different sedimentation processes related to the talus cone above.

ERU III was only observed in section ERT (b). This unit is characterised by high values (>14 kΩm) at the top 2–5 m with lower values (5–14 kΩm) below this surface layer. At the letter f and g (Figure 5b), CRSF deposits have been mapped in the field. The decrease of resistivity with depth can be explained by the typical grain-size distribution of massive CRSF deposits with large boulders on the surface and an increase in finer matrix material and moisture with depth (Ostermann et al., 2012).

While the resistivity values are generally lower within the section c-d in the same profile, this unit is also correlated to FA III. Here, up to 20 cm deep surface water was observed in the field indicating a high water content in the subsurface, leading to a decrease in the resistivity.

ERU IV occupies the last 100–200 m of the 900-m-long ERT profile (a) (a–c; Figure 5a), where the electrical resistivity values lie between 50 and 350 Ωm. This section of the profile is part of the basin upriver of the high-elevated valley-fill sediments below Mannen. We suggest this unit to be silty sediments deposited either in a calm water environment or as overbank fluvial deposits similar to FA IV (Groover et al., 2016).

GPR

The surface along most of the GPR profile (Figures 1b and 5c) is characterised by agriculturally used lawn. A chaotic boulder field confining the Rauma River along this section generated a minor knickpoint in the longitudinal profile (Figure 3), which is often the consequence of CRSF into rivers (Korup et al., 2010; Ouimet et al., 2007). The characteristic reflection configurations and analogies to the ERT units allowed defining four main radar units (RU): RU I-IV, where RU III and IV correspond to FA III and IV, respectively.

RU 0 is defined by an area below 45 m a.s.l. where the signal strength decreases abruptly and no clear reflectors are distinguishable. The upper limit of this unit lies only a few metres lower than the upper limit of ERU 0 (50–60 m a.s.l.) wherefore we interpret this unit tentatively as bedrock.

RU I is defined by a partly clear reflecting boundary that can be followed throughout large parts of the profile (thick line, Figure 5c). This reflector becomes rather indistinct at the boundary to RU III and the reflectors within this unit are rather chaotic and unclear. Because of the lack of homogeneous reflectors, we interpret this unit as valley-fill sediments that have been deformed by the impact of the CRSF (cf. Blikra et al., 2006).

The RU II unit is characterised by steeply downstream dipping (35–45°) generally parallel reflectors. Analogue reflectors are commonly observed in deltaic environments (Eilertsen et al., 2011). Considering the fjord-valley setting we interpret RU II to represent delta foreset deposits, indicating a previous sea- or lake-level at ca. 55 m a.s.l.

RU III represents the central section of the GPR survey, where we observe distinguishable parable-shaped reflectors in different elevations. RU III is interpreted as CRSF deposits (FAIII; cf. Schleier et al., 2016), which can be observed at the surface only a few metres north-east of the profile.

The RU IV unit is characterised by smooth parallel and continuous horizontal reflectors with varying thickness. While these characteristics are common for stratified sediments, it is difficult to distinguish between different possible deposition processes. Similar reflectors have been observed for delta bottom/topsets (Eilertsen et al., 2011), lake deposits (Storms et al., 2012) and flood plains (Hansen et al., 2009). Because of its location and our field observations, we interpret this unit as fluvial overbank flow sediments correlating with FA IV.

Quaternary geology map

The Quaternary geology map displays the dominance of CRSF deposits and their position relative to the stratified drift. The volumes of the individual CRSF events vary between 0.05 (Lobe 5) and 1.95 Mm³ (Lobe 4, Figure 6). The small rock-slope failure Lobe 5 is characterised by a clast-supported chaotic block field with large angular boulders (3–6 m) and little to no matrix exposed at the surface. Considering its short run-out length, Lobe 5 probably represents a large rock-fall event without major disintegration. It exceeds the active extensive talus slope by >150 m. According to a previous GPR survey (Tønnesen, 2009), the CRSF deposits 6a and 6b are much larger than the superficial deposits indicate. The study suggests that the deposits continue below the Rauma River and that the volume is thus much larger than our estimated 0.43 mm³.

Figure 6.

Quaternary geology map of the study site in the lower Romsdal Valley below the Mannen rock-slope instability. The individual apparent exposure ages are stated with 1σ uncertainties. The CRSF deposits are numbered for further discussions.

Geochronology

We have determined the apparent exposure ages of 13 boulder samples using the ¹⁰Be-isotope (Figure 6; Table 2). The locations of the two samples with the oldest apparent ¹⁰Be ages are adjacent to each other and give a mean exposure age of 9.39 ± 0.64 ka (error-weighted mean with 1σ uncertainty). The majority of the deposit lies below the marine limit, and while only the highest boulders were sampled, shielding by seawater may have reduced the ¹⁰Be production rate. Thus, assuming no inheritance, these dates are interpreted to be minimum limiting ages. The other 11 exposure ages range from 4.75 ± 0.33 ka to 5.93 ± 0.39 ka. Considering sample MANN-26 as a statistical outlier (beyond the coefficient of variation of the mean of the others) owing possibly to inherited ¹⁰Be isotopes from pre-failure production, the range is 4.75 ± 0.33 ka to 5.12 ± 0.36 ka and the ages are indistinguishable within their 1σ uncertainties. Excluding MANN-26, the mean ages for deposits 4a, 5 and 6 (Figure 6) are 4.91 ± 0.30 ka, 4.96 ± 0.33 ka and 4.95 ± 0.31 ka, respectively. The single sample from Lobe 4c gives an age of 4.98 ± 0.34 ka, which lies within the standard deviation of Lobe 4a and, based on their close proximity and stratigraphic relationship, we recommend considering them as one event with a mean age of 4.93 ± 0.30 ka in the following discussions. The clearly distinguishable deposits of several CRSF events with indistinguishable ages within one standard deviation indicate temporal acluster of multiple failures from the same slope 4.9 ± 0.6 ka ago.

Table 2.

Analytical data and calculated exposure ages with the LSDn scaling scheme.

Sample name	Latitude (dd)	Longitude (dd)	Altitude (m a.s.l.)	¹⁰Be concentration (10⁴ at/g)	1σ analytical unc. (10⁴ at/g)	Shielding correction	Age (ka)	Age unc. internal (ka)	Age unc. external (ka)	Error weighted mean with int. (ext.) unc.
MANN-07	62.46503	7.793195	70	21.30	0.75	0.9128	4.75	0.12	0.33	4.95 ± 0.10 (0.31)
MANN-10	62.46522	7.796929	63	22.00	0.69	0.9173	5.00	0.16	0.34
MANN-23	62.46579	7.795452	68	22.10	0.84	0.9152	5.12	0.20	0.36
MANN-26*	62.46529	7.795589	68	26.00	0.77	0.9165	5.93^*	0.18	0.39
MANN-31	62.46634	7.788837	111	22.80	0.80	0.9100	5.03	0.17	0.35	4.96 ± 0.15 (0.33)
MANN-32	62.46655	7.788266	110	21.70	1.21	0.9084	4.81	0.18	0.39	4.96 ± 0.15 (0.33)
MANN-35	62.46807	7.786672	118	45.40	2.33	0.9059	9.92	0.27	0.79	9.39 ± 0.30 (0.64)
MANN-36	62.46801	7.786364	117	41.80	1.69	0.9054	9.12	0.52	0.66	9.39 ± 0.30 (0.64)
MANN-28	62.46924	7.783871	138	23.40	0.79	0.9033	5.11	0.37	0.35	4.93 ± 0.09 (0.30)
MANN-37	62.47001	7.786400	95	21.10	0.69	0.8946	4.70	0.16	0.32
MANN-38	62.47067	7.787071	88	21.60	0.82	0.8901	5.05	0.17	0.36
MANN-39	62.47122	7.784973	78	21.20	0.73	0.8856	4.88	0.16	0.36
MANN-40	62.47038	7.782797	117	22.30	0.75	0.8956	4.98	0.20	0.34

Shielding values include the topographic shielding as well as shielding by snow. Sample marked with * is defined as a statistical outlier.

Poor estimations of partial cosmic ray shielding by snow and vegetation provide an unconstrained source of error in our ages. Our shielding estimations for snow, with an average of 37 cm snow depth for a 7-month snow season and an average density of 0.3 g/cm³, yielded a value of 0.999 which has a small effect on the absolute exposure ages. The effect of the sparse birch tree forest that covers the sampled CRSF deposits can roughly be estimated. Plug et al. (2007) show that the shielding effect in forests is dependent on stem thickness and tree height, sample location, succession rate and age. Considering the generally small stem diameters (5–20 cm) and forest density, we expect the shielding effect to be smaller than the numerically estimated shielding of <2.25% for Acadian forest in Nova Scotia (Plug et al., 2007).

Potential inheritance of ¹⁰Be in each boulder depends on its depth below the cliff face prior to failure and the rate of cliff retreat (frequency of mass wasting). The study area has relatively small CRSF volumes (0.05–1.95 mm³). For a conservative realistic scenario of 7 ka pre-failure exposure (4.9 ka subtracted from an exposure history of ~12 ka after deglaciation), the effect of inheritance for 10 and 5 m depth below the cliff face are ca. 0.8% and 1.5%, respectively (Hilger et al., unpublished data). However, the average effect becomes >6% for depths smaller than 2 m. But because of the existence of an outlier (MANN-26) that possibly came from a depth <2 m, and otherwise uniform ages, we expect that most of the boulders originate from greater depth. Because shielding by vegetation and inheritance impart opposite effects on exposure ages (decreasing or increasing apparent exposure time), and because they are of similar effect (a few percentage), and considering the tight distribution of the boulder exposure ages, it is possible that the two factors effectively cancel each other. Therefore, we have not adjusted the ages for either factor and interpret the measured exposure ages as the timing of the CRSF events.

Discussion

Timing of CRSF cluster from Mannen

Our study indicates multiple early post-glacial rockslide events and a CRSF cluster during the mid-Holocene. Because of the high marine limit, the sampling of stratigraphically low deposits was restricted to boulders on higher elevations in the study site. Consequently, we are lacking absolute exposure ages for the CRSF deposits 1 and 3 and are dependent on geomorphologic and stratigraphic observations relative to the dated deposits. The geomorphological setting of the deposits clearly indicates that the CRSF events were post-glacial. If the rock-slope failures were deposited supra-glacially, they would have been transported down valley to form moraine ridges with characteristically uniform boulder lithology, as observed by Schleier et al. (2015). Such discontinuous rock-avalanche deposits are very distinctive from intact CRSF lobes deposited in an ice-free valley. In the Romsdal Valley no analogue moraine ridges were observed, and all our CRSF deposits under the Mannen instability form continuous lobate landforms with clear runouts into or across the valley.

Stratigraphic observations (Figure 4) help constraining the timing of CRSF Lobe 1. While boulders from the Lobe 1 event overlie sandy to gravelly stratified drift (FA I), fine-grained stratified sediments (FA IV) also cover the bouldery CRSF material at this location (GPR; Figure 5). The deltaic foreset structures of RU II are observed at the same elevation as the chaotic CRSF structures, which lie downriver of RU II and seem to ‘truncate’ the delta. Both units are covered by the bank overflow sediments (FA/RU IV). The relative stratigraphy of these units narrows the failure timing for CRSF Lobe 1 to between 12 ka, when the valley became ice free, and 10.5 ka, when the coastline dropped below the recent riverbed of 50 m a.s.l. (Figure 7). Observations of abandoned erosional channels through the CRSF deposits several metres above today’s river level support this interpretation.

Figure 7.

Local sea-level curve approximated after Svendsen and Mangerud (1987) (a) and approximated Holocene temperature anomalies for southern Norway (Lilleøren et al., 2012) (b) above apparent exposure ages as individual probability density functions (PDF) and stacked PDF (c).

The adjacent CRSF deposit (Lobe 2) is the oldest of our dated events with an apparent exposure age of 9.39 ± 0.64 ka (Table 2). The approximated sea-level curve after Svendsen and Mangerud (1987) indicates that the location was effectively above sea-level 11 ka ago (Figure 7a). The age therefore may represent the time of failure and not the timing of sea-level drop. According to the morphology in the DEM, it is possible that this deposit underlies CRSF Lobe 1, which would mean that the real age lies in the upper range of the uncertainties and has thus been deposited into a shallow fjord or high river level, shortly after deglaciation.

To identify CRSF Lobe 3, we had to rely on the high-resolution DEM and relief analysis, as the deposit is hardly distinguishable from the surrounding landforms. In most places, it could only be mapped by the characteristic chaotic boulder fields covering the surface. This indicates that the deposits have been modified by the same erosional processes, as the underlying sediments. We therefore suggest that this deposit is older or the same age as CRSF Lobe 1 and 2. Thus, according to the regional deglaciation, the composite stratigraphy and morphology, the reconstructed curve of sea-level drop and the TCN ages, we can place three CRSF events (Lobes 1-3) from Mannen into a 2000 year time period between 12 and 10 ka ago.

Ten of our 13 ¹⁰Be ages fall into the time range of 4.5–5.5 ka (Figure 7c) considering their external errors. The overlapping ages of the different deposits indicate that there have been several failures from the same slope within a couple of 100 years at most, witnessing a ‘geological crisis’ during this time period. Based on the morphology, it is not clear if the deposits 4a-c are only one or up to three individual events. The same applies for the CRSF deposits 6a and 6b. The ‘geological crisis’ thus included three to six failures from the same slope. The phenomenon of spatial rock-slope failure clusters and multiple failures from the same slope has been observed worldwide (Jarman et al., 2014; Orwin et al., 2004). In Norway this has happened at the Loen site, which has failed repeatedly within a few decades in the early 20th century. Together with a large scale rock-slope failure in Tafjord, CRSF became the natural hazards in Norway with the highest death toll (Grimstad and Nesdal, 1991; Hermanns et al., 2006a; Reusch, 1907).

Possible conditioning for multiple CRSFs at Mannen

Some of the regional and local clusters of CRSF in the Karakoram, the Andes, the Alps and the Scottish Highlands are discussed to be conditioned by tectonic or isostatic uplift and related stresses and seismic activity (Ballantyne et al., 2014; Hermanns and Strecker, 1999; Hewitt et al., 2011; Köpfli et al., 2017). However, in regions with recent low seismicity, this connection is often ambiguous and temporal rock-avalanche clusters are also linked to climatic changes and increased precipitation (Sanchez et al., 2010; Zerathe et al., 2014). Studies about rock-avalanche clusters in the Alps suggest that lithology and the structural predisposition is the most important long-term control on rock-avalanching (Hermanns et al., 2006a; Ostermann and Sanders, 2017), while seismicity is often the trigger with climate conditions as a second-order control.

The deposits at Mannen cluster not only in space but also in time, in contrast to many regional clusters where CRSFs seem to have happened throughout the Holocene with rather low recurrence intervals implying a relaxation of the rock slope after failure (Schleier et al., 2015, 2016). This compares with the increasing number of studies globally which show that one slope of the same mountain can fail repeatedly within only a few years and decades (Crosta et al., 2017; Hermanns et al., 2001, 2004; McSaveney, 2002; Plafker and Ericksen, 1978). Hermanns et al. (2006b) argue that sudden stress release due to a failure causes a reorganisation of the stress field and can thus have a destabilising effect on the rock slope, which is in agreement with the structural simulations by Crosta et al. (2017).

In the Romsdal Valley, the regional clustering of CRSF within a geological unit supports the strong pre-conditioning based on the geological and structural setting. Considering that neo-tectonic activity has so far not been demonstrated in western Norway, the relatively short recurrent interval indicates that driving factors other than tectonic activity may play a significant role in this region.

The age of the first three CRSF at the Mannen rock-slope instability closely post-dates the local deglaciation (12.8–11.7 ka) and coincides with the main peak of rock-avalanche activity in Norway (Böhme et al., 2015; Hermanns et al., 2017). Thus, sudden failure was most likely conditioned by the paraglacially-induced stress increases in the over-steepened slopes during and immediately following deglaciation.

The later timing of the mid-Holocene cluster 5 ka ago invites for a speculation of how changing climate conditions in the Holocene may have contributed to these events. There are two main climatic factors which can contribute to Coulomb failure by decreasing effective shear strength owing to reduced coefficient of friction, increased pore-water pressure, increased slopes and redistribution of centre of mass: (1) regional precipitation changes (amount and type) and (2) temperature changes leading to changes in possible permafrost conditions. There is evidence from several studies in western Norway that a climatic deterioration initiated about 6 ka ago, after a long warm period of the Holocene thermal optimum (HTO). Glacier growth (Nesje et al., 2001) and changing vegetation (Barnett et al., 2001) indicate a generally cooler and wetter climate in this period, and studies documenting Holocene debris flows, snow avalanches, and flooding events suggest a strong seasonality with severe winters and warm summers at this time (Blikra and Nemec, 1998; Blikra and Selvik, 1998; Vasskog et al., 2011). Precipitation between 6 and 5 ka ago was 170% greater than during the reference period 1961–1990 (Bøe et al., 2006). Lilleøren et al. (2012) approximated the temperature anomalies based on published climate proxies for southern Norway for the last 10 ka compared to the same reference period (Figure 7b), which was about 1°C colder than mean annual air temperatures (MAAT) today. The MAAT during and after the HTO was warmer than during the reference period, with very mild winters and warm summers, driven by higher solar radiation due to the Earths orbital position. Cooling from 6 ka ago was most likely driven by low winter temperatures, while the summer temperatures decreased less extremely, reflecting a strong seasonality. The documented glacier growth in southern Norway (Nesje et al., 2001) indicates additional high precipitation rates in winter after 6 ka ago. The timing of our ‘geological crisis’ with at least three CRSF within a short period of time coincides with this period of strong winter temperature decrease, high precipitation rates and strong seasonality, following the HTO with high air temperatures.

Not much is known about the rock mechanics and the history of stream discharge and fluvial dynamics at Mannen. The mountain section seems to be generally very dry where no surface discharge is observed and precipitation and snow-melt water drains by baseflow through fractures underground. Continuous monitoring at the most active part of the Mannen instability ‘Veslemannen’ reveals that rock mass deformation is very sensitive to precipitation and is stable during the winter season (Oppikofer et al., 2013). This could imply that increasing precipitation in the mid-Holocene together with warm summer temperatures led to widespread and repeated rock-slope destabilisation at Mannen. A continued climate warming combined with a more pronounced seasonality leading to increased snow melt could thus cause sudden destabilisation of the Mannen rock-slope in the future causing one or multiple CRSF events.

The other first-order control is the regional degradation of permafrost during the HTO and after the 6 ka cooling period. Thermal measurements today in crevasses and the back scarp of the Mannen instability along with regional permafrost mapping indicates that Mannen is situated at the present mountain permafrost limit in the area (Gisnås et al., 2013; Steiger et al., 2016; Westermann et al., 2013). When the Mannen area became ice-free after the YD, the climate was 7–9°C cooler than today based on Greenland ice-core analysis (e.g. North Greenland Ice Core Project Members, 2004). First approximations through a modelling approach suggests the build-up of several tens to hundreds of metres of permafrost in mostly snow-free rock walls during the period of deglaciation until MAAT reached similar levels to today (ca. 10 ka ago; Myhra et al., 2017). Rock joint weakening due to permafrost aggradation and degradation during this 2 ka period after deglaciation may have played a role in addition to debuttressing for the first three CRSFs at Mannen (Krautblatter et al., 2013). The permafrost certainly degraded during the HTO, but the degradation rate is depending on cracks and the ice content in the steep slopes because of thermal inertia. Thawing permafrost is widely recognised as an important factor for CRSFs (Blikra and Christiansen, 2014; Fischer et al., 2006, 2012) due to melting of ice-bonds in cracks and generally weakening of tensile and compressive strength in rock masses (Krautblatter et al., 2013; Murton et al., 2006) and has to be taken into account when discussing possible former rock slide conditioners in high-mountain environments. However, a more substantial conclusion on this can only be drawn through sensitivity studies using coupled thermo-mechanical models (Grämiger et al., 2017).

Conclusion

Below the Mannen instability in the lower Romsdal Valley, western Norway, a cluster of at least six post-glacial CRSF deposits complements a complex valley-fill stratigraphy. The present landforms are the result of the concurrence of sedimentation processes connected to deglaciation, isostatic rebound and sea-level drop, and mass wasting from the slopes. Prominent steep steps in the relief, parallel to the valley, are evidence for erosional processes either by strong tidal currents or fluvial incision. These processes have modified both stratified drift and three of the rock-slope failure deposits, supporting the stratigraphically derived time constraints of failure timing.

A set of 13 exposure ages together with sedimentologic and morphologic analyses allowed for the age determination of the six to nine distinct CRSF events. They divide into two periods of CRSF activity, one shortly after deglaciation and one 5.5–4.5 ka ago, where multiple CRSF from the same slope occurred within a short period of time. The fact that one slope fails repeatedly with recurrence intervals of a few years or decades has been observed before and must be considered for future failure scenarios. Debuttressing is a probable conditioner for the early multiple failures between 12 and 10 ka, that coincide with a major peak in rock-avalanche activity in Norway. The timing of a mid-Holocene cluster with three to six individual CRSF events has been connected to climate variation during the Holocene, especially in relation to a climatic deterioration at the end of the Holocene Climate Optimum. Higher precipitation connected to a strong seasonality, temperature changes and rock mass strength alterations related to permafrost degradation are possible climatic conditions responsible for the mid-Holocene crisis at Mannen.

Supplemental Material

1m-hillshade-Lower_Romsdalen_below_Mannen – Supplemental material for Multiple rock-slope failures from Mannen in Romsdal Valley, western Norway, revealed from Quaternary geological mapping and 10Be exposure dating

Supplemental material, 1m-hillshade-Lower_Romsdalen_below_Mannen for Multiple rock-slope failures from Mannen in Romsdal Valley, western Norway, revealed from Quaternary geological mapping and 10Be exposure dating by Paula Hilger, Reginald L Hermanns, John C Gosse, Benjamin Jacobs, Bernd Etzelmüller and Michael Krautblatter in The Holocene

Supplemental Material

Additional_comments_on_the_reported_uncertainties_for_apparent_10Be_exposure_ages – Supplemental material for Multiple rock-slope failures from Mannen in Romsdal Valley, western Norway, revealed from Quaternary geological mapping and 10Be exposure dating

Supplemental material, Additional_comments_on_the_reported_uncertainties_for_apparent_10Be_exposure_ages for Multiple rock-slope failures from Mannen in Romsdal Valley, western Norway, revealed from Quaternary geological mapping and 10Be exposure dating by Paula Hilger, Reginald L Hermanns, John C Gosse, Benjamin Jacobs, Bernd Etzelmüller and Michael Krautblatter in The Holocene

Supplemental Material

Tabular_summary_of_sedimentary_facies_associations_and_correspondent_ERT_and_GPR_units – Supplemental material for Multiple rock-slope failures from Mannen in Romsdal Valley, western Norway, revealed from Quaternary geological mapping and 10Be exposure dating

Supplemental material, Tabular_summary_of_sedimentary_facies_associations_and_correspondent_ERT_and_GPR_units for Multiple rock-slope failures from Mannen in Romsdal Valley, western Norway, revealed from Quaternary geological mapping and 10Be exposure dating by Paula Hilger, Reginald L Hermanns, John C Gosse, Benjamin Jacobs, Bernd Etzelmüller and Michael Krautblatter in The Holocene

Footnotes

Acknowledgements

The TCN sample preparation was mostly completed by the first author at CRISDal at Dalhousie University under supervision of G. Yang. Field assistance was provided by F. Magnin, B. Altena and S. Westermann (all University of Oslo) and B.R. Snook (NTNU). Discussions about the mapping investigations with R. Eilertsen, L. Rubensdotter and L. Hansen (all NGU) improved the study significantly. We thank G. Gilbert for reviewing and improving the manuscript significantly before submission and thank S. Zerathe and two anonymous reviewers for their suggestions that allowed improving the former version of the manuscript.

Funding

The study is part of the project ‘CryoWALL – Permafrost slopes in Norway’ (243784/CLE) funded by the Research Council of Norway (RCN). Additional funding was provided by the Norwegian Geological Survey, Trondheim, and the Department of Geosciences, University of Oslo. JCG acknowledges support for the TCN laboratory from Canada Foundation for Innovation (21305 and 36158), NSERC and NSRIT grants.

ORCID iD

Reginald L Hermanns

References

Agliardi

Zanchi

Crosta

(2009) Tectonic vs. gravitational morphostructures in the Central Eastern Alps (Italy): Constraints on the recent evolution of the mountain range. Tectonophysics 474: 250–270.

Aizebeokhai

(2010) 2D and 3D geoelectrical resistivity imaging: Theory and field design. Scientific Research and Essays 5: 3592–3605.

Allen

Gruber

Owens

(2009) Exploring steep bedrock permafrost and its relationship with recent slope failures in the Southern Alps of New Zealand. Permafrost and Periglacial Processes 20: 345–356.

Ballantyne

Stone

(2013) Timing and periodicity of paraglacial rock-slope failures in the Scottish Highlands. Geomorphology 186: 150–161.

Ballantyne

Sandeman

Stone

et al . (2014) Rock-slope failure following Late Pleistocene deglaciation on tectonically stable mountainous terrain. Quaternary Science Reviews 86: 144–157.

Ballantyne

Stone

Fifield

(1998) Cosmogenic Cl-36 dating of postglacial landsliding at the Storr, Isle of Skye, Scotland. The Holocene 8: 347–351.

Barnett

Dumayne-Peaty

Matthews

(2001) Holocene climatic change and tree-line response in Leirdalen, central Jotunheimen, south central Norway. Review of Palaeobotany and Palynology 117: 119–137.

Blikra

Christiansen

(2014) A field-based model of permafrost-controlled rockslide deformation in northern Norway. Geomorphology 208: 34–49.

Blikra

Nemec

(1998) Postglacial colluvium in western Norway: Depositional processes, facies and palaeoclimatic record. Sedimentology 45: 909–959.

10.

Blikra

Selvik

(1998) Climatic signals recorded in snow avalanche-dominated colluvium in western Norway: Depositional facies successions and pollen records. The Holocene 8: 631–658.

11.

Blikra

Longva

Braathen

et al . (2006) Rock slope failures in Norwegian fjord areas: Examples, spatial distribution and temporal pattern. In: Evans

Mugnozza

Strom

et al . (eds) Landslides from Massive Rock Slope Failure (NATO Science Series, vol. 49). Dordrecht: Springer, pp. 475–496.

12.

Blikra

Majala

Anda

et al . (2016) Fare-og risikoklassifisering av ustabile fjellparti – Faresoner, arealhåndtering og tiltak. Report no. 77–2016, September. Oslo: NVE (in Norwegian).

13.

Bøe

A-G

Dahl

Lie Nesje

ØA

(2006) Holocene river floods in the upper Glomma catchment, southern Norway: A high-resolution multiproxy record from lacustrine sediments. The Holocene 16: 445–455.

14.

Böhme

Oppikofer

Longva

et al . (2015) Analyses of past and present rock slope instabilities in a fjord valley: Implications for hazard estimations. Geomorphology 248: 464–474.

15.

Clavero

Sparks

Huppert

et al . (2002) Geological constraints on the emplacement mechanism of the Parinacota debris avalanche, northern Chile. Bulletin of Volcanology 64: 40–54.

16.

Corner

(2006) A transgressive-regressive model of Fjord-Valley fill: Stratigraphy, facies and depositional controls. In: Dalrymple

Leckie

Tillman

(eds) Incised Valleys in Time and Space. Broken Arrow, OK: SEPM (Special Publication no. 85), pp. 161–178.

17.

Cossart

Braucher

Fort

et al . (2008) Slope instability in relation to glacial debuttressing in alpine areas (Upper Durance catchment, southeastern France): Evidence from field data and 10Be cosmic ray exposure ages. Geomorphology 95: 3–26.

18.

Crosta

Hermanns

Dehls

et al . (2017) Rock avalanches clusters along the northern Chile coastal scarp. Geomorphology 289: 27–43.

19.

Dortch

Owen

Haneberg

et al . (2009) Nature and timing of large landslides in the Himalaya and Transhimalaya of northern India. Quaternary Science Reviews 28: 1037–1054.

20.

Eilertsen

Corner

Aasheim

et al . (2006) Valley-fill stratigraphy and evolution of the Målselv Fjord Valley, Northern Norway. In: Dalrymple

Leckie

Tillman

(eds) Incised Valleys in Time and Space. Broken Arrow, OK: SEPM (Special Publication no. 85), pp.179–195.

21.

Eilertsen

Corner

Aasheim

et al . (2011) Facies characteristics and architecture related to palaeodepth of Holocene fjord-delta sediments: Facies characteristics and architecture. Sedimentology 58: 1784–1809.

22.

Eilertsen

Corner

Hansen

(2015) Using LiDAR data to characterize and distinguish among different types of raised terraces in a fjord-valley setting. GFF 137: 353–361.

23.

Evans

Clague

(1994) Recent climatic change and catastrophic geomorphic processes in mountain environments. Geomorphology 10: 107–128.

24.

Fischer

Kääb

Huggel

et al . (2006) Geology, glacier retreat and permafrost degradation as controlling factors of slope instabilities in a high-mountain rock wall: The Monte Rosa east face. Natural Hazards and Earth System Sciences 6: 761–772.

25.

Fischer

Purves

Huggel

et al . (2012) On the influence of topographic, geological and cryospheric factors on rock avalanches and rockfalls in high-mountain areas. Natural Hazards and Earth System Sciences 12: 241–254.

26.

Gisnås

Etzelmüller

Farbrot

et al . (2013) CryoGRID 1.0: Permafrost distribution in Norway estimated by a spatial numerical model. Permafrost and Periglacial Processes 24: 2–19.

27.

Gosse

Phillips

(2001) Terrestrial in situ cosmogenic nuclides: Theory and application. Quaternary Science Reviews 20: 1475–1560.

28.

Grämiger

Moore

Gischig

et al . (2017) Beyond debuttressing: Mechanics of paraglacial rock slope damage during repeat glacial cycles. Journal of Geophysical Research: Earth Surface 122: 1004–1036.

29.

Grimstad

Nesdal

(1991) Loen rockslides: A historical review. In: Barton

Stephansson

(eds) Rock Joints. Rotterdam: Balkema, pp. 1–6.

30.

Groover

Burgess

Howle

et al . (2016) Electrical Resistivity Investigation of Fluvial Geomorphology to Evaluate Potential Seepage Conduits to Agricultural Lands Along the San Joaquin River, Merced County, California, 2012–13. Report No. 2016–5172. Reston, VA: U.S. Geological Survey.

31.

Hansen

Beylich

Burki

et al . (2009) Stratigraphic architecture and infill history of a deglaciated bedrock valley based on georadar, seismic profiling and drilling. Sedimentology 56: 1751–1773.

32.

Hermanns

Longva

(2012) Rapid rock-slope failures. In: Clague

Stead

(eds) Landslides. Cambridge: Cambridge University Press, pp. 59–70.

33.

Hermanns

Strecker

(1999) Structural and lithological controls on large quaternary rock avalanches (Sturzstroms) in arid northwestern Argentina. Geological Society of America Bulletin 111: 934–948.

34.

Hermanns

Blikra

Naumann

et al . (2006a) Examples of multiple rock-slope collapses from Köfels (Ötz valley, Austria) and western Norway. Engineering Geology 83: 94–108.

35.

Hermanns

Fauqué

Wilson

CGJ

(2015) ³⁶Cl terrestrial cosmogenic nuclide dating suggests Late Pleistocene to Early Holocene mass movements on the south face of Aconcagua mountain and in the Las Cuevas–Horcones valleys, Central Andes, Argentina. Geological Society 399: 345–368.

36.

Hermanns

Niedermann

Garcia

et al . (2001) Neotectonics and catastrophic failure of mountain fronts in the southern intra-Andean Puna Plateau, Argentina. Geology 29: 619.

37.

Hermanns

Niedermann

Garcia

et al . (2006b) Rock avalanching in the NW Argentine Andes as a result of complex interactions of lithologic, structural and topographic boundary conditions, climate change and active tectonics. In: Evans

Mugnozza

Strom

et al . (eds) Landslides from Massive Rock Slope Failure. Dordrecht: Springer, pp. 497–520.

38.

Hermanns

Niedermann

Ivy-Ochs

et al . (2004) Rock avalanching into a landslide-dammed lake causing multiple dam failure in Las Conchas valley (NW Argentina) ? Evidence from surface exposure dating and stratigraphic analyses. Landslides 1: 113–122.

39.

Hermanns

Schleier

Böhme

et al . (2017) Rock-avalanche activity in W and S Norway peaks after the retreat of the Scandinavian Ice Sheet. In: Mikoš

Vilímek

Yin

et al . (eds) Advancing Culture of Living with Landslides. Cham: Springer, pp. 331–338.

40.

Hermanns

Trauth

Niedermann

et al . (2000) Tephrochronologic constraints on temporal distribution of large landslides in northwest Argentina. Geology 108: 35–52.

41.

Hewitt

Clague

Orwin

(2008) Legacies of catastrophic rock slope failures in mountain landscapes. Earth-Science Reviews 87: 1–38.

42.

Hewitt

Gosse

Clague

(2011) Rock avalanches and the pace of late quaternary development of river valleys in the Karakoram Himalaya. Geological Society of America Bulletin 123: 1836–1850.

43.

Høgaas

Hansen

Rinstad

et al . (2012) Database for registrering av marin grense (MG) i Norge. Report No. 2012.063, December. Trondheim: Geological Survey of Norway (in Norwegian).

44.

Holm

Bovis

Jakob

(2004) The landslide response of alpine basins to post-Little Ice Age glacial thinning and retreat in southwestern British Columbia. Geomorphology 57: 201–216.

45.

Hughes

ALC

Gyllencreutz

Lohne

ØS

et al . (2016) The last Eurasian ice sheets – A chronological database and time-slice reconstruction, DATED-1. Boreas 45: 1–45.

46.

Ivy-Ochs

Kober

(2008) Surface exposure dating with cosmogenic nuclides. Quaternary Science Journal 57: 179–209.

47.

Ivy-Ochs

Martin

Campedel

et al . (2017) Geomorphology and age of the Marocche di Dro rock avalanches (Trentino, Italy). Quaternary Science Reviews 169: 188–205.

48.

Jarman

Calvet

Corominas

et al . (2014) Large-scale rock slope failures in the eastern Pyrenees: Identifying a sparse but significant population in paraglacial and parafluvial contexts. Geografiska Annaler: Series A, Physical Geography 96: 357–391.

49.

Jia

Leith

Krautblatter

(2017) Path-dependent frost-wedging experiments in fractured, low-permeability granite. Permafrost and Periglacial Process 28: 698–709.

50.

Jia

Xiang

Krautblatter

(2015) Quantifying rock fatigue and decreasing compressive and tensile strength after repeated freeze-thaw cycles. Permafrost and Periglacial Process 26: 368–377.

51.

Köpfli

Grämiger

Moore

et al . (2017) The Oeschinensee rock avalanche, Bernese Alpes, Switzerland: A co-seismic failure 2300 years ago? Swiss Journal of Geosciences 111: 205–219.

52.

Korup

(2004) Geomorphometric characteristics of New Zealand landslide dams. Engineering Geology 73: 13–35.

53.

Korup

Densmore

Schlunegger

(2010) The role of landslides in mountain range evolution. Geomorphology 120: 77–90.

54.

Krautblatter

Leith

(2015) Glacier- and permafrost-related slope instabilities. In: Huggel

Carey

Clague

et al . (eds) The High-Mountain Cryosphere: Environmental Changes and Human Risks. Cambridge: Cambridge University Press, pp. 147–165.

55.

Krautblatter

Funk

Günzel

(2013) Why permafrost rocks become unstable: A rock-ice-mechanical model in time and space. Earth Surface Processes and Landforms 38: 876–887.

56.

Leith

Moore

Amann

et al . (2014) In situ stress control on microcrack generation and macroscopic extensional fracture in exhuming bedrock. Journal of Geophysical Research: Solid Earth 119: 594–615.

57.

Lilleøren

Etzelmüller

Schuler

et al . (2012) The relative age of mountain permafrost – Estimation of Holocene permafrost limits in Norway. Global and Planetary Change 92–93: 209–223.

58.

Longva

Blikra

Dehls

(2009) Rock Avalanches: Distribution and Frequencies in the Inner Part of Storfjorden, Møre og Romsdal County. Report No. 2009.002, 15 March. Trondheim: Geological Survey of Norway (in Norwegian).

59.

Lønne

(1995) Sedimentary facies and depositional architecture of ice-contact glaciomarine systems. Sedimentary Geology 98: 13–43.

60.

McSaveney

(2002) Recent rockfalls and rock avalanches in Mount Cook National Park, New Zealand. In: Evans

Degraff

(eds) Catastrophic Landslides: effects, occurrence, and mechanisms. Boulder, CO: Geological Society of America, pp. 35–70.

61.

Moreiras

Hermanns

Fauqué

(2015) Cosmogenic dating of rock avalanches constraining quaternary stratigraphy and regional neotectonics in the Argentine Central Andes (32° S). Quaternary Science Reviews 112: 45–58.

62.

Murton

Peterson

Ozouf

J-C

(2006) Bedrock fracture by ice segregation in cold regions. Science 314: 1127–1129.

63.

Myhra

Westermann

Etzelmüller

(2017) Modelled distribution and temporal evolution of permafrost in steep rock walls along a latitudinal transect in Norway by CryoGrid 2D. Permafrost and Periglacial Processes 28: 172–182.

64.

Nagelisen

Moore

Vockenhuber

et al . (2015) Post-glacial rock avalanches in the Obersee Valley, Glarner Alps, Switzerland. Geomorphology 238: 94–111.

65.

Nesje

Matthews

Dahl

et al . (2001) Holocene glacier fluctuations of Flatebreen and winter-precipitation changes in the Jostedalsbreen region, western Norway, based on glaciolacustrine sediment records. The Holocene 11: 267–280.

66.

NGU (2018) Kvartærgeologiske Kart (Løsmassekart). Available at: http://geo.ngu.no/kart/losmasse/ (accessed 17 April 2018) [in Norwegian].

67.

North Greenland Ice Core Project Members (2004) High-resolution record of Northern Hemisphere climate extending into the last interglacial period. Nature 431: 147–151.

68.

NVE (2018) NVE Atlas. Available at: https://atlas.nve.no/Html5Viewer/index.html?viewer=nveatlas# (accessed 13 February 18) [in Norwegian].

69.

NVE, met.no and Kartverket (2018) seNorge. Available at: http://www.senorge.no/ (accessed 17 April 2018).

70.

Oppikofer

Saintot

Otterå

et al . (2013) Investigations on Unstable Rock Slopes in Møre og Romsdal – Status And Plans After Field Surveys in 2012. Report No. 2013.014, 13 November. Trondheim: Geological Survey of Norway (in Norwegian).

71.

Orwin

Clague

Gerath

(2004) The Cheam rock avalanche, Fraser Valley, British Columbia, Canada. Landslides 1: 289–298.

72.

Ostermann

Sanders

(2017) The Benner pass rock avalanche cluster suggests a close relation between long-term slope deformation (DSGSDs and translational rock slides) and catastrophic failure. Geomorphology 289: 44–59.

73.

Ostermann

Ivy-Ochs

Diethard

et al . (2016) Multi-method (14C, 36Cl, 234U/230Th) age bracketing of the Tschirgant rock avalanche (Eastern Alps): Implications for absolute dating of catastrophic mass wasting. Earth Surface Processes and Landforms 42: 1110–1118.

74.

Ostermann

Sanders

Ivy-Ochs

et al . (2012) Early Holocene (8.6ka) rock avalanche deposits, Obernberg valley (Eastern Alps): Landform interpretation and kinematics of rapid mass movement. Geomorphology 17: 183–293.

75.

Ouimet

Whipple

Royden

et al . (2007) The influence of large landslides on river incision in a transient landscape: Eastern margin of the Tibetan Plateau (Sichuan, China). Geological Society of America Bulletin 119: 1462–1476.

76.

Palacky

(1988) Resistivity characteristics of geologic targets. In: Nabighian

(ed.) Electromagnetic Methods in Applied Geophysics, Theory: Investigations in Geophysics. Tulsa, OK: SEG Books, pp. 52–129.

77.

Penna

Hermanns

Niedermann

et al . (2011) Multiple slope failures associated with neotectonic activity in the Southern Central Andes (37°–37°30′S), Patagonia, Argentina. Geological Society of America Bulletin 123: 1880–1895.

78.

Plafker

Ericksen

(1978) Nevados Huascarán Avalanches, Peru. In: Voight

(ed.) Developments in Geotechnical Engineering, Rockslides and Avalanches, vol. 1. New York: Elsevier, pp. 277–314.

79.

Plug

Gosse

McIntosh

et al . (2007) Attenuation of cosmic ray flux in temperate forest. Journal of Geophysical Research: Earth Surface 112: F02022.

80.

Reusch

(1907) Skredet i Loen 15. Januar 1905. Norges Geologiske Undersøkelse, Aarbok 1907. Kristiania: Aschehoug & Co., 20 pp (in Norwegian with English summary).

81.

Roberts

(2003) The Scandinavian Caledonides: Event chronology, palaeogeographic settings and likely modern analogues. Tectonophysics 365: 283–299.

82.

Saintot

Dahle

Derron

M-H

et al . (2012) Large gravitational rock slope deformation in Romsdalen valley (Western Norway). Revista de la Asociación Geológica Argentina 69: 354–371.

83.

Sanchez

Rolland

Corsini

et al . (2010) Relationships between tectonics, slope instability and climate change: Cosmic ray exposure dating of active faults, landslides and glacial surfaces in the SW Alps. Geomorphology 117: 1–13.

84.

Schleier

Hermanns

Gosse

et al . (2016) Subaqueous rock-avalanche deposits exposed by post-glacial isostatic rebound, Innfjorddalen, Western Norway. Geomorphology 289: 117–133.

85.

Schleier

Hermanns

Rohn

et al . (2015) Diagnostic characteristics and paleodynamics of supraglacial rock avalanches, Innerdalen, Western Norway. Geomorphology 245: 23–39.

86.

Soldati

Corsini

Pasuto

(2004) Landslides and climate change in the Italian Dolomites since the Late glacial. CATENA 55: 141–161.

87.

Steiger

Etzelmüller

Westermann

et al . (2016) Modelling the permafrost distribution in steep rock walls. Norwegian Journal of Geology 96: 329–341.

88.

Storms

JEA

de Winter

Overeem

et al . (2012) The Holocene sedimentary history of the Kangerlussuaq Fjord-valley fill, West Greenland. Quaternary Science Reviews 35: 29–50.

89.

Stroeven

Hättestrand

Kleman

et al . (2016) Deglaciation of Fennoscandia. Quaternary Science Reviews 147: 91–121.

90.

Sturzenegger

Stead

Gosse

et al . (2015) Reconstruction of the history of the Palliser Rockslide based on 36Cl terrestrial cosmogenic nuclide dating and debris volume estimations. Landslides 12: 1097–1106.

91.

Svendsen

Mangerud

(1987) Late Weichselian and holocene sea-level history for a cross-section of western Norway. Journal of Quaternary Science 2: 113–132.

92.

Tønnesen

(2009) Georadarmålinger ved Rønningen og Horgheim i Romsdalen for undersøkelse av løsmassetyper i dalbunnen under det ustabile fjellpartiet Mannen. Report No. 2009.062, 11 November. Trondheim: Geological Survey of Norway (in Norwegian).

93.

Topping

(1993) Paleogeographic reconstruction of the Death Valley extended region: Evidence from Miocene large rock-avalanche deposits in the Amargosa Chaos Basin, California. Geological Society of America Bulletin 105: 1190–1213.

94.

Trauth

Alonso

Haselton

et al . (2000) Climate change and mass movements in the NW Argentine Andes. Earth and Planetary Science Letters 179: 243–256.

95.

Tveten

Lutro

Thorsnes

(1998) Geologisk kart over Norge, berggrunnskart Ålesund, 1:250,000. Trondheim: Geological Survey of Norway.

96.

Vasskog

Nesje

Støren

et al . (2011) A Holocene record of snow-avalanche and flood activity reconstructed from a lacustrine sedimentary sequence in Oldevatnet, western Norway. The Holocene 21: 597–614.

97.

Westermann

Schuler

Gisnås

et al . (2013) Transient thermal modeling of permafrost conditions in Southern Norway. The Cryosphere 7: 719–739.

98.

Zerathe

Lebourg

Braucher

et al . (2014) Mid-Holocene cluster of large-scale landslides revealed in the Southwestern Alps by 36Cl dating. Insight on an Alpine-scale landslide activity. Quaternary Science Reviews 90: 106–127.

99.

Zimmerman

Evenson

Gosse

et al . (1994) Extensive Boulder erosion resulting from a range fire on the type-pinedale Moraines, Fremont Lake, Wyoming. Quaternary Research 42: 255–265.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

4.72 MB

0.17 MB

0.22 MB