Keywords

1 Landslides Along the Slope Off Vesterålen

Landslide processes off the Norwegian margin are abundant, with the Storegga Slide being the most prominent example (e.g. Haflidason et al. 2004). In this study, we focus on clusters of small-scale landslide features on the gently dipping slope (~3–4°) off Vesterålen, Northern Norway (Fig. 21.1). The landslides in one cluster have their headwalls at around the 500 m isobaths, whereas the second cluster is located in slightly deeper water areas. The spatially-isolated landslides occur in an area essentially devoid of topographic relief and disconnected from both canyon systems and the glacial termini further upslope (Laberg et al. 2007). The volume of failed material varies between 8.5·106 and 16.0·106 m3. The run-out distances of these landslides do not exceed 3350 m (L’Heureux et al. 2013). The landslides (SL1-SL4, SLA, SLW) cut the slope-parallel stratified glacio-marine silty to sandy clays on two well-defined slide planes (Rise et al. 2012). The sliding surfaces coincide with spatially continuous high-amplitude reflections (Rise et al. 2012; L’Heureux et al. 2013, Fig. 21.3, horizons a and c). It is unclear, however, whether the sliding planes of the slides are different in composition or have happened at the same time or not. Based on an already existing multidisciplinary database (see Vanneste et al. 2014), the soils consist of predominantly normally consolidated clays to silty clays with intermediate plasticity. Triaxial test results show intervals with strain-softening behaviour. A comprehensive slope stability assessment by L’Heureux et al. (2013) (i) attests that the slope is per se stable under static conditions and (ii) indicates that external seismically induced load and post-seismic pore pressure accumulation are the most relevant triggers for landslide and deformation processes.

Fig. 21.1
figure 1

Bathymetric map of the study area on the gently dipping slope off Vesterålen (www.mareano.no). Isolated slope failure features are aligned parallel along 500 and 800 isobath

Full size image

2 Methods

2.1 Sub-bottom Mapping

In addition to the first sub-bottom mapping in 2010 (Rise et al. 2012), a complementary survey in summer 2013 was acquired using a parametric sub-bottom profile r system (TOPAS PS18) operated from R/V G.O. Sars (Univ. Bergen). Focus of the survey was set on the mapping of the landslide complexes SL1–SL4 (Fig. 21.1; Haflidason et al. 2010).

2.2 CPTU Investigation of Slope Sediments

During the cruise POS472 (MARUM/Univ. Bremen) with R/V Poseidon, geotechnical in situ data were collected using the high-resolution static Geotechnical Offshore Seabed Tool (GOST) developed at MARUM. The system is equipped with a 5 cm2 electrical piezocone with a pore pressure port (CPTU). After landing the GOST on the seafloor, the probe is pushed hydraulically with a constant velocity of 2 cm/s into the sediments. Cone resistance qc, sleeve friction fs and excess pore water pressure response Δu2 are recorded with a sampling frequency of ~30 Hz, which results in a vertical resolution of ~2 mm. Excess pore water pressure is measured differentially against the water filled rod directly behind the cone in u2 location. For further details, see Jorat et al. (2014).

Processing of the CPTU data and the derivation of strength properties was accomplished according to standard practice summarized in Lunne et al. (1997). Undrained shear strength su was calculated using the static cone factor Nkt =12, 15 and 17 for normally consolidated clayey sediments (Lunne et al. 1997). The normalised undrained shear strength (su/σ′v0) can be used to determine the state of consolidation. For normally-consolidated soil, this ratio typically falls in the range of 0.2–0.3 (Karlsson and Viberg 1967). A ratio exceeding 0.3 may indicate overconsolidation.

Identification of soil types was carried out following the soil behaviour type classification by Ramsey (2002; SBTRamsey) and Robertson (1990; SBTRobertson) using the normalised cone resistance Qt [(qt − σvo)/σ′vo] and the pore pressure parameter Bq [Δu2/(qt − σvo)] with the vertical total stress (σvo) and the effective vertical stress (σ′vo) (Lunne et al. 1997).

2.3 Pseudo-static Factor of Safety (FoS)

The 1D undrained, infinite slope stability analysis by Morgenstern and Price (1965) defines the pseudo-static factor of safety (FoS) – the ratio of resisting forces to driving forces during earthquake shaking. Seismic -induced shear stresses are considered to be constant over the period of shaking (Hampton et al. 1996). If FoS >1, the sediments are assumed to be stable. FoS <1 indicates permanent deformation and/or failure.

FoS is defined as

$$ FoS=\frac{s_u\ }{\sigma_{\mathrm{v}0}^{\prime }\ \left( \sin \left(\alpha \right) \cos \left(\alpha \right)+k\left(\frac{\gamma }{\gamma^{\prime }}\right){ \cos}^2\left(\alpha \right)\right)} $$

with α = the slope angle, γ′ = the effective unit weight and the seismic coefficient k. Given the fact that soil slopes are not rigid and the peak ground acceleration (PGA) lasts for a very short period of time, seismic coefficients used in practice correspond to lower PGA values (Kramer 1996). Here, k was calculated with 0.3 · PGA [g] recommended by Marcuson (1981) and following the proposition by Hynes-Griffin and Franklin (1984): k = 0.5 · PGA [g]. The latter criteria is suggested to be appropriate for most slopes by Kramer (1996).

We run the analysis using a PGA of 0.051 g and 0.224 g. These values correspond to potential seismic events offshore Vesterålen with a 90 % probability of no exceedance for 475 years and 10,000 years recurrence period, respectively (NORSAR 1998).

3 Results

Geophysical as well as in situ geotechnical CPTU data presented here (Figs. 21.2 and 21.3) focus on the characterization of the uppermost strata of the slope in terms of sediment-mechanical properties and slope stability (Fig. 21.1).

Fig. 21.2
figure 2

Sub-bottom TOPAS profile crossing the location of the CPTU transect along the intact portion of the slope. Location of the seismic profile is shown in Fig. 21.1. Position as well as penetration of three CPTU are illustrated by blue lines. Dashed red lines represent the sliding planes according to Rise et al. (2012) and L’Heureux et al. (2013)

Full size image
Fig. 21.3
figure 3

Sediment -mechanical characterisation of intact slope sediments: (a) Position and penetration depth of GeoB18638 on the TOPAS line GOS_13_JUL_009; (b). Yellow-shaded areas indicate mechanically weaker zones (MWZ). (c) Undrained shear strength su was calculated with Nkt = 12 (green),15 (black) and 17 (red). Dashed lines are indicative of the state of consolidation based on su/σ′υo. Soil behaviour types were determined by Robertson (1990) and Ramsey (2002). Colour-coding of the soil types: green = Clay , NC (Ramsey), silty clay (Robertson), yellow: silty sand to sand, red: sensitive fine-grained material. Please note that the CPTU and su data are clipped for better illustration

Full size image

The TOPAS data illustrate that the sediments are well stratified with continuous strata packages of different reflectivity (Fig. 21.2). The scars of the different slide complexes clearly cut the slope sediments and terminate at different levels, typically on spatially-continuous reflections. Since the new TOPAS data closely resemble those presented by L’Heureux et al. (2013), we refrain from showing them here.

The CPTU penetrated between 7 and 19.8 m and profiled (i) intact material (GeoB18633, −38, −39) (Fig. 21.2) and (ii) different morphological parts of SL3 (GeoB18634, −35, −40) (Fig. 21.1) The CPTU located upslope (GeoB18633, −34) were stopped at 7 m and 9 m, when qc reached 20 MPa and 44 MPa, respectively to avoid bending or breaking of the rods. Availability of reliable fs data is limited as the sleeve friction jacket was blocked during some measurements.

3.1 Sedimentological and Geotechnical Characterisation of Slope Sediments

The correlation of geophysical and in situ CPTU data identifies three different sediment-mechanical units in the upper 20 m of slope sediments (Fig. 21.3).

Unit I has a well pronounced slope-parallel, distinct reflection pattern, in which corrected cone resistance qt reaches peaks up to 0.93 MPa. Excess pore pressure Δu2 rises up to 0.12 MPa. qt of the background sediment increases linearly to 0.35 MPa, which corresponds to su of 18 kPa (Fig. 21.3c). With the normalized undrained shear strength up to 0.45 (based on su with Nkt = 15), the sediments appear slightly overconsolidated.

The reflections within Unit II appear less coherent than those of Unit I. A locally distributed coarse layer is detected between 10.5 m and 11 m with a peak in qt reaching 12 MPa. A peak of that order does not appear in GeoB18639. The linear trend of su is decreasing marginally and approaches values typical for normally consolidated sediments (Fig. 21.3c). Immediately underneath the distinct peaks of qt we find intervals with exceptional low qt. These mechanically weaker zones (MWZ) are shaded in yellow in Fig. 21.3b.

Unit III is characterized by the alteration of coarse layers (high qt and low Δu2 values) with relative weaker zones (qt values = 0.58 MPa, Fig. 21.3b). In this unit, again the geophysical data reveal clearly accentuated slope-parallel layering, su of MWZ decreases to 23 kPa minimum. The material is normally consolidated (Fig. 21.3c).

Excess pore pressure response Δu2 increases linearly with depth to 330 kPa. Δu2 lows coincide with qt-peaks, which corresponds to dilative behaviour induced by shearing of the penetrating cone in coarser-grained material.

Two different types of soil type behaviour classification were used (SBTRamsey, SBTRobertson) to characterize the thin-layered strata with qt-values >1500 kPa as sandy layers. The background sediment is characterized by SBTRamsey as normally consolidated clay and by SBTRobertson as silty clay. Based on SBT the MWZ in Unit II and III correspond partly to sensitive fine-grained material (Fig. 21.3c).

3.2 Pseudo-static Slope Stability Analysis

Results of the FoS analysis show that the PGA of the seismic event with 475 years-recurrence time (PGA = 0.051 g) is not sufficient to destabilize the slope sediment in any depth level (FoS >2; Fig. 21.4). When assuming an earthquake with a PGA of 0.224 g (i.e. 10,000 year return period event) the FoS comes close to 1 (meta-stable) or falls below unity within the sediment-mechanical Unit II and III for cases where su is estimated with Nkt = 15, respectively. These failure depths correlate to the MWZ identified by means of CPTU tests in Fig.21.3b.

Fig. 21.4
figure 4

FoS was calculated for the CPTU profile GeoB18638. The lower boundary represents the FoS calculated with k = 0.5 * PGA [g], the upper boundary relates to k = 0.3 * PGA [g]. Slope angle α is 3°. su derived with Nkt = 15 was used for the calculation of FoS with a PGA of 0.224 g (10,000-years recurrence period). The black line represents the FoS related to a PGA = 0.051 g (475-years recurrence period). Yellow-shaded portions correlate to MWZ (see Fig. 21.3)

Full size image

4 Discussion and Outlook

Along the stable slope offshore Vesterålen, static CPTU data reveal the occurrence of mechanically weaker zones (MWZ) by the drop of qt-values (Fig. 21.3b). Within the sediment-mechanical Unit III, some MWZ coincide with layers where sensitive fine-grained sediments were classified by SBTRobertson and SBTRamsey (Fig. 21.3; Units II, III).

The 1D infinite, undrained pseudo-static stability calculation presents several critical slope layers (meta-stable to unstable) in Unit II and III, indicating permanent deformation or failure in case additional dynamic load is induced by an earthquake (Fig. 21.4). According to L’Heureux et al. (2013), a 10,000 year-seismic event with a PGA of 0.224 g is necessary to produce failure or permanent deformation along the slope. Such seismic events, which historically had magnitude M5.7 may have been frequent due to the glacio-isostatic uplift after the last glaciation along the slope off Vesteralen (Bungum et al. 1991; Olesen et al. 2013).

Given the fact that the stability analysis does not account for earthquake-induced excess pore pressure and inherent shear strength reduction due to accumulation of deformation , this undrained pseudo-static analysis is only a first-order approximation providing a relative stability index for the upper slope sediments. The seismic reflectors b and c (Figs. 21.2 and 21.3a) represent the sliding surfaces for SL1-SL4 and SLA, respectively (L’Heureux et al. 2013; Rise et al. 2012). The results of our analysis reveals the occurrence of critical layers, which correspond to reflector b (~10 mbsf) and c (~13 mbsf) (Fig. 21.4). The analysis attests that the mechanically weaker zones (MWZ) correlate well with the portions where FoS ≤1 (Fig. 21.4). Thus, the mechanically weak layer s can be considered as one important pre-condition for landslide activity along the gently dipping slope. External trigger (e.g. large magnitude seismic event), however, is necessary to generate landslide activity along the slope off Vesterålen.

In conclusion, the integration of in situ CPTU data with geophysical data improves soil characterization and hence foster a better understanding of the pre-conditioning factors for slope instability at the upper continental slope off Vesterålen.

Based on these preliminary in situ CPTU data, further (2D) analysis is necessary to understand i) the liquefaction potential of the coarse-grained layers and ii) the influence of the stratified fine-grained – sandy sequence in Unit III on the dissipation behaviour of seismic-induced excess pore pressure , and iii) the factors, which control the stability of the RWZ in deeper portion (13 m–18 m, Unit III; Fig. 21.4). Risk assessment for the present-day slope requires these studies for the slope off Vesterålen, where the opening for offshore oil and gas exploration is controversially debated (Misund and Olsen 2013).