Keywords

1 Introduction

During the last decades large submarine landslides have been widely discovered in Polar Regions. The Norwegian margin has been subject of a comprehensive study motivated by the occurrence of gas and oil fields associated to nearby landslides. Large and medium-size landslides are well documented (i.e. the Trænadjupet Slide and the Andøya Slide), including the largest known submarine landslide, the Storegga Slide with a volume of 5600 km3 (Haflidason et al. 2005; Laberg et al. 2000; Laberg et al. 2002). The thick deposits accumulated during glacial and interglacial cycles have been subject to ice sheet dynamics, loading and unloading by the grounded ice sheet, glacio-eustatic sea-level variations, glacio-isostatic rebound and associated seismicity (Bugge et al. 1987; Mulder and Moran 1995). Continental margin development must have played a significant role in controlling the migration of interstitial fluids, determining the occurrence of sediment instability when combined with depositional oversteeping (Dimakis et al. 2000). The aim of this study is therefore to: (1) characterize the compression and permeability characteristics of glacial-deglacial-interglacial marine sediments of a polar continental margin; (2) reconstruct the paleohydrogeological evolution; and (3) determine how continental margin architecture and physical properties couple to control the location of submarine slope instability.

The study area (Fig. 9.1) is located in the Storfjorden Trough Mouth Fan (TMF) south of the Svalbard archipelago. This TMF covers an area of about 40.000 km2 and has a radius of about 190 km, developed concentrically off the Storfjorden trough.

Fig. 9.1
figure 1

Location of the study area. Bathymetric shaded relief showing submarine landslides beyond the shelf break. Seismic lines 1 and 2 (in pink) extracted from Faleide et al. 1996 [6] and ITEG08-09 (in green) from EGLACOM cruise report (REL OGS 2008/111). Green dots are gravity cores. Red dot is the Site ODP 986

Full size image

2 Data and Methods

Consolidation and permeability tests were performed using a GDS Rowe & Barden-type Consolidation cell equipped with three 2 MPa advanced pressure/volume controllers. Atterberg limit were also determined. The liquid limit was determined using an 80 g, 60° apex fall cone device, while the plastic limit was estimated from the hand rolled thin thread of sediment method (Karlsson et al. 1977).

The stratigraphy and 2D architecture of the different units used to perform the hydrogeological modeling correspond to the seismic units defined in between seismic reflections R1-R7 described in Faleide et al. 1996. These seismic reflections were picked on seismic line ITEG08-09 acquired during the Italian cruise EGLACOM (Fig. 9.2). The R7 and the OB (top of Oceanic Basement) reflectors correspond to the projected position of these reflectors on line ITEG08-09 made from the two seismic lines north and south of Storfjorden included in Faleide et al. (1996) as the line ITEG08-09 did not have enough penetration to image these reflectors. The velocities for the time to depth conversion have been approximated from ODP Leg 162 holes 986C and 986D sonic data.

Fig. 9.2
figure 2

(a) uninterpreted and (b) interpreted EGLACOM ITEG08-09 seismic line used for “BASIN” modelling. (c) ages for seismic reflectors [11]

Full size image

Using the Finite Element Software “BASIN” (Bitzer 1996; 1999) continental margin hydrogeological modelling has been carried out to simulate the fluid migration and pore pressure development. In this study pore pressure is also described in terms of overpressure (λ), defined as (Flemings et al. 2008):

$$ \uplambda ={{{\left( {\mathrm{P}-{{\mathrm{P}}_{\mathrm{h}}}} \right)}}\ \left/{{\left( {{\upsigma_{\mathrm{v}}}-{{\mathrm{P}}_{\mathrm{h}}}} \right)}} \right.} $$
(9.1)

where P is pore pressure, Ph is the hydrostatic pressure and σv is lithostatic or total stress. The initial thickness (Hi) of different strata used as input for the model was calculated using van Hinte’s decompaction equation (Van Hinte 1978):

$$ {{\mathrm{H}}_{\mathrm{i}}}={{\mathrm{H}}_{\mathrm{f}}}\left[ {\left( {1-{\phi_{\mathrm{f}}}} \right)\ /\ \left( {1-{\phi_{\mathrm{i}}}} \right)} \right] $$
(9.2)

where ϕ i is the initial porosity, ϕ f is the present-day porosity and Hf is the present sediment thickness.

The total length of the modeled transect is around 162 km and the mesh nodes are equally spaced every 4 km. Ice-induced stresses or erosion by ice have not been considered. Sedimentary facies are often represented by a mixture of sediment types, whose composition will vary according to the relative abundance of each sediment type for a given area and unit. Physical properties are also averaged according to the sediment mixture. Time intervals were extracted from Knies et al. (2009) (Fig. 9.2c).

3 Results

Consolidation and permeability testing were performed on two sediment types: laminated sediments interpreted as plumites, and glacigenic debris flows (GDFs) (Table 9.1). Pre-consolidation pressures indicate normally consolidated sediments for plumites and GDFs (Table 9.1). Plumites also have higher initial hydraulic conductivity (4.8 × 10−8 m/s compared to 2.9 × 10−8 m/s) and compressibility (0.36 versus 0.19) than GDFs. The specific storage and permeabilities show also a more marked decrease with increasing stress for plumites than for GDFs (Table 9.1 and Fig. 9.3). Geotechnical tests (Fig. 9.3) show that climatically controlled sedimentation changes on polar continental margins produce sediments with contrasting physical properties. The steeper trend in the virgin consolidation and theoretical permeability lines of plumites clearly highlights these differences (Fig. 9.3). Tests performed on split plumites cores SVAIS-02 and SVAIS-03 indicate higher water content and lower shear strength than the GDFs for a similar burial depth (Lucchi et al. 2012).

Table 9.1 Most important parameters derived from geotechnical tests in this study
Full size table
Fig. 9.3
figure 3

Consolidation tests: (a) effective stress/void ratio relationship for the tested samples. (b) Flow-through permeability versus void ratio. Extrapolation to 1 kPa is used to determine initial parameters for the simulation. Greenish lines correspond to plumites and reddish to GDFs

Full size image

The parameters obtained from consolidation/permeability testing (see Table 9.2) have been used as input for the hydrogeological modeling. The last output stage in terms of physical properties and sedimentological/architectural characteristics from the model are presented in Fig. 9.4. At the end of the simulation (present day), the minimum porosities are around 10 % in the continental shelf and upper slope, mainly corresponding to the units consisting of tills and glacigenic debris flows. Hydraulic conductivities for these units have values between 10−12 and 10−10 m/s, but the values start decreasing at ∼1.2 Ma. Overpressure reaches values of ∼0.15 near the shelf break and upper slope where the sediment thickness is maximum. A sensitivity analysis showed that hydraulic conductivity is the parameter that has larger influence on the resulting overpressure than other the parameters involved in the calculations as porosity or specific storage.

Table 9.2 Parameters used for “BASIN” and fluid flow modeling
Full size table
Fig. 9.4
figure 4

Margin stratigraphic and hydrodynamic modeling with “BASIN” at final simulated present condition. (a) Margin stratigraphy according to seismic units described by Faleide et al. (1996). (b) Facies composition. (c) Fractional porosity. (d) Log hydraulic conductivity. (e) Excess pore pressure (kPa). (f) Overpressure (λ)

Full size image

4 Discussion

“BASIN” modeling allows us to investigate how physical properties, sedimentology and stratigraphic architecture couple to control margin hydrogeology and fluid flow pathways. The simulation shows that low porosities (10–30 %) develop in most of the continental margin, specially the continental shelf and upper-middle slope, due to consolidation and the initial low porosity of till and glacigenic debris flow sediments.

Accumulation of GDFs on the upper-middle continental slope, which has low initial permeability and compressibility, determines the formation of fluid flow divergence beneath the continental shelf and lower slope. Most of the margin shows vertical fluid flow which is a normal situation for a fluid flow trend in a continental margin dominated by consolidation processes. The low porosity and permeability of the tills act as an impervious boundary which results in shallow moderate pore pressure and high overpressure. A more marked reduction in permeability of hemipelagic sediments due to their higher compressibility induces high pore pressures at depth but only moderate overpressure.

Onset of glaciation between 1.8 and 1.6 Ma (Butt et al. 2000; Laberg et al. 2010) provides abundant glacigenic sediments (tills and GDFs) to the continental shelf and upper slope, particularly since intensification of glaciation from 0.99 Ma (∼R4) (Knies et al. 2009). These sediments control to a large extent overpressure development. A maximum value of 0.5 is reached on the continental shelf, while the shelf edge and upper slope (where most of the landslides occur in that area) display a lower maximum value of 0.15.

High overpressure in glaciated continental margins may favor onset of slope failure (Kvalstad et al. 2005) and recent and buried landslides have also been identified off Storfjorden (Fig. 9.1) (Lucchi et al. 2012; Rebesco et al. 2012). A preliminary analysis of the factor of safety (FoS) using Eq. (9.3) (Flemings et al. 2008):

$$ FoS=\frac{{\tan {\phi_f}}}{{\sin \phi \cdot \cos \phi }}\left( {\cos^2 \phi -\lambda } \right) $$
(9.3)

where ϕ f is the friction angle (set constant at 28º) and ϕ the slope, shows that overpressure significantly reduced the FoS (FoS > 1 stable and FoS < 1 unstable), but not enough to induce slope failure. Along the reflector R2 (0.5 Ma) (where the largest slope failure is rooted), the steepening during margin development induced a slight decrease of the FoS in the upper slope since its deposition to present day, but an increase in some areas of the lower slope (Fig. 9.5). Close to the seafloor (layer immediately beneath) the FoS decreases with time in the slope area, but the lower values of FoS are still above 9. This fact shows that the slope has a higher control on the stability of the continental slope than overpressuring in the study area. At the maximum slope angle of 3º in the model, a minimum value of overpressure of 0.89 is needed for the slope to fail, highlighting that external mechanisms are needed to induce failure.

Fig. 9.5
figure 5

(a) Evolution of the safety factor on the layer immediately below the seafloor. (b) Evolution of the safety factor on the R2 surface (∼0.5). Dashed lines represent the position of the shelf edge at different times during simulation

Full size image

Weak layers are often involved in slope failure of Arctic and peri-Arctic regions (Vanneste et al. 2006). The high water content interglacial sediments that occur beneath rapidly deposited tills and GDFs can develop locally high overpressure zones which cannot be resolved in our model. Future work also needs to explore the effects of ice sheet erosion and isostatic compensation due to ice advance and retreat.

5 Conclusions

Consolidation tests of glacial and glacially influenced sediments of the upper slope of the Storfjorden TMF, western Barents Sea show that plumites (glacial melt-water plumites) have high void ratios, consolidation coefficients and permeabilities with respect to glacigenic debris flows at initial deposition conditions.

The compressibility and permeability values, together with stratigraphic input from seismic reflection profiles, were used in “BASIN” modeling. The modeling shows that onset of glacial sedimentation has a significant role in developing permeability barriers (tills) and/or allowing pore water drainage (GDFs). Occurrence of relatively uncompressible GDFs on the continental slope controls the margin fluid migration pathways.

Present day overpressure reaches a maximum value of 0.5 in the shelf area, 0.15 in the upper slope and 0.23 in the lower slope. These values of overpressure cannot trigger submarine landslides alone, but reduce the FoS of the margin by up to one half. Given the relatively reduced overpressure the FoS is mainly controlled by the evolution of the slope gradient during the margin development. An overpressure of more than 0.89 is needed to bring the FoS below 1. Also trigger mechanisms as glacio-eustatic rebound earthquakes can be considered.