Elsevier

Geomorphology

Volume 208, 1 March 2014, Pages 34-49
Geomorphology

A field-based model of permafrost-controlled rockslide deformation in northern Norway

https://doi.org/10.1016/j.geomorph.2013.11.014Get rights and content

Abstract

Knowledge about the detailed processes linked to the existence of permafrost in rockslide fractures is sparse. Large parts of the Jettan rockslide are located right below the discontinuous permafrost limit in the arctic part of the alpine landscape of northern Norway. Combining four years of meteorological and rockslide deformation data with temperature measurements from different parts of open fractures, shallow bedrock boreholes and air, as well as daily snow cover observations, allows a detailed identification of the key processes involved. These field data are the basis for the development of a permafrost controlled rockslide model. Seasonally, the deformation has a very distinctive pattern with high deformation starting abruptly right after snowmelt in May, and lasting until snow isolation in winter. Then there is a gradual transition to medium deformation as the ground is cooled further for another 1–2 months. Finally, the winter period, when maximum snow occurs in the fractures, is characterized by limited or almost no deformation. The primary controlling deformation process is meltwater percolation into fractures in summer with significant refreezing, ice formation and temperature increase in the lower part of the fractures from − 1 °C to 0 °C. Sporadic permafrost exists below the discontinuous permafrost limit, and may extend into open fractures and sliding planes. Another primary process is the significant cold air accumulation in fractures in early winter, due to the Balch effect, which significantly cools the fracture and surrounding rock promoting permafrost development. Finally, the cold air effect is stopped by snow isolation once enough snow has accumulated in the fracture by late winter. The deformation itself is thought to be controlled by changing shear strength of the brecciated sliding planes due to either changing ice temperatures and/or variations in water infiltration to the unsaturated sliding zones. The overall system is very locally controlled driving itself, and the effect of a future climate change can thus be of minor importance.

Introduction

Rockslides pose a significant geohazard in the cold, periglacial fjord landscapes of northern Norway. The Jettan rockslide at Nordnes at 69°30′N (Fig. 1) may cause serious consequences to the inhabited fjord areas due to its potential of generating destructive tsunamis (Kristensen and Blikra, 2011). A clear seasonal deformation pattern has been identified for the Jettan rockslide, which has been demonstrated not to be controlled by rainfall and snowmelt alone (Nordvik et al., 2010).

Large parts of the north Norwegian fjord landscapes are located in the arctic zone with altitudinal permafrost. The permafrost regional limit descends from ca. 990 m a.s.l. in the west, to 550 m in the interior in the eastern part of northern Norway (Christiansen et al., 2010). This means that the Jettan rockslide area, extending from 400 to 800 m a.s.l., is located at the regional modern permafrost limit. However, the irregular topography of the rockslide area, with deep open fractures, a complex fracture geometry and snow accumulation pattern, creates special local terrain and subsurface conditions. This complex topographic setting strongly suggests that permafrost can be locally present, which can affect the rockslide deformation. Studies of processes and driving mechanisms of rockslides in such environments are very limited.

The seasonal deformation of large landslides is mainly thought to be controlled by rainfall causing increased water pressure (e.g. Crosta and Agliardi, 2003, Sartori et al., 2003, Hong et al., 2005), a factor assumed also as fundamental for several rockslide events in Norway (Bjerrum and Jørstad, 1968). Snowmelt is also thought to be an important factor increasing the water level in fractures and thus controlling displacements, as assumed to be the case for the Rio Colorado Rockslide in Chile (Casassa and Marangunic, 1993), the rockslope at Affliction Creek in British Colombia (Bovis, 1990) and documented displacements based on the monitoring data from the Åknes rockslide in western Norway (Grøneng et al., 2011). Alternatively, Watson et al. (2006) interpreted the cyclic deformation of the Checkerboard Creek landslide in British Colombia as being controlled by a thermal response of the rock mass due to seasonal temperature variations. However, detailed process understanding is lacking, and the displacements in all other previously investigated sites can also be linked to the variations in water pressure.

Even though permafrost and its potential influence on landslides have been assumed to be important (e.g. Davies et al., 2001, Harris et al., 2001, Geertsema et al., 2006, Gruber and Haeberli, 2007, Sosio et al., 2008, Huggel et al., 2010), detailed field studies on the subject are still sparse. However, the slide scar of the Felik landslide in the Aosta valley in Italy has been observed to have fractures completely filled by ice down to more than 30 m depth (Bottino et al., 2002). The review of recent large slope failures in ice and rocks in Alaska, New Zealand and the European Alps (Huggel et al., 2010) indicates that periods of high air temperatures, melting of snow and ice and rapid thaw reduce the overall slope strength. Also, studies of periglacial landforms such as rock glacier dynamics provide some background for the state of knowledge concerning the seasonal dynamics in landforms characterized by open fractures and large pore spaces. A clear seasonal rhythm has been identified in the European Alps for rock glacier movement rates. The highest velocities were recorded mainly between summer and early winter; while the lowest velocities were generally observed in spring or early summer (Delaloye et al., 2010). To some degree seasonal changes in rock glacier kinematics appear to primarily reflect permafrost temperature variations, mainly considered to be thermally-induced (Delaloye et al., 2010). Davies et al. (2001) found minimum shear strength of frozen rock joints at temperatures between − 1.5 °C and 0 °C using centrifuge modeling. Also other authors have collected various indications that the deformation of frozen debris or rock glaciers increases with increasing ground temperatures up to 0 °C (e.g. Hoelzle et al., 1998, Frauenfelder et al., 2003, Ikeda et al., 2003, Kääb et al., 2007).

Since 2007, continuous rockslide deformation measurements have been carried out in different parts of the Jettan rockslide using various methods, enabling quantification and characterization of the deformation style (Nordvik et al., 2010). Contemporarily, continuous records of air, rock, snow and ice temperatures have been collected in and around the rockslide area. Meteorological data were recorded close to the unstable area. This, in combination with direct field observations of ice in the deeper part of fractures, has enabled the identification of the key seasonal frozen ground processes controlling the rockslide deformation.

The main scientific aim of this work is to obtain a better understanding of the seasonally active processes of the complex terrain and permafrost environment in the fractures of the Jettan rockslide. Here we focus on data collected from early autumn 2007 until the end of summer 2011, providing a four-year data record.

Section snippets

The Jettan rockslide study area

The Jettan rockslide is part of several large instabilities along a steep plateau of the Nordnesfjellet Mountain at the east side of the fjord Storfjorden, extending up to about 900 m a.s.l. (Fig. 1). Among the instabilities, the largest active movements have been documented on the Jettan rockslide (Fig. 2). Geological studies were initiated by the Geological Survey of Norway (NGU) in 1999 (Blikra and Longva, 2000, Braathen et al., 2004), and have been continued using various investigation

Ground deformation

Extensive surface displacement monitoring was established in 2007 (Blikra et al., 2009, Nordvik et al., 2010), consisting of 11 crackmeters, 15 tiltmeters and three lasers all installed in the Jettan rockslide area. In 2010 a DGPS network consisting of 11 antennas and three extensometers was added, providing valuable 3D movement vectors. The noise level can be relatively large for GPS data, and movement pattern can be difficult to see if the deformation is low. All GPS antennas are located in

Ground deformation

As displacement patterns at the Jettan rockslide was earlier analyzed including lasers and crackmeters in a 16-month period (2007–2009) (Nordvik et al., 2010), we focus here on a two year period from summer 2009 to summer 2011. The data presented include mainly sensors close to the main active fracture and the high-velocity northern part. The overall displacement data clearly document that only smaller parts of the unstable structure have active displacement (Fig. 2). The GPS data and the

Seasonal processes in fractures

The deformation at the Jettan rockslide has a significant seasonal pattern, and we propose a model divided into four seasonal stages (Fig. 16). The model is focused on the seasonal process changes in the fractures, and is discussed below.

Conclusions

The unique field data set collected for several years in the periglacial environment of the Jettan rockslide in northern Norway has enabled a detailed understanding of the interactions between permafrost and deformation processes. The permafrost controlled rockslide model we present clearly shows that the primary controlling factors are the spring snowmelt, the early to middle winter cold air accumulation into the fractures, and the middle to late winter snow isolation of the fractures. The

Acknowledgments

Our best thanks to ‘Nordnorsk fjellovervåking’, the monitoring center for the Jettan rockslide and its crew, for drilling the three 2.5 m boreholes, good assistance with logistics during field work, permitting us to use their equipment and facilities on Nordnes, and providing data included in this paper. Thanks to Håvard Juliussen for assisting with the setup of some of the temperature installations. Many thanks to Ulrich Neuman, Geokolibri, and to Markus Eckerstorfer for carrying out the BTS

References (44)

  • M.J. Bovis

    Rock-slope deformation at Affliction Creek, southern Coast Mountains, British Colombia

    Can. J. Earth Sci.

    (1990)
  • A. Braathen et al.

    Rock-slope failures in Norway; type, geometry, deformation mechanisms and stability

    Nor. J. Geol.

    (2004)
  • D. Carrera et al.

    Analysis of groundbased Lidar data from Troms County (Norway)

  • G. Casassa et al.

    The 1987 Río Colorado rockslide and debris flow, Central Andes, Chile

    Bull. Eng. Geol.

    (1993)
  • H.H. Christiansen et al.

    The thermal state of permafrost in the Nordic area during the International Polar Year 2007–2009

    Permafr. Periglac. Process.

    (2010)
  • G.C. Crosta et al.

    Failure forecast for large rock slides by surface displacement measurements

    Can. Geotech. J.

    (2003)
  • M.C.R. Davies et al.

    The effect of rise in mean annual temperature on the stability of rock slopes containing ice-filled discontinuities

    Permafr. Periglac. Process.

    (2001)
  • J.F. Dehls et al.

    Extensive InSAR observations of the Jettan rockslide in northern Norway, using Radarsat-2, TerraSAR-X, and corner reflectors

  • R. Delaloye et al.

    Overview of rock glacier kinematics research in the Swiss Alps. Seasonal rhythm, interannual variations and trends over several decades

    Geograph. Helv.

    (2010)
  • R. Frauenfelder et al.

    Rockglacier occurrence and related terrain parameters in a study area of the eastern Swiss Alps

  • H.M. French

    The Periglacial Environment

    (2007)
  • S. Gruber et al.

    Permafrost in steep bedrock slopes and its temperature related destabilization following climate change

    J. Geophys. Res.

    (2007)
  • Cited by (0)

    View full text