Bjorn Heincke, NGU- Geological Survey of Norway (Norway)
Thomas Günther, GGA-Leibniz Institute for Applied Geosciences (Germany)
Jan Steinar Rønning, NGU- Geological Survey of Norway (Norway)
Einar Dalsegg, NGU- Geological Survey of Norway (Norway)
Guri Venvik Ganerød, NGU- Geological Survey of Norway (Norway)
Despite of large methodological improvements of shallow geophysical methods in the past two decades, applied geophysics on rockslides is still a challenging task. Many factors like fracture zone distribution, water content, flow paths, characteristics of sliding plane(s) and other geological settings can affect the stability and movements of slopes making rock slides to highly complex research objects. Unfortunately, individual geophysical methods often suffer from limited resolution and difficulties to "translate" the determined physical model to actual geotechnical parameters required to understand mechanisms responsible for slope instability. Moreover, inexpensive 2-D investigations are usually performed that are inherently limited to resolve simple subsurface conditions.
Accounting for these limitations we present an application of a structural joint inversion to 3-D seismic refraction tomography and DC resistivity data collected on a huge slope instability - the Åknes rockslide in Western Norway. Structural joint inversion means that two (or more) inversions are linked to each other by identifying and adjusting structural similarities. Otherwise these inversions are independent and determine different model parameters; in our case seismic velocity and resistivity. We use the approach from Gnther and Rcker (2006) in which the structural link is accomplished by mutually controlled smoothing constraints based on the principles of robust modelling. Because such structural information constraints both inversions, their resolution is improved and finally the obtained models are in better accordance with each other than for individual inversions.
The volume of the Åknes landslide is estimated to be 30-40 million m3 and a potential failure would generate a hazardous tsunami wave in the Sunnylvsfjord below. A multidisciplinary research project has been initiated in 2003 to understand the internal processes and to build up an early warning system. Our seismic survey was located in the upper part of the unstable rock mass where debris coverage limits mapping of fracture zones. Altogether 96 geophones were placed along 4 crossing profiles. Three profiles (Q1-Q3) ran along slope and one perpendicular to slope direction (P1). Receiver spacing was 10 m for Q2 and 5 m for the other profiles. Most of the 163 shots were evenly distributed over an area of 250 x 250 m. Nine intersecting 2-D geoelectric profiles cover the complete unstable area. Six of them crossed the seismic investigation area. Electrode spacing was 20 m. Wenner and dipole-dipole configurations were measured.
With a single 3-D seismic inversion main shallow fracture zones were detected in the surveyed area. Resistivity variations determined from single 2-D and 3-D geoelectric inversions give indications for the water distribution. Because the joint inversion links both velocity and resistivity structure, its results give a better understanding in which way fractures act as flow paths.