Analyses of past and present rock slope instabilities in a fjord valley: Implications for hazard estimations of large rock slope failures M. Böhme* 1,2 , T. Oppikofer 1 , O. Longva 1 , M. Jaboyedoff 3 , R. L. Hermanns 1 , M.-H. Derron 3 *Corresponding author, email: martina.bohme@ngu.no 1 Geological Survey of Norway, Leiv Eirikssons vei 39, 7040 Trondheim, Norway, 2 Norwegian University of Science and Technology, 7491 Trondheim, Norway 3 Center for Research on Terrestrial Environment, University of Lausanne, 1015 Lausanne, Switzerland NH33A-1635 References Hermanns, R. L., Oppikofer, T., Anda, E., Blikra, L. H., Böhme, M., Bunkholt, H., Crosta, G. B., Dahle, H., Devoli, G., Fischer, L., Jaboye- doff, M., Loew, S., Sætre, S., Molina, F. Y., 2012. Recommended hazard and risk classification system for large unstable rock slopes in Norway. Tech. Rep. 2012.029, Geological Survey of Norway. Longva, O., Blikra, L. H., Dehls, J., 2009. Rock avalanches - distribution and frequencies in the inner part of Storfjorden, Møre og Romsdal County, Norway. Tech. Rep. 2009.002, Geological Survey of Norway. Oppikofer, T., December 2009. Detection, analysis and monitoring of slope movements by high-resolution digital elevation models. Ph.D. thesis, University of Lausanne, Switzerland. Geiranger Fjøra Tafjord Stranda Åknes 390000 390000 395000 395000 400000 400000 405000 405000 410000 410000 415000 415000 6885000 6890000 6890000 6895000 6895000 6900000 6900000 6905000 6905000 6910000 6910000 6915000 6915000 Fjord deposits (Age) 0-1,000 years BP 1,000-5,000 years BP 5,000-8,000 years BP 8,000-9,000 years BP 9,000-10,000 years BP 10,000-11,000 years BP 11,000-12,500 years BP Potential instabilities (Volume) 0.05 - 0.1 million m³ 0.1 - 0.3 million m³ 0.3 - 1 million m³ 1 - 3 million m³ 3 - 10 million m³ 10 - 32 million m³ ± Tafjord Norddals Fjord Sunnylvs Fjord Geiranger Fjord Storfjord 9,800 0 2,000 4,000 6,000 8,000 10,000 12,000 0 5 10 15 20 Age [calibrated 14 C years BP] Cumulative frequency Age interval Data Exhaustion model (R²=0.89) Steady state decline model (R²=0.94) Constant frequency model for deposits <10,000 years BP (R²=0.92) 0 2,000 4,000 6,000 8,000 10,000 12,000 0 20 40 60 80 100 120 Age [calibrated 14 C years BP] Cumulative frequency Age interval Data Exhaustion model (R²=0.94) Steady state decline model (R²=0.95) Constant frequency model for deposits <9,000 years BP (R²=0.92) 0 2,000 4,000 6,000 8,000 10,000 12,000 0 5 10 15 20 25 Age [calibrated 14 C years BP] Frequency per millennia 0 2,000 4,000 6,000 8,000 10,000 12,000 0 0.5 1 1.5 2 2.5 3 3.5 4 Age [calibrated 14 C years BP] Frequency per millennia 0 2,000 4,000 6,000 8,000 10,000 12,000 Age [calibrated 14 C years BP] Total volume per millennia [10 6 m 3 ] 1 10 100 0.01 0.1 Denudation rate per millennia [mm/year] 0−9,000 years BP 9,000−12,500 years BP 0.1 1 10 100 Volume [10 6 m 3 ] 0.01 0.1 1 Normalized cumulative frequency (V≥x) Volume [10 6 m 3 ] 0 2,000 4,000 6,000 8,000 10,000 12,000 0.1 1 10 Age [calibrated 14 C years BP] 0.1 0.01 Denudation rate per millennia [mm/year] Total volume per millennia [10 6 m 3 ] 0−10,000 years BP 10,000−12,500 years BP 0.1 1 10 0.1 1 Normalized cumulative frequency (V≥x) 0.01 0.1 1 10 100 0.1 1 Volume [10 6 m 3 ] Normalized cumulative frequency (V≥x) Taord deposits (80% of volume) Ancient scars Lognormal distribution Power-law model (for V≥0.3) 0.01 0.1 1 0.1 1 Volume [10 6 m 3 ] Normalized cumulative frequency (V≥x) Taord deposits < 10,000 years BP (80% of volume) Potential instabilities (smallest scenarios) Potential instabilities (largest scenarios) Lognormal distribution Power-law model (for V≥0.3) 414000 414000 415000 415000 416000 416000 417000 417000 418000 418000 6903000 6903000 6904000 6904000 6905000 6905000 6906000 6906000 6907000 6907000 6908000 6908000 Rockslide scars Potential instabilities Fjord deposits 0-1,000 years BP 1,000-5,000 years BP 5,000-8,000 years BP 8,000-9,000 years BP 9,000-10,000 years BP 10,000-11,000 years BP with names 49 48 47 47b H1 46 46b 46c 45 44 44b 44c 43c 43 43b H2 H3 Northeastern flank of Tafjord 10 6 m 3 10 6 m 3 S i H i x f (V ≥ x) f (V ≥ x, 1) 1.08(0.73, 1.30) × 10 -3 9.01(4.99, 12.85) × 10 -4 6.82(2.7, 11.94) × 10 -4 3.94(0.63, 8.73) × 10 -4 1.82(0.08, 5.62) × 10 -4 5.59(0.04, 37.85) × 10 -5 6.20(4.65, 7.61) × 10 -3 4.23(2.79, 6.08) × 10 -3 2.37(1.23, 3.68) × 10 -3 7.92(2.36, 17.96) × 10 -4 1.79(0.25, 6.90) × 10 -4 1.96(0.10, 16.80) × 10 -5 Background For quantitative hazard estimations it is necessarry to define the magnitude-frequency distribution and a temporal model of the landslide frequency. This is often complicated for large rock slope failures due to the lack of signifi- cant numbers of large rock slope failures in inventories of a given homogeneous region or sparse information about their timing. Owing to the existence of one of the most complete rock slope instability inventories worldwide, which is based on the unique combination of independently mapped fjord deposits (Longva et al., 2009), ancient rockslide scars and potential instabilities (Oppikofer, 2009), regional hazard values are assessed. Furthermore, it is attempted to assign semi-quantitative hazard values to a number of potential instabilities in the same study area. Temporal analysis of ord deposits Volume-frequency analysis Use of frequencies of past failures to assess hazard → Given that the inventory is complete → Verified fjord deposit inventory allowing for hazard calculations f(V≥x) - normalized cumulative frequency of landslice size f(T) - frequency of landslide occurrence for time period T (in years) n - number of potential instabilities within the considered volume class x 1 ≤V≤x 2 Conclusions This study presents a potential method to quantitatively assess hazard values for large rock slope instabilities in a limited region in western Norway. The assigned semi-quanti- tative hazard values for each potential instability may form the input for quantitative risk assessments and help on decisions regarding land-use planning or building applications. It needs to be investigated in how far the results can be transferred to other regions in Norway, but also worldwide. Regions with a similar geology, a similar glacial history and a similar topography should show similar magnitude-frequency relations. The temporal distribution of rock slope failures following a deglaciation, could be investigated in this study, and the results are valuable for other deglaciated regions in the world. Inventories Rockslide deposits on the ord bottom for the Stor - ord region (Longva et al., 2009) • Based on swath bathymetry and reflection-seismic profiles • 108 rock avalanche deposits • Relative slide chronostratigraphy • Surficial area and volume estimations Rockslide scars and potential instabilities in the Taord region (Oppikofer, 2009) • Based on fieldwork, DEM and aerial photograph analysis • 17 scars of ancient rockslides • 17 potential rock slope instabilities • Volume estimations • Large frequency directly after deglaciation • Large uncertainties due to large age intervals • Constant frequency model following a rapid response • Constant yearly frequencies: Storfjord region: 0.0063±0.0013 Tafjord region: 0.0011±0.0002 Storfjord region • Largest total volumes failed directly after deglaciation • No rock slope failure >5x10 6 m 3 during last 10,000 years • Constant total volumes per millenia: Storfjord region (from 9,000 years BP): 4.4±0.5x10 6 m 3 Tafjord region (from 10,000 years BP): 1.9±0.2x10 6 m 3 • Same geomechanical, geomorphological and geographic setting for all three in- ventories in Tafjord • Underlying processes can be assumed identical • Consistency in the distributions may serve as a measure of correctness of the proportions of different volumes within each inventory Hazard calculations are based on a lognormal distribution, including 95% confidence bounds in brackets Assignment of hazard values to single instabilities → Susceptibility S i for each instability necessary → Qualitative susceptibility based on Hermanns et al. (2012) resulting in ranking r i of instabilities Implications for hazard Tafjord region Norway Sweden Denmark Almost identical distributions Different distributions for V>0.2x10 6 m 3 Power-law exponents: 0.48±0.14 0.42±0.08 Power-law exponents: 0.57±0.18 1.09±0.48 0.81±0.32