Sweeney, C. G, Brideau, M. A, Augustinus, P. C & Fink D. (2013) Lochnagar landslide-dam – Central Otago, New Zealand: geomechanics and timing of the event Proc. 19 th NZGS Geotechnical Symposium. Ed. CY Chin, Queenstown Lochnagar landslide-dam, Central Otago, New Zealand: geomechanics and timing of the event C G Sweeney School of Environment, University of Auckland, Auckland, New Zealand c.sweeney@blackrockmining.net (Corresponding author) M A Brideau School of Environment, University of Auckland, Auckland, New Zealand MBrideau@bgcengineering.ca P C Augustinus School of Environment, University of Auckland, Auckland, New Zealand p.augustinus@auckland.ac.nz D Fink ANTARES-AMS, Australian Nuclear Science and Technology Organisation, Sydney, Australia Keywords: landslide dammed lake, kinematic analysis, distinct element modelling, 10 Be dating ABSTRACT The stability of natural rockslide dams with respect to sudden breaching is a major safety issue in mountain areas, although unbreached rockslide-dammed lakes such as Lochnagar in Central Otago, New Zealand, may persist in the landscape for millennia. Consequently it is important to attempt to understand the mode and drivers of these failures as they may impact on our expanding populations. The Lochnagar landslide-dam is located within the steep schistose mountains of the Southern Alps of New Zealand. During an extensive study of the site, structural and Schmidt hammer measurements were taken from the immediate area of the failure, as well as samples collected for laboratory analysis. Point load estimates of the uniaxial compressive strength were of 85 MPa perpendicular and 18 MPa parallel to the schistosity. The rock mass quality was estimated using the Geological Strength Index (GSI) with values of 35-45 observed. These results were used in a series of numerical modelling techniques: kinematic analysis (DIPS), limit equilibrium (Swedge) and distinct element modelling (3DEC). The results of the numerical modelling suggest that a wedge failure with toe buckling or ploughing through its rockmass is the likely failure mechanism. An observed fault zone at the base of the landslide may have preconditioned the slope to failure by weakening the toe. Preliminary ages from a program of Terrestrial Cosmogenic Nuclide (TCN) 10 Be exposure age dating, indicate that the slide mass that formed the dam is of at least early Holocene age and therefore glacial retreat after the LGM could also be a contributing factor. 1 INTRODUCTION Landslide-dam lakes are a natural phenomenon throughout the world, the drivers of which are becoming increasingly important to understand (Hermanns, 2011) as these failures continue to encroach on the expanding population - given the potential for catastrophic outburst floods and debris flows from rockslide-dammed lakes (Korup, 2011). Natural landslide-dams form as a consequence of rockslope instability in mountainous terrain. The stability of natural rockslide-dams, with respect to sudden breaching, is a major safety and security issue in mountain areas, such as Lochnagar in Otago (Evans et al., 2011). Korup (2002) reviewed relevant research on landslide-dams in New Zealand and found that in most cases the studies did not progress beyond the descriptive stage, with a significant lack of knowledge on the processes that controlled the formation, failure, and instability of landslide-dams (Korup, 2002). A regional study undertaken by Korup (2011) on landslides and landslide-dams in the Central Southern Alps found that 43 (from a national inventory of 240) rock avalanches exist. However only two dams within this study area still retain lakes, of which Lochnagar is the largest with a lake volume of 10 8 m 3