The Queenstown Area is a major tourist destination in the lower South Island and has approximately 16,000 permanent residents. Residential development occurs predominantly on flat-lying alluvial terraces, and elsewhere on steep till-covered bedrock slopes modified by glacial ice and freeze-thaw processes. The steep slopes surrounding the Queenstown area are predisposed to instability due to inherent weakness of the Otago Schist due to lithotype variation, foliation attitude, foliation shears, and rock mass discontinuities. Pressure for residential expansion has resulted in an increasing number of development proposals on the lower slopes of Queenstown Hill.

The primary aims of this research were to develop a detailed geomorphic and geotechnical characterization of the schist within the Queenstown Hill Landslide, in response to increased development in the surrounding area. This study investigates the geomorphology and geotechnical properties of the landslide through mapping, paired with rock mechanics testing of the underlying schist. Four 25 m boreholes drilled into the landslide provided an opportunity to characterize recovered rock material and to evaluate the rock mass within landslide, as subsurface investigations were not previously available to inform investigations along the Frankton Arm.

The Queenstown Hill Landslide is a 50-75 m deep compound translational rockslide comprising interlayered quartzofeldspathic and semi-pelitic schist having an average foliation attitude dipping subparallel to the slope at 15-25° SSW. The main landslide body is interpreted as a compound rockslide with a bi-linear rupture surface, while the geometry of the eastern reactivation zone suggests a planar rockslide. The failure surface is interpreted as being subparallel to foliation, structurally controlled along foliation shears, and with block release defined by the intersection of steep to subvertical joint sets with multiple pre-existing foliation shear zones 0.1 to 1.5 m thick.

Geomorphic mapping identified five zones within the landslide from distinct surface morphologies: (1) a joint controlled headscarp; (2) the main landslide body with a subdued hummocky topography and well established drainage; (3) an extensional zone characterized by a graben and large tension cracks > 5 m deep, up to 100 m long; (4) a reactivation zone interpreted as a separate phase of movement; (5) a complex undulating toe zone bounded by compressional features but with no clear breakout.

Two interlayered schist lithotypes were mapped across the landslide and logged in the boreholes, these being a medium grey quartzofeldspathic schist and a dark green-grey semi-pelitic schist. Both lithotypes belong to the chlorite zone of the greenschist facies but exhibit variations in mineral assemblage, textural zone and weathering. A general predominance of the more micaceous dark green-grey semi-pelitic schist was recognized in the boreholes; although an uneven distribution was not observed in the field as the lithotypes are interlayered.

Laboratory testing of the physical and mechanical properties of the schist lithotypes showed significant variations in geotechnical properties and demonstrated a clear relationship between mineralogy and texture. On average, the medium grey quartzofeldspathic schist, with moderately developed foliation (Textural Zone III) is 1.5 times stronger than the dark grey-green semi-pelitic schist, with well-developed foliation (Textural Zone IV). Laboratory results indicate a low porosity (3.0 %) and a weak to moderately strong compressive strength (11.3 - 61.8 MPa) with low deformation modulus (0.9 - 41.1 GPa).

Lithological variations of the schist and geomorphic setting are likely to have a direct geotechnical influence on slope stability and foundation design, this is supported by kinematic analysis. Kinematic analysis demonstrates that failure is unlikely in moderately inclined (20°-35°) areas, other than localized toppling along foliation with subvertical joints acting as a releasing surface. In areas where the slope angle or cutback exceeds 50°, the number of potential planar and wedge failures increases and continue to increase as the angle approaches 70-85°. Kinematic analysis has identified that the reactivation and extensional zones of the landslide are most susceptible to failure, especially planar failure along foliation and joint sets orientated parallel to the reactivation headscarp.

A detailed ground model developed from the results in this study suggests present movement within the landslide is unlikely. However, the complexity of the landslide warrants site-specific geotechnical investigations to avoid failures during construction. Future development within/surrounding the extension and reactivation zones should be avoided until further investigations are undertaken to refine our current understanding of stability.