Escarpment stability: processes and modes of failure

Abstract

The aim of this research is to understand escarpment stability; specifically, the controlling factors, processes and modes of failure. The escarpment investigated is referred to as the Kokoamu Bluff and is located two kilometres southeast of Duntroon in the Waitaki Valley, North Otago, New Zealand. The features observed on the escarpment and their spatial relationships are used to build up a picture of formative processes and modes of failure. Some of these features indicate the failure mechanisms that have been operating (e.g. slump blocks) while others can be interpreted as affecting the susceptibility of the escarpment to failure (e.g. erosional notches and tension cracks). Three modes of contemporary failure and transport have been developed based on these observed features, thereby providing an explanation of landform features at this site. The modes of failure can be shown to be interdependent and sequential. In other words, the occurrence of a given mode of failure provides the necessary pre-conditions for a different mode of failure to occur. A hypothetical geomorphic evolution model of the Waitaki River environment is also proposed to show how the cycle of escarpment development and retreat might have taken place at this location. Stability and sensitivity analysis is used in order to determine the stability of the escarpment and to establish which factors have the most influence on stability. Stability analysis is also used to test the validity of the modes of failure established by field observation. The results indicate that each of the nine models tested by stability analysis is capable of producing failure (when the margin of error arising from natural variability in strength parameters is taken into account). Factors of safety range from 0.129 to 2.228. Both field observation and stability analysis demonstrate that: Rotational sliding dominates processes responsible for escarpment development by displacing and transporting caprock blocks. Rotational sliding can occur solely in the basal slope unit underlying the caprock, removing basal support from the caprock. Stability analysis alone suggests that if tension cracks in the limestone caprock are filled with water, block toppling can occur without the aid of rotational sliding. However observation of recent block toppling suggests that it tends to occur in areas of the escarpment where rotational slumping has previously taken place. Vertical aerial photographs, digital photographs, pollen analysis of buried soils and [historic] climate data are used to determine whether scarp development is relict or currently active and to assess the frequency and magnitude of mass movement processes both spatially and temporally. Vertical aerial photographic and climatic data analysis show that an extreme rainfall event in 1938 triggered two large failures resulting in distinctive flow lobes appearing in the basal slope. While no specific events of block toppling could be identified in the historical records, it is clear from field evidence that block failure has occurred in the recent past (one such event occurred during the course of this study). Measurements of the visible surface area (i.e. the exposed face) of 78 detached blocks along the basal slope indicate block sizes ranging from 0.05 m2 to 58 m2, with the majority being smaller than 28 m2. Observations indicate that the escarpment is not of uniform age and its development appears episodic, relating to different modes of failure. It is argued that landsliding and block toppling are active formative processes and of sufficient magnitude and frequency to represent a significant hazard to any development on or in the vicinity of the escarpment. All of the results indicate that the processes leading to scarp development at this location are mass movement driven and episodic and not dependent on slow gradual weathering processes.