Forensic analysis of rockfall scars

Abstract

We characterise and analyse the detachment (scar) surfaces of rockfalls to understand the mechanisms that underpin their failure. Rockfall scars are variously weathered and comprised of both discontinuity release surfaces and surfaces indicative of fracturing through zones of previously intact rock, known as rock bridges. The presence of rock bridges and pre-existing discontinuities is challenging to quantify due to the difficulty in determining discontinuity persistence below the surface of a rock slope. Rock bridges form an important control in holding blocks onto rockslopes, with their frequency, extent and location commonly modelled from the surface exposure of daylighting discontinuities. We explore an alternative approach to assessing their role, by characterising failure scars. We analysed a database of multiple rockfall scar surfaces detailing the areal extent, shape, and location of broken rock bridges and weathered surfaces. Terrestrial laser scanning and gigapixel imagery were combined to record the detailed texture and surface morphology. From this, scar surfaces were mapped via automated classification based on RGB pixel values. Our analysis of the resulting data from scars on the North Yorkshire coast (UK) indicates a wide variation in both weathering and rock bridge properties, controlled by lithology and associated rock mass structure. Importantly, the proportion of rock bridges in a rockfall failure surface does not increase with failure size. Rather larger failures display fracturing through multiple rock bridges, and in contrast smaller failures fracture occurs only through a single critical rock bridge. This holds implications for how failure mechanism changes with rockfall size and shape. Additionally, the location of rock bridges with respect to the geometry of an incipient rockfall is shown to determine failure mode. Weathering can occur both along discontinuity surfaces and previously broken rock bridges, indicating the sequential stages of progressively detaching rockfall. Our findings have wider implications for hazard assessment where rock slope stability is dependent on the nature of rock bridges, how this is accounted for in slope stability modelling, and the implications of rock bridges on long-term rock slope evolution.