TEMPORAL ANALYSIS OF 4D ROCKFALL ACTIVITY AND PATTERNS OF EROSION FROM AUTOMATED HOURLY-RESOLUTION LASER SCANNING MONITORING

Rock-slope failures are hazardous erosional processes. Conventionally, our understanding of the timing and nature of rockfall activity has been hindered by the frequency of data acquisition, the limited duration of monitoring datasets and/or limitations to computational capacity that hinder efficient processing of large laser-scanning datasets. Using a unique, extensive and high-resolution 3D dataset based on 1 h change detections of the cliff-face topography undertaken on a vertical cliff (Whitby, North Yorkshire, UK), we examined the timing of rockfall activity during three years of monitoring (2017 – 2019). We developed a novel computational routine to automatically process over 20,000 scans, yielding 14,741 fully automatic pairwise change detection measurements. We detected and characterised over 8,300 rockfalls, equating to a total volume loss of ~178 ± 71 m³ and ~0.03 m of overall cliff retreat. Our results indicate that the long-term erosional signature is controlled by infrequent large rockall events (≥1.0 m³) that drive erosive jumps or abrupt changes in the volume loss that punctuate periods of steady baseline erosional activity. Significantly, we identified periods of post-failure quiescence that scale in duration with the preceding volumetric rockfall magnitude, indicative of a correlation between failure magnitude and relaxation time associated with the preceding failure event. As such, our research holds implication for surface hazard, risk mitigation and, in particular, the processes inducing progressive rock fracture and resultant deformation that precede rock-slope failure.