Remote sensing and seismic data integration for the characterization of a rock slide and an artificially triggered rock fall

On May 5th, 2013 a planar rock slide (similar to 450 m(3)) occurred in the village of La Riba (NE Spain), which forced the closure of the road C-240z for 6 months. This slide left a hanging block (similar to 130 m(3)) suspended on the slope forcing a controlled blasting, followed by rock slope stabilization works. The volume of rock displaced during the both events was deduced from LiDAR and photogrammetry data following two approaches: subtracting pre- and post-event data and reconstructing the volume by fitting planes on the structural surfaces after a structural analysis of the slope. Information about the natural rock slide was obtained from the records of two permanent broadband seismic stations located 10 km from the site. From these seismic records, the existence of a rock slide was confirmed and its time of occurrence was determined, information that would be otherwise unknown. In addition, despite the small volume displaced during the event, its location was deduced from a single seismic station analysis. The blasting process was recorded with two high-definition (HD) video cameras and by two temporary seismic stations deployed close to the site (< 100 m). Both the seismic and video recordings enabled us to reconstruct the trajectories and propagation details of the blasted rock blocks, involving material of different size sliding on the slope, suspended in the air or bouncing and impacting along the slope and on the road. Potential and seismic energy ratios (E-s/E-p) for each event were calculated from seismic data analysis in order to investigate the possibility of estimating properties of the rockfalls, primarily volume. The potential energy of both events was deduced from the volumes calculated using remote sensing methods and ranged between 189 and 201 MJ for the natural rock slide and between 48 and 54 MJ for the artificially triggered rockfall. The seismic energy was calculated following two approaches; estimating pseudo local magnitudes and by classical wave propagation theory, obtaining E-s values ranging from 2.0 x 10(-1) MJ to 4.4 x 10(-1) MJ for the natural rock slide and from 4.5 x 10(-3) to 9.1 x 10(-3) MJ for the artificial event. We estimated ranges of E-s/E-p ratios between 1.5 x 10(-7) and 5 x 10(-3) for the natural rock slide and between 8.5 x 10(-5) and 1.1 x 10(-4) for the artificial rockfall. The comparison of the volumes calculated using these ratios with the realistic volumes estimated from remote sensing data analysis, show that the seismic method is far less reliable for this task, specifically for small volumes (< 500 m(3)) at long distances (> 10 km). Partially, because only a part of the released energy is transmitted into the ground as seismic energy, and partially because the recorded seismic signal is highly dependent on the event characteristics and the geotechnical conditions of the ground materials. Nevertheless, seismic data is very well suited to detect and characterize in detail both rockfall events of different nature and size. Merging and integrating remote sensing techniques such as LiDAR or photogrammetry with seismic measurements should allow the implementation of rockfall early warning systems.