Dynamic rockfall risk analysis

Rockfall is a dangerous hazard on steep and susceptible slope. However, it is not easy to identify potential rockfalls on cliff mountains. Traditionally, rockfall risk analysis ignores the temporal evolution of risk due to the changes of the elements at risk, or changes in the annual probability. This may lead to an underestimation of risk in time. In this paper, we present an innovative approach for a dynamic risk analysis, and we demonstrate this approach for the ShenXianjJu scenic area case study, where rockfall represent a threat for thousands of tourists each year. Moreover, we studied the role of rock mass quality in controlling the rockfall potential and rockfall volume, since this may lead to significant changes of rockfall risk in space. We describe the use of Unmanned Aerial Vehicle, Terrestrial LiDAR and detailed field surveys to identify 34 potential rockfalls on slopes where historical rockfalls have occurred. Within ignimbrite and rhyolite rock masses rockfall blocks range in size between large and very large. Andesite derived rockfalls are characterized by medium-size to large block size. Faulted and fractured ignimbrite and rhyolite rock masses within the fault damage zone exhibit small-very small rockfall block sizes. Based on the identified potential rockfalls in the study area, we quantified the dynamic risk by considering the temporal–spatial changes of tourist activity. To demonstrate the methodology, two potential rockfalls on two heavily-used tourist routes were selected. For these scenarios we quantified the annual probability of occurrence, the reach probability, the dynamic temporal-spatial probability, and the vulnerability of tourists. The dynamic temporal probability was also calculated considering different visiting periods (e.g., working days, weekends, holidays), showing significant changes in the risk level among the different visiting periods and with time. Some of the investigated scenarios were within the ALARP zone in the first years (i.e. PLL on workday in 2017–2019), and subsequently became unacceptable (i.e. PLL on workday in 2020). As a conclusion, the research demonstrates that the traditional static risk analysis would lead to a significant rockfall risk underestimation.