The importance of monitoring interval for rockfall magnitude-frequency estimation.

Abstract

The frequency distribution of rockfall sizes ('magnitude-frequency') is important for erosion and hazard modeling. This typically follows a power law, with few larger rockfalls compared to more numerous small events. Advances in LiDAR hardware and algorithms have improved our ability to detect small rockfalls, which in sum contribute significantly to overall erosion. However, improvements in spatial resolution have outstripped improvements in the temporal resolution of monitoring. If the interval between surveys exceeds the return interval of rockfalls, neighbouring rockfalls within a single interval are incorrectly recorded as one. We examined the timescales of rockfall occurrence to identify suitable monitoring intervals to discretize rockfalls. Monitoring interval has a considerable impact on the frequency distribution of measured rockfall volumes. An order of magnitude increase in rockfall numbers and a threefold decrease in mean volume is observed over hourly intervals, compared to 30 d. Below ~12 h, these changes increase nonlinearly with more frequent monitoring. Similarly, average rockfall size measured over timescales < 4 h falls is comparable to the scale of individual discontinuities, indicating that fragmented detachments may drive much of the increase in small events. Such analysis is required to constrain the timescales of rockfall evolution and to attribute specific drivers.