Evaluating postglacial and modern rock-fall rates in Yosemite Valley, California

Abstract:

Rock falls in Yosemite Valley, California, are a driver of geomorphic change and pose significant hazard and risk to its many visitors. We investigated the pace of rock falls and cliff retreat over a range of time scales by quantifying rock loss from two adjacent granitic cliffs in Yosemite Valley, the southeast faces of El Capitan and Middle Brother. Terrestrial laser scanning and historical Structure-from-Motion photogrammetry provide a ~40-year (1976-2017) record of rock loss from the cliffs, whereas talus volumes that have accumulated on the flat valley floor since deglaciation offer a 15,000-year record. These data allow us to compare decadal-scale volumes and rates against longer-term postglacial values. Normalizing rock-fall volumes by the contributing cliff area provides spatially and temporally averaged cliff retreat rates, a simple metric for comparison. For El Capitan, the ~40-year rock-fall volume is 16,653 ± 1,455 m³, which yields a cliff retreat rate of 0.8 mm/yr. The debulked (porosity corrected) postglacial talus volume beneath El Capitan is 7.81 ± 2.35 million m³, which also yields a retreat rate of 0.8 mm/yr. Identical modern and postglacial rates indicate that the rock-fall activity observed at El Capitan over the past ~40 years is broadly representative of long-term activity. In contrast, at Middle Brother, the 40-year rock-fall volume is 60,133 ± 7,096 m³, yielding a retreat rate of 1.6 mm/yr, double the modern rate observed at El Capitan. However, the debulked talus volume of 4.51 ± 1.13 million m³ beneath Middle Brother yields a retreat rate of 0.3 mm/yr, far slower than the modern rate and slower than that observed at El Capitan. This suggests that Middle Brother has been experiencing a pulse of increased rock-fall activity over the past 40 years that substantially outpaces postglacial rates. The data also appear sensitive to the measurement interval. The modern retreat at El Capitan of 0.8 mm/yr includes a cluster of large rock falls that occurred in 2017. If the analysis is instead limited to 1976-2016, a reduction of just one year, the cliff retreat rate declines to 0.3 mm/yr, similar to the postglacial rate at Middle Brother. As previously shown for coastal cliffs, our results highlight the importance of event clustering on measured rock-fall and cliff retreat rates, and further support caution when extrapolating short-term measurements to longer-term hazard assessments. Our results also show that adjacent cliffs with similar geology can have very different rock-fall activity over both short and long timescales, offering new lines of research.