ROCK PILLARS AS GROUND MOTION CONSTRAINTS IN OREGON AND CALIFORNIA, USA

Fragile geologic features, such as precariously balanced rocks (PBRs), have been used to constrain the intensity of past earthquake shaking. Here, we investigate the utility of another category of fragile geologic features, rock pillars, in Oregon and California. Rock pillars differ from PBRs because they are materially connected to bedrock at their bases, implying that rock pillars respond to shaking as cantilever beams rather than unrestrained rigid bodies. In this study, we first analyze the seismic response of rock pillars at the Trona Pinnacles, which are located in southern California, less than 30 km from the epicenter of the 2019 M7.1 Ridgecrest mainshock. We deployed broadband seismometers and recorded aftershocks (<M2.5) simultaneously at rock pillars and far-field bedrock. We then conducted finite-element simulations of the pillars’ response using photogrammetric shape models and compared the results. Our simulations capture the first vibrational mode and the magnitude of amplification. Both the numerical simulations and semi-analytic methods indicate that peak stresses occur near the base of the pillars during shaking. We document relative weakness at the base of rock pillars in the field, using a rebound hammer, and we postulate that earthquake damage accumulates there on a geologic time scale. Next, we present an updated and expanded survey of rock pillars and their quasi-static ground motion constraints from Oregon. Empirical ground motion constraints are particularly valuable in the Pacific Northwest, where a major subduction zone earthquake has not occurred in written history. Most of these rock pillars are sea stacks, and their fragility ages may be estimated as a function of their distance from the coastal cliffs. We estimate tensile strength in situ, using a rebound hammer. We provide context for the quasi-static results by comparing them with predictions from kinematic earthquake simulations and ground motion prediction equations. Among the sea stacks old enough to have survived multiple megathrust earthquakes, we find that nearly all the predictions and quasi-static constraints overlap within uncertainty. We conclude that rock pillars show promise for investigating both past earthquake shaking and its spatial variability.