CHARACTERIZING PRECARIOUSLY BALANCED ROCKS IN THE EASTERN U.S. FOR ESTIMATING GROUND MOTIONS

Fragile geologic features like precariously balanced rocks (PBRs) can be used to constrain strong ground shaking at a site over the past few thousands to tens of thousands of years. Determining the ground motions that would topple a balanced rock places an upper limit on the strength of ground shaking since the rock became balanced. Such rocks could be especially important in the eastern U.S.. where a lack of fault scarps limits paleoseismic investigations and where relatively low seismic attenuation suggests earthquakes could topple relatively distant rocks. In the northeastern U.S., dozens of glacial erratics were placed in balanced positions as the continental ice sheet melted 19,000 to 14,000 years ago. Balanced rocks in the southeastern U.S. were formed by erosion, and their ages are therefore more problematic. Here I describe a pilot study of three PBRs formed by erosion in the Blue Ridge of Virginia. These rocks are 110 to 150 km from the Central Virginia seismic zone, and about 90 km from the Giles County seismic zone, thus potentially providing upper limits on earthquake magnitudes within those zones. Photogrammetry methods were used to create detailed 3-D models of the PBRs to compute the locations of the centers of mass relative to their bases, with the resulting angle being crucial for estimating fragility. Topography also is known to amplify seismic waves, so modeled relations between hill geometry and amplification were used to estimate the topographic effect. Once the level of ground shaking needed to topple the most fragile rock was determined, a ground-motion model (GMM) was used to estimate the sizes of earthquakes in the surrounding area that would produce that ground motion. Preliminary results indicate maximum earthquake magnitudes of about M6.5 to M7 in the Central Virginia seismic one over the past few thousand years, which is consistent with the M~6.2 to M~6.5 earthquake(s) estimated from paleoliquefaction studies being near the maximum that have occurred in that time. It also is unlikely that an earthquake larger than about M6.5 has occurred in the Giles County seismic zone during that time. Estimating the ages of the PBRs in Virginia will require detailed dating, but the methods can be applied to glacial erratics in the northeastern U.S., where ages are better known.