SHAKING DISTRIBUTION OF THE 23 AUGUST 2011 MW5.8 MINERAL, VIRGINIA, EARTHQUAKE (Invited Presentation)

The intensity data collected by the USGS "Did You Feel It?" (DYFI) Web site for the Mw5.8 Mineral, Virginia, earthquake, is unprecidented in its spatial richness, with over 140,000 responses received. These data can be used to investigate the nature of wave propagation and site response in the eastern US. The earthquake is further enormously valuable as a calibration event to compare against the intensity distributions of historical earthquakes for which instrumental data are not available. The overall distribuiton of intensities is consistent with earlier suggestions that wave propagation is more efficient along the (SW-NE) strike of the prevailing tectonic fabric than across it, although weakly felt shaking was reported as far west as Wisconsin and Minnesota. Consider the distribution of intensities, I calculate modified Mercalli intensity residuals relative to an azimuthally-isotropic decay curve using the intensity attenuation relation of Atkinson and Wald (2007), and consider their spatial distribution. The distribution of residuals provides clear evidence for significant site response, for example along Chesapeake Bay and Delaware Bay, with intensities amplified by as much as two MMI units. Significant amplification is also suggested for parts of the District of Columbia, east of Capitol Hill. At large distances mainshock shakign was felt at many locations at distances up to 1000 km, and some locations as far as ~1500 km. The radius of the felt distribution is roughly comparable to that of the 1895 Charleston, MO, earthquake, suggesting roughly comparable magnitudes for the two events. I further present a quantitative analysis of the DYFI data to show how reporting and sampling biases arise when intensities are assigned subjectively based on archival accounts of historical or modern earthquakes. I show these biases commonly inflate intensity values by a full unit or more, in particular in cities where light damage occurred. thus, intensity maps for historical earthquakes generally imply more widespread damage patterns than are revealed by modern earthquakes of comparable magnitude. However, potential damage from long-period shakign effects will not be captured fully by the intensity distributions for historical earthquakes due to the paucity of large structures in historical times.