Initial Assessment of the Intensity Distribution of the 2011Mw 5.8 Mineral, Virginia, Earthquake

The intensity data collected by the U.S. Geological Survey (USGS) “Did You Feel It?” (DYFI) Website (USGS, DYFI; http://earthquake.usgs.gov/earthquakes/dyf i/events/se/ 082311a/us/index.html, last accessed Sept 2011)for theMw 5.8 Mineral, Virginia, earthquake, are unprecedented in their spatial richness and geographical extent. More than 133,000 responses were received during the first week following the earthquake. Although intensity data have traditionally been regarded as imprecise and generally suspect (e.g., Hough, 2000), there is a growing appreciation for the potential utility of spatially rich, systematically determined DYFI data to address key questions in earthquake ground-motions science (Atkinson and Wald, 2007; Hauksson et al., 2008). DYFI intensities for the Mineral, Virginia, earthquake provide a unique opportunity to explore the variability of wave propagation and site amplification along the Atlantic seaboard of the United States, a region where instrumental recordings of moderate and large earthquakes are limited. The raw DYFI data suggest some general patterns, including not only significant apparent site response along Chesapeake Bay, in parts of the District of Columbia (D.C.), and elsewhere, but also source directivity and more efficient propagation along the predominantly southwest–northeast-striking tectonic fabric. These suggested patterns can be explored further by considering intensity residuals relative to an established intensity ground-motionprediction equation (e.g., Atkinson and Wald, 2007). The Mw 5.8 Mineral, Virginia, earthquake occurred at 17:51:04.59 UTC on 23 August 2011, one of the largest earthquakes in the region in historical times and the largest earthquake to strike the central/eastern United States (CEUS) in 70 yr. Shaking was widely felt through several major metropolitan areas, including the greater Washington, D.C., region, Philadelphia, and parts of New York State. The overall felt extent of the earthquake was enormous, with perceptible shaking reported as far west as Minnesota and as far south as Florida. To the northeast it was felt as far as Fredericton, New Brunswick, Canada (http://www.earthquakescanada.nrcan.gc. ca; last accessed March 2012). The felt extent was significantly bigger than that of the 1897 Giles County,Virginia, earthquake (Bollinger and Hopper, 1971; Nuttli et al., 1979; Bollinger and Wheeler, 1983; Fig. 1a,b, present article). Nuttli et al. (1979) determine mb 5.8 for this earthquake; the National Earthquake Information Center (NEIC) lists this event as M 5.9, the magnitude type not specified (http://earthquake.usgs. gov/earthquakes/states/events/1897_05_31.php; last accessed September 2011). Instrumental recordings of the Mineral earthquake will provide valuable additional constraints for central/eastern North America ground-motion relations and potential site response. There is a unique opportunity to consider these data in combination with the intensity data collected for the earthquake. Compared to the relatively limited number of instrumental recordings of the earthquake, well-calibrated modified Mercalli intensity (MMI) values calculated from responses to the USGS Community Internet Intensity Map, also known as “Did You Feel It?,” Web site (Wald et al., 1999) provide orders of magnitude better spatial sampling. Although not an instrumental measure of ground motions, Atkinson and Wald (2007) show that DYFI intensities provide a surprisingly stable indicator of quantitative ground-motion parameters such as peak ground acceleration (PGA). The DFYI response for the Mineral, Virginia, earthquake was unprecedented, with over 133,000 responses in the first week reported from a total of over 8600 ZIP codes in the United States and towns in Canada. DYFI responses can be geocoded to improve the spatial resolution of locations. However, for most of the area spanned by this data set population density is high enough that intensity responses averaged within ZIP codes provide good spatial correlation to the location of the actual reporting sites. Geocoded locations can also only be determined for a subset of the total responses. I thus used the averages within ZIP codes in our subsequent analysis. The DYFI-intensity locations, geocoded or otherwise, are fundamentally imprecise, which results in limitations in the ability to explore site response. Nonetheless, the unprecedented spatial richness of the DYFI data set for this earthquake provides a unique opportunity to explore the factors that influence wave propagation and therefore shaking intensity for a central/eastern North America earthquake, and to provide insight into analyses that might profitably be investigated further with available instrumental data.