Historical Seismic Activity in the Central United States

The expected recurrence of the 1811-1812 New Madrid sequence about every 500 years (Tuttle et al., 2002) dominates the U.S. Geological Survey (USGS) estimates of seismic hazard over much of the central United States (CUS) (Harmsen et al., 1999; Frankel et al., 2000, 2002). While earthquakes in the New Madrid seismic zone (NMSZ) apparently dominated CUS seismicity in the first two decades of the 19th century, more moment magnitude M 5.0 and larger earthquakes have occurred in the Illinois basin in southern Illinois and Indiana in the past 150 years than in the NMSZ (Bakun and Hopper, 2004a). Paleoseismic studies ( e.g., Munson et al. , 1997) have shown, moreover, that large Holocene earthquakes have occurred in the Illinois basin where 5.0≤ M ≤6.0 historical earthquakes have occurred. M ≥ 5.0 historical earthquakes have also occurred in Kansas, Kentucky, Nebraska, and Oklahoma. Thus, a consistent evaluation of historical seismic activity occurring throughout the CUS is important in estimating the seismic hazard. Bakun et al. (2003) developed a model specifically designed to estimate the location and M for historical earthquakes in eastern North America (ENA) and analyzed the 5 January 1843 and 31 October 1895 M 6 CUS earthquakes. Bakun and Hopper (2004b) used this model to estimate the location and M for the three largest events in the 1811-1812 NMSZ earthquake sequence. Bakun et al. 's (2003) model can be used to analyze historical CUS earthquakes with only a few intensity assignments and includes empirical site corrections to account for anomalous site effects, such as the enhanced shaking effects in 1811-1812 in communities built on weak ground along waterways (Hough et al., 2000). Bakun and Hopper (2004a) applied Bakun et al. 's (2003) model to M < 6 historical CUS earthquakes and compiled a catalog of historical seismicity wherein …