Fig. 1. Concentric circles showing estimated locations of energy sources (i.e. energy centers) for large prehistoric earthquakes. The estimated moment magnitude, M, for a prehistoric earthquake is located by the circle. Sites of liquefaction associated with the paleo-earthquakes are shown in Fig. 4. Prehistoric energy sources are mainly from Munson et al. (1997), Pond (1996), Obermeier (1997), and McNulty and Obermeier (1997). Epicenters of historic earthquakes are shown for the time period 1804–1992. A star represents magnitude of 5 or higher. A solid circle represents magnitude between 4.5 and 5. The plus symbol represents magnitude between about 2.3 and 4.5. Historic epicenters are from US Geological Survey/NEIC Global Hypocenter Data Base CD-ROM (Version 3.0). States of Illinois and Indiana are located in the outline of the conterminous states of the USA.

Fig. 2. Approximate limits of New Madrid seismic zone and Wabash Valley seismic zone. New Madrid seismic zone is the source area of the great New Madrid, Missouri, 1811–1812 earthquakes; the region continues to have many small earthquakes and some slightly damaging earthquakes. Wabash Valley seismic zone is a weakly defined zone of historic seismicity having infrequent small to slightly damaging earthquakes.

Fig. 3. Generalized locations of major fluvial and slackwater deposits near the confluence of the Wabash and Ohio Rivers. Dotted areas show late Wisconsinan outwash sand and gravel, as well as Holocene alluvial sand and gravel with subordinate silt and clay. Cross-hatched areas show late Wisconsinan slackwater deposits, which are mainly clay and silt but locally sand. From Gray et al. (1991).

Fig. 4. Overview showing locations of paleoliquefaction sites (darkened circles) in southern Indiana and Illinois. Maximum dike width at a site is indicated by diameter of solid circle. The survey of stream banks typically was done using a boat for continuous examination of the banks. In general, at least 10% of the lengths of the rivers had freshly eroded exposures. Only exceptionally were there no fresh exposures of mid-Holocene or older sediments within a 20 km length of a river, although at places there were no exposures for longer distances along the Wabash and Ohio Rivers. Liquefaction sites plotted on the map generally have at least several dikes, and many sites have tens of dikes. Dike width was measured at least 1 m above the base of the dike. Liquefaction sites are bounded for specific earthquakes. Shaded areas show regions of shallow bedrock with limited exposures of liquefiable sediments, where amplification of bedrock motions was probably very small, causing a much-reduced likelihood for forming liquefaction features. Ages are radiocarbon years (uncalibrated yr BP). Figure modified from Munson et al. (1997).

Limits of liquefaction for separate earthquakes in Indiana are based on data collected mainly by P.J. Munson, C.A. Munson, R.C. Garniewicz and S.F. Obermeier. Interpretations shown in Indiana are chiefly by P.J. Munson. In Illinois, data were collected mainly by S.F. Obermeier, W.E. McNulty, P.J. Munson, R.C. Garniewicz, E.R. Hajic, M.P. Tuttle and W.J. Su. Interpretations shown in Illinois are chiefly by S.F. Obermeier.

Fig. 5. Diagrammatic vertical section showing the general characteristics of buried sand- and gravel-filled dikes with their vented sediments along the Wabash River. Source beds are Holocene point-bar deposits or late Wisconsinan braid-bar deposits, which are overlain by fine-grained overbank sediment. The sediment in source beds directly beneath dikes shows evidence of flowage into dikes. In many dikes the gravel size and content decrease upward. The extreme upper part of a dike often is widened and shows evidence of ground shattering. The column on the right side of the figure contains pedological abbreviations.

Fig. 6. Schematic vertical section showing idealized dikes cutting through a host of silt and clay strata, and also showing the overlying vented deposits. The relations shown are very commonly observed in the meizoseismal zone of the great 1811–1812 New Madrid earthquakes. All dikes are tabular in plan view, whether or not they vented to the surface. Width of a dike can much exceed the value of 15 cm shown in the figure, at sites of severe lateral spreading. (A) Stratigraphy of dike with sediment vented to the surface; (B) dikes that pinch together as they ascend; and (C) dike characteristics in fractured zone of weathering, in highly plastic clays. Schematic depictions of the influence of depth of burial and weathering on prehistoric dikes in the study area are shown in Obermeier et al. (1993), Fig. 5.

Fig. 7. Moment magnitude versus maximum distance to surface evidence of liquefaction effects. Solid line shows bound for maximum epicentral distance of liquefaction features, for worldwide, shallow-focus earthquakes (<50 km). Darkened band shows maximum distance (plan view) from energy source (i.e. energy center) for earthquakes in the study region, which was developed from the study of historic earthquakes in Indiana and Illinois and in the nearby New Madrid seismic zone (Fig. 2).

Fig. 8. Peak accelerations as a function of hypocentral distance, for M 7.5, M 7.8 and M 8.0 earthquakes in the study area. Peak accelerations determined from back-calculation (open dots) and seismological modeling (solid line). Back-calculations were performed using the energy-based method of Pond (1996) at sites of liquefaction for the earthquake of 6100 yr BP. Curves that are based on modeling are from relations developed by Boore and Joyner (1991) and by Atkinson and Boore (1995). The best fit between back-calculated and modeled accelerations is for a M 7.8 earthquake. From Pond (1996), Fig. 6.22.

Fig. 9. Dike widths versus distance from the energy source (i.e. energy center) for the earthquake of 6100 yr BP. Widths shown are the sum of dike widths observed in a bank exposure, in which the exposure has from one to many dikes. Data are from all sites in Indiana, and from sites in Illinois along the Wabash River. From Pond (1996), Fig. 6.2; modified from Munson and Munson (1996).