ei .-~ ENGINEERING GEOLOGY ELSEVIER Engineering Geology 46 (1997) 235-258 Tertiary and Quaternary tectonic faulting in southernmost Illinois W. John Nelson *, F. Brett Denny, Joseph A. Devera, Leon R. Follmer, John M. Masters Illinois State Geological Survey, 615 E. Peabody, Champaign, IL 61820, USA Received 20 April 1996; accepted 30 November 1996 Abstract Tertiary and/or Quaternary tectonic faulting is documented in three areas of southernmost Illinois: the Fluorspar Area Fault Complex (FAFC) in Pope and Massac Counties, the Ste. Genevieve Fault Zone (SGFZ) in Alexander and Union Counties, and the Commerce Fault Zone (CFZ) in Alexander County. In the FAFC, faults that strike NE and NNE displace Mounds Gravel (late Miocene to early Pleistocene) and, locally, the Metropolis terrace gravel (Pleistocene; pre-Woodfordian). No Woodfordian or younger deposits are deformed. Faults typically outline narrow, linear grabens that formed under tension with a component of strike slip. North-south to NW-trending vertical faults near the southeast end of the SGFZ displace Eocene sediments. Again, faults outline narrow grabens and show indications of strike slip. Deformed Quaternary sediments have not been observed. The CFZ, which trends northeast, displaces Mounds Gravel in Illinois and units as young as Peoria Silt (Woodfordian) in Missouri. Quaternary movement has been interpreted as right-lateral strike-slip. The CFZ coincides with a subtle gravity and magnetic lineament and seems to reflect a major feature in the basement. Surface expression in Illinois is subtle, but marie and ultramafic intrusions, hydrotliermal alteration and small faults align with the Commerce geophysical lineament. Earthquake foci in Missouri and Illinois lie on or close to the CFZ; some focal mechanisms fit the fault trend. Among these structures, only the CFZ exhibits slip that conforms to the current stress field (principal compressive stress axis E - W to ENE-WSW). Possibly, the stress field changed during Neogene time. Alternatively, high fluid pressures or local stress concentrations may have induced slip on less favorably oriented fractures. Tighter constraints are needed on timing, magnitude, and direction of Neogene displacement. © 1997 Elsevier Science B.V. Keywords: Faults; Grabens; Neotectonics; Neogene; Tertiary; Quaternary; Illinois; Mississippi embayment I. Introduction Southern Illinois, one o f the m o s t intensely faulted areas in the Midcontinent, lies immediately n o r t h o f the N e w M a d r i d seismic zone (Fig. 1 and * Corresponding author. Tel.: + 1 217 244 2428; fax: + 1 217 244 7004; e-mail: jnelson@geoserv.isgs.uiuc.edu 0013-7952/97/$17.00 © 1997 Elsevier Science B.V. All rights reserved. PH S0013-7952 (97)00005-7 Fig. 2). Small to m o d e r a t e earthquakes are a c o m m o n occurrence in this area. Tertiary and possible Q u a t e r n a r y faulting have long been k n o w n in K e n t u c k y ( R h o a d e s and Mistier, 1941) and Missouri (Grohskopf, 1955) directly adjacent to Illinois. Ross (1963) discussed evidence for tectonism involving units as y o u n g as the M o u n d s Gravel (Pliocene to early Pleistocene) in southern- 236 ~ John Nelson et al. /' Engineering Geology 46 (1997) 235-258 .+. -y. o.. ÷ ..~ ~,o~ ! earthquake epicenter ._/ ,-r" -p ILLINOIS I"-" -,?. ~ - 2-..= OZARK ~" DOME - - " ~' " " _." .-- ~ • ...,~ - " "=- ' ,,...• -- ..~.. ., ~ • " _ ._ Rough Creeff"- ~- ---- --'-----~-=- • " O + c" . . . . ~ .- -.~ - _ - - ' / _ = 4- ":" • '.-~'.." ,~ N e w M a d r i d imic Zone '~_+ ? _/ Fig. 1. M a p showing m a j o r structural features and New M a d r i d Seismic Zone. W. John Nelson et al. / Engineering Geology 46 (1997) 235-258 "v- m *~,n. i 237 on ,0thro,n ~,ao IqAI~I.'ION • norrr~ ~Jlt, I~II ancl l)ar on downthrown side • earmcltake epicemer mentloned in text • sludy ~ ' m mentloned in text 0 20mi 0 / 25km ¢. Fig. 2. Map of southern Illinois showing selected faults and locations of sites mentioned in this report. most Illinois. In spite of these findings, an opinion persists that Illinois has been tectonically inactive since late Mesozoic time. A comprehensive study (Kolata et al., 1981) found no evidence for postCretaceous tectonic faulting in southern Illinois. New geological mapping has re-opened the casebook. We found tectonic faults that clearly offset Tertiary and, locally Quaternary deposits in several areas of southernmost Illinois. Also, recent collaborative investigations by the US Geological Survey and Missouri Department of Natural Resources, Division of Geology and Land Survey, have discovered Quaternary tectonic deformation in the Benton Hills of southeast Missouri, immediately west of southern Illinois (Harrison and Schultz, 1994b; Harrison et al., 1995; Hoffman et al., 1996; Palmer et al., 1996). This paper presents our initial findings on faults that displace Tertiary and Quaternary sediments in southernmost Illinois. 1.1. Location and structural setting The study area lies on the east flank of the Ozark Dome and the southern margin of the Illinois Basin (Fig. 1). Outcropping sedimentary rocks range from Ordovician through Pennsylvanian and dip regionally northeast and north from the dome into the basin. Weakly lithifled Cretaceous and early Tertiary sediments overlap Paleozoic bedrock at the northern edge of the Mississippi Embayment. We observed post-Cretaceous faulting in three areas of Illinois. The Fluorspar Area Fault Complex in southeastern Illinois and the adjacent part of Kentucky contains northeast-trending faults that offset sediments as young as Quaternary. The southeast end of the Ste. Genevieve Fault Zone in southwestern Illinois locally displaces Cretaceous and Eocene sediments. 238 W. John Nelson et al. / Engineering Geology 46 (1997) 235-258 The northeast-trending Commerce Fault Zone displaces Quaternary loess in Missouri and crosses into Illinois. All three represent complex crustal fracture zones that have undergone recurrent episodes of movement since early Paleozoic time. Tertiary and Quaternary tectonic faults are elusive. Outcrops are sparse and typically obscured by vegetation and slumped material. Distinguishing tectonic structures from products of landsliding, solution collapse and other nontectonic processes can be problematic. Large gaps exist in the stratigraphic record. Most boreholes in the Embayment area lack reliable records. New drilling, trenching, and geophysical studies are needed to verify features observed in outcrops. 2. Huorspar Area Fault Complex The Fluorspar Area Fault Complex (FAFC) is a family of fault systems in the fluorspar-mining district of extreme southeastern Illinois and the adjacent part of western Kentucky. The FAFC contains faults of diverse attitude, structure, and age. The principal fault zones strike northeast, outlining a series of horsts and grabens (Fig. 2). The precursor to the FAFC was the Reelfoot Rift, a failed intracratonic rift that formed during the breakup of North America in latest Precambrian to early Cambrian time (Ervin and McGinnis, 1975; Hildenbrand, 1985; Thomas, 1991). Subsequently, faults in the rift underwent recurrent movement under different stress fields (Kolata and Nelson, 1991; Nelson, 1991; Hildenbrand and Hendricks, 1995). Five fault zones in the FAFC in Illinois displace Cretaceous and younger sediments. From northwest to southeast they are the Lusk Creek, Raum, Hobbs Creek, and Barnes Creek Fault Zones and the Rock Creek Graben (Fig. 2). In addition, published maps indicate that Tertiary and early Quaternary sediments are displaced in the Lockhart Bluff Graben in Kentucky. 2.1. Lusk Creek Fault Zone The Lusk Creek Fault Zone forms the northwest border of the FAFC (Fig. 2). The Lusk Creek strikes northeast and its larger faults dip southeast. The zone varies from a few meters to about 500 m wide and comprises subparallel, steeply dipping (60 ° or greater) normal and reverse faults. Net throw on Paleozoic rocks across the fault zone is 100-250 m down to the southeast. The Lusk Creek Fault Zone is a major element in regional tectonics. It is directly in line with the northwest margin of the deep Cambrian graben (Reelfoot Graben) within the Reelfoot Rift (Hildenbrand and Hendricks, 1995), and also in line (although strike trends differ slightly) with the zone of intense seismic activity northeast of New Madrid (Fig. 1). At its northeast end the Lusk Creek joins the Rough Creek-Shawneetown Fault System, which forms the northern boundary of the Rough Creek Graben (Fig. 2). The Lusk Creek and Rough Creek-Shawneetown have similar structural histories. Multiple episodes of displacement are evident from stratigraphic and structural relationships: ( 1) normal faulting during development of the Reelfoot Rift, as shown by thicker Cambrian section southeast of the fault zone on a seismic reflection profile (Bertagne and Leising, 1991; their Fig. 15-4); (2) renewed normal faulting during mid-Chesterian (Late Mississippian) and Morrowan (Early Pennsylvanian) time, as shown by abrupt thickening of units of these ages within the graben (Weibel et al., 1993; Nelson, 1996a); (3) reverse faulting that raised the southeast side of the fault zone after Pennsylvanian time (Weibel et al., 1993); (4) later normal faulting lowered the southeast block again, below its original position (Weibel et al., 1993). Any of these four episodes of slippage may have included a component of strike slip, although dip-slip movements appear to have dominated. Pliocene to early Pleistocene(?) displacement in the Lusk Creek Fault Zone is evident near New Columbia in northern Massac County (Fig. 3). A narrow northeast-trending graben contains McNairy Formation (Upper Cretaceous) and Mounds Gravel (Pliocene to early Pleistocene?) downdropped against Mississippian sandstone. This graben lies about 1.5 km northwest of the main faults of the zone, which are concealed by Quaternary alluvium. W. John Nelson et al. I Engineering Geology 46 (1997) 235-258 239 ~46" ..... ...... ,..... . , . . . ...... ....., "...... ,..... ...... ...,.."i.:i."i.::."i.:::'i.:?:,!.:::..i.:!...i.::.. :: .- : - . . : : - . . . : : . : . / ' / . . i . ~ ~ "" : "7 : :: .... !.::.. i,i,~(-;:~ .:-: -": ..4 ~~ :.~t'.. .. • ::::::::::::::::::::::: • ,.....:.....:.... :: 7. :---.:-- .:-::::"::::'::...-'-:..'I: ':..' :: ':..": ':.':: ':.. :: ':.-' :: ' :V:::~":":::"~i:::~: ~ ~..f...~o ...- /7",.--., 0 0 ..." . . ~. ~:z.:..'5.~:-.~:t.?....:.::.:-i.::.:-:../ .-" . <" I .f~ .. Y" ...~ F--2Z-'I l,,lolocone alluvium i "~ J ~ ~i'iocene ~ .'4 -:.7 ...-" .-' • ....~.~" ~ r~,.,.,~~,,, ~ ,,o,~ ~,,., vails ~ ~ Mct,liry r-ormai~ ~----~ W ~ ~ 1 : ~ ~ ..% O.Smi 0.5 I km ,,,lO ® smkn and dip of b e d l g hoazontaJ bedding Fig. 3. Map of the New Columbia structure. In cross-section the New Columbia structure forms a narrow "V" (Fig. 4). Mississippian strata are horizontal and at nearly the same elevation on either side of the graben. Bedrock dips steeply (up to 55 °) inward along the marginal faults. Within the graben, unlithified sand and clay of the McNairy is steeply dipping and contorted, complexly faulted, and riddled with clay-filled fractures. Steeply dipping Mounds Gravel, cemented by iron oxide, crops out along the axis of the structure. The Mounds is downthrown 30-45 m relative to its position outside of the graben. Outside the graben, Mounds rests directly on Mississippian rocks, with no McNairy between. This observation implies two episodes of faulting: the first post-McNairy and pre-Mounds, the second post-Mounds. Undeformed, horizontally bedded, unlithified sand (New Columbia Sand) overlies faulted and steeply dipping older units in the north part of the New Columbia structure (Fig. 3). The sand is fine to coarse-grained and crossbedded to plane-laminated; clay balls and rip-up clasts are present. A truncated soil profile near the top is overlain by the Peoria Silt (late Wisconsinan). We interpret the New Columbia Sand as fluvial sediment that was deposited prior to incision of the Cache Valley. The time of initial incision of the Cache Valley is 240 14~ John Nelson et al. / Engineering Geology 46 (1997) 235 258 Northwest Southeast New Columbia structure Lusk Creek ' . - ° . • . . . • Mdr older Mis=dssippian rocks o oar Mlaissippian rocks A I 1 ---~ Holocenealluvium ~-~q ~ C~Jr~is Sand ~ W~t B ~ n Ssl~m~e ~'1'~ Downeys Btuff and Renm.dtLimutono 0 I I 0 .5 .5 I I 1 I 1 mt I 1.5 km Fig. 4. Cross-section of the New Columbia structure. uncertain, but it is certainly pre-Woodfordian and probably early Pleistocene (Masters and Reinertsen, 1987; Esling et al., 1989; Follmer et al., in press). Thus, the last tectonic movement of the New Columbia structure probably took place in Pliocene to early Pleistocene time. 2.2. Raum Fault Zone The Raum Fault Zone is 2-3 km southeast of and parallel to the Lusk Creek Fault Zone (Fig. 2). The northeast part of the Raum dips northwest and may intersect the Lusk Creek deep within Paleozoic rocks. To the southwest in Massac County the Raum widens to nearly 2 km and its net throws changes to the southeast. At least two episodes of post-Pennsylvanian, pre-Cretaceous movement are known: reverse faulting that raised the northwest block, followed by normal faulting that lowered the northwest block (Weibel et al., 1993). Strike-slip motion also may have occurred along this fault zone. Post-Cretaceous movement took place at a site called Reineking Hill in Massac County (Fig. 5 and Fig. 6). There, a narrow graben containing McNairy Formation (Upper Cretaceous) is downdropped at least 45 m into Mississippian bedrock. The Reineking Hill structure eroded to a narrow, linear valley that strikes N20°E. Floored with Quaternary alluvium, the valley is 2 km long and 100-200 m wide. No Tertiary sediments are present, and Quaternary sediments are not visibly deformed. The time of faulting therefore cannot be constrained better than post-Cretaceous, preQuaternary. The McNairy is no longer exposed at Reineking Hill, but formerly was visible in a railroad cut (now removed) and a clay pit (now flooded). W.A. White (Illinois State Geological Survey, emeritus), who visited the clay pit while it was active, identified the clay as McNairy, and described the east and south walls of the pit as near-vertical fault surfaces (field notes by White (1961) in open files of the Illinois State Geological Survey, and oral communication to Nelson, 1991 ). 241 W. John Nelson et aL / Engineering Geology 46 (1997) 235-258 88*45" a : ~ . . . . ~ -'... ~..~. ". • .'.' ... :. [(:~ Hobm~ ~ttuUn .'.'-.'.. HUl • ..-.-... ~ - - - ' ~ U~er pope Grouo ~ 7 " ~ Tar S¢~ngsto Hlm:llngbu~ ~ W~t ~ F~ Do~mo~EU ~ RemuUUmbrae Sandmne ~ - - ~ - .......... ~Jlt, cllmlx~ m Inl~r~l anclclaaN ~ c~-.~,dUd • Qal hoCmmm ~Kkmng 0 [ o 0.5 I nJ a i ; ' 1.$km Fig. 5. Map of part of the Raum Fault Zone, showing the Reineking Hill structure (from Nelson, 1996b). Lamar (1927) and Lamar and Sutton (1930) previously described deformed McNairy in the railroad cut, but attributed deformation to slump or creep. Slump or creep, however, could not explain how the McNairy, which is not present on nearby hills, was displaced at least 10 m below the valley floor. The Reineking Hill structure closely resembles the New Columbia structure. There is little displacement across the structure, and Mississippian rocks are nearly horizontal on either side. Along the bordering faults the bedrock dips steeply inward. Tilted bedding and primary joints in bedrock strike parallel with the Reineking Hill structure. The Reineking Hill graben and several nearby faults strike north-northeast, oblique to the overall northeast trend of the Raum Fault Zone (Fig. 5). This pattern suggests an extensional duplex formed by left-lateral wrench faulting. However, the fault pattern alternatively might reflect step-over between parallel, overlapping pre-Cretaceous faults. Drillers' logs of three water wells indicate the Reineking Hill graben may continue 9-10 kin to the southwest. The logs show the top of Paleozoic bedrock (beneath probable McNairy Formation) downdropped 30-60 m relative to nearby wells. We plan to conduct additional drilling to try to determine whether any strata younger than the McNairy are deformed in the Reineking Hill structure. 242 W. John Nelson et al. / Engineering Geology 46 (1997) 235-258 Northwest Southeast Reineldng Hill Cache Valley Km aal • . - ~ , I Km ~ , .. -...~ : - . . , -~.-:.~.~- ,---- , I ~/"f--;, • . .,/rl o,o.,. -.;- : . , _, Km / . - • ~'\. I Ilk * " i McNaio/ForrnalJon 0 ~Tar Springsto Har~a"~rg ~ Golcor't~iFormmion ~ We~'t B=clen ~ i e Omene~Bluffand R ~ . • ' ~ , ' . * . • "% * , • * "- ', o .i • , M / 0 ~ -. ~ older Misalssippian rocks i m , • -" ". / I I ~.,~-- i" -1/.. 4,r. Mi.~si~ippian rocks • .5 ..... I I .5 1 mi I I I 1 1.5 k m ~ Fig. 6. Cross-section of the Reineking Hill structure. 2.3. Hobbs Creek Fault Zone The Hobbs Creek Fault Zone, 2-3 km southeast of the Raum Fault Zone, consists mostly of highangle normal faults having throw down to the northwest. The Lusk Creek, Raum, and Hobbs Creek zones together outline a graben (Dixon Springs Graben) (Fig. 2). The youngest known displacement on the Hobbs Creek took place near Massac Creek in Massac County, where a narrow graben underlies a linear valley that trends N20°E (Fig. 7). Ross (1963, 1964) reported MeNairy Formation and Mounds Gravel with steep dips along this valley. Kolata et al. ( 1981, p. 19) restudied the site, noting that water well records indicated a bedrock low along the valley. They performed seismic-refraction and earth-resistivity surveys, but the results were inconclusive. Kolata et al. summarized, "The available data do not indicate whether this structure was caused by tectonic faulting, landsliding, solution collapse, or a combination of these processes". New mapping (Nelson, 1996b) shows that the tilted Mounds Gravel described by Ross is part of a northwest-dipping block more than 600 m long and 100 m wide and downdropped at least 20 m. The tilted block is directly in line with faults that displace Mississippian bedrock to the north. The McNairy Formation within the structure has strongly contorted bedding and is riddled with small faults and elastic dikes. The size and linearity of the downdropped block rule out landsliding or solution collapse as possible causes. A water well directly in line with the downthrown Mounds logged abnormal strata. The well was drilled to +80 feet (+24 m) bottom-hole elevation without reaching bedrock, whereas nearby wells encountered bedrock at elevations of + 325 to +360 feet (+99 to + l 1 9 m ) . The Weaver borehole (Figs. 7 and 8) is a continuously-cored test hole 301 feet (91.8 m) deep adjacent to the anomalous water well. The core showed 15 feet (4.5 m) of loess and alluvium, overlying multi-colored silt to coarse sand with thin layers W. John Nelson et al. / Engineering Geology 46 (1997) 235-258 • . :--" :,,: "..:J''....:-t1,:,.:-" :.,:.-" •:_...:..-•~.''." ~._.;.._..;~ ".%._.-+~_- r •Holocenealluvium ". ~ PtdmC~ terraces =2 I'ci:.".'.:i~".'.:i i..: :: I~ " ' ~ ""~.'~,¢."~" • • ~ 243 ".'iS.""'-.. • "~.'~" , ":..: Mct~ry ~ Tar Springsto Ha~nsburg ~ Vtl(16tBIItl~lltrl~ o ~ ..L- . . . . . . . . f ~ t , ~ whoro tnl~mla and dottadwhere c,~-~,,,~mled ,>,,'to s t ~ m'mct~pof ~ • h~e, em:t~ 0 3~ wmmrw~l, ~ I:~lm~ <130 !~.-OTm" of ta~ af ~ o:e .:.:; in f ~ t • ° I1~1t'0¢k top b~Ow ekMltion of 130 fqmt • °° ° Q° ° °o *° ° . ; , ~ . : ' . .. 7 . ..'~_. QTm." . ~:::=-:~,;':'; . . . . . . . . . . . . . . . . . . . . . . . . ::o 17"15" -1-:-:-2-2-.". .." • i::-:(.-:_=_:~:4_-:-_=~o:; •" .: • ~ .. ,, " ~::::::::. .............. _........... : - - - - - : - = - - - : ~ _ - : : Z - : : : : - ~ : - = - - : : : : .. f~-=_---_::_:--_::il ~=-:::'-"---:" 0 ---------.--------- ........... 0.5 ~ , 1 -_ -_-_- - . . . . . . . . . . . . . O~ I 88°40 ' Fig. 7. Map of the Massac Creek structure (Hnbbs Creek Fault Zone)• ,' -i 1.5 ICe -~ -__:: 244 W. John Nelson et al. / Engineering Geology 46 (1997) 235-258 Northwest Southeast Weaver ...~_,~m -~'ff- QTm borehole Qal i ~ i -"'¥.M~-'~- 7~ ~,,'- older Miuissipp~n rocks , ~p.'J.~-'.:J ~~'~ I ~ Q " ~ Hoiocane alluvium I 0 ~ 0 Quaternary and upper Tertiary, undiffer~a~aid ~ McNairy F o ~ ~Tar Mississippian betow Mdr I .5 .5 I I 1 1 mi I 1 1.5 km Springs to Hatclim~urg ~=~'~ ~ Formation ~ West Baden Sarclstone ~ Oowneys Bluffand Rlmauit Limestone Fig. 8. Cross-section of the Massac Creek structure. and lenses of chert gravel from 15 to 108 feet (4.5-33.0 m). Chert pebbles have bronze patina diagnostic of the Mounds, but the unit contains much more sand and silt than usual Mounds. From 108feet (33.0m) to the bottom of the Weaver borehole the core sampled a unit previously unknown in Illinois. It is mostly light gray to yellow, very fine to fine-grained sand that is well sorted and composed of quartz with little or no mica. Scattered small, rounded granules of quartz and chert are present. Interbeds of silt are present, becoming thicker and more numerous downward. The silt is medium to dark olive and brownish gray and contains abundant coarse fibrous, peaty plant material. The lower 70 feet (21 m) of the core contains layers of tough, black, highly organic clay that can be called gyttia or sapropel. Deformed sediments occur throughout the core. Steeply tilted and contorted laminations, breccia zones, clastic dikes (?), and chaotically intermixed sand, silt, and clay are present. Palynological analysis was performed on two samples from the core by R.J. Litwin and N.O. Frederiksen, both of the US Geological Survey. The samples came from depths of 145.3-146.0 and 298.6-299.0 feet. Both samples contained diverse pollen assemblages of nearly "modern" aspect, representing a variety of hardwood trees, conifers, grasses and herbs. R.W. Litwin (written communication 1995) stated, "The abundance of the alga Botryococcus in the (upper) sample suggests that this sample was deposited in standing to sluggish water, such as a channel cut-off, marsh, lake, or pond". Litwin interpreted the age of both samples as late Miocene to early Pleistocene, the older sample being most likely Pliocene and the younger sample Pleistocene. Frederiksen (written communication, 1996) differed in age interpretations, reporting the lower sample probably represents a W. John Nelson et al. /Engineering Geology 46 (1997)235-258 late Miocene flora, while the upper sample "may not be younger than Pliocene, but it could be Miocene, probably no older than late Miocene". The discrepancy in age assignments is a topic for further study. Thus the Massac Creek graben probably began to sink during or before late Miocene time, and it persisted at least into Pliocene time. The graben became a swampy depression that trapped sediment, pollen, and plant debris. Following deposition of the unnamed unit, the graben continued to subside and received a sandy facies of the Mounds Gravel (Pliocene to early Pleistocene). The base of the Mounds is downdropped about 190 feet (57 m) relative to outcrops outside the structure. 2.4. Barnes Creek Fault Zone The Barnes Creek Fault Zone is named here for Barnes Creek in Massac County. Weller and Krey (1939) and Baxter et al. (1967) previously mapped part of the fault zone but did not name it. The Barnes Creek Fault Zone strikes N40°E across Massac, Pope, and Hardin Counties, a distance of about 40 km (Fig. 2). Along most of its length it is a zone less than 0.6 km wide of subparallel faults that outline grabens or, less commonly, horsts. Throw across the zone is less than 30 m in most places, but some grabens are downdropped 100 m or more. There is little drag, and fault slices are horizontal or gently tilted. Primary joints in bedrock strike parallel to the zone; secondary joints are perpendicular to the faults. Faults that offset Quaternary strata are exposed in two places. Site A is along Barnes Creek in the SW~ NW~ SE~, Sec. 9, T15S, R5E, Massac County (Fig. 9). The Metropolis gravel (informal name) is faulted against clay and sand of the Cretaceous McNairy Formation at Site A (Fig. 10). The Metropolis gravel is crudely stratified and composed of rounded chert pebbles in a matrix of gray to brown silt and clay. Some pebbles have the brown patina typical of the Mounds Gravel, but most are bleached, worn and pitted. The Metropolis therefore is interpreted as reworked Mounds Gravel and must be younger than the Mounds. 245 Faults at Site A are vertical or high-angle normal (Fig. 10). Throws range from a few cm to more than 1.5 m (the maximum vertical exposure in the streambank). Most faults strike N10°E to N40°E, but one NW-trending fault is exposed. Small grabens and elastic dikes in the McNairy are filled with Metropolis gravel (Fig. 10 and Fig. 11 ). Dikes and grabens range from a few cm to 0.9 m wide. The wider grabens look like the New Columbia and Reineking Hill structures in miniature (Fig. 11). Narrow ones are simply infilled tension fractures. All structures are truncated at the top by horizontal, undisturbed Holocene sand and silt. Seismic-reflection and georadar profiles were run along the north bank of Barnes Creek (Sexton et al., 1996). Georadar profiles gave information to depths of 10-15 m, whereas seismic profiles yielded data to roughly 100 m, mainly in Paleozoic bedrock. The seismic data indicated numerous high-angle faults, some of which outline grabens and others form apparent flower structures. Radar profiles showed numerous small faults in the upper McNairy and probably the lower part of the Quaternary sediments. The uppermost, Holocene radar reflectors are not displaced. Some faults interpreted through geophysics match exposed faults, while others are concealed. Site B (the Midway Site) is along a creek in the SW] SW~ NW 1, Sec. 36, T14S, R5E, Massac County (Fig. 9). Steeply dipping normal and reverse faults strike N20°E and outline a complex horst or fractured anticline in the McNairy Formation and Metropolis gravel (Fig. 12). As at Site A, narrow fissures or grabens filled with clay and Metropolis gravel intrude the McNairy. All these structures are truncated by horizontal, unfaulted Holocene sediment. The Barnes Creek Fault Zone crosses the Cache Valley in Pope County (Fig. 9). The Cache Valley is a broad lowland that was occupied by the Ohio River during most of Pleistocene time. The Ohio shifted into its present channel between 8000 and 25 000 years ago (Weller, 1940; Masters and Reinertsen, 1987; Esling et al., 1989). Surface features, such as meander scrolls and terraces, in the Cache Valley near the fault zone were examined on aerial photographs and in the field. No offsets 246 IV. J o h n N e l s o n et al. / E n g i n e e r i n g G e o l o g y 46 ( 1 9 9 7 ) 2 3 5 258 J~CKSON :,POPE i •.. +,/ - "1t > " . .." ..,:" Reinetdng .. ~.~ I~l'ti/Iote .. ..., ,,::" ". 3~te .'-;~'" ~" ..¢'ite " ~_~:.. A " "::'" ." .:" ~.';". ..'%..'-"'-~'" B "" "-, 0 • ~'-" " I "'" ~ . 5km II 0 ": " "" 3 mi • Mallard Creek I 88o45 ' 88=30, Fig. 9. Map showing Barnes Creek fault zone, Massac Creek structure, Reineking Hill structure and Rock Creek graben. Significant outcrops and drillholes are indicated. Faults in Kentucky are from Amos (1966); faults north of Cache Valley are from Weller and Krey (1939). East BARNES ................ CREEK, SITE A ........................ H~al|trdtmn - ...................................................... ................. ...... " _'. ......!;...~...p.~ : "_ . ,...... .*. _. . . . . . . . . . . ~.. ' " • t '¢ . I ~ " ' ' " / - ' - /;, - - ,/. , ~ ~ ' ~ - ~ , 10ft I ~ I ~ ' ' . ' ~ . ~ • ". • ~ - . . . . . - . . ~ ~ : I 5 I : "7".---!. ~ I ". ' ." ' . " ~ " . ~ ' 0 0 - /,-'.'-''.- ....,.*..... ....- ' ,. ' o - ".,.- ' /., ...'-_./.. .... • : .• . ~" 7- " , ." ." . -* .= .-~"--.. • ~ . . . . • ,,.~_,,=.. • . . . ~ h • ~ I , . . . u , ~ a r v ~ ..: : . r ~ . ~ . - . .: : ...:.: V l ~ "• ' .". .,' . . . . ' ~~. .:. ..' .". :. ."' . ; r ' -'. .' ~' " ': " ~/ : ~ t~ ' • west " ' ' .,., . ' ... ' . • _ " ":.'" ' :-I ' - I . • . "~ ~/.~--~.~ - ~ _. "-~, A . J ' - ' ". . . . ' , , - ~ - , - , = ,.,t,,.,-,, 3m Fig. 10. Sketch of structure exposed in south bank of Barnes Creek at Site A. See Fig. 9 for location. or disturbances attributed to tectonic activity were found. Five test holes were drilled straddling the Barnes Creek Fault Zone just west of Homberg, on a prominent terrace flanking the northeast side of the Cache Valley (Fig. 13). Units drilled include the Parkland Sand, an aeolian deposit of Woodfordian age, the Henry Formation, slightly older Woodfordian fluvial sediments; and the informally-named Homberg sequence (Fig. 13). The Homberg sequence is composed of interbedded gray to reddish brown silt and clay. It contains two well-developed paleosols, one at the top and the second 5-11 m lower. Radiocarbon dating on two samples of plant material from the upper paleosol yielded dates of 18 160+190 and 19 640+210 years (Chao-Li Liu, Illinois State Geological Survey, written communication, 1995). These dates match the Gardena soil, an informallynamed paleosol of early Woodfordian age. The 247 W. John Nelson et al. / Engineering Geology 46 (1997) 235-258 Fig, 11. Photograph of small pull-apart graben at Site A, Barnes Creek. The filling is Metropolis gravel downdropped into sand of the McNairy Formation. BARNES CREEK FAULT East z \ ~ " / " . MIDWAY EXPOSURE West / / " / \ - slumped -- grovel ,/ , / ", Metronolis / " N \ / material s i [ 5ft I. \ • [ 0 / . . . J \ / - A"----L / " ' ~ . . .'. _.V~-~.• "/:--~ ~-~" • • . . • • \ ZONE: t o p of b a n k , ,." Holocene .-..-....-.. . . °.-.... . . . ,. , ",, "~"" ~ . " / " I P clay.filled \ fissures Fig. 12. Sketch of structure exposed in south bank of creek at Site B along the Barnes Creek fault zone. See Fig. 9 for location. lower paleosol in the H o m b e r g sequence probably is the Farmdale Soil, developed in Altonian (middle Wisconsinan) sediments. The cross-section (Fig. 13) indicates no significant vertical offset o f any sampled unit across the Barnes Creek Fault Zone. This constrains the latest movement in the H o m b e r g area to pre-Farmdale (before 25 000 years ago). To summarize, last movement on the Barnes Creek Fault Zone is pre-Farmdale and postMetropolis gravel. Age of the Metropolis gravel is poorly constrained. The gravel forms a dissected terrace at 410-420 feet (125-128 m) west of Site A. This feature, the Metropolis terrace of Alexander and Prior (1968), is composed of reworked "Lafayette" (Mounds) Gravel and is 248 W. John Nelson et al. ,. Engineering Geology 46 (1997) 235 258 BARNES (;;REEK FAULT ZONE: mLLH 380- ASS 9 ..~ . . : .J} .". "i- ' ! : peoria 8 ~ 10 ! ~-'- ~ : 6 ..... :':. I p ~ l +....... : . ". " , ' 1 ~ " I- " - 1 : . : ; ; : : . ." .I- ..'. .'l -+ G co --" -- :-l:.l'i"v " - - ° - - I- - - ~ . . # - . " . . ... .. . .. . . . I.. .. . . . ...... ::-~~ >320- ~ . ~ + , , / . 7 . al ( - - ~j;.. l --+ - - "~" - -+ • radiocarbon samples no horizontal scale ' ~ '... ~ _- _-_ o =0- " "" ~ ,_ "+-~-;.+.'7..~ i , + . -- - ,-o."I,,,E-,H.=.¢.-~ - 5-'11 oi/ __-£ + . . . , , ~ ~ ~ i _ -~c = r- "1 - ' - 41+= -+ = Fault Zone Fig. 13. Interpretive cross-section based on drilling at Homberg. overlain by Peoria Silt (Woodfordian). The Metropolis terrace is higher than, and therefore older than, the Brownfield terrace, which has a Woodfordian radiocarbon age of 13000 to 14 000 years (Alexander and Prior, 1968). Further study is required to pin down the age of the Metropolis. The Barnes Creek Fault Zone shows elements of both extensional (normal) and wrench faulting; unfortunately we found no piercing points or kinematic indicators in Cretaceous or younger materials. Slickensides and mullion in an underground fluorspar mine near the northeast end of the fault zone indicate dextral oblique slip. These slip indicators, however, might be as old as late Mississippian. The fault zone probably has undergone two or more episodes of movement since that time. 2.5. Rock Creek Graben The Rock Creek Graben is one of the major grabens in the FAFC. It extends from Union County, Kentucky into Illinois on a heading of $55°W, then curves to $25°W, crosses back into Kentucky, and finally re-enters Illinois (Fig. 2). High-angle normal faults and a few reverse faults displace Pennsylvanian and older rock, with throws locally more than 600m (Baxter and Desborough, 1965; Trace and Amos, 1984). We mapped the Rock Creek Graben where it crosses the margin of the Mississippi embayment in Pope County (Fig. 9). Previous geological maps by Weller and Krey (1939), Ross (1964) and Amos (1966) show faults, but indicate no displacement of units younger than Mississippian. In Pope County the Rock Creek Graben is composed of two smaller, parallel grabens about 3 km apart (Fig. 9). The western graben is bordered by high-angle normal faults that bear dipslip striations and mullion. Throw ranges up to 230 m. The eastern graben curves from N10°E on the south to N50°E where it crosses the Ohio River. On the Ohio River bluff just south of Bay City, the southeastern fault displays drag folds opposite to observed throw, and a series of tight anticlines and synclines parallel the northwest side. These features suggest two or more episodes of displacement with reversals of throw, or sinsitral strike-slip. The sharp jog in the fault zone at Bay W. John Nelson et al. / Engineering Geology 46 (1997) 235-258 City would make a restraining bend for sinsitral movement. Outcrops of folded and faulted McNairy Formation (Cretaceous) occur along Mallard Creek in line with the western graben (Fig. 9). Folds are gentle (dips less than 30°); folds and faults strike northeast, parallel to faults in bedrock. Narrow clay dikes or clay-filled joints in the McNairy also trend N20-50°E. One clay-lined fault in the McNairy is truncated at the base of overlying gravel. The gravel, composed of wellrounded gray to black chert pebbles in a matrix of sand and lacking the Mounds patina, resembles Eocene gravel elsewhere in the northern Mississippi Embayment. The McNairy-Mounds contact does not change elevation across the western graben north of Mallard Creek. In the same area the McNairy-Mississippian contact also appears not to be offset, Some Mounds exposures near Mallard Creek contain narrow clay veins or clay-lined fractures that strike in a variety of directions. Whether these joints were created by tectonic stress, or nontectonic factors such as creep, is not known. A suggestion of tectonism involving Mounds Gravel was found in a small gravel pit near Sevenmile Creek, about 5.5km southwest of Mallard Creek. Here the Mounds dips about 4 ° northwest and contains numerous near-vertical, clay-lined joints or fissures that strike N20-50°E. Logs of nearby water wells indicate that the Mounds-McNairy contact gently dips northwest. Sevenmile Creek is near the projected trace of the Rock Creek Graben, but no data on bedrock structure are available here. 2.6. Lockhart Bluff Graben Published geological maps show faults displacing units as young as the Mounds Gravel in the Lockhart Bluff Graben, Livingston County, Kentucky. The relevant maps cover the Little Cypress (Amos and Wolfe, 1966), Smithland (Amos, 1967), and Burna (Amos, 1974) quadrangles. The Lockhart Bluff Graben is a major graben in the FAFC, comparable to the Dixon Springs and Rock Creek Grabens (Fig. 2). Faults along 249 its southeast margin have throws as great as 900 m, the largest in the FAFC. The fault pattern is intricate. Near the Cumberland River the graben bends from S30°W to ST0°W; it sharply narrows and curves south again near the Tennessee River. Some faults on the geological maps show "scissoring", or reversal of displacement. The major fault on the southeast border of the graben consistently displaces Paleozoic rocks down to the northwest (Amos and Wolfe, 1966; Amos, 1967). However, the McNairy Formation and "continental deposits" (Mounds Gravel) are downthrown to the northwest in some places, and downthrown to the southeast in others. Apparent strike-slip displacement of the Mounds Gravel is mapped about 2130m from the north line and 3050m from the east line of the Little Cypress Quadrangle (Amos and Wolfe, 1966). The fault strikes northeast and offsets the contact of the Mounds with older units about 120 m in a right-lateral sense. Kinematic studies are needed to determine whether such anomalies resulted from lateral motion or from recurrent episodes of dip slip. 3. Ste. Genevieve Fault Zone The Ste. Genevieve Fault Zone (SGFZ) extends more than 190 km across southeast Missouri into southernmost Illinois (Fig. 1). It is a complex zone of high-angle faults that have undergone repeated episodes of movement. The SGFZ may have originated as a crustal plate boundary or suture zone during the Proterozoic (Heigold and Kolata, 1993). Faulting during late Middle to Upper Devonian time displaced rocks down to the southwest in Missouri (Weller and St. Clair, 1928; Nelson and Lumm, 1985). The Devonian faulting included a component of left-lateral strike slip, which formed a transtensional graben in the Minnith Quadrangle (Schultz and Harrison, 1994). During the Pennsylvanian Period, the block southwest of the SGFZ was uplifted along high-angle reverse faults (Weller and St. Clair, 1928; Nelson and Lumm, 1985). Strike-slip faulting of uncertain magnitude and timing also has been postulated (Clendenin et al., 1989; Harrison and Schultz, 1994a) 250 W. John Nelson et aL/ Engineering Geology 46 (1997) 235-258 We mapped the southeast part of the SGFZ in Illinois, where it approaches the Mississippi Embayment. Here the SGFZ changes from a narrow zone of large displacement (over 600 m) to a widely bifurcating set of much smaller faults. Some faults in this area displace Cretaceous and Tertiary sediments. 3.1. Iron Mountain Fault Zone The Iron Mountain Fault Zone in Union County is a branch of the SGFZ (Devera and Nelson, 1995). The Iron Mountain structure is about 5 km long, changing trend from NW-SE at the north end to nearly N-S at the south end (Fig. 14). The zone is 120-760 m wide and contains parallel faults that outline narrow downdropped blocks (Fig. 15). A few short, perpendicular cross-faults also are mapped. Throws range to as much as about 100 m. The western large fault displaces Tertiary gravel on the east against Devonian and Mississippian bedrock on the west. The cut bank of Clear Creek at the north end of Iron Mountain shows steeply dipping Tertiary conglomerate adjacent to Middle Devonian limestone. Outcrops of silicified, limonitic fault breccia along the crest of Iron Mountain contain both angular clasts of Paleozoic chert and well rounded Tertiary pebbles. Clay was mined from Tertiary sediments along the east face of Iron Mountain during the early 20th Century. Geologists who visited the mines found clay deposits in deep vertical-walled depressions in Mississippian limestone. The limestone walls were linear and vertical, running mostly N-S and E-W. Clay was mined both from open pits and from shafts as deep as 30 m (St. Clair, 1917; Parmelee and Schroyer, 1921; Lamar, 1948). Although St. Clair (1917) described a fault, striking east-west and having about 4.5 m of throw down to the south, the clay-filled depressions were generally interpreted as sinkholes that formed shortly before or during deposition of the clay. The clay pits are all inaccessible today, and no natural exposures were found. Factors that favor tectonic grabens over solution collapse include: ( 1 ) walls of clay deposits were linear, nearly vertical, and parallel or normal to mapped faults; (2) large magnitude of downdrop (Tertiary sediments are absent from nearby hills as much as 200 m higher in elevation); (3) lack of reported solution features, rubble zones or chaotic structure; and (4) lack of any similar sediment-filled sinks elsewhere in the region. The youngest Tertiary deposit along the Iron Mountain Fault Zone is gravel, composed of wellrounded gray to black chert pebbles that lack polish or patina. The matrix is white to gray quartz and chert sand. This gravel is similar, but coarser than Eocene gravel of southern Illinois (Kolata et al., 1981) western Kentucky (Olive, 1980) and southeastern Missouri (Johnson, 1985; Harrison, in press). Perhaps gravel along the Iron Mountain Fault Zone is coarser because it was deposited closer to the source area. Two separate palynological analyses of lignite from wastepiles of abandoned clay pits indicate Eocene age (Aureal T. Cross, 1984, oral communication to Nelson; D.J. Nichols, 1993, written communication). White to red water-bearing sand, reported to underlie the clay (St. Clair, 1917) may be McNairy. The Mounds Gravel does not occur in the area. 3.2. Other faults Other faults displace Cretaceous and early Tertiary strata in southern Union and northern Alexander Counties. These faults trend in a north-south belt more or less in line with the Iron Mountain Fault Zone. Details are given by Nelson and Devera (1995) and Nelson et al. (1995a,b). Most post-Cretaceous faults in the area strike N-S to N20°W, and are linked by short E-W cross-faults. Fault surfaces are nearly vertical and lack slickensides or breccia. Various Devonian and Mississippian bedrock units are displaced, along with the Tuscaloosa Formation (Cretaceous), the McNairy Formation (Cretaceous, overlying Tuscaloosa), and Eocene(?) gray to black chert gravel. A well-exposed fault is in a cutbank of Cooper Creek (Fig. 2; SW¼ SW¼ NE¼, Sec. 1, T14S, R2W, Mill Creek Quadrangle). A backhoe trench along the hillside south of the creek revealed a zone of parallel, high-angle normal and reverse faults that W. John Nelson et al. / Engineering Geology 46 (1997) 235-258 251 ~ ' ~ - - ] Holocene alluvium T'-" ~ Mi~isn-Chesterian , - J L _ .......... fault, rw~se fautt, and dott~:l where concealed o J o i 0.5 0.5 I 1 1 mi I 1'.5 Icm .'"" ~ ~ ,... ..." . ."" Qai TKU 89"22"30" 89"15" Fig. 14. M a p of the Iron M o u n t a i n fault zone. 252 HL John Nelson et al. / Engineering Geology 46 (1997) 235-258 Southwest Northeast Iron Mountain Fault Zone Qal ,~luaan and Omovio~ I 0 , .5 " .,'.- , 1 km '-~ Holocenealluvium ~'~ Tertiaryand Cretaceous ~ Mississippian~riarl • ol_'j . ~s~ppian-Valmeyeran =='~ Middle Devonian ~.=~ LowerDevonian Fig. 15. Cross-sectionof the Iron Mountain fault zone. strike N-S to N10°W and juxtapose silicified Paleozoic bedrock with gravel interpreted as Tuscaloosa Formation (Fig. 16). Overlying Quaternary colluvium, alluvium and loess are not displaced. Some north-trending faults contain grabens filled with megabreccia of angular bedrock fragments up to several meters across. The Cape Road Fault in Union County has a mega-breccia zone as wide as 0.6 km. Similar breccias in the Thebes Gap area are interpreted as filling pull-apart grabens produced by divergent wrench faulting (Harrison and Schultz, 1994b). In Alexander and Union Counties no Cretaceous or younger materials were found in breccia. In fault exposures throughout the area Quaternary sediments are not disturbed. Many stream banks and artificial cuts show undeformed loess, colluvium and alluvium truncating faults or strongly deformed bedrock. Perhaps the finest such exposure was along the Cape Road Fault in the NE~ SWJ, Sec. 17, T l l S , R2W, Union County. The Peoria, Roxana and Loveland Silts, along with two distinct layers of pre-Loveland alluvium, are horizontal and undisturbed above a chaotic megabreccia of Paleozoic rocks. West 0 * i • Fig. 16. Sketch of the fault zone trenched at Cooper Creek, based on a photo mosaic. W. John Nelson et al. / Engineering Geology 46 (1997) 235-258 4. Commerce Fault Zone The Commerce Fault Zone (CFZ) is a northeasttrending structure in Scott County, Missouri and Alexander County, Illinois (Fig. 1 and Fig. 2). The CFZ coincides with a major geophysical lineament, displaces 10-25 ka Woodfordian loess, and possibly is still generating earthquakes. Stewart (1942), Stewart and McManamy (1944) and Grohskopf (1955, pp. 26-28) reported units as young as Mounds Gravel are folded and faulted near Commerce, Missouri (on the Mississippi River). Stewart (1942) described a fault displacing Pleistocene loess about 6 km southwest of Commerce. Johnson (1985), Harrison and Schultz (1994b) and Harrison (in press) mapped a complex array of faults in the Thebes Quadrangle. The CFZ strikes N50-55°E and lies along the southeast escarpment of the Benton Hills in Missouri. It displaces Mounds Gravel in both Missouri and Illinois. Numerous faults that also displace the Mounds branch NNE off the CFZ. These faults reflect multiple episodes of displacement dating back at least to the Ordovician (Harrison and Schultz, 1994b). The most recent episode involved right-lateral strike slip. Trenches were dug in 1995 and 1996 across the English Hill Fault, a NNE-striking splay off the CFZ. They reveal a series of horsts, grabens and thrust faults that displace Pleistocene units as young as the Peoria Silt (Woodfordian; 10-25 ka). As many as five episodes of tectonic deformation may have taken place after deposition of the Mounds Gravel (Palmer et al., 1996). Fault patterns fit both transtensive and transpressive, dextral strike-slip. Seismic profiles show that the faults dip steeply and extend to depths of at least several kilometers (Harrison et al., 1995; Hoffman et al., 1996; Palmer et al., 1996). The CFZ coincides with a northeast-trending zone of subtle gravity and magnetic highs, the Commerce Geophysical Lineament (CGL). The CGL extends more than 400 km from northeast Arkansas to southern Illinois and possibly beyond (Hildenbrand and Hendricks, 1995; Langenheim and Hildenbrand, 1996). Modeling indicates the source of the CGL may be igneous intrusions of 253 intermediate composition (Langenheim and Hildenbrand, 1996). The CGL parallels the Reelfoot Rift, and likely represents a basement shear zone that developed during rifting. Although no continuous fracture zone can be mapped in Illinois, a series of structural features lie along the CGL (Fig. 2). In Alexander County the CFZ extends directly from the Thebes Quadrangle (Harrison, in press) across the Tamms Quadrangle (Devera and Follmer, in preparation), southeastern Mill Creek Quadrangle (Nelson et al., 1995b), and into the southwestern Dongola Quadrangle (Nelson and Follmer, in preparation). The northeast-striking faults are near-vertical and create strikingly linear valleys and bluff lines. Paleozoic bedrock along the faults is intensely silicified. Coincident positive gravity and magnetic anomalies are centered on the CFZ and CGL in Alexander County (Heigold, 1976; Heigold et al., 1993; Hildenbrand et al., 1993). The anomalies probably represent a mafic igneous intrusion, which was a likely heat source for hydrothermal activity and silicification in the southern Illinois tripoli district (Berg and Masters, 1994). Northeastward in Union, Johnson, and Saline Counties, small faults that strike NE and N - S lie along the CGL (W.J. Nelson, unpublished mapping). The New Burnside Anticline and associated faults also follow the CGL. Faults along the New Burnside strike northeast and dip steeply. Most are dip-slip normal faults, but some exhibit signs of an earlier episode of reverse faulting. On coalmine highwalls these faults do not displace Pleistocene loess (Nelson et al., 1991 ). In central Saline County a northeast-trending fault, unusual for the area, lies amidst a swarm of NW-trending ultramafic dikes. The dikes, radiometrically dated early Permian, are part of the Cottage Grove Fault System: an east-trending dextral wrench fault (Nelson and Krausse, 1981; Nelson and Lumm, 1987). These mantle-derived intrusions are concentrated where the Cottage Grove crosses the CGL. The Omaha Dome in northwest Gallatin County contains ultramafic sills and lies close to the CGL. Finally, faults in the Wabash Valley Fault System 254 I4( John Nelson et al. / Engineering Geology 46 (1997) 235-258 change trend from N-S to about N30°E where they cross the CGL. The overall picture for the CGL in southern Illinois is of a long-lived, deep crustal fracture zone. The CGL provided a conduit for magma and hydrothermal fluids, and influenced but did not dominate near-surface structures. To date, evidence of Quaternary deformation is confmed to a small area near the Mississippi River. 4.1. Seismicity along the CGL Harrison and Schultz (1994b) identified 12 earthquakes of Mb 3.0 or greater that possibly are related to slippage along the CFZ. Nine of those quakes occurred in Missouri and three in Illinois. A fourth earthquake on or near the CFZ in Illinois took place after Harrison and Schultz' paper went to press. The epicenter of the Tamms earthquake of 1965 (Mb 3.8) lies approximately 2 km southeast of the CFZ (Fig. 2). The unusually shallow focal depth of 1.5 km (Herrmann, 1979) lies near the base of Ordovician bedrock. Focal analysis by Zoback (1992) indicates the most likely solution was strikeslip along a surface trending N80°W/70°NE, which would be antithetic to the CFZ. A second earthquake near Tamms (Mb 3.3) was part of the same event and was cited by Harrison and Schultz (1994b). No focal solution is available. The Harrisburg earthquake of 1984 (Mb 4.1) had an epicenter about 1.5 km northwest of a fault mapped by Nelson and Lumm (1986) from borehole data. The mapped fault is a NE-striking normal fault that has about 20 m of throw down to the northwest. Focal depth of the quake was about 2 km (about Upper Ordovician depth). Focal analysis (Otto Nuttli, personal communication, 1984) indicated slip on a high-angle normal fault striking ENE - - a close match for the mapped fault. The most recent earthquake close to the CFZ in Illinois occurred near Dongola in February 1994 (Mb 4.2). The focal depth of 16 km is deep into Precambrian basement. The focal-plane solution indicates the fault was oriented either N20°E/ 70°SE with right-lateral slip or N59°W/62°SW with left-lateral slip (Robert HeHii~ann, written com- munication, 1995). The former could fit the CFZ, the latter would be an antithetic fracture. More study of the CFZ clearly is in order. Further trenching is planned near Thebes, Illinois where Quaternary terraces terminate abruptly along mapped faults. Another promising area is in Saline and Gallatin Counties, where thick lacustrine silt and clay of the Equality Formation (Wisconsinan age) occur near the CFZ. Trenching here might show faulting or earthquake-related liquefaction features. 5. Discussion Tectonic faulting of Miocene to early Pleistocene sediment is widespread in the FAFC in southeastern Illinois and nearby in Kentucky. Neogene faults typically strike N20°E to N40°E and outline narrow linear grabens. Fissures or open joints filled with downdropped sediments occur in the Barnes Creek Fault Zone. These structures imply NW-SE extension. Several structures bear evidence of strike-slip movement, but the direction and magnitude of lateral movement are not known. No piercing points or reliable slip indicators have been identified yet in the area. The FAFC has undergone repeated episodes of faulting, with reversals in the sense of displacement, dating back to Cambrian time. All Neogene faults are probably older faults that have been reactivated. The youngest unit deformed by the FAFC is the Metropolis gravel. Its age is not well constrained, but probably it is Illinoian or older. Undeformed New Columbia sand in the New Columbia structure most likely is pre-Illinoian. Wisconsinan and younger sediments are not displaced anywhere, to our knowledge. Faults near the southeast end of the SGFZ in Illinois displace Cretaceous and Eocene strata. No younger units are known to be affected; several exposures clearly show undeformed Illinoian and younger sediments overlying large faults that offset older rocks. Tertiary faults mostly strike N-S to ~-SSE, dip nearly vertically, and appear to be pull-apart grabens. As in the FAFC, features suggesting strike slip are present, but the direction and amount of lateral movement cannot be deter- W. John Nelson et al. / Engineering Geology 46 (1997) 235-258 mined. The SGFZ has undergone multiple tectonic events under different stress fields. The youngest documented faulting in the area took place on the English Hill Fault, which branches off the CFZ. Woodfordian (10-25ka) loess is displaced in Missouri a few miles from the Illinois border. Multiple episodes of movement took place, the most recent being dextral. The CFZ lies along a magnetic anomaly believed to represent a major fracture zone in the basement. Surface expression in Illinois is subtle, but mafic and ultramafic intrusions, intense hydrothermal activity, and faults follow the magnetic lineament. Several earthquake foci in Illinois and Missouri are on or very near the CFZ; focal analysis indicates that at least some may have originated in the CFZ. However, the level of seismic activity on the CFZ is not higher than background throughout southern Illinois. 5.1. Stress field and faulting The maximum horizontal compressive axis of the modern stress field in southeastern Illinois and the adjacent part of Kentucky is oriented nearly east-west. This trend is shown by numerous lines of evidence, including earthquake focal mechanisms (Hei-~mann, 1979; Zoback and Zoback, 1991), borehole breakouts (Dart, 1985), straingauge and overcoring measurements (Ingram and Molinda, 1988), and roof-failure trends in underground mines (Nelson and Bauer, 1987; Ingram and Molinda, 1988). Elsewhere in Illinois, the maximum stress axis is oriented about N50°E to N70°E, as it is in most of the Midcontinent of North America. The maximum horizontal compressive stress axis in the New Madrid seismic zone is oriented about N45°E to N80°E, the average orientation being ENE. Focal-plane solutions indicate that northeast-trending faults in the New Madrid seismic zone are undergoing right-lateral slip with a reverse component, whereas north-south to north-northwest-trending faults are undergoing primarily reverse slip. Fault motions in the New Madrid area thus are consistent with the measured stress regime (Zoback and Zoback, 1981, 1991; Rhea et al., 1994). 255 Among Neogene faults in southern Illinois, only the Commerce Fault Zone exhibits the slip that is expected under the modern stress field. The Commerce strikes northeast and shows right-lateral displacement, like northeast-striking faults in the New Madrid seismic zone. Neogene faults in the FAFC strike NNE to NE, so they also would be expected to undergo rightlateral slip under the contemporary stress field. These faults appear to be strike-slip or obliqueslip (direction unknown) with a large component of extension. The extensional component is not a good match for the current stress field. Tertiary extensional grabens in the SGFZ strike nearly perpendicular to the present-day maximum compressive stress axis. These structures clearly are inconsistent with the modern regional stress field. Several explanations are available for the discrepancy of fault and stress orientation. One is that the stress field may have changed orientation from Tertiary to present. The relatively high present-day strain rate and lack of pervasive deformation imply that the New Madrid seismic zone is a young feature - - at most, a few million years old (Schweig and Ellis, 1994). The center of seismic activity possibly shifted from southern Illinois to the New Madrid area as the stress field rotated. Unfortunately, little evidence is available on Neogene stress fields. Engebretson et al. (1985) proposed that the relative motions of the North American and Pacific plates changed roughly 5 million years ago, during the Pliocene (Engebretson et al., 1985). Such a change in plate movement and interaction might have changed the stress regime of the continental interior. Another explanation is that high fluid pressure may have triggered slippage of unfavorably oriented faults. Fluid pressures commonly are high in compressional regimes, where high-angle reverse and strike-slip faults predominate (Sibson, 1990). Seismicity in the New Madrid Seismic Zone today is partly attributed to high fluid pressure (McKeown and Diehl, 1994), Other factors that may contribute to slippage on unfavorably oriented faults are local variations in friction along faults and variable local stress fields near fault intersections (Zoback, 1992). Also, slippage along undulating faults may create local, 256 W. John Nelson et al. / Engineering Geology 46 (1997) 235-258 residual stress that differs markedly from regional stress. For example, in southern California the leftlateral Cleghorn Fault strikes nearly parallel to the adjacent, right-lateral San Andreas Fault (Saucier et al., 1992). Resolving which of these factors are at work in southern Illinois will require further study. In particular, closer constraints are needed on the timing of movement and the magnitude and direction of slippage on Neogene faults. Acknowledgment This research was partially supported by the US Geological Survey (USGS), Department of the Interior, under USGS award number 1434-95-G-2525. The views and conclusions contained in this document are those of the authors and should not be interpreted as necessarily representing the official policies, either expressed or implied, of the US Government. References Alexander, C.S., Prior, J.C., 1968. The origin and function of the Cache Valley, southern Illinois. In: Bergstrom, R.E. (Ed.), The Quaternary of Illinois. University of Illinois, College of Agriculture, Special Publication 14, pp. 19-26. Amos, D.H., 1966. Geologic map of the Golconda Quadrangle, Kentucky-Illinois, and the part of the Brownfield Quadrangle in Kentucky. US Geological Survey, Map GQ-546, scale 1:24,000. Amos, D.H., 1967. Geologic map of part of the Smithland Quadrangle, Livingston County, Kentucky. US Geological Survey, Map GQ-657, scale 1:24,000. Amos, D.H., 1974. Geologic map of the Burna Quadrangle, Livingston County, Kentucky. US Geological Survey, Map GQ-1150, scale 1:24,000. Amos, D.H., Wolfe, E.W., 1966. Geologic map of the Little Cypress Quadrangle, Kentucky-Illinois. US Geological Survey, Map GQ-554, scale 1:24,000. Baxter, J.W., Desborough, G.A., 1965. Areal geology of the Illinois fluorspar district, Part 2, Karbers Ridge and Rosiclare Quadrangles. Illinois State Geological Survey, Circular 385, 40 pp. with map, scale 1:24,000. Baxter, J.W., Desborough, G.A., Shaw, C.W., 1967. Areal geology of the Illinois fluorspar district. Part 3. Herod and Shetlerville Quadrangles. Illinois State Geological Survey, Circular 413, 41 pp. with map, scale 1:24,000. Berg, R.B., Masters, J.M., 1994. Geology of microcrystalline silica (tripoli) deposits, southernmost Illinois. Illinois State Geological Survey, Circular 555, 89 pp. Bertagne, A.J., Leising, T.C., 1991. Interpretation of seismic data from the Rough Creek graben of western Kentucky and southern Illinois. In: Leighton, M.W., Kolata, D.R., Oltz, D.F., Eidel, J.J. (Eds.), Interior Cratonic Basins. American Association of Petroleum Geologists, Memoir 51, pp. 199-208. Clendenin, C.W., Niewendorp, C.A., Lowell, G.R., 1989. Reinterpretation of faulting in southeast Missouri. Geology 17 (3), 217-220. Dart, R., 1985. Horizontal stress directions in the Denver and Illinois basins from the orientations of borehole breakouts. US Geological Survey, Open-File Report 85-733, 41 pp. Devera, J.A., Nelson, W.J., 1995. Geologic map of the Cobden Quadrangle, Illinois. Illinois State Geological Survey, Map IGQ-16, scale 1:24,000. Devera, J.A., Follmer, L.R., in preparation, Geologic map of the Tamms Quadrangle, Alexander and Pulaski Counties, Illinois. Illinois State Geological Survey, Map IGQ Series, scale l:24,000. Engebretson, D.C., Cox, A., Gordon, R.C., 1985. Relative motions between Oceanic and Continental plates in the Pacific Basin. Geol. Soc. Am. (Special Paper) 206, 57. Ervin, C.P., McGinnis, L.D., 1975. Reelfoot rift: reactivated precursor to the Mississippi embayment. Geol. Soc. Am. Bull. 86, 1287 1295. Esling, S.P., Hughes, W.B., Graham, R.C., 1989. Analysis of the Cache Valley deposits in Illinois and implications regarding the late Pleistocene-Holocene development of the Ohio River Valley. Geology 17, 434437. Follmer, L.R., Masters, J.M., Su, W.-J., in press, Surficial geology of the Mermet Quadrangle, southern Illinois. Illinois State Geological Survey, Open File Series, scale t:24,000. Grohskopf, J.G., 1955. Subsurface geology of the Mississippi Embayment of southeast Missouri. Missouri Geological Survey, 37 (2nd Series), 133 pp. Harrison, R.W., Schultz, A., 1994a. Geologic map of the Coffman 7.5-minute quadrangle, Ste. Genevieve County, Missouri. US Geological Survey, Open-File Report 94-419, scale 1:24,000. Harrison, R.W., Schultz, A., 1994. Strike-slip faulting at Thebes Gap, Missouri and Illinois: implications for New Madrid tectonism. Tectonics 13 (2), 246-257. Harrison, R.W., Hoffman, D., Palmer, J.R., Vaughn, J.D., Schultz, A., 1995. Late Quaternary deformation on the English Hill Fault, southeast Missouri. Geol. Soc. Am. (Abstracts with Programs) 27 (6), A-389. Harrison, R.W., in press, Geologic map of the Thebes Quadrangle, Illinois and Missouri. US Geological Survey, Open File Series, scale 1:24,000. Heigold, P.C., 1976. An aeromagnetic survey of southwestern Illinois. Illinois State Geological Survey, Circular 495, 28 pp. Heigold, P.C., Kolata, D.R., 1993. Proterozoic crustal boundary in the southern part of the Illinois Basin. Tectonophysics 217, 307-319. Heigold, P.C., McGinnis, L.D., Ervin, C.P., Kucks, R.P., 1993. W. John Nelson et al. / Engineering Geology 46 (1997) 235-258 Simple Bouguer gravity anomaly map of Illinois. Illinois State Geological Survey, Illinois Map 1, 1 sheet, scale 1:500,000. Herrmann, R.B., 1979. Surface focal wave mechanisms for eastern North American earthquakes with tectonic implications. J. Geophys. Res. 24 (87), 3543-3552. Hildenbrand, T.G., 1985. Rift structure of the northern Mississippi Embayment from the analysis of gravity and magnetic data. J. Geophys. Res. 90 (B14), 12607-12622. Hildenbrand, T.G., Hendricks, J.D., 1995. Geophysical setting of the Reelfoot rift and relations between rift structures and the New Madrid Seismic Zone. US Geological Survey, Professional Paper 1538-E, 30 pp. Hildenbrand, T.G., Kucks, R.P., Heigold, P.C., 1993. Total intensity magnetic anomaly map of Illinois. Illinois State Geological Survey, Illinois Map 2, 1 sheet, scale 1:500,000. Hoffman, D., Palmer, J., Vaughn, J.D., Harrison, R., 1996. Late Quaternary surface faulting at English Hill in southeastern Missouri (abstract). Seismol. Res. Lett. 67 (2), 41. Ingram, D.K., Molinda, G.M., 1988. Relationship between horizontal stresses and geologic anomalies in two coal mines in southern Illinois. US Bureau of Mines, Report of Investigations RI 9189, 18 pp. Johnson, W.D. Jr., 1985. Geologic map of the Scott City Quadrangle and part of the Thebes Quadrangle, Scott and Cape Girardeau Counties, Missouri. US Geological Survey, Map MF-1803, scale 1:24,000. Kolata, D.R., Nelson, W.J., 1991. Tectonic history of the Illinois basin. Am. Assoc. Petroleum Geologists Memoir 51,263-283. Kolata, D.R., Treworgy, J.D., Masters, J.M., 1981. Structural framework of the Mississippi Embayment of southern Illinois. Illinois State Geological Survey, Circular 516, 38 pp. Lamar, J.E., 1927. Cretaceous, Tertiary and Quaternary deposits of extreme southern Illinois. unpublished ms. in ISGS Library (Lamar Ms. #46, pages not numbered). Lamar, J.E., 1948. Clay and shale resources of extreme southern Illinois. Illinois state Geological Survey, Report of Investigations 128, 107 pp. Lamar, J.E., Sutton, A.H., 1930. Cretaceous and Tertiary sediments of Kentucky, Illinois, and Missouri. Am. Assoc. Petroleum Geologists Bull. 14 (7), 845-866. Langenheim, V.E., Hildenbrand, T.G., 1996. The Commerce Geophysical Lineament - - its source, geometry, and relation to the Reelfoot Rift and New Madrid Seismic Zone. Seismol. Res. Lett. 67 (92), 43. Masters, J.M., Reinertsen, D.L., 1987. The Cache Valley of southern Illinois. Geological Society of America, Centennial Field Guide, North-Central Section, pp. 257 262. McKeown, F.A., Diehl, S.F., 1994. Evidence of contemporary and ancient excess fluid pressure in the New Madrid Seismic Zone of the Reelfoot rift, central United States. US Geological Survey, Professional Paper 1538-N, 24 pp. Nelson, W.J., 1991. Structural styles of the Illinois Basin. Am. Assoc. Petroleum Geologists Memoir 51,209-261. Nelson, W.J., 1996a. Recurrent faulting in Fluorspar District and Wabash Valley, southern Illinois Basin (abstract). Geo- 257 logical Society of America, Abstracts with Programs, 1996 Annual Meeting, Denver, Colorado. Nelson, W.J., 1996b. Geologic map of the Reevesville Quadrangle, Illinois. Illinois State Geological Survey, Map IGQ-17, scale 1:24,000. Nelson, W.J., Bauer, R.A., 1987. Thrust faults in southern Illinois basin - - result of contemporary stress? Geol. Soc. Am. Bull. 98 (3), 302-307. Nelson, W.J., Devera, J.A., 1995. Geologic map of the Jonesboro and Ware Quadrangles, Illinois. Illinois State Geological Survey, Map IGQ-14, scale 1:24,000. Nelson, W.J., Krausse, H.-F., 1981. The Cottage Grove Fault System in southern Illinois. Illinois State Geological Survey, Circular 522, 65 pp. Nelson, W.J., Lumm, D.K., 1985. Ste. Genevieve Fault Zone, Missouri and Illinois. Illinois State Geological Survey, Contract/Grant Report 1985-3, 94 pp. Nelson, W.J., Lumm, D.K., 1986. Geologic map of the Rudement Quadrangle, Saline County, Illinois. Illinois State Geological Survey, Map IGQ-3, scale 1:24,000. Nelson, W.J., Lumm, D.K., 1987. Structural geology of southeastern Illinois and vicinity. Illinois State Geological Survey, Circular 538, 70 pp. Nelson, W.J., Devera, J.A., Jacobson, R.J., Lumm, D.K., Peppers, R.A., Trask, B., Weibel, C.P., FoUmer, L.R., Riggs, M.H., Esling, S.P., Henderson, E.D., Lannon, M.S., 1991. Geology of the Eddyville, Stonefort, and Creal Springs Quadrangles, southern Illinois. Illinois State Geological Survey, Bulletin 96, 85 pp. Nelson, W.J., Devera, J.A., Masters, J.M., 1995a. Geology of the Jonesboro 15-minute Quadrangle, southwestern Illinois. Illinois State Geological Survey, Bulletin 101, 57 pp. Nelson, W.J., Devera, J.A., Masters, J.M., 1995b. Geologic map of the McClure and Mill Creek Quadrangles, Illinois. Illinois State Geological Survey, Map IGQ-15, scale 1:24,000. Nelson, W.J., Follmer, L.R., in preparation, Geologic map of the Dongola Quadrangle, Alexander, Pulaski and Union Counties, Illinois. Illinois State Geological Survey, Map IGQ Series, scale 1:24,000. Olive, W.W., 1980. Geologic maps of the Jackson Purchase region, Kentucky. US Geological Survey, Map 1-1217, 1 sheet plus 11-page text. Palmer, J.R., Harrison, R.W., Hoffman, D., Vaughn, J.D., 1996. Neotectonic history of the Benton Hills, southeast Missouri. Geol. Soc. Am. (Abstracts with Programs, Southeastern Section) 28 (2), 40. Parrnelee, C.W., Schroyer, C.R., 1921. Further investigations of Illinois fire clays. Illinois State Geological Survey, Bulletin 38, pp. 272-417. Rhea, S., Wheeler, R.L., Tarr, A.C., 1994. Map showing seismicity and sandblows in the vicinity of New Madrid, Missouri. US Geological Survey, Map Mf-2264-A. Rhoades, R.F., Mistier, A.J., 1941. Post-Appalachian faulting in western Kentucky. Am. Assoc. Petroleum Geologists Bull. 25 (11 ), 2046-2056. 258 I~ John Nelson et aL / Engineering Geology 46 (1997) 235-- 258 Ross, C.A., 1963. Structural framework of southernmost Illinois. Illinois State Geological Survey, Circular 351, 28 pp. Ross, C.A., 1964. Geology of the Paducah and Smithland Quadrangles in Illinois. Illinois State Geological Survey, Circular 360, 32 pp. St. Clair, S., 1917. Clay deposits near Mountain Glen, Union County, Illinois. Illinois State Geological Survey, Bulletin 36, 71 83. Saucier, F., Humpreys, E., Weldon II, R., 1992. Stress near geometrically complex strike-slip faults: application to the San Andreas Fault at Cajon Pass California. J. Geophys. Res. 97 ([14), 5081-5094. Schultz, A., Harrison, R.W., 1994. Geologic map of the Minnith 7.5-minute quadrangle, Ste. Genevieve and Perry Counties, Missouri. US Geological Survey, Open File Report 94-421, scale 1:24,000. Schweig, E.S., Ellis, M.A., 1994. Reconciling short recurrence intervals with minor deformation in the New Madrid seismic zone. Science 264, 1308--1311. Sexton, J.L., Henson, H., Koffi, N.R., Coulibaly, M., Nelson W.J., 1996. Seismic reflection and georadar investigation of the Barnes Creek area in southeastern Illinois (abstract handed out at meeting). Seismol. Res. Lett. 67 (2), 27. Sibson, R.H., 1990. Rupture nucleation on unfavorably oriented faults. Bull. Seismol. Soc. Am. 80 (6), 1580--1604. Stewart, D.R., 1942. Mesozoic and Cenozoic geology of southeastern Missouri: unpublished manuscript at Missouri Geological Survey in Rolla. Stewart, D.R., McManamy, L., 1944. Early Quaternary or late Tertiary folding in the vicinity of Commerce, Scott County, southeastern Missouri. Missouri Academy of Science Proceeding, 10 (1); also unpublished manuscript at Missouri Geological Survey in Rolla. Thomas, W.A., 1991. The Appalachian-Ouachita rifled margin of southeastern North America. Geol. Soc. Am. Bull. 103 (3), 415 431. Trace, R.D., Amos, D.H., 1984. Stratigraphy and structure of the western Kentucky fluorspar district. US Geological Survey, Professional Paper l151-D, 41 pp. with map, scale 1:48,000. Weibel, C.P., Nelson, W.J., Oliver, L.B., Esling, S.P., 1993. Geology of the Waltersburg Quadrangle, Pope County, Illinois. Illinois State Geological Survey, Bulletin 98, 41 pp. Weller, J.M., 1940. Geology and oil possibilities of extreme southern Illinois. Illinois State Geological Survey, Report of Investigations 71, 71 pp. and 1 plate. Weller, S., Krey, F.F., 1939. Preliminary geolgic map of the Mississippian formations in the Dongola, Vienna and Brownfield Quadrangles. Illinois State Geological Survey, Report of Investigations 60, 11 pp. with map, scale 1:62,500. Weller, S., St. Clair, S., 1928. Geology of Ste. Genevieve County, Missouri. Missouri Bureau Geol. Mines 22 (2nd Series), 352. Zoback, M.L., 1992. Stress field constraints on intra-plate seismicity in eastern North America. J. Geophys. Res. 97, 11761 11782. Zoback, M.D., Zoback, M.L., 1981. State of stress and intraplate earthquakes in the United States. Science 213, 96 110. Zoback, M.D. and Zoback, M.L., 1991. Tectonic stress field of North America and relative plate motions. In: Slemmons, B. et al. (Eds.), Neotectonies of North America. Geological Society of America, Decade Map, Vol. 1, pp. 339-366.