Randel Tom Cox

The Mississippi Embayment, North America: a first order continental structure generated by the Cretaceous superplume mantle event

Abstract The Mississippi Embayment of North America, a northward extension of the Gulf of Mexico coastal plain, is a southwestward-plunging trough containing ∼1.5 km of Cretaceous and Cenozoic sediments. The Embayment is underlain by the early Paleozoic Mississippi Valley graben basement fault complex. Previous authors have attributed Embayment subsidence to the opening of the Gulf of Mexico. However, the Embayment subsided 60 million years after cessation of the sea-floor spreading in the Gulf. We have previously argued that the Mississippi Embayment formed as a result of the westward passage of faulted crust (Mississippi Valley graben) over the Bermuda hotspot in mid-Cretaceous. More recently published age data clarify age progressive (northwest-to-southeast) mid-Cretaceous volcanism that crosses the Mississippi Embayment, beginning ∼115 Ma in eastern Kansas and ending ∼65 Ma in central Mississippi. This line of volcanism coincides with the predicted Bermuda hotspot path and has isotopic signatures consistent with a mantle hotspot source. We propose that during mid-Cretaceous, the weak crust of the Mississippi Valley graben complex was uplifted 1–3 km as it passed over the Bermuda plume, and this upland was eroded. As the Mississippi Valley graben complex moved west of the hotspot, it subsided, and the eroded region became a topographic low that filled with fluvio-marine sediments, the Mississippi Embayment. Supporting evidence for mid-Cretaceous uplift and erosion of the Embayment region includes: (1) an angular unconformity on pre-Late Cretaceous rocks with ∼2 km eroded at mid-Cretaceous along the hotspot path; (2) a broad anticline in the Embayment at mid-Cretaceous (revealed by unfolding the down-warped basal Late Cretaceous unconformity); (3) exhumation and weathering of mid-Cretaceous plutons before burial by Late Cretaceous sediments; and (4) a mid-Cretaceous change in the northern part of the Gulf of Mexico sedimentation from a continuous carbonate platform to a large influx of deltaic clastics. We now suggest that magmatic activity and pronounced uplift in the Mississippi Valley graben region may have been a result of increased hotspot flux of the typically weak Bermuda hotspot during the Cretaceous superplume mantle event (∼120–80 Ma).

Hotspot origin of the Mississippi embayment and its possible impact on contemporary seismicity

Abstract Previous authors have related the Late Cretaceous/early Tertiary subsidence of the Mississippi embayment to the opening of the Gulf of Mexico, but the Gulf opened earlier in Triassic/Jurassic time. We offer an alternative hypothesis that development of the embayment was coeval with the passage of the Mississippi Valley graben system over the Bermuda hotspot about 90 Ma. Several lines of evidence of significant uplift of the embayment axis accompanying mid-Cretaceous magmatism and prior to Late Cretaceous subsidence support this proposal. First, reactivation of the Pascola arch in the northern embayment is recorded by flanking deposits of basal Upper Cretaceous gravel. Second, beneath a regional mid-Cretaceous unconformity, subcrops of Jurassic and Early Cretaceous strata define a pronounced southwest-plunging arch in the southern embayment. This arch is collinear with an arch revealed in Paleozoic rocks after restoration to mid-Cretaceous structural geometries. Third, a deep weathering profile on mid-Cretaceous alkalic plutons along the western embayment margin is nonconformably overlain by Paleocene sediments, and rapid mid-Cretaceous cooling of these intrusions has been interpreted from apatite fission tracks. Moreover, exploratory holes along the embayment axis encountered similar weathered alkalic intrusions nonconformably overlain by basal Upper Cretaceous strata. Fourth, there was an anomalous influx of clastic sediment into the northern Gulf of Mexico during mid-Cretaceous time, and subsequent clastic facies patterns suggest the Mississippi River drainage began to enter the Gulf in the Late Cretaceous. Passage of the Mississippi Valley graben over the Bermuda hotspot during elevated hotspot activity of Cretaceous time may have significantly weakened the previously rifted lithosphere. Rifted continental margin at Charleston, South Carolina, also passed over this hotspot in latest Cretaceous time. Similarly, the St. Lawrence rift system passed over the Great Meteor hotspot during the Cretaceous. It is important to note that these rift systems are the principal loci of strong seismicity in eastern North America, and thus weakening by increased Cretaceous hotspot activity may be an important common factor for these seismic rift zones.

Identification of possible Quaternary deformation in the northeastern Mississippi Embayment using quantitative geomorphic analysis of drainage-basin asymmetry

To investigate neotectonism in the Mississippi Embayment east of the New Madrid seismic zone, we identified geomorphic domains that show evidence of ground tilting during Quaternary time. Transverse basin profiles were converted to two-dimensional vectors that denote channel position with respect to basin divides. These basin-asymmetry vectors record the net direction and degree of lateral migration of trunk streams. More than 2500 vectors were measured and spatially averaged within 400 km 2 bins. This field of 300 mean vectors delineates several domains that show preferred directions of stream migrations possibly driven by ground tilting. The timing of stream migration was interpreted using across-valley distributions of Quaternary alluvial terraces. Comparison of our mean vector field with subsurface structures suggests that some domain boundaries may be related to reactivated faults. Late Quaternary activity is suggested for two northeast-striking faults of the southeastern Reelfoot Rift margin. We acquired two seismic profiles showing near-surface faulting beneath scarps that follow the domain boundary associated with one of these northeast-striking faults (Big Creek fault zone). Reelfoot thrust seismicity ends on the south against this fault, suggesting that the rift margin has dextral slip accommodating northeastward movement of the thrust hanging wall. Our vector field also suggests late Quaternary movement on the Reelfoot thrust and on two other northwest-striking faults, here termed the Hatchie River fault and the Wolf River fault. Several other weak domains may imply minor elements of neotectonism. Our results demonstrate that morphometric analysis of drainage-basin asymmetry can be an effective reconnaissance tool within neotectonic settings.

Reelfoot rift and its impact on Quaternary deformation in the central Mississippi River valley

Geophysical and drill-hole data within the Reelfoot rift of Arkansas, Tennessee, Missouri, and Kentucky, USA, were integrated to create a structure contour map and threedimensional computer model of the top of the Precambrian crystalline basement. The basement map and model clearly defi ne the northeast-trending Cambrian Reelfoot rift, which is crosscut by southeast-trending basement faults. The Reelfoot rift consists of two major basins, separated by an intrarift uplift, that are further subdivided into eight subbasins bound by northeast- and southeast-striking rift faults. The rift is bound to the south by the White River fault zone and to the north by the Reelfoot normal fault. The modern Reelfoot thrust fault, responsible for most of the New Madrid seismic zone earthquakes, is interpreted as an inverted basement normal fault. Geologic interpretation of 5077 shallow borings in the central Mississippi River valley enabled the construction of a structure contour map of the Pliocene‐Pleistocene unconformity (top of the Eocene‐base of Mississippi River alluvium) that overlies most of the Reelfoot rift. This map reveals both river erosion and tectonic deformation. Deformation of the Pliocene‐Pleistocene unconformity appears to be controlled by the northeast- and southeast-trending basement faults. The northeast-trending rift faults have undergone and continue to undergo Quaternary dextral transpression. This has resulted in displacement of two major rift blocks and formation of the Lake County uplift, Joiner ridge, and the southern half of Crowley’s Ridge as compressional stepover zones that appear to have originated above basement fault intersections. The Lake County uplift has been tectonically active over the past ~2400 yr and corresponds with a major segment of the New Madrid seismic zone. The aseismic Joiner ridge and the southern portion of Crowley’s Ridge may refl ect earlier uplift, thus indicating Quaternary strain migration within the Reelfoot rift.

Neotectonics of the southeastern Reelfoot rift zone margin, central United States, and implications for regional strain accommodation

The northeast-striking New Madrid fault system of central North America has been described as a right-lateral strike-slip system with a left-stepping restraining-bend thrust. The fault system has one of the highest rates of seismic energy release in an intraplate setting, and it has been regarded as a zone of significant earthquake hazard. The New Madrid fault system is the central part of the wider Reelfoot rift fault system, a northeast- striking basement fault zone. Although component faults of the New Madrid system are well defined by microseismicity, recent geodetic surveys suggest that little if any strain is accumulating on the principal (southern) strike-slip arm of the fault system. Seismicity and geologic data show that the “restraining-bend” thrust continues past the southern strike-slip arm to the southeastern Reelfoot rift margin. Thus, we suggest that earthquakes defining the southern arm of the New Madrid fault system are primarily aftershocks of an earthquake on that arm during the last sequence of great earthquakes (A.D. 1811– 1812), and it is the southeastern rift-margin fault system that is currently accommodating right-lateral strain along the boundary of the thrust block. This interpretation is consistent with recent geodetic results. The southeastern rift margin coincides with a 150-km-long linear topographic scarp from near Memphis to the Tennessee-Kentucky line, and S-wave reflection profiles, auger data, and a trench excavation reveal late Wisconsin–early Holocene surface faulting and late Holocene liquefaction associated with this fault-line scarp. Variation in sense of throw along strike and flower-structure geometry suggest that this is a strike-slip fault. Recognition of this rift margin as an important element of active tectonism in the Mississippi embayment has broad implications for assessment of the seismic hazard of this and similar intraplate settings. Temporal shifts in strain accommodation may give rise to short-term seismicity patterns and/or geodetic velocities that do not reveal long-term tectonic patterns.

Tectonic geomorphology of the southeastern Mississippi Embayment in northern Mississippi, USA

We analyzed drainage-basin geometry in part of the southeastern Mississippi Embayment near the New Madrid seismic zone in order to detect areas of lateral stream migration that could indicate recent tectonism. Toward this goal, a field of drainage-basin asymmetry vectors was generated from a digital terrain model and compared to geologic structure, lithofacies, seismicity, remotely sensed lineaments, and major drainage divides. Transverse topographical drainage-basin asymmetry (T-index) results define geomorphic domains in northern Mississippi (33° to 35°N, 88.5° to 90.25°W) that may reflect deep crustal blocks bounded by active faults or flexures. T-index measurements (a proxy for lateral stream migration) of more than 4500 second-order drainage basins produced a two-dimensional field of 282 spatially averaged vectors from which we interpreted 12 domains. Some domain boundaries correspond to mapped faults of two Paleozoic fault systems in our study area, the northeast-striking Mississippi Valley graben system and the northwest-striking Ouachita thrust belt and foreland faults. Several other prominent domain boundaries may be previously unmapped faults (flexures?) parallel to these systems. This interpretation is supported by satellite image analysis of the northeastern study area, which reveals lineaments consistent with our domain boundaries that have strikes that reflect the regional subsurface structural grain.

Quaternary faulting in the southern Mississippi embayment and implications for tectonics and seismicity in an intraplate setting

The recently recognized Saline River fault zone in the southwestern Mississippi embayment (strike = 135°) is characterized by moderate historic seismicity. We document paleoseismicity (possibly strong) on the Saline River fault zone that is outside the known region of neotectonism in the northern embayment. Six surface faults were excavated within the Saline River fault zone, and all faults displace marine Eocene units and fluvial Pliocene–Pleistocene units; five show post-Wisconsin loess movement (thermoluminescence [TL] age = 23.6 ± 3 ka), and three deform middle to late Holocene eolian silt (TL age = 5.1 ± 0.6 ka and 3.6 ± 0.5 ka) as fault-tip flexures. Along the principal excavated fault, fault-plane braiding and plunging drag folding suggest a strong component of left slip, and subsidiary faults show both normal and reverse displacement (<1m) of Pliocene−Pleistocene, Wisconsin, and Holocene units. We interpret these faults as components of a left-slip flower structure. A shallow seismic reflection profile acquired across the principal fault and a subsidiary fault shows Eocene stratigraphy separated in a reverse sense below 120 m depth and separated in a normal sense above 120 m, supporting a strike-slip interpretation. In addition, the only focal plane mechanism available in the area indicates left slip with a small reverse component. We conclude that the Saline River fault zone is a Paleozoic–Mesozoic basement fault zone reactivated as a left-slip system in a Quaternary east-west compressive stress field. Field evidence suggests that earthquakes with several meters of strike slip may be characteristic, and the width of a liquefaction field on the Saline River fault zone suggests an event of magnitude 5.5–6. Considering a similarity of strike and sense of Holocene movements of the Meers fault of southern Oklahoma and of the Saline River fault zone, there may be additional active fault zones that pose seismic hazards concealed by sediments in southern North America.

Preliminary Assessment of Sand Blows in the Southern Mississippi Embayment

Trenching of sand blows in two liquefaction fields on late- to mid-Holocene alluvium within the southern Mississippi Embayment was conducted to determine the origin of the sand blows and their possible relationship to New Madrid seismic zone events. Trench stratigraphy and crosscutting relationships show evidence of multiple events of seismically induced sand venting. These two fields of moderate to intense liquefaction are 175 km and 240 km southwest of the southern limit of similarly intense sand blows within the New Madrid seismic zone (NMSZ) liquefaction field. The southernmost field (Ashley County, Arkansas) is ∼25 km in diameter, and sand blows are densely spaced. Sand blows are more sparsely spaced in the larger (∼30 km by 45 km) northern field (Desha County, Arkansas). Various correlations of liquefaction events between these two fields and between each of these fields and the NMSZ can be accommodated by our radiocarbon and infrared stimulated luminescence (IRSL) ages of sediments predating and postdating vented sand deposits in our trenches. These ages show a major liquefaction event at both the Ashley and Desha County fields between 4600 and 5500 years ago and a major event at the Ashley County field approximately 700 years ago. The early venting episode is older than the documented NMSZ chronology, and so a NMSZ seismic source cannot be assessed. The age of the later venting episode falls between two postulated NMSZ events, so it may reflect seismicity from a different source zone. Our trench logs show at least three other moderate to minor sand-venting episodes at the Ashley County field and two at the Desha County field, and one may be the same event at both fields about 2200 years ago. If these sand blows formed during large NMSZ events (≥175 km northeast), the area of strong ground motion for this seismic source has been underestimated. Alternatively, if these liquefaction fields formed due to local seismicity, then additional significant seismic source zones are present in the Mississippi Embayment outside the NMSZ. Each of the two fields sits astride seismicity alignments associated with the margins of a previously identified active fault block between the Arkansas River fault zone that projects through the Desha County field and the Saline River fault zone that projects through the Ashley County field.

Crustal velocity structure associated with the eastern Tennessee seismic zone: Vp and Vs images based upon local earthquake tomography

We present three-dimensional P and S wave velocity models for the active eastern Tennessee seismic zone (ETSZ) using arrival time data from more than 1000 local earthquakes. A nonlinear tomography method is used that involves sequential inversion for model and hypocenter parameters. We image several velocity anomalies that persist through most of the inversion volume. Some anomalies support the presence of known features such as an ancient rift zone in northern Tennessee. Other anomalies reveal the presence of basement features that can be correlated with regional gravity and magnetic anomalies. We image a narrow, NE-SW trending, steeply dipping zone of low velocities that extends to a depth of at least 24 km and is associated with the vertical projection of the prominent New York-Alabama magnetic lineament. The low-velocity zone may have an apparent dip to the SE at depths exceeding 15 km. The low-velocity zone is interpreted as a major basement fault juxtaposing Granite-Rhyolite basement to the NW from Grenville southern Appalachian basement to the SE. Relocated hypocenters align in near-vertical segments suggesting reactivation of a distributed zone of deformation associated with a major strike-slip fault. We suggest that the ETSZ represents reactivation of an ancient shear zone established during formation of the super continent Rodinia.

Seismicity of the New Madrid Seismic Zone Derived from a Deep-Seated Strike-Slip Fault

A conceptual three-dimensional flower structure model of strike-slip faulting is proposed to explain the occurrence of earthquakes in the New Madrid seis- mic zone (NMSZ) and to illustrate the potential rupture faults for the 1811-1812 earth- quake sequences. The proposed NMSZ model is based on elastic dislocation theory and concepts of material failure under a stress field. Using a conceptual model of a strike-slip subsidiary fault array, we identify tectonic features (geological structures) that are oriented properly relative to regional stresses and classify the regions where stresses might be expected to be amplified. The brittle upper crust in the vicinity of the NMSZ is modeled as a uniform over- burden with a horizontal-basal surface, which rests on a horizontal ductile lower crust that is cut by a vertical, northeast-striking right-lateral strike-slip shear zone. We acknowledge that many favorably oriented preexisting faults have been exploited as components of the flower structure. The brittle overburden material is subject to sim- ple shearing stress parallel to the deep-seated lower crustal shear zone, and preexisting faults of the Reelfoot rift system give the upper crust a mechanical anisotropy (planes of weakness striking northeast) that is the correct orientation for development of P shear faults. The deep-seated fault movement deforms the overlying upper crust that controls the structural geometry, the modern seismicity, and the large earthquake sequences in the NMSZ. The three-dimensional NMSZ model offaulting developed in this study shows that the Bootheel and Big Creek lineaments, inferred to be two subparallel P shear faults rooted in a deep-seated fault in the lower crust, are significant in shaping the geometry of the NMSZ. These series of faults produce a large-scale flower structure in cross section. The proposed NMSZ model uses the intersections of the deep-seated fault and the two subparallel P shear faults for the locations of the 1811 and 1812 earth- quakes. The model gives rise to a predictable pattern of surface deformation that is in good agreement with the observed seismicity patterns in the region.