Geomorphology, active tectonics, and landscape evolution in the Mid-Atlantic region

This field trip partially or wholly traverses four regional geologic and physiographic provinces of the Commonwealth of Virginia (Fig. 1). From west to east, these are the Ridge and Valley (Day 1, Stops 1-3); Blue Ridge (Day 1, Stop 4); Piedmont (Days 1 and 2, Stops 5-11), and innermost Coastal Plain along the Fall Zone (Day 3, Stop 12). The Ridge and Valley along with the Blue Ridge are part of the Appalachian foreland, with the Blue Ridge being the farthest-traveled crystalline thrust sheet to be included in the deformation of the eastern part of the Appalachian basin. In contrast, the Piedmont is part of the former Appalachian hinterland that includes rocks that were buried to great depth during orogenesis, and which experienced extension and igneous intrusions during the opening of the Atlantic Ocean in the Mesozoic. A relatively thin veneer of unconsolidated Coastal Plain sediments has onlapped the Piedmont to varying degrees during late Mesozoic-Cenozoic subsidence. The current Coastal Plain is restricted to that region east of the Fall Zone.

Virginia is part of the mid-Atlantic eastern North American (ENAM) passive margin intraplate setting that includes the Appalachian Mountains, Appalachian foreland, and the archetype Atlantic rift-passive margin (Fig. 2) (reviewed in Sheridan and Grow, 1988; Faill, 1997a, 1997b; Wilson, 1966; Oliver et al., 1983). Geologic heterogeneity of the Atlantic passive margin lithosphere is both the result and consequence of diverse tectonic events experienced over the past billion years, including Grenville and Appalachian compressive orogenesis (Faill, 1997a, 1997b), rifting and opening of the Atlantic Ocean (Withjack et al., 2012; Schlische et al., 2002), massive igneous activity associated with the Central Atlantic Magmatic Province (CAMP; Marzoli et al., 1999; Hames et al., 2003; Hutchinson, 2005), and unsteady Cenozoic epeirogeny (Pazzaglia and Brandon, 1996; Gallen et al., 2013). The passive margin geodynamic evolution is superposed on a lithosphere that marks the transition from fully continental Precambrian Grenville basement in the west, to fully oceanic Jurassic Atlantic sea floor in the east (e.g., Nettles and Dziewonski, 2008; Grand, 1994; van der Lee, 2001; van der Lee and Frederiksen, 2005; Darbyshire et al., 2007).

The Appalachians are underlain by relatively thin (~80 km) lithosphere (van der Lee, 2002; Rychert et al., 2005, 2007). Recent tomography of the shallow upper mantle constructed from USArray teleseismic data reveals two low velocity zones in eastern North America, one in New Hampshire and the other in west-central Virginia (Schmandt and Lin, 2014), both of which are coincident with the most recent episodes of volcanism on the passive margin in the Cretaceous (Chu et al., 2013) and Eocene (Mazza et al., 2014), respectively. Our field-trip route is fully within the west-central Virginia mantle low velocity zone, and that coincidence may have links to post-orogenic tectonic and geomorphic processes we are investigating during the trip.

Appalachian crust varies in thickness from ~47-56 km under the high standing parts of the topography, specifically the Blue Ridge, and generally thins eastward into the Piedmont (Hawman et al., 2012; Wagner et al., 2012); however, crust up to 60 km thick underlies parts of the western Piedmont in the southern Appalachians (Wagner et al., 2012). The preservation and non-uniform buoyancy of these crustal roots might be explained by long time scale-dependent changes in the metamorphic grade in the deep crust (Fischer, 2002), and to varying degrees, it is possible that the deep crustal roots may predate the formation of the Appalachians. Bouguer and free-air gravity maps of eastern North America (Simpson et al., 1987) indicate a major gravity low that roughly coincides with the NE-SW strike of the highest-standing Appalachian topography along the drainage divide (Fig. 2). This gravity low, traditionally interpreted to represent an Appalachian crustal root, has a mostly flat isostatic residual, consistent with compensated topography (Simpson et al., 1986); however, there remain several isostatic residual highs that have been interpreted as crust or lithosphere of varying density (Simpson et al., 1986). Long-term dynamic or static crustal buoyancy could be envisioned to have a major impact on mean topography and, correspondingly, the geomorphic processes driving movement of the drainage divide and major rivers that we will be investigating at Stops 1 and 2.

Much of the eastern North American lithosphere and crust is the product of the Proterozoic Grenville orogeny associated with the assembly of the Rodinia supercontinent, and crust of this age constitutes the basement of the modern passive margin (Whitmeyer and Karlstrom, 2007; Thomas, 2006). Opening of Iapetus in the late Proterozoic resulted in two continental rifts, the Catoctin rift and Rome trough that ultimately subsided and were covered by a thick wedge of passive margin siliciclastics and carbonates. We will see these rocks and detritus eroded from them at Stops 3, 4, and 5. Low-magnitude, but persistent seismicity (Fig. 2) remains concentrated along the structures that define the rifts.

The Appalachians were constructed on top of the Catoctin rift-Rome trough system and passive margin following a protracted period of collisional tectonics during the Paleozoic and closing of Iapetus by the Permian (Hibbard et al., 2006; Fig. 3). The traditional interpretation holds that the three great clastic wedges preserved in the Appalachian foreland are related to three pulses of orogenesis during the Paleozoic-Late Ordovician Taconic orogeny, the Late Devonian Acadian orogeny, and the Pennsylvanian-Permian Alleghenian orogeny.

By the Early Permian (ca. 280 Ma), the Appalachians were a lofty mountain chain modeled to be similar in mean elevation, relief, and width to the modern central Andes (Slingerland and Furlong, 1989). Deformation during the Alleghenian orogeny propagated far westward into the foreland, imbricating much of the former foreland basin, and shedding a thick molassic wedge west across the craton (Riggs et al., 1996; Rahl et al., 2003; Hegarty et al., 2007) that exhumed the hinterland. That wedge was responsible for up to 11 km of burial in the anthracite fields of eastern Pennsylvania (Levine, 1986), 4 km in central West Virginia (Reed et al., 2005), and up to 2 km of burial in the midcontinent (Hegarty, et al., 2007). In the hinterland, Mesozoic rift basin sediments fill and onlap a Piedmont basement that was exhumed to the same structural level in both the hanging wall and footwall of the rift-margin faults, suggesting that the former lofty Appalachian topography had been reduced to several hundred meters or less of local relief when rifting began ca. 200 Ma.

Extension during the Mesozoic reactivated many of the preexisting Grenville and Appalachian structures, producing a series of wide, deep fault-bounded rift basins from northern Florida to the Grand Banks of Canada (Olsen et al., 1996; Withjack et al., 1998; Withjack and Schlische, 2005; Faill, 2003; Thomas, 2006). Rifting was roughly coincident with one of the most voluminous but short-lived volcanic events in Earth’s history, the Central Atlantic magmatic province, or Central Atlantic magmatic province volcanics and intrusives. After rifting, the rift basins underwent significant erosion (locally >5 km) (e.g., Malinconico, 2010). Much of this erosion occurred soon after breakup, producing a pronounced unconformity between the synand postrift rocks. Additionally, significant deformation occurred after rifting, folding, and tilting the synrift strata (e.g., Withjack et al., 1998). Like their post-Rodinian predecessors, low-magnitude, but persistent, seismicity (Fig. 2) is located along the flanks of the rift basins today (Seeber and Armbruster, 1988; Wheeler, 2006; Thomas, 2006). Although not a stop on this field trip, much speculation surrounds the evidence for offset of Quaternary deposits along the Mountain Run fault (Pavlides, 1994; Bobyarchick, 2015), one of the rift basin bounding faults of the Culpepper basin in central Virginia.

Rifting rejuvenated the topography and opened up new basins to the east of the foreland, which initiated a reversal of Appalachian drainage from the Paleozoic slope towards the foreland (west) to one that was split between the old west-flowing rivers and the newly formed Atlantic slope (east-flowing) drainages (Judson, 1975). The formerly low-standing Appalachian basin became a relatively high-standing region and portions of the foreland and Blue Ridge experienced a new pulse of erosion during the Late Jurassic and Early Cretaceous (ca. 140-150 Ma), also recorded by apatite fission track cooling ages (Miller and Duddy, 1989; Roden and Miller, 1989) and delivery of siliciclastic detritus to Atlantic shelf-slope basins (Poag, 1985, 1992; Poag and Sevon, 1989; Figs. 2B and 2C).

Correspondingly, the synand postrift geologic, tectonic, and geodynamic development of eastern North America is preserved as a sedimentologic and stratigraphic archive in several shelf-slope basins (Fig. 2). The long-term depositional and subsidence history of the 400-km-long, 100-km-wide, and up to 18-km-deep Baltimore Canyon trough has been particularly well studied (Karner and Watts, 1982; Poag, 1985, 1992; Poag and Sevon, 1989; Steckler et al., 1988, 1999). Collectively, the Baltimore Canyon trough contains siliciclastic sediment equivalent to ~4 km of rock (Hulver, 1997) removed from an area spanning the modern central and New England Appalachian Atlantic slope. We will see the feather edge of this siliciclastic wedge on the inner Coastal Plain at Stop 12.

The field trip begins in the Valley and Ridge geologic province (Ridge and Valley physiographic province) at Blacksburg, Virginia. This province is characterized by clastic and calcareous sedimentary rocks of Cambrian to Mississippian age (Fig. 4) that collectively record the transition from an early Paleozoic passive margin to middle-late Paleozoic convergent margin culminating in Alleghenian orogenesis (Butts, 1940; Rodgers, 1970; Hatcher, 2010). Alleghenian foreland deformation affecting these rocks including cleavage, folds, regional faults, and variable anchizone metamorphism is the consequence of emplacement of the Blue Ridge-Piedmont megathrust sheet to the east (Hatcher, 2002).

Throughout the drive from Stops 1 through 4, we will view the classic Appalachian topographic fabric arising from the longterm dynamic adjustments of geomorphic form to process operating on rocks of variable resistance and structure. For example, southwest of Blacksburg, interbedded sandstone and siltstone with 3-15 m-thick beds of quartz-pebble conglomerate comprise the lower part of the Mississippian Price Formation and hold up Price Mountain (Bartholomew and Lowry, 1979). These rocks, in the core of the Price Mountain anticline, are exposed in the Price Mountain window beneath the Staunton-Pulaski fault; carbonates and shale of Cambrian Rome-Waynesboro and Elbrook formations comprise the hanging wall of the thrust sheet, and underlie the rolling hills to the north and south (Fig. 5). Northward from Roanoke, the field-trip route through the Valley and Ridge generally follows Cambrian to Ordovician carbonates, and the topography is dominated by rolling hills and karst. The exception is Purgatory Mountain, near Buchanan and the James River. Here, Ordovician Martinsburg Formation cores an anticline beneath the Staunton-Pulaski fault, with Silurian Tuscarora sandstone comprising the limbs (Spencer, 1968).

Throughout much of Virginia, the Blue Ridge-Valley and Ridge boundary is mapped as a major décollement (Harris and Milici, 1977; Evans, 1989). South of the James River, the Blue Ridge fault serves as the structural contact between the provinces, with Cambrian Shady-Tomstown and Rome-Waynesboro formations in the footwall. North of the James River, these units are mapped in stratigraphic continuity with cover and basement rocks of the Blue Ridge (e.g., VDMR, 1993), and the significant structural transition occurs at the Midvale-South River fault, with Cambrian Elbrook Formation in the footwall (Wilkes et al., 2007).

Stuarts Draft (Stop 3), ~15 km southeast of Staunton, Virginia, is situated at the boundary between the Blue Ridge and Valley and Ridge, but the nature of the geologic contact between the provinces is obscured. Here, the oldest formations of the Valley and Ridge, the Cambrian Shady-Tomstown and Rome-Waynesboro formations, are mostly covered by thick, north-facing alluvial fans at the foot of the mountain front coalesced into a bajada piedmont. Bailey et al. (2002) suggest stratigraphic continuity from Blue Ridge to Valley and Ridge strata along an east-west contact south of Stuarts Draft and beneath the bajada, although northeast-southwest segments of the boundary remain faulted. In contrast, Heller et al. (2014) state that there is no evidence for a significant reverse fault at the contact, but a transverse fault is one of several possible structural solutions. Detailed mapping (Carter and Heller, unpublished, in progress USGS mapping) is inconclusive; the nature of the boundary here will likely remain ambiguous without significant boring and correlation, or seismic profiling.

The Blue Ridge, the topic of Stop 4, is cored by Mesoproterozoic basement rocks representing the original crust of the Laurentian (ancestral North American) continent. This basement is overlain by a thick sequence of metasedimentary and metavolcanic cover rocks during extension and breakup of the Rodinian supercontinent in the Neoproterozoic to earliest Cambrian. All Blue Ridge rocks were significantly deformed during multiple episodes of Paleozoic orogeny.

The Blue Ridge province changes significantly in topographic form from north to south in Virginia (Figs. 1 and 5). In northern Virginia, Blue Ridge topography consists of a high (maximum elevation of ~1230 m) NE-trending linear ridge, just a few kilometers wide, which in effect is the easternmost range of the Valley and Ridge province (Hack, 1982). Between this main ridge, and the low-lying topography of the Piedmont to the east, is a series of lower-elevation, NE-trending foothill ridges. South of Roanoke, however, the Blue Ridge is a broad upland plateau, 10s of kilometers wide, with a steep southeast-facing escarpment known locally as the “blue wall.” This change in topographic expression is due in part to the fact that the rocks and structures associated with the hanging wall of the Blue Ridge persist east into what is typically recognized as Piedmont topography, but also in part to the location of the drainage divide between the west-flowing Mississippi drainage streams and the east-flowing Atlantic slope streams, the topic of our investigations at Stop 1.

In Maryland and northern Virginia, the Blue Ridge geologic province is defined as an orogen-scale, northwest-vergent, northeast-plunging reclined anticlinorium (Cloos, 1947; Griffin, 1971). Here, Neoproterozoic to Cambrian metamorphosed sedimentary and volcanic “cover” rocks of the Swift Run, Catoctin, Unicoi, and Chilhowee Group formations comprise west limb stratigraphy, proximal to Mesoproterozoic “basement,” including orthogneisses and metamorphosed granitoid rocks of the Shenandoah massif (Rankin, 1976) in the core (Fig. 4). Eastern limb stratigraphy includes rocks of the Catoctin Formation and Lynchburg Group (Nelson, 1962). Map- to outcrop-scale folds parasitic to the regional anticlinorium deform the basement-to-cover stratigraphic sequence. It is the regional attitude of resistant rocks of the west limb cover sequence—Chilhowee Group quartzites and Catoctin basalt—that forms the high linear backbone of the “Blue Ridge” at these latitudes (Fig. 5).

Regional Paleozoic faults also cut and duplicate elements of the anticlinorium in the region of this field trip. The Rock-fish Valley high-strain zone (Fig. 5) is an ~1- to 6-km-wide, northeast-trending zone of greenschist-facies mylonite and protomylonite that extends more than 200 km near the axial trace of the anticlinorium (Bartholomew et al., 1981; Bailey and Simpson, 1993; Southworth et al., 2009). Younger brittle faults of the Blue Ridge thrust system root into the high-strain zone from the west as proposed by Spencer (1995). Bartholomew (1983) suggested nearly 50 km of Paleozoic displacement across the Rock-fish Valley zone, but Evans (1991) noted that displacement must diminish to the northeast, toward the terminus of the anticlinorium in northern Virginia and Maryland.

Very locally, diabase dikes of earliest Jurassic age (Kunk et al., 1992) intrude Blue Ridge rocks. These dikes were emplaced in the Blue Ridge during continental extension and the opening of the Atlantic Ocean in the Mesozoic.

South of Roanoke and Stop 1, the Blue Ridge geologic province quickly transitions from an anticlinorium to a stack of imbricated thrust sheets (Hatcher, 1978, 1989). In the western Blue Ridge, Chilhowee Group cover rocks are through-going, and overlie isolated inliers of Mesoproterozoic basement rocks (Rankin et al., 1972). The eastern Blue Ridge is a fault-bounded geologic terrane comprised of high-metamorphic-grade sedimentary and volcanic rocks deposited east of the Laurentian continental margin from the Neoproterozoic to early Paleozoic (Espenshade et al., 1975). These rocks were significantly metamorphosed, deformed, and transported westward onto the Laurentian margin along major faults during Paleozoic orogenesis. The high plateau of the Blue Ridge in southern Virginia and northwestern North Carolina has developed above these rocks, and it is through these rocks that the Blue Ridge escarpment continues its westward migration via erosional dissection (Spotila et al., 2004).

Given the location of the 2011 Mineral, Virginia, earthquake, the geologic framework and Paleozoic to neotectonic history of the central Piedmont province has drawn much new focus. The Virginia Piedmont (Fig. 6), at the latitude of field-trip Stops 6 through 11, is a Paleozoic amalgamation of complex terranes and structures of varying age sutured to North America, and deformed during the building of the Appalachian Mountains (Horton et al., 1989; Hibbard et al., 2006). From west to east, these include the Mountain Run fault zone, the Potomac terrane, the Chopawamsic terrane, the Spotsylvania lineament, and the Goochland terrane. Other tectonstratigraphic elements include suites of Silurian-Ordovician to Carboniferous felsic to intermediate plutons and post-Taconian successor basin synclinoriums. Mesozoic diabase locally intrudes these Piedmont rocks and structures, and rift basins flank the Piedmont both to the east and west. In this summary, we focus on the major terranes and their bounding faults with the goal of providing a context for which crustal structures may be seismogenic and linked to the Mineral earthquake (Shah et al., 2015) and related CVSZ earthquake activity.

The Mountain Run fault zone is a NE-trending, variably wide (< ~1 km) zone of Paleozoic ductile shear that defines the geologic boundary between the Blue Ridge and Piedmont provinces at the latitude of our field trip (VDMR, 1993; Bailey et al., 2006; Figs. 5 and 6). It is continuous with the Brevard fault zone of the Carolinas, forming part of an orogen-parallel 1000-km- long multi-deformed transprovince shear zone (Bobyarchick, 1999a, 1999b, 1999c). Overprinting the Paleozoic NW-directed thrust (Conley, 1987) or dextral oblique structure (Bobyarchick, 1999b) is evidence for down-to-southeast Mesozoic extension (Evans and Milici, 1994) and near Everona, Virginia (Fig. 6), a set of NW-dipping reverse faults that juxtapose saprolite over Pliocene or younger stream gravel and associated colluvium, with ~1.5 m of vertical offset (Pavlides et al., 1983; Crone and Wheeler, 2000; Bobyarchick, 2015). The Everona fault has no geomorphic expression (Pavlides et al., 1983; Manspeizer et al., 1989), but two sets of scarps along the margins of the strikingly linear valley of Mountain Run may be related to Quaternary faulting (Crone and Wheeler, 2000); Pavlides (1994) argued that the rugged topography of both scarps indicates relative youth.

Metaclastic rocks of the Potomac terrane (Drake, 1989; Horton et al., 1989) comprise the hanging wall of the Mountain Run fault zone. These low-grade metagraywackes and phyllites locally contain exotic blocks of amphibolite, metagabbro, and altered ultramafic rocks, which led Pavlides (1989) to interpret the complex as a mélange. Intrusive plutons as old as ca. 473 Ma and detrital zircon grains as young as ca. 500 Ma indicate that at least some parts of the Potomac terrane were deposited in the Late Cambrian to Early Ordovician (Hughes et al., 2014a, 2014b). We will see one of the plutons that intrude the Potomac terrane at Stop 6, Day 2. Although the Potomac terrane is of Laurentian affinity (Hughes et al., 2014b), it is part of the Piedmont province because it was originally an accretionary complex, well outboard of the Laurentian margin.

The Chopawamsic Formation (Southwick et al., 1971) is a greenschist-facies (west) to amphibolite-facies (east) interlayered metavolcanic, volcaniclastic, and sedimentary unit (Burton et al., 2015a, 2015b), and interpreted to be an island-arc-accretionary prism sequence outboard of the Laurentian margin (Pavlides, 1981; Horton et al., 1989; Hughes et al., 2014c). High-resolution U-Pb zircon TIMS (thermal ionization mass spectrometry) geochronology on metavolcanic and metaigneous rocks from the formation (Coler et al., 2000; Horton et al., 2010; Hughes et al., 2013, 2014c) indicates an Ordovician age between ca. 470 and ca. 450 Ma. The boundary between the Chopawamsic Formation and Potomac terrane rocks to the west is the Chopawamsic fault, an obvious target of investigation as a seismogenic structure in the CVSZ. The 440 Ma Silurian Ellisville Granodiorite intrudes both terranes and stitches the Chopawamsic fault (Hibbard et al., 2010; Hughes et al., 2013; Hibbard and Karabinos, 2013), which agrees with regional cross-cutting relationships and constrains the timing of arc accretion to the Ordovician. Alleghenian deformation (Burton et al., 2015b) folded rocks of the Chopawasmic Formation and overlying Quantico Formation, the folds being expressed in the subtle contrasts in lithologic resistance and topographic grain of the central Virginia Piedmont, illustrated well in LiDAR DEMs (light detection and ranging digital elevation models) flown shortly after the 2011 earthquake (Fig. 7).

Unconformable above Chopawamsic Formation rocks are quartzite, mica schist, and rare metafelsite of the Ordovician (ca. 448 Ma, Horton et al., 2010) Quantico Formation (Burton et al., 2014, 2015a). Its southern equivalent, the Arvonia Slate, is interpreted as a Silurian successor basin (Stose and Stose, 1948; Smith et al., 1964) because it rests unconformably above the ca. 444 Ma Carysbrook pluton (Sinha et al., 2012). Bailey et al. (2008) report detrital zircons from the Arvonia as young as 390 Ma, but Hughes et al. (2013) have now dated the Ellisville pluton as young as ca. 437 Ma and interpreted the Quantico Formation similarly as a Silurian successor basin. However, the Horton et al. (2010) date suggests that the Quantico is not a successor basin, as its age cannot be older than the plutonic rock on which it should unconformably rest. In the context of the Mineral earthquake and field-trip Stops 6 through 11, the seismogenic significance of this debate is clear: If the Quantico Formation is a Silurian successor basin, then it is a shallow crustal feature occupying a surface synform that does not extend to great depth. In contrast, if it is a quartzite-mica schist unit within the Chopawamsic Formation, its contacts with the Chopawamsic Formation are likely to be faults with an orientation and dip consistent with the fault plane that ruptured during the Mineral earthquake. Conceivably, compressive strain may be localized at the fault contact between the relatively incompetent phyllite of the Chopawamsic Formation and the more rigid quartzite of the Quantico Formation.

The 15-km-wide, NE-trending Spotsylvania fault zone forms the boundary between Chopawamsic terrane to the northwest and Goochland terrane to the southeast (Pavlides, 1989; Spears et al., 2004). First recognized as a sharp geophysical lineament (Neuschel, 1970; Zietz et al., 1977), it is now interpreted to be a zone of Paleozoic high strain with as much as 80-300 km SW-directed Paleozoic dextral transpressional displacement (Bailey et al., 2004), overprinted by Mesozoic brittle faults and marked by extensive siliceous breccia (Spears and Bailey, 2002). Amphibolite- to granulite-facies rocks of the Goochland terrane sit in the hanging wall of the Spotsylvania fault zone. From metamorphic and field relationships (Farrar, 1984) and geochemical similarities (Owens and Tucker, 2003; Owens and Samson, 2004), these rocks were thought to be a fragment of wholly Mesoproterozoic Laurentian continental crust, either overridden during Paleozoic orogenesis and re-exhumed during Mesozoic crustal extension (Glover et al., 1997; Farrar, 2001), or rifted but reattached during the Paleozoic (Bartholomew and Tollo, 2004; Owens and Samson, 2004). Although some rock units in the terrane are Mesoprotoerozoic, such as the ca. 1023-146 Ma State Farm Gneiss (Owens and Tucker, 2003) and the ca. 1045 Ma Montpelier Anorthosite (Aleinikoff et al., 1996), new geochronologic data suggest much of the terrane is Devonian. At least part of the regionally extensive Maidens Gneiss crystallized at ca. 405-384 Ma (Owens et al., 2004, 2010), with an overprint of amphibolite- to granulite-facies metamorphism at 380 Ma (Shirvell et al., 2004).

Atlantic Coastal Plain stratigraphy in eastern-central Virginia ranges from Cretaceous to Holocene (Mixon et al., 1989), and west of the Fall Line, these strata onlap a weathered paleotopography of the Piedmont. The Fall Line is the boundary, at the land surface, between the westernmost conterminous Coastal Plain units (mostly of estuarine to nearshore marine origin) and Piedmont rocks (Carter et al., 2007a). East of the Fall Line, Coastal Plain sediments form an eastward-thickening contiguous wedge of fluvial, deltaic, estuarine, and marine sediments that extends out to the continental shelf (Bayer and Milici, 1987). West of the Fall Line, mostly fluvial clay, sand, and gravel are the proximal, up-dip equivalents of Miocene to Pleistocene Inner Coastal Plain marine units to the east (Fig. 8). These innermost Coastal Plain units form discontinuous deposits that thin to the west for up to 30 km, capping higher elevations above Piedmont rocks, an example of which we will see at Stop 12. They serve as important stratigraphic and geomorphic markers for the numerous local and regional Cenozoic faults that here deform the easternmost Piedmont and innermost Coastal Plain.

Coastal Plain geology dictates application of two important stratigraphic tenets—superposition and geomorphic succession (Berquist and Goodwin, 1989). East of the Fall Line, Cretaceous to Neogene units follow basic stratigraphic superposition as older Cretaceous deposits (Potomac Group) of fluvial to deltaic origin are buried beneath younger marine to nearshore units of Paleogene and Neogene age. Pliocene to Pleistocene fluvial to shallow marine units deposited during sea-level regression (Bacons Castle, Windsor, and younger formations), however, follow principles of allostratigraphy as these units occupy stair-step terraces and plains marked by scarps and treads that are incised and etched into the landscape along the James and Chickahominy Rivers, and major tributaries (Fig. 8); thus, older units occupy positions higher on the topography than younger units. West of the Fall Line, the oldest fluvial gravel deposits cap the Midlothian upland (Johnson and Peebles, 1983) whereas younger gravels cap lower hills across the landscape, or are inset along the James River and its major tributaries (Carter et al., 2007a, 2007b; Fig. 8).

West of Richmond, sand and gravel of the up-dip nearshore-facies of the Pliocene Yorktown Formation (upper Chesapeake Group) abut the Chippenham scarp (Fig. 5; Thornburg scarp of Mixon, 1978), which separates the Midlothian upland from the Richmond plain (Johnson and Peebles, 1983) at an elevation of ~240 ft (~70 m) above sea level (Johnson and Peebles, 1984; Goodwin, 1980; Carter et al., 2007b). West of the scarp, gravel and sand deposits cap low hills and mantle slopes along the James and Chickahominy rivers, first-order major tributaries, and interfluves between these systems, at discrete elevations of 200 to 290 ft (61-88 m), 130 to 200 ft (40-61 m), and ~120 ft (~36 m) above present sea level. These deposits consist mostly of rounded quartz and quartzite pebbles, with many containing the trace fossil Skolithos, derived from the Cambrian Antietam Formation currently exposed only on the west flank of the Blue Ridge. The number of Skolithos-bearing clasts increases up section. The oldest gravels contain just a few Skolithos- bearing clasts whereas younger mid- and low-level gravels contain many more Skolithos-bearing clasts. Recent detailed mapping demonstrates lateral continuity of these deposits eastward onto the Richmond Plain and along the major streams and rivers with, respectively, Yorktown strata (upper strath), Pliocene Bacons Castle Formation (middle bench), and lower Pleistocene to upper Pliocene Windsor Formation (lower terrace), providing an absolute correlation between these innermost Coastal Plain units (Carter et al., 2007a). The presence and relative abundance of Skolithos-bearing clasts in these units factor into correlation and ultimately need to be explained in terms of unroofing of the Blue Ridge rocks where they are sourced in and around field-trip Stops 3 and 4.

Deposits of gravel and sand, generally interpreted to be fluvial to fluvial-deltaic in origin (Hack, 1955; Goodwin and Johnson, 1970; Goodwin, 1970, 1980; Weems, 1981, 1986), that cap the highest hills and the relatively flat Midlothian upland at elevations from 240 to more than 350 ft above present sea level west of the Fall Line have been variously referred to as the Appomattox (McGee, 1888), Columbia (Shaler and Woodworth, 1899), Lafayette (Shattuck, 1906; Darton, 1911), Brandywine (Clark, 1915; Wentworth, 1930), Citronelle (Doering, 1960), Midlothian (Mathews et al., 1965; Goodwin and Johnson, 1970), and Bon Air (Johnson et al., 1987) gravels. Pazzaglia (1993) correlated these gravels with the Bryn Mawr Formation in Pennsylvania. Several ages for these and similar highest-level gravels in the Mid-Atlantic region have been proposed, ranging from Cretaceous to Pleistocene (Darton, 1911; Wentworth, 1930; Hack, 1955; Goodwin and Johnson, 1970; Owens and Minard, 1979; Johnson et al., 1987; Pazzaglia et al., 1997), with most clustering about the Miocene.

Goodwin and Johnson (1970) noted the local occurrence of a significant dark-gray clayey silt beneath the highest-level gravels in the Richmond area. Though sparse, unidentified mollusk and bivalve fossils have been found in the unit, and more recently Weems et al. (2012) report late middle Miocene dino-flagellates, which places a maximum age of ca. 12 Ma on the overlying high-level gravels and a general down-dip correlation to the Choptank Formation.

The character and distribution of the Piedmont rocks beneath the thin Coastal Plain cover are important to understanding neo-tectonic deformation in the innermost Coastal Plain. From west to east, these Piedmont units include rocks of the easternmost Goochland terrane and Hylas fault zone, the Petersburg Granite, Mesozoic basin sedimentary and igneous rocks, and brittle fault zones that bound the basins and overprint older rocks and structures (Fig. 6).

The Hylas zone separates easternmost Goochland terrane rocks (Maidens Biotite Gneiss) from Petersburg Granite and rocks of the Richmond/Taylorsville Mesozoic basins to the east (Fig. 6). Mylonitic fabric in the biotite gneiss is the regional product of Hylas fault zone Paleozoic deformation, ductile dextral transpression, along this orogen-scale structure, which also exhibits significant younger Mesozoic to Cenozoic brittle overprint (Bobyarchick and Glover, 1979; Hollis and Bailey, 2012; Bailey et al., 2015).

Petersburg Granite underlies a large region of central-eastern Virginia (Fig. 6). Its outcrop belt is bounded to the north and west by the Richmond-Taylorsville Triassic basins and to the east by Coastal Plain deposits. Northeast- and northwest-trending joint sets developed in the underlying granite (Dailide and Diecchio, 2005) extend upward, and parallel sets developed in overlying innermost Coastal Plain units, particularly Miocene clayey silt both east and west of the Fall Line (Carter et al., 2007a). Swarms of chalcedony- and quartz-filled fractures, many with slickenlined surfaces, also occur throughout the granite. Some of these fractures show evidence for either influencing deposition of younger gravel deposits (zones of scour and deposition along these fractures), or reactivation and offset (Carter et al., 2007b).

The Mesozoic Richmond-Taylorsville basin is one of more than two-dozen fault-controlled rift basins filled with fluvial to lacustrine Newark Supergroup sedimentary rocks, conglomerate, sandstone, shale, and coal, that formed along the eastern North American margin during Mesozoic continental extension. The structure of the Richmond-Taylorsville basin is generally considered to be a half-graben or tilted fault block (Wilkes and Lasch, 1980). Its western boundary is a system of high-angle faults that overprinted earlier ductile fabrics within the Alleghenian Hylas shear zone (Bobyarchick and Glover, 1979). Its eastern margin is considerably complex, with the stratigraphic unconformity above Petersburg Granite being faulted in many places (Goodwin et al., 1986; Carter, 2010).

Stratigraphic offsets in sediments throughout the innermost and upper Coast Plain in eastern Virginia attest to significant Cenozoic faulting along the Atlantic margin, and in light of the 2011 earthquake, are no longer geologic curiosities. Most of these occur on the eastern fringe of the CVSZ, but modern reactivation of these structures would be potentially devastating given their proximity to high-density East Coast population centers and critical infrastructure in Washington, D.C., Fredericksburg, Richmond, and Virginia Beach. All of the significant faults in inner Coastal Plain strata near the Fall Line are rooted in Piedmont basement, and most are reactivated Paleozoic to Mesozoic structures.

The Stafford fault system in northeastern Virginia near Washington, D.C. (Fig. 6; Powars et al., 2015), is arguably the most recognized. This system of mostly down-to-east normal faults strikes into the CVSZ zone from the north, but includes NW-SE-oriented cross faults, too. Faults of the Stafford system displace units as young as Pleistocene (Mixon, 1978). Dischinger (1987) documents deformation as young as Pliocene along the Dutch Gap fault, directly east of Richmond (Fig. 6). Here, NW-vergent reserve motion on the Dutch Gap fault and conjugate structures contrasts with its presumed root zone that coincides to the north with the western boundary of a buried Mesozoic basin, although kinematic and temporal relationships between the west-vergent Dutch Gap fault and the assumed down-to-east normal fault bounding the buried basin are unknown (Dischinger, 1987; Wilkes et al., 1989; Carter et al., 2007a). Similar significant Pliocene deformation occurs on the Shockoe fault system (Carter et al., 2007a); Providence Forge (Gilmer and Berquist, 2012); Malvern Hill (Berquist and Gilmer, 2014); and several faults along the Fall Zone south of Petersburg, Virginia, and west of the Surry Nuclear Power facility near Virginia Beach (Berquist and Bailey, 1999; Weems and Lewis, 2007; Weems et al., 2010).

West of Richmond, zones of silicified cataclasite bound, in part, the easternmost edge of the Richmond basin, or continue through the Petersburg Granite. These zones, which also include intense swarms of chalcedony- and quartz-filled joints and fractures, likely formed in the Mesozoic. Bobyarchick (1978) correlates 220 m.y. laumontite mineralization in vein fillings with quartz-breccia silicification in the Hylas zone west of the Richmond basin, and Carter et al. (2007a, 2007b) document brecciated clasts of Jurassic diabase in one silicified cataclasite zone along the Fall Line east of the basin. Direct evidence for younger Cenozoic faulting is rare. Johnson et al. (1987) document only one fault that cuts through the Petersburg Granite and into clayey silt of the highest-level Bon Air gravels, with 3.5-5 m of vertical displacement. Similarly, a difference of ~60 m in elevation between the base of middle Miocene clayey silt mantling the Midlothian upland and equivalent strata exposed in Richmond provides clear evidence for significant vertical offset, warping, or tilting here (Fig. 8), and as noted by Weems and Edwards (2007), for slightly older Miocene strata north of Richmond along the Fall Zone. Carter et al. (2007b) and Carter (2009) give sparse and circumstantial evidence for offset of segments of Mesozoic silicified breccia zones based on distribution of Pliocene fluvial deposits.

Field-trip participants are likely familiar with the venerable arguments of the cycle of erosion (Davis, 1899) and dynamic equilibrium (Hack, 1960) that were developed in the Appalachian landscape. As with our discussion of the bedrock above, our aim here is to provide a contextual framework, based on key recent scholarship, that helps us interpret a venerable landscape in new ways. In this respect, the interested reader is directed to work by Pazzaglia and Brandon (1996), Pazzaglia et al. (2010), Portenga et al. (2013), and McKeon et al. (2014) for more thorough reviews of these basic long-term landscape evolution arguments. Corollary studies addressing short-term rates of erosion for the Mid-Atlantic region are found in Sevon (1989), Langland and Haney (1997), Conrad and Saunderson (1999), and Walter and Merritts (2008). Excellent examples of how paleorelief could be inferred from geomorphic data are found in McKeon et al. (2014), Gallen et al. (2013), and Miller et al. (2013).

The central Virginia landscape preserves a Miocene to Holocene record of unsteady colluvial, fluvial, debris-flow and aeolian deposition as well as pedogenesis. Locally in the Shenandoah Valley or other regions in eastern North America underlain by carbonates, residuum and unconsolidated deposits of Miocene age or older are known to occur (Traverse, 1955; Pierce, 1965; Hack, 1965; Christopher et al., 1980; Bikerman et al., 1999; Wallace and Wang, 2004) and we will have the opportunity to visit some of these sites at Stop 3.

The Ridge and Valley and Blue Ridge have the highest relief and steepest local slopes in the central Appalachians, which correspondingly result in some of the more energetic surface processes, including mass movements and periglacial activity during Pleistocene cold periods. Generally speaking, topography and relief is well adjusted to rock type in these provinces with rocks more resistant to weathering underlying higher standing topography (Hack, 1979). However, there are examples of where this is not true, particularly where transients, such as knickpoints, are moving through the landscape (Harbor et al., 2005; Gallen et al., 2013). Similarly, recent compilations of bedrock erosion rates (Fig. 9) suggest that there is locally poor correspondence between rock type and erosion rate (Portenga et al., 2013).

One of the more intriguing topics of long-term landscape evolution in the Blue Ridge and Ridge and Valley concerns the development and expansion of the mid-Atlantic slope drainages, namely, the Delaware, Susquehanna, Potomac, James, and Roanoke. This field trip will visit locations at Stops 1 and 2 where there is morphological and depositional evidence of drainage captures that document the slow, but unsteady movement of the Appalachian divide (Figs. 2 and 7) westward across the Blue Ridge and into the Ridge and Valley (Davis, 1903; Johnson, 1907, 1931; Meyerhoff and Olmsted, 1936; Thornbury, 1965; Meyerhoff, 1972; Hack, 1975, 1979, 1982; Pazzaglia and Brandon, 1996; Harbor, 1996; Pazzaglia and Gardner, 2000; Spotila et al., 2004; Gunnell and Harbor, 2010; Prince et al., 2010, 2011; Bossu et al., 2013; Gallen et al., 2013).

At the scale of the entire eastern North American margin, there is good sedimentological (Scholle, 1977, 1980; Poag, 1985, 1992; Poag and Sevon, 1989), mineralogic (Smith, 1980), and thermochronologic (Spotila et al., 2004; Naeser et al., 1999, 2004, 2006) evidence that the eastern continental divide made a major westward jump in the early or middle Miocene. The heavy mineral suites from Jurassic through Pleistocene sediments in the Baltimore Canyon trough show relatively unweathered pyroxene and amphiboles, and notably, a high percentage of rounded zircon, sphene, and staurolite in the younger part of the section that contrasts sharply with stratigraphically older suites that are characterized by tourmaline, garnet, epidote, and chlorite (Smith, 1980). A reasonable interpretation is that the youngest suite, from sediments of Miocene to Pleistocene in age, has a significant Ridge and Valley provenance that has liberated reworked heavy minerals from foreland basin sandstones, mixed with relatively unweathered glacial material, whereas sediments older than Miocene are dominated by heavy minerals with a Piedmont and Blue Ridge provenance. In this context, the huge increase in sediment flux to the Baltimore Canyon trough in the Miocene, typically interpreted as a change in rock uplift or climate (Pazzaglia and Brandon, 1996; Figs. 2B and 2C), might more simply reflect a large westward jump in the continental divide, driven by poorly constrained dynamic and flexural isostatic forces (Pazzaglia and Gardner, 1994; Forte et al., 2007, 2010; Moucha et al., 2008; Rowley et al., 2013; Liu, 2014), and the subsequent rapid incision of the newly acquired Atlantic slope landscape (Harbor, 1996; Gunnell and Harbor, 2010).

As rivers carve into the Blue Ridge and Ridge and Valley landscape with the expansion of Atlantic slope drainages, hill-slopes are steepened, accelerating creep and mass movement processes as important agents of erosion. In particular, debris flows associated with infrequent, but high-magnitude catastrophic storms have been documented to deeply incise older Pleistocene periglacially generated deposits and transport sediment to the foothills of the ridges, where it accumulates in alluvial fans (Kochel, 1987, 1988, 1990; Eaton et al., 2003). The episodic, but high-magnitude storm events, the recurrence interval of which is ~3000 yr, are estimated to transport half of all of the regolith that accumulated in high gradient, low-order drainage basins during the late Pleistocene and Holocene (Eaton et al., 2003).

Related rates of hillslope weathering, regolith production, and creep transport are emerging from process geomorphic studies in the Shale Hills Critical Zone Observatory in central Pennsylvania (Ma et al., 2010, 2013; West et al., 2013; Fig. 9). Here, the relief and sedimentary rock substrate compare reasonably well to the Virginia Ridge and Valley rocks. Soil production has been calculated using a novel U-series technique (Dosseto et al., 2008; Ma et al., 2013), as well as measuring inventories of meteoric cosmogenic ¹⁰Be (West et al., 2013) along a ridge to valley bottom soil catena. The rates of soil production vary from ~45 m/m.y. at the ridge top to ~17 m/m.y. for the toe slope. The faster ridge-line rates compare well to those calculated by Braun (1989) estimated from the volume of fill in colluvial hollows along the periglacial fringe of Pennsylvania (Fig. 9). Particularly for those parts of the landscape that are in the latitude or elevation-defined zone impacted by Pleistocene periglacial processes, ridge lines may be lowering faster than the rest of the landscape, resulting in a reduction of relief for these regions. West et al. (2013) also show that hillslope aspect plays an important role in the transport efficiency of regolith given the similar fluxes on opposing slopes of unequal gradient.

These relatively rapid rates of sediment production and transport contrast with what we know about the long-term, background rates of weathering, erosion, and landscape lowering (Price et al., 2008; Hancock and Kirwan, 2007; Portenga et al., 2013; Fig. 9). Portenga et al. (2013) report an average ridge-line erosion rate of exposed bedrock of ~9 m/m.y. and landscape-scale averages of all bedrock outcrops of 6 m/m.y., which compares favorably to those reported by Hancock and Kirwan (2007). Similarly, geochemical mass balances for watersheds in the Pennsylvania and Virginia piedmonts show that the rate of chemical dissolution of noncarbonate bedrock accompanying its conversion to saprolite is occurring between ~4.5 and 6 m/m.y. (Pavich et al., 1985, 1989; Price et al., 2008). These rates account for one-third to one-half of the long-term, watershed-averaged, cosmogenic radionuclide-determined rates of ~ 10-20 m/m.y. obtained by sampling channel alluvium (Reuter, 2005; Matmon et al., 2003).

The well-known Cretaceous lignite at Pond Bank, Pennsylvania (Pierce, 1965), and similar residual deposits preserved in sinkholes of the Shenandoah Valley (reviewed in Hack, 1965), place some constraints on how much relief, erosion, and chemical dissolution is possible in the Blue Ridge and Ridge and Valley provinces over millions of years. The Pond Bank lignite contains upper Cretaceous (early Campanian, ca. 80 Ma) terrestrial pollen, distinctly lacking any marine palynomorphs, encased in part by residuum derived from the host carbonate bedrock. This biostratigraphic age overlaps with the upper part of the Potomac Group lithostratigraphy represented by the Raritan and Magothy Formations in New Jersey and eastern Maryland and the Patapsco Formation in western Maryland, and is at least in part consistent with the lignite common in the Magothy Formation. The highest sea levels in the Cretaceous are Cenomanian in age (ca. 91-92 Ma; Sahagian et al., 1996), coincident with Raritan-Patapsco Formation deposition; thus it is possible that the Pond Bank lignite represents the fluvial, upstream equivalent of these Coastal Plain deposits.

Preservation of the Pond Bank lignite illustrates the peculiar way in which Cenozoic sediments are preserved in the Shenandoah Valley landscape. The juxtaposition of a forested quartzite ridge (Blue Ridge) against texturally and compositionally immature, sandy carbonates drives dissolution of those carbonates, production of an insoluble residue, and accommodation space due to karst subsidence. Sediment washed off of the western flank of the Blue Ridge is trapped by that subsidence, and a stratigraphy of Cenozoic hillslope erosion is preserved. The Pond Bank deposit, and related deposits that we will visit at Stuarts Draft, Virginia (Stop 3), is preserved at depth near the base of a thick colluvial and alluvial fan stratigraphic package. That package thickens upwards, reflecting the relative contributions of sediment supply and subsidence. At the same time, dissolution of the carbonate to produce a residual deposit including saprolite proceeds beneath the lignite. Measurements of carbonate dissolution in Pennsylvania (White, 1984, 2000) range from ~8-30 m/m.y., values consistent with the long-term average determined by the thermochronology. Using an insoluble residue content of 10% for the carbonates at Pond Bank, Pierce (1965) inferred that at least 430 m of carbonates have weathered beneath the gravel to produce the residuum below the lignite-bearing beds. Thus, the stratigraphic column in this setting is oldest in the middle, at the horizon that preserves the Late Cretaceous lignite. That stratigraphy youngs both upward through the alluvial deposits toward the surface, and downward through the residuum toward the bedrock-saprolite interface.

In summary, all available data point to erosion and landscape change in the Mid-Atlantic region, including central Virginia, to be generally slow, on the order of ~5-30 m/m.y. Embedded in those slow long-term rates are non-uniformities linked to rock type, the distinction between fluvial and hillslope environments, and the distinction between ridge tops and toe slopes. Furthermore, erosion and landscape change is unsteady, driven by unsteady changes in climate and base level. There are spatial and temporal transients in the landscape, namely in the form of river channel knickpoints, and these take millions if not tens of millions of years to propagate through Appalachian watersheds, imparting their base level signal to the hillslopes. The result is a largely transient landscape that still reflects basic differences in rock type and proximity to base level change, but which also is slowly and dynamically adjusting to changes in climate and base level.

The Virginia Piedmont at the latitude of this field trip is drained by four major river systems: the Roanoke, James, Rappahanock, and Potomac (Figs. 5 and 6). Of these, only the Rappahannock River does not breach the Blue Ridge. Headwaters of the North and South Anna rivers (Figs. 5 and 6) lie in the western Piedmont between the James and Rappahannock rivers.

Atlantic slope drainages are segmented, containing concaveup graded reaches with intervening steep, convex reaches containing one or more knickpoints (Fig. 10). Where these knick-points are not fixed by some particularly resistant bedrock, they are presumed to have been caused by unsteady or impulsive base level fall at or near the Fall Zone, and have subsequently migrated upstream (Reusser et al., 2004, 2006; Miller et al., 2013). In our field-trip area, all of the streams have experienced a base level fall of ~100 m in the past ~12 m.y., based on the projection of the uppermost reaches of the drainages, over channel knickpoints, downstream to the Fall Zone where the projection intersects bio-stratigraphically constrained fluvial deltaic and shallow marine deposits of the Choptank Formation (Fig. 10; Weems et al., 2012). Correspondingly, river incision into the Virginia Piedmont at the Fall Zone since the late middle Miocene is ~10 m/m.y. as a long-term average, but evidently can be as rapid as ~250 m/m.y. (Reusser et al., 2004), where it is measured at knickpoints and over shorter time spans.

Assuming that the prominent knickpoint at ~160 m (~525 ft) on Piedmont streams like the South Anna represents the elevation that the post-Choptank base level fall has propagated into the landscape, then higher elevation knickpoints, and expansion of rivers like the James across the Blue Ridge and into the Ridge and Valley, must be older than ca. 12 Ma if those knickpoints are in fact related to base level fall on the Atlantic slope drainages. This simple logic fails if the landscape through which the knickpoints are migrating is not uniformly uplifting. If it is tilted, warped, or offset by faults as we intend to discuss on this field trip, then interpretation of knickpoint elevation as a crude proxy for timing of base level fall is not valid.

Alluvium (unconsolidated clay, silt, sand, and gravel of Neogene to Quaternary age) occurs along all or portions of these river systems, typically at leeward-side point bars and horseshoe bend cutoffs. Alluvium in these areas ranges from a few meters to tens of meters thick. Topographic side slopes and hilltops are typically underlain by saprolite beneath colluvial and residual soil horizons. Saprolite can range from a few meters to tens of meters thick. These topographic settings are also typically mantled by thin deposits (typically less than 0.5 m thick) of pebble-, cobble-, and boulder-size polycrystalline vein quartz, which may have accumulated in situ from weathering of surrounding bedrock, but more likely are colluvial, having been transported and concentrated down-slope, albeit locally, by creep (Malenda et al., 2014).

Demonstrable alluvial terrace deposits consisting of sub- to well-rounded pebbles and boulders of vein quartz, as well as exotic lithologies including Skolithos-bearing quartzite from the Blue Ridge and Valley and Ridge provinces, also cap topography and mantle side slopes in the vicinity of these major streams, including those that head east of the Blue Ridge front. These deposits occur up to many tens of meters above modern alluvial drainages. Along the James River in the western Piedmont, Hancock and Harbor (2002) used cosmogenic ¹⁰Be to date terrace deposits ~60 m to 75 m above the modern river channel. Their data yielded ages of ca. 1.1 Ma to 1.3 Ma, with incision rates of ~45 m/m.y. between the lower strath and the modern river (Fig. 9). New OSL (optically stimulated luminescence) data from terraces along the South Anna River in the epicentral area of the 2011 earthquake (this guide) corroborate rapid, recent incision along Piedmont rivers. A more through summary of these data and resulting incision rates can be found in Day 2 stop descriptions.

The M 5.8 Mineral, Virginia, earthquake occurred at 17:51:04 UTC (1:51:04 p.m. EDT) on 23 August 2011. Its epicenter was located at 37.905°N, 77.975°W (WGS84), with a focal depth of ~8.0 km (Chapman, 2013; Fig. 6). This event occurred ~61 km northwest of Richmond and ~135 km southwest of Washington, D.C., in a diffuse zone of intraplate seismicity referred to as the Central Virginia seismic zone (Bollinger, 1969, 1973a, 1973b). Slip occurred on a shallow reverse fault having a moment tensor solution of N28°E, dip 50°, and rake 113° (Chapman, 2013). The rapid deployment of instruments following the main shock resulted in nearly 400 well-recorded and well-located aftershocks of M_W > 1.8, with moment tensors computed for the largest, which illuminated the causative fault of this earthquake (Horton and Williams, 2012; Horton et al., 2015). These data indicate a previously unrecognized 10-km-long zone of rupture that strikes about N30°E and dips east-southeast ~45° (Ellsworth et al., 2011), now termed the Quail fault zone (Horton et al., 2015). Smaller aftershock clusters illuminate active subsidiary faults (Horton et al., 2015).

The Quail fault zone is within the Chopawamsic Formation, between the Chopawamsic fault to the northwest and the Spotsylvania fault zone to the southeast (Spears, 2012). The fault zone projects to the surface near and/or gently east of the contact between rocks of the Chopawamsic Formation and Ellis-ville plutonic rocks (Burton et al., 2014, 2015a, 2015b). Trenching at one locality along the Chopawamsic-Ellisville contact in 2012-2014 exposed a zone of Paleozoic shear overprinted by brittle deformation (Burton et al., 2014). This study also notes a zone of mineralized breccia along the contact and slickenside striations observed at various locations along the contact in float. This faulted contact has been interpreted to offset two 2-km-long Jurassic dikes ~250-275 m, indicating post-Jurassic movement along this structure.

The trip begins at the crest of the Appalachian Mountains at the eastern continental divide in Blacksburg, Virginia (Stop 1, Figs. 11 and 12). This stop, which gives us our only view of the topographic character of the divide and Ohio Valley basins to the west, allows us to make three important observations that are indicative of landscape disequilibrium. First, the topography of the divide is very subdued (Stop 1A). The divide does not correspond to the highest peaks or ridgelines in the area. Second, the divide is asymmetric, separating lower-relief basins of the Gulf of Mexico side from rugged headwater basins of the Atlantic slope (Stop 1B). Finally, the divide has been the site of major deposition in the past few million years, evident in nearby preserved fluvial terraces (Stop 1C).

If the topography of the Appalachian Mountains existed in a form of erosional equilibrium, several key characteristics would be expected. These include alignment of the drainage divide to the spine of the highest, most rugged topography; the deepest crustal root (Hawman et al., 2012); and the most resistant rock types (Hack, 1960, 1973, 1982). Erosion rates and mean slopes should be reducing gradually with time (Ahnert, 1970; Baldwin et al., 2003). In contrast, there is abundant evidence of topographic disequilibrium in the Appalachian Mountains. At Stop 1, we will explore key evidence of this disequilibrium as expressed in the eastern continental divide migration. A number of exogenic factors have been proposed for divide migration in this region (indicated above); however, it is also possible that the drainage capture and divide retreat are primarily endogenic.

The eastern continental divide is asymmetric in this region, separating the low-relief, higher elevation New River and Blue Ridge uplands from the steeper basins of the Atlantic slope. The New River upland is located ~3500 km from ultimate base level, whereas Atlantic streams must flow only ~500 km to reach base level. This, in combination with resistant, more deeply exhumed strata of the Appalachian foreland basin, may be enough to result in a setting that favors irregular basin reorganization (Hack, 1973). A similar idea has been proposed for the evolution of the Blue Ridge escarpment further south, where low temperature thermochronometry suggests the divide retreats away from a welt of locally intense exhumation that slowly migrates westward from the Piedmont (Spotila et al., 2004). The erosion pattern and topography of the escarpment are similar to other great escarpments on other (albeit more recently rifted) passive margins (Spotila et al., 2004), suggesting the Blue Ridge escarpment may even be continuing to topographically adjust to rift-flank uplift. In any case, climate-intensified erosion and mantle-driven uplift need not be invoked to drive the observed drainage basin migration, because the topographic juxtaposition at the asymmetric divide itself provides ample potential energy to the system to force punctuated, piecemeal retreat and the resulting evidence for topographic disequilibrium that can be observed today.

Exit the Holiday Inn Christiansburg (37.162259°N, 80.427292°W) parking area by turning left (northwest) onto Bradley Drive; proceed 0.03 mile northwest on Bradley Drive to Peppers Ferry Road; turn right (northeast) onto Peppers Ferry Road and proceed 0.8 mile to U.S.-460 West via the ramp to Blacksburg/Bluefield WV; proceed 2.2 miles on U.S.-460 West to Exit 5A-5B for U.S.-460 BUS East/U.S.-460 BUS West toward Blacksburg/Smart Road; exit right (northwest) and stay right to continue on Exit 5A, following signs for U.S.-460 BUS East/ Industrial Park Road; turn left (west) onto Industrial Park Road and proceed 0.2 mile to Ramble Road/Research Center Drive; turn right (north) onto Ramble Road/Research Center Drive and proceed 0.7 mile to Stop 1A at VA Tech Airport.

The eastern continental divide cuts through Blacksburg, Virginia, across a low-relief, undulating topography that is rather inconspicuous. The divide has such low relief that the runway of the Virginia Tech regional airport actually sits directly on it. Here, the divide is sporadically capped by deeply weathered fluvial gravels derived from the Blue Ridge to the southeast (Ward et al., 2005). Some of these are visible in a red bluff across the runway. (If possible, depending on time and accessible cuts, which vary seasonally and with construction, we will stop to examine one of these terrace deposits along the way to Stop 1B.) Gentle topography continues westward from the divide to the New River, although the upland surface is incised over 50 m by the New River and its tributaries, possibly due to late Cenozoic climate change (Prince and Spotila, 2013). To the east, the topography very quickly transitions to steep slopes that fall into Cedar Run, a tributary of the Roanoke River that has recently cut through a resistant ridge that forms the foundation for the divide to the north and south (Fig. 12). The divide and headwaters at this location are thus in a transitional state and will evolve, as Cedar Run further incises and advances west.

On the way to Stop 1B, we will cross a broad field just west of the airport on Research Center Drive that affords a good view of the surrounding landscape. The low-relief surface that makes up the surrounding New River upland plain is situated on easily weathered and eroded Cambrian shale and carbonate of the upper plate of the Pulaski thrust sheet (Bartholomew and Lowry, 1979). From this vantage point, however, we can see the more characteristic ridges that occur in the typical siliciclastic-rich units of the Valley and Ridge section. This includes Brush Mountain and Gap Mountain to the northwest (Fig. 11), which are narrow ridges held up by (respectively) Devonian quartz pebble conglomerate and sandstone and Silurian quartzite. These ridges continue for great distances along strike and merge with the classic ridge and valley terrain in Pennsylvania.

From Stop 1A at VA Tech Airport, proceed 1.4 miles on Tech Center Drive/Research Center Drive (Research Center Drive becomes Tech Center Drive at 0.1 mile) to Southgate Drive; turn right (east) onto Southgate Drive and proceed 0.6 mile to Airport Road; turn right (south) onto Airport Road and proceed 0.6 mile to Hubbard Street (Airport Road becomes Hubbard Street at sharp left bend); proceed east for 0.3 mile on Hubbard Street to South Main Street; cross South Main Street onto VA-603/Ellett Road; proceed southeast on VA-603/Ellett Road for 1.5 miles to Stop 1B at Cedar Run.

Although the headwaters of Cedar Run are low relief, because they rest on inherited, transient topography formerly of the New River upland, the stream rapidly steepens into a broad knick zone, roughly coincident with the Yellow Sulfur fault (Bartholomew and Lowry, 1979), which drops ~200 m over ~3 km to the floor of Ellett Valley to the east (Fig. 13). This is thus a classic “leaky” knickpoint that is presumably migrating relatively rapidly to the east relative to the background erosion rate. Along the knick zone, the stream sits in a v-shaped gorge and exhibits a tilted, planar but stepped bedrock morphology.

The stepped nature of the steep run is partly because it follows the shallow dip of the underlying strata. The steep channel consists of a broad 2-3-m-wide fluted bedrock plane that is only locally incised. The bedrock is comprised of dolomite and limestone of the Ordovician Knox Group. The paucity of quartz-ridge bedrock in the catchment may contribute as a lack of erosion tools for the channel bed, which may partly explain the rarity of potholes and the minimal channel incision. The stream is fed by springs that drain the karstified bedrock, the occurrence and chemistry of which result in heavy vegetation in the gorge. This vegetation is characteristically lacking in rhododendron due to the neutral to slightly alkaline soils from the carbonate bedrock. The springs also result in widespread tufa deposition, which occurs along banks and builds out tufa falls from the channel bed itself right below springs (a large tufa waterfall occurs several hundred meters downstream of Stop 1B).

Cedar Run provides a glimpse at the active erosional front, where Atlantic slope streams are capturing and pushing into water sheds draining to the Gulf of Mexico. After leaving Stop 1B, we will drive to the bottom of Ellett Valley (Fig. 14) and continue to Stop 1C via North Fork Road (VA-603) and Den Hill Road (SR-641). Along the way, notice how broad and flat the valley bottom is, but also how low it sits relative to the remnants of the New River upland surface that tower all around (e.g., Paris Mountain to the east; Fig. 11).

From Stop 1B at Cedar Run, proceed southeast on VA-603/ Ellett Road for 2.0 miles to SR-641/Den Hill Road; continue south onto SR-641/Den Hill Road and proceed 4.1 miles to U.S. -11/U.S.-460/Roanoke Street; turn right (southwest) onto U.S.-11/ U.S.-460/Roanoke Street and proceed 1.5 miles southwest on U.S. -11/U.S.-460/Roanoke Street to Stop 1C at Crab Creek.

Relict alluvium is common atop the low-relief topography of the New River upland and the eastern continental divide itself, which is indicative of basin inversion and recent drainage divide migration. An excellent example of this occurs at the Crab Creek terrace in Christiansburg, Virginia (Fig. 15) (Prince et al., 2011). Crab Creek is a tributary of the New River. Just south of it, the Roanoke River appears to have “taken a bite” out of the eastern continental divide via punctuated, piecemeal stream capture and westward headwater retreat.

Atop the divide and throughout the immediate area are weathered terrace remnants (shown as circles in Fig. 15) that contain clasts with provenance indicative of drainage capture. These deposits contain abundant rounded clasts of Blue Ridge origin (metaquartzite, quartz mylonite, and clear and blue vein quartz), as well as chert clasts from the eastern Valley and Ridge (Fig. 15). The deposits are highly weathered, but clearly distinguishable from the clay-rich saprolite derived from local carbonates, which they are deposited atop. Paleobasin reconstruction of the distinctive clast assemblage at Crab Creek, as well as deposits at Fishers View (Fig. 15), suggest an origin >20 km to the east. Based on the topography, including a high-elevation, low-relief relict surface atop Poor and Bent Mountains and the longitudinal profiles of streams (nos. 2 and 3 in Figs. 15 and 16A, which align with the Crab Creek profile [no. 1]), the eastern continental divide was therefore positioned farther east, as approximated as the line of circles in Figure 15. This large drainage capture would have been the result of southward erosion by the Roanoke River. Tributaries of the Roanoke River display grapnel-like morphologies in mapview and steep knickpoints in profile, consistent with rapid headward retreat (Figs. 13 and 16).

The timing of this capture must postdate the Crab Creek deposits, which are unfortunately not dated. Based on the similar degree of weathering and preservation of dated New River terraces (Ward et al., 2005), Prince et al. (2011) suggested that this capture was likely in the past few million years. The capture is interpreted to be piecemeal and punctuated, in the sense that the divide was likely quasi-stable and slowly retreating until a divide was breached from the side, enabling capture of a large upstream area (Fig. 17). The sudden increase in stream power would have resulted in rapid topographic adjustment, as has been envisioned elsewhere atop the divide along the Blue Ridge escarpment (Prince et al., 2010). The process is continuing today, following the path shown as the wide, shaded, dashed line in Figure 15. The Roanoke River appears to be taking advantage of an erodible shale unit (the Rome Shale), resulting in the westward “bite” in the divide. The next major capture will occur when the Roanoke River reaches the Little River, and soon thereafter (<5 m.y.?) when a 7000 km² portion of the New River basin is captured at Graysontown, Virginia (Prince et al., 2011).

From Stop 1C at Crab Creek, proceed west 1.6 miles on U.S.-11/U.S.-460/Roanoke Street; turn right (north) to merge onto I-81 North; proceed 72.7 miles on I-81 North to Exit 191 (on left) for I-64 West toward Charleston, West Virginia; proceed 0.9 mile on I-64 West to Exit 55 for U.S.-11 toward VA-39/ Lexington/Goshen; turn left (southeast) onto U.S.-11 South and proceed 0.6 mile to Stop 2 at Lexington.

Stop 2 is located in the remains of a topographically inverted river terrace ~60 m above the Maury River, a tributary of the James River (Fig. 18). The alluvium, most of which has been removed for construction purposes, varies from sand to 20 cm cobbles. Along with minor chert from the Cambrian and Ordovician carbonates, the alluvium is primarily gravel and cobbles derived from the Silurian-age sandstone that crops out to the west of the Great Valley. It is several meters thick along the southeastern forested edge of the borrow pit. That thickness, however, is highly suspect and variable given that at least 20 m of the underlying Ordovician Edinburg Formation is completely weathered to a silty saprolite. Along U.S.-11, adjacent to this borrow pit, saprolite exposed during roadwork surrounded a cave with both stalactites and alluvial cobbles. Only near the road level was the bedrock unweathered. The argillaceous limestone and interbedded shale of the Edinburg Formation will preserve the elevation of the original alluvial deposit better than a limestone (Sherwood et al., 1987), but lowering of the surface is expected during the extended period of time that it takes to invert topography. An inverted terrace of similar height just across the Maury River and slightly upstream is also capped by gravels and underlain by thick saprolite. Alluvial cobbles comprise the float that encircle these and other high elevation terraces of the Maury River, indicating substantial erosion of the surface and/or edges of the terrace deposits. Large sinkholes are common at this elevation near the river and its terraces (Fig. 18B).

The western end of the outcrop consists of thin gravel over saprolite showing pedogenic alteration (clay influx and extensive mottling). Most of this part of the deposit has been removed or these gravels have been eroded from the nearby terrace deposit and now form a thin layer over the saprolite. The middle region has a fine-grained soil presumably formed in the upper portions of the original floodplain soil, plus completely weathered cobbles from the coarse layer below. The eastern end of the outcrop consists of cobble-gravel with deep red pedogenic clay filling the spaces between clasts. Some parts of the deposit appear to have imbrication that is too steep to the west, which supports collapse of the supporting material as saprolite weathers and/or sinkholes form.

The soil developed in these terraces indicates substantial weathering and development of iron- and clay-rich subsoils. The Rockbridge County soil survey (Cook, 2014) places these high, valley-center terraces in the Shottower Series. These soils are characterized by clay to sandy clay loam texture, and Munsell color of 10R to 2.5YR in the Bt horizon. In this exposure, erosion of the soil surface (naturally and by humans) is causing the surface horizons (thin A and underlying E) to reform in the upper B horizon, indicating that erosion is limiting the maximum development of the B horizon. At depth in the B horizon, pebbles and cobbles float in the finer grained alluvium and pedogenic clay. Pebbles highest in the B horizon tend only to be the smaller, dark-red, hematite-cemented Silurian Rose Hill sandstone. Only deeper in the profile is quartz arenite of the Silurian Tuscarora and Keefer formations intact enough to be removed from the outcrop. Some of the quartz sandstone clasts are weathered such that they can be crushed by hand once the case-hardened outer rind is cracked with a hammer, equal to weathering category 4 for clasts in fans near Stuarts Draft (Whittecar and Duffy, 2000). Even chert is crumbly at this degree of weathering.

The degree of weathering and soil development is consistent with middle Pleistocene or older age, given the clay, clay films, and color (Pazzaglia and Gardner, 1993; Engel et al., 1996; Elliott, 1998). Applying regional river incision rates of ~25 m/m.y. (Granger et al., 1997; Miller et al., 2013; Sevon, 1989) yields an age just greater than 2 Ma. However, geochemical, clastic, and paleomagnetic stratigraphy in a cave just upstream (Knapp et al., 2004) suggest instead a higher incision rate of 150 m/m.y. and a therefore considerably younger age. Moreover, this incision rate falls within a range (65-150 m/m.y.) determined from deposits in a cave on a downstream tributary to the Maury East-house, 1998). On a James River terrace also 60 m above the river, Erickson (1998) determined a cosmogenic isotope burial age of 380 ± 80 ka, which yields an incision rate of 160 ± 40 m/m.y.

A more rapid incision rate is consistent with passage of a knickpoint through the James River system, which has been described by Miller et al. (2013) for the Susquehanna River and Harbor et al. (2005) for the Maury River. The Susquehanna River erosion rate above the knickpoint is 5-30 m/m.y. but increases to 100-150 m/m.y. below. Upstream of this terrace, the Maury River profile shows knickpoints that total 50-70 m of relief (Fig. 19B), suggesting that incision below this terrace is completely consistent with the transient passage of knickpoints. Terraces 50-60 m above the Maury River continue to Goshen Pass, where the Maury cuts through the Silurian units at the valley edge and the knickpoint propagation slows (Harbor et al., 2005). The James River profile above the terrace dated by Erickson (1998) has similar knickpoints (Figs. 18A, 19C), but terraces farther upstream are unmapped.

Accepting the hypothesis that knickpoint migration is responsible for the unsteady incision by the Maury River and that the terrace is the same age as the dated James River terrace, the migration of the knickpoints from the Stop 2 terrace to the current position of the knickpoint is 0.1-0.13 m/y (100-130 km/m.y., Fig. 19C). This propagation rate is fast but in line with the compilation of Loget and Van Den Driessche (2009). Using the same method for the knickpoint in the James River, the knickpoint propagation rate calculates to 0.18-0.26 m/y (180-260 km/m.y.) on a stream with a slightly larger drainage basin. If knickpoints cross the Valley and Ridge section of the upper James River so rapidly, they likely propagated across the Piedmont as well. Along James River in the Piedmont, both upstream and downstream from an inverted terrace mapped by Giannini (1984) near Howardsville (Fig. 18A), Hancock et al. (2004) dated terraces using cosmogenic isotope inventories in highly bioturbated, red (2.5YR to 10R) thick argillic (40%-60% clay) soils. The age of these terraces lying ~60 m above the James River are 1.2-1.1 Ma. This same 60 m of incision could be a result of the same knick-point system that passed through the Maury, but the correlation of terraces and knickpoints is difficult through the steeper, terrace-poor Blue Ridge section of the James River. If the terraces in both regions are truly the result of one series of knickpoints, 200 km of upstream migration from the Howardsville terrace (1.16 Ma) to the upper knickpoint in the Maury River at Goshen Pass occurred at 0.17 m/y (170 km/m.y.). The presence of knick zones in the Maury River between these two terraces questions this hypothesis, but rearrangement of drainage in the valley could be responsible for other profile anomalies. Clearly, more dating of terraces and quantification of erosion rates is necessary to test these ideas.

The rapid knickpoint migration estimates are somewhat greater than those reported in the Susquehanna River basin (Miller et al., 2013), where knickpoints travel between 220 and 560 km in 4-19 Ma. These rates are much higher than those reported by Gallen et al. (2013), where southern Blue Ridge knickpoints may have traveled less than 40 km since the Miocene. Similarly, the rates of knickpoint retreat on the South Anna River have been modeled to fall between 7 and 14 km/m.y. (this guide). Erosion might be more efficient in the James River basin than the Susquehanna, because it is overall a steeper network due to the capture of the upper James River through the Blue Ridge (Thompson, 1939; Dietrich, 1959; Harbor, 1996; Gunnell and Harbor, 2010), and the rates are faster than what is modeled on the Piedmont given the much smaller size of the South Anna River. The profile of the upper James River (Fig. 19B) has hundreds of meters of additional relief on knickpoints that end in low-relief summits in the Blue Ridge (Pedlar River) and Valley and Ridge. Although undated, this capture event likely occurred after that of the Potomac River in the Miocene (Naeser et al., 2001, 2005), and might be linked to the Pliocene increase of sediment volume near the mouth of the James (Poag and Sevon, 1989). Like the Callasaja River of Gallen et al. (2013), recession of knickpoints is dramatically slower in the headwaters, so evidence of the disequilibrium erosion following capture persists.

Before the passage of the Quaternary knick zones through the valley, but after the erosion produced by capture, erosion rates were slower. Cobbles up to 65 m below the surface of fan/ pediments that flank the Blue Ridge just north of Buena Vista display weathering up to stage 4 of Whittecar and Duffy (2000) and contain Pliocene-age plant fossils (Leonard, 1963). This storage of alluvial material requires actively subsiding basins, the accommodation space likely being produced by dissolution rates of the subjacent carbonates that are in excess of the average surface lowering rate of ~25 m/m.y., but potentially as low as 5-10 m/m.y. (Duxbury et al., 2015; Portenga et al., 2013). These recently exposed, but formerly deeply buried cobbles lie at approximately the same elevation as the high elevation Maury River terraces but are likely 1-3 m.y. older.

Acknowledgments. David Harbor would like to thank the students and faculty of the Keck Geology Consortium project who worked in the area during 1997-1998, particularly Peter Erickson, Dylan Easthouse, Carrie Elliott, Thomas Gardner, and Dorothy Merritts, for some of the data and ideas presented herein; mistaken interpretations are his alone.

From Stop 2 at Lexington, proceed 0.6 mile northeast on U.S.-11 North to the ramp onto I-64 East; turn right (southeast) onto I-64 East ramp to merge onto I-64 East toward I-81 North; proceed 0.5 mile on I-64 East to I-64 East/I-81 North exit (on left); proceed 22 miles on I-64 East/I-81 North to Exit 213 for U.S.-11 toward U.S.-340/Greenville; turn right (south) onto U.S -11 South to Greenville; proceed 1.5 miles on U.S.-11 South to Indian Ridge Road on left; turn left (east) onto Indian Ridge Road, and take an immediate right (south) onto Main Street; proceed 0.2 mile south on Main Street to Greenville School Road on left; turn left (east) onto Greenville School Road and proceed 1.7 miles to Cold Springs Road; turn left (northeast) onto Cold Springs Road and proceed 3.0 miles to Boxley Road (entrance to Rockbridge Stone Products) on right; turn right (south) into Rockbridge Stone Products and proceed 0.6 mile to Stop 3A at Rockbridge Stone Products.

From Stop 3A at Rockbridge Stone Products, proceed 0.6 mile north to SR-608/Cold Springs Road; turn right (east) onto SR-608/Cold Springs Road and proceed 3.8 miles to Howardsville Turnpike; turn right (south) onto Howardsville Turnpike and proceed 0.4 mile to Lake Road on right; turn right (south) onto Lake Road and proceed 0.3 mile to entrance to Acres Sand and Stone and D.M. Conner Sand and Gravel on left; turn left (southeast) onto entrance road and proceed 0.5 mile to Pine Trail (entrance to Acres Sand and Stone) on right; turn right (south) onto Pine Trail; proceed 0.6 mile to parking lot of Stop 3B at Acres Sand and Stone.

Surficial deposits of the central Virginia Blue Ridge and Valley and Ridge provinces in this region (Fig. 20) are as varied and complex as the underlying bedrock. Talus, boulder streams, and related colluvial deposits occur on very steep slopes; debris-flow deposits occur in steep, higher-order stream valleys; debris and alluvial fans form at the base of stream valleys and older fan deposits are preserved on interfluves; alluvium and terrace deposits are preserved along larger streams; and a few stranded remnants of riverine gravels occur both in high mountain wind gaps and above valley river networks. The most obvious surficial deposits in the vicinity of Big Levels are the broad alluvial fans at the foot of the mountain front, which coalesce into one large bajada (Fig. 21). Sourced from mountain streams that drain mostly Antietam quartzite (Chilhowee Group), at least four generations of fan deposits create the landform.

Morphologically, the youngest alluvial fan deposits are within or adjacent to active stream valleys. The upstream portions of some deposits probably resulted from debris flows, based on material size and sorting. Intermediate-age deposits include well-developed alluvial fans that extend for several kilometers into the Shenandoah Valley. Several of these deposits are moderately incised and modified by karst. Older deposits are elevated and dissected remnants, with hummocky surfaces and highly incised drainage. The sources of these fans are the major stream valleys. Kennedy Creek, Coles Run, and Johns Run feed younger alluvial fans, whereas Loves Run and Stony Run were source streams for older deposits (Fig. 21). Lesser deposits appear to source from smaller drainages along the mountain front. Lithologically, the fans consist mostly of subangular to subrounded cobbles and small boulders of Antietam orthoquartzite and quartz sandstone in a brown to orange-brown, clayey sandy loam to sandy loam matrix. Clasts of other bedrock lithologies are rare. Fan surfaces of all ages are mantled with relatively unweathered cobbles and boulders of Antietam quartzite, particularly near the mountain front. These unweathered cobbles and boulders form the source material for the recent alluvial fill within the modern stream floodplains.

Whittecar and Duffy (1992) mapped in detail deeply weathered fan deposits exposed in the Acres sand and gravel pit. Map units were differentiated on degree of weathering of finegrained sandstone clasts. Clasts in younger deposits are fresh and unweathered; progressively older deposits contain clasts that crumble at the touch. Recently, Heller et al. (2014) dated two samples of deeply weathered fan material exposed in the Acres pit at depths of 15-20 m below the surface (Fig. 22), using the cosmogenic ²⁶Al/¹⁰Be burial decay method of Granger et al. (1997) and Granger and Muzikar (2001). The samples yielded ages of 6.9 ± 1.7 m.y. from a higher fan layer (F2 of Whittecar and Duffy, 1992) and 7.94 ± 2.4 m.y. from a lower unit (F1 of Whittecar and Duffy, 1992).

It is not clear if these deposits represent a deeper part of the fan mapped at the surface that is sourced from Coles Run, or the remnants of an earlier fan complex that may have a different source. Sinkholes on the surface of older fan deposits indicate carbonate at depth; solution of these rocks continues to provide accommodation, and may allow for the long-term (> 6 m.y.) accumulation of sediments here. Contributory to this discussion are unpublished data from a peat deposit newly discovered by Carter and Heller during mapping in an adjacent source stream. Exposed in the bed of Mill Creek, the peat is ~1 m thick. It overlies alluvial strata consisting of rounded and imbricated pebbles and cobbles of quartzite, siltstone, and shale, and underlies an alluvial-colluvial deposit of subrounded to angular pebbles, cobbles and boulders of mostly quartzite in a finer-grained matrix of sand, silt, and clay. A sample was dated using the ¹⁴C method at greater than 50,000 yr B.P. (Jack McGeehin, USGS, 2014, written commun.). Pollen analysis shows the peat to contain an abundant arboreal assemblage dominated by Jack pine, with auxiliary oak, spruce, birch, and juniper, as well as minor fir, chestnut, larch, aspen, and hickory; pollen indicates a pre-Holocene depositional age (Ron Litwin, USGS, 2014, personal commun.). Combined, these new age data suggest Big Levels fan systems are extremely long-lived.

The longevity and unusual size of this fan complex stem from a combination of stratigraphic and structural influences on the topography. The distal end of this fan complex contains several large remnants of sloping surfaces comprised of old (F1) deposits (Whittecar and Duffy, 2000). Deeply weathered deposits reported in other Appalachian fans in this topographic setting have little or no surface expression (e.g., Grote, 2006). The preservation of particularly thick, old fan deposits and ancient fan surfaces in this area may result from dissolution of carbonate rocks within the Shady Formation and adjacent units that underlie many of the fans along the west edge of the Blue Ridge in Virginia. Locally those beds are credited for the sinkhole ponds and wetlands that pock the upper half of fans (Whittecar and Duffy, 2000), and for the water wells that reach 85 m depth in fans in this region (Hinkle and Sterrett, 1976; Simmons, 1988) and more than 120 m elsewhere in Virginia (King, 1950). In Virginia, Maryland, and Pennsylvania, much of the buried karstic surface below fans on the west slope of the Blue Ridge resembles a solutional trough (Chichester, 1996), an evolving accommodation space that could collect sediment for many millions of years. Accordingly, Whittecar and Duffy (2000) suggest that the oldest remnant fan surfaces are best preserved in areas underlain by non-carbonate rocks where little dissolution produces no new accommodation space. These fans sit near the drainage divide between the James and Potomac rivers, and in another case of topographic inversion, source few large streams to drive their erosion. Finally, an additional contributing factor to the size and longevity of this fan complex is the unusually large expanse of Antietam quartzite exposed in the Big Levels portion of Blue Ridge. In most areas of central and northern Virginia, this unit dips west at high angles and forms long narrow ridges; in the Big Levels area, the gentle dips and open folds that developed on the Antietam Formation in the zone between the overlapping terminations of two regionally extensive faults created a large cobble-generating plateau. The massive volume of quartzite cobble fan deposits that surround the Big Levels massif filled the adjacent carbonate trough and spread across the floor of the Shenandoah Valley.

From Stop 3B at Acres Sand and Stone, proceed north on Pine Trail for 1.1 miles to Lake Road; turn right (northwest) onto Lake Road and proceed 0.3 mile to Howardsville Turnpike; turn left (northwest) onto Howardsville Turnpike and proceed 1.9 miles to U.S.-340 North (Howardsville Turnpike becomes Draft Avenue after 0.4 mile); turn right (northeast) onto U.S.-340 North and proceed 5.3 miles to I-64 East; turn right (southeast) to merge onto I-64 East; proceed 5.1 miles southeast on I-64 East to Exit 99 for U.S.-250 toward Afton/Waynesboro; turn right (south) onto U.S.-250 East toward Afton/Charlottesville and proceed 0.2 mile to SR-610 on right; turn right (southwest) onto SR-610 and proceed 0.1 mile to entrance to Blue Ridge Parkway on left; turn left (east) and immediate right (south) onto Blue Ridge Parkway and proceed 0.2 mile south on Blue Ridge Parkway to Stop 4 at Afton Overlook on left.

Afton overlook on the Blue Ridge Parkway (Fig. 23) lies at an elevation of 626 m near the local crest of the Blue Ridge. The overlook provides an excellent venue to review the main topics of the first day and synthesize observations with more regional data on Appalachian landscape evolution (Fig. 9). Specifically, the comparison of long-term and short-term rates of erosion, and where these rates are distributed through the Appalachians with regard to bedrock, will be used to stimulate discussion on transient versus steady-state features in the landscape. Recent studies have taken advantage of new developments in the application and interpretation of low-temperature thermochronology data and have profoundly changed the way we look at origin of topography in the modern mountain range. In the White Mountains of New Hampshire (Roden-Tice et al., 2012) and the Blue Ridge Mountains of western North Carolina (McKeon et al., 2014), models indicate prolonged periods (~50 m.y.) of accelerated exhumation in valley floors relative to neighboring ridges and summits during the Mesozoic, long after tectonic deformation related to rifting. These spatially variable exhumation rates generated kilometer-scale relief in both areas, nearly matching the relief of the modern landscape and indicating that there is no need, and in fact it may be completely incorrect, to connect modern topography with the Paleozoic orogen. It is also interesting to note that the central Appalachians have the oldest apatite fission track ages in the range (and therefore the least postrift exhumation), but they are also the area with the most evidence for drainage divide migration (Prince et al., 2010, 2011) and rapid river incision (Ward et al., 2005; Miller et al., 2013). By combining these observations, it appears that the Appalachian landscape is neither uniformly decaying, nor finished evolving and as a result is something disconnected from its start as a collisional orogen. Despite sedimentary and geomorphic evidence for considerable unsteadiness, no Cenozoic cooling ages have been found for low-temperature thermochronometers (McKeon et al., 2014, and references therein; Roden-Tice et al., 2009, 2012), which means that no part of the landscape has experienced erosion and exhumation significant enough to reset ages during this time (~2-3 km for apatite U-Th/He). These observations do indicate that the evolution of the modern Appalachians has on the whole been rather slow; however, the techniques used lack the resolution to address the nuances of the processes.

The view to the east from the overlook is the erosional remnant of the breached Blue Ridge anticlinorium (Fig. 24). Resistant rocks here on the west limb of the anticlinorium, the Catoctin greenstone and Chilhowee Group quartzites and shales, dip steeply to the northwest or are overturned to the southeast (Gathright et al., 1977). The linear range of monadnocks on the farthest horizon to the east near Charlottesville is also underlain by Catoctin, but on the east limb of the anticlinorium. Metasedimentary and metavolcanic rocks of the Lynchburg Group (along the lower hills in front of the Catoctin ridges on the horizon) and Mesoproterozoic basement (beneath the Ragged Mountains to the southeast and in the valley below) comprise the excavated core.

The 1- to 3-km-wide Rockfish Valley high-strain zone occupies the floor of Rockfish Valley, below the overlook. Some suggest this zone of ductile deformation in the core of the Blue Ridge anticlinorium is a significant orogen-scale structure, separating Blue Ridge Mesoproterozoic basement into two distinct terranes with contrasting crystallization and metamorphic histories (Bartholomew and Lewis, 1984). Basement to the west of the shear zone comprises the Pedlar terrane, dominated by granulite-facies gneisses and high-temperature charnokitic intrusions; shallow-crustal amphibolite-facies rocks of the Lovingston terrane crop out to the east. Others attribute lithologic differences across the fault zone simply to the effects of varying post-Grenvillian regional metamorphic hydration (e.g., Evans, 1991). Greenschist-facies ductile mylonite in the high-strain zone is likely a product of Paleozoic orogenesis (Neoacadian to earliest Alleghenian); Jenkins et al. (2012) cite ⁴⁰Ar/³⁹Ar white mica age-spectra from penetratively deformed basement across the central Virginia Blue Ridge to conclude that Blue Ridge high-strain zones developed during the Mississippian. Earlier ⁴⁰Ar/³⁹Ar studies by Wooten et al. (2005) from Blue Ridge high-strain zones ~90 km to the northeast indicated early Alleghenian (320-260 Ma) cooling through ~350 °C. Regardless of age or tectono-orogenic significance, these mica-rich mylonitic rocks are extremely erodible, which allows the Rockfish River (a first-order branch of the James River) and its tributaries to carve out the core of the anticlinorium here.

We compare Blue Ridge erosional dissection along zones of lithologic incompetence here, with a similar factor facilitating Blue Ridge escarpment retreat ~160 km to the southwest in southern Virginia. There, the erosive power of Rock Castle Creek, a third-order tributary of the Roanoke River (via the Dan and Smith river systems) is entrenching a deep canyon into the Blue Ridge escarpment (Fig. 5), along a mica-rich mylonitic fault zone between two major eastern Blue Ridge units (Carter, 2012). Current models for Blue Ridge escarpment evolution (Harbor, 1996; Prince et al., 2010; Spotila and Prince, 2012) presuppose that over time, continued excavation along Rock Castle Creek will dissect a large massif from the escarpment, forcing the escarpment to retreat to the northwest and leaving behind an isolated monadnock east of the creek. This process will thus produce a landscape similar to the geomorphic lay of the Rockfish Valley below Afton overlook today.

From Stop 4 at Afton overlook, proceed 0.2 mile north on Blue Ridge Parkway to SR-610 on left; turn left (west) and make an immediate right (north) onto SR-610 and proceed 0.1 mile to U.S.-250; turn right (southeast) onto U.S.-250 East and proceed 4.6 miles to Stop 5 at Lebanon church on left.

In this road cut along U.S.-250 in front of Lebanon Church (Fig. 23), Nelson (1962) photodocumented a reverse fault (Fig. 25) that displaced crystalline bedrock ~1.5 m over terrace gravels (Crone and Wheeler, 2000), but a half-century of road maintenance and tree growth have now completely obscured the structure. The fault here is one of several that cut this deposit at the eastern foot of Blue Ridge (White, 1950). White (1952) mentions two small faults 4 km to the east at Yancey Mills that displace fluvial gravel up to 1 m. Prowell (1983) gives the orientation of the reverse fault here at Lebanon Church as N20°E, with 57° dip to the southeast.

Alluvial-colluvial material at this site consists of matrix-supported clasts of mostly rounded to subrounded pebbles and cobbles of greenstone, with few subangular to subrounded clasts of quartz, in a clay-rich matrix (mapped as terrace and alluvial-fan deposits by Gathright et al., 1977). Morphology suggests a complex alluvial fan system derived from Catoctin greenstone and deposited at the foot of the Blue Ridge in this area; east-flowing Stockton Creek has reworked and redeposited fan material at the toe of the fan complex into stream-parallel terraces, and possibly mixed material other than greenstone and quartz into the deposit (for example, Prowell reports graywacke clasts too; Prowell, unpublished data, in Crone and Wheeler, 2000). Thickness of the faulted deposit is unknown, but based on Nelson’s (1962) photograph, the gravel is at least 3 m thick.

Age of faulting is speculative, as is its origin. Prowell (1983) determined a Miocene to Pliocene age for these “high-level fluvial gravels,” but Crone and Wheeler (2000) suspect the deposit could be as young as Pleistocene, from Prowell’s (unpublished data) description of other fluvial gravels in this part of the Appalachians, and a 1:24,000-scale topographic map-based geomorphic analysis of the site. The landscape position, deposit weathering, and strath elevation above the valley bottom are all suggestive of at least a middle Pleistocene or older age. The fault has no geomorphic expression on existing 1:24,000 topographic maps, and LiDAR is not currently available east of the Blue Ridge in this region of Virginia. Although the structure is a NW-vergent reverse fault, its position at the foot of the Blue Ridge (and at the toe of a regional alluvial fan complex) does not rule out genesis from post-depositional non-seismogenic landslides and slumping.

On the west flank of the Blue Ridge, 20 km to the north of this locality, the Harriston fault (Wieczorek et al., 2004) cuts Neogene alluvium and colluvium blanketing the foot of the Blue Ridge (Doctor et al., 2014). The Harriston fault, which shows as a linear ridge on regional LiDAR imagery, has an orientation of N20°W that parallels joints and Mesozoic dikes in Valley and Ridge rocks, as well as cave passages beneath Cave Hill along the northerly trend of the fault (Doctor et al., 2014).

From Stop 5 at Lebanon Church, proceed 3.2 miles east on U.S.-250 East to 1-64 East; turn right (east) onto 1-64 East ramp, merge left onto 1-64 East, and proceed 28.9 miles to Exit 136 for U.S.-15 toward Gordonsville/Palmyra; turn left (northeast) onto U.S.-15 North and proceed 0.4 mile to Spring Creek Parkway; turn left (northwest) onto Spring Creek Parkway and proceed 0.1 mile to Wood Ridge Terrace; turn left (southwest) onto Wood Ridge Terrace and proceed 0.1 mile to entrance to Best Western Plus Crossroads Inn & Suites on left.

On the second day of the trip we will observe the geologic and geomorphic evidence for river incision into the Virginia Piedmont and Inner Coastal Plain, much of it in the vicinity of the 2011 Mineral earthquake, with the goal of using stratigraphic and geomorphic markers to document differential rock uplift.

Exit Best Western Plus Crossroads Inn & Suites parking area by turning right (northwest) onto Wood Ridge Terrace and proceed 0.1 mile to Spring Creek Parkway; turn right (southeast) onto Spring Creek Parkway and proceed 0.1 mile to U.S.-15; turn left (northeast) onto U.S.-15 North and proceed 6.6 miles north on U.S.-15North to VA-22 East; turn right (southeast) onto VA-22 East and proceed 2.3 miles to entrance to Virginia Vermiculite, LLC on left; turn left (north) into entrance to Virginia Vermiculite, LLC and proceed 0.5 mile to Stop 6 at Virginia Vermiculite.

There are two goals at this stop (Fig. 26). First, it offers an excellent opportunity to observe a very well-exposed and deep weathering profile into the rocks of the Virginia Piedmont in an operating vermiculite quarry. Second, a rare exposure of an intact South Anna River terrace is exposed in an abandoned highwall of the quarry.

The quarry operations are mining vermiculite from the Green Springs Pluton, an early Paleozoic (432 Ma; Sinha et al., 2012) diorite intrusion into the main body of the Potomac mélange (Rossman, 1991). The pluton is flat-topped and its roof is coincident with the present land surface. The diorite is deformed, fractured, and sheared. The vermiculite is likely a product of wall-rock metasomatism of biotite. The diorite is deeply weathered into a saprolite that extends for tens of meters in depth, and is particularly well developed along the fractures and shear zones. The saprolite is massive and featureless near the land surface, but becomes progressively more structured toward the bedrock interface. Solid bedrock is rarely observed, except in the deepest part of the pit, but the transition to solid bedrock is marked by a zone of corestone weathering where relatively intact cores of metagabbro are surrounded by grus formation along fractures (Fig. 27).

The conversion of bedrock to saprolite is an important rate limiting step in the erosion and transformation of the Appalachian landscape. Important studies on the Virginia Piedmont, albeit mostly in more quartz-rich schists, using chemical mass balance (Cleaves et al., 1970; reviewed in Cleaves, 1993) and meteoric ¹⁰Be cosmogenic techniques (Pavich et al., 1985, 1989) conclude that Appalachian Piedmont saprolite is not wholly relict from some past chemical weathering environment, but rather, it is being generated throughout the Pleistocene and Holocene at unsteady rates. The long-term average rate of conversion of bedrock to saprolite probably ranges from ~4.5-20 m/m.y., but most studies argue for the slower end of those rates (Fig. 9). The saprolite is converted into soil near the land surface and ultimately surface processes such as creep and overland flow remove the soil. In those parts of the Piedmont, namely the broad upland interfluves, where the conversion of rock to saprolite, saprolite to soil, and surface erosion all proceed at about the same rate, the landscape is approaching a steady-state relief. In other places, namely near a major river valley responding to base level fall at the Fall Zone, the rates of erosion (incision in this case) may be faster than the generation of saprolite, and relief is increasing (Costa and Cleaves, 1984). The Virginia Piedmont shows that the production of saprolite, often associated with low-relief, tropical landscapes, may in fact be throttled by the ability of a landscape to vigorously deliver carbonic acid and remove dissolved products within an active hydrologic system, which are characteristics of an incised landscape with locally high relief (Cleaves, 1993).

The abandoned eastern highwall of the quarry offers a rare exposure of a terrace of the South Anna River (Fig. 28). The base of the terrace is a distinct strath, with < 1 m relief, carved into the corestone-weathered diorite. The strath is littered with a single-clast thick layer of subrounded-rounded quartzite cobbles ~4-8 cm in diameter with local outsized ~40 cm clasts. The quartzite clasts are clearly not locally derived from the diorite, but rather must be sourced up-basin in the Piedmont, including the foothills of the Blue Ridge. The cobbles are both clast- and matrix-supported where the matrix is locally derived diorite grus. Curiously, a small number of the cobbles are faceted, resembling ventifacts, and most of them have a smooth, polished patina surface (inset, Fig. 28). The cobble layer is buried by ~0.5 m of pebbly alluvial channel facies sand composed of locally derived grus and up-basin quartz-rich material. The pebbly sand is conformably overlain by an overbank facies of clayey fine sand and silt ~1 m thick. An alternately gleyed and orange-brown soil with an argillic B horizon is developed through this alluvial package. The alluvial deposits are truncated and unconformably overlain by one or more colluvial wedges of matrix-supported locally and exotically derived angular clasts. The base of the colluvial facies is marked by a stone line. Reddish-brown soils with argillic and cambic B horizons are developed through the colluvial package.

Quartz and feldspar sand at the base of the alluvial facies were sampled for OSL and infrared stimulated luminescence (IRSL) geochronology, respectively. Not surprisingly, the feldspar, which is locally sourced from the diorite, returned a saturated age of 42.1 ± 10.7 ka. The exotically sourced quartz, however, returned a finite age of 69.2 ± 16 ka. Based on mapping through two quadrangles (Malenda et al., 2014) and geochronology of key deposits (Malenda, 2015), the alluvium here is determined to be terrace Qt4. The full South Anna fluvial stratigraphy and geochronology will be presented at Stop 9. The elevation of the Qt4 strath here places it ~3 m above the South Anna channel (Fig. 26) for an incision rate of ~43 m/m.y.

From Stop 6 at Virginia Vermiculite, LLC, proceed 0.5 mile to VA-22; turn left (east) onto VA-22 East and proceed 8.9 miles to VA-208/Courthouse Road on right; turn right (southwest) onto Courthouse Road and proceed 1.1 miles to SR-646/ Yanceyville Road on left; turn left (southeast) onto SR-646/Yanceyville Road and proceed 1.8 miles to SR-604/Roundabout Road on right; turn right (southwest) onto SR-604/Roundabout Road and proceed 3.5 miles to Table Top Alley on left; turn left (northeast) onto Table Top Alley and proceed 1.4 miles to Stop 7 at the Rogers property.

The goal of this stop is to observe the Qt4 terrace near the epicenter, but in the footwall of the Quail fault. The site is located a few kilometers downstream of a major bedrock step in the South Anna longitudinal profile called the Byrd Mill knick-point (Figs. 10, 29). At Byrd Mill, where VA Route 649 crosses the river, the South Anna River channel rises ~2 m in elevation, with the top of the knickpoint at 94.5 m (310 ft). The Byrd Mill knickpoint is the crest of a broader knick zone that stretches to Yanceyville, ~4 valley kilometers downstream from the present location (Figs. 10, 29).

At Stop 7, there is a distinct terrace tread at ~94 m (310 ft) that aligns with the top of the Byrd Mill knickpoint ~6 valley kilometers (~9 river kilometers) upstream. The elevation alignment opens the possibility that this terrace was formed by headward propagation of the Byrd Mill knickpoint and rapid abandonment of the former South Anna floodplain. Knickpoint regression modeling parameterized by an assumed post-Yorktown Formation age (2 Ma) for the top of the knick zone at 52 m (Fig. 10) indicates that the Byrd Mill knickpoint is migrating through the South Anna River trunk channel at a rate between ~7-14 km/m.y. (Malenda, 2015). That would place the knickpoint at this present location ~1 m.y. ago, and predicts that the age of the terrace should be ~1 m.y.

The Rogers terrace (Qt4) tread is underlain by intact, fining-up fluvial facies characterized by a vein quart and quartzite, open-framework, sandy-cobble channel facies ~1-1.5 m thick, overlain by 1-2 m of yellow fine sandy silt and clay overbank facies, with possible aeolian inputs. The soil formed through this alluvium appears to be cumulic (Fig. 30). It is distinctly brownish-yellow to orange-brown in color and locally gleyed, particularly at the cobble-overbank contact. The upper ~30 cm is a brownish-yellow weak argillic horizon that grades down into an orange-brown massive argillic horizon with a weak, gleyed, prismatic structure. Throughout this part of the Virginia Piedmont, this particular alluvial soil is commonly, but not exclusively, mapped as the Masada series.

This terrace was sampled for OSL and IRSL geochronology from a pit hand dug to a depth of 1.8 m. The samples were taken from the base of the overbank facies. The OSL date is saturated at ~50 ka, but the IRSL date returned a finite age of 81.5 ± 14.4 ka. Within error, this terrace and the terrace observed at Stop 6 are the same age, have similar soil characteristics, and are correlated as Qt4. The base of this terrace lies at an elevation of ~93 m (305 ft), which is ~10 m above the channel, for an incision rate of ~110 m/m.y. The much faster rate of incision here, as compared to Stop 6, may be related to the fact that this part of the South Anna River channel is steeper, being part of the knick zone between Byrd Mill and Yanceyville. However, the 81.5 ka age and the soil stratigraphy and morphology make it clear that the terrace is not related to the passage of the Byrd Mill knickpoint, although passage of smaller knickpoints that are ~1 km upstream probably contributed to the accelerated incision observed at this location. Another factor that needs to be considered, and will be addressed more directly at the next stops, is the possibility that the deeper incision is being caused by active tectonics and uplift of Piedmont rocks in the vicinity of the Mineral earthquake.

From Stop 7 at the Rogers property, proceed 1.4 miles southwest on Table Top Alley to SR-604/Roundabout Road; turn right (north) onto SR-604/Roundabout Road and proceed 2.2 miles to SR-714/Horseshoe Farm Road on left; turn left (northwest) onto SR-714/Horseshoe Farm Road and proceed 1.1 miles to entrance to Horseshoe Farm on left; turn left (south) into Horseshoe Farm and proceed 0.8 mile south to Stop 8 at Horseshoe Farm.

Since the 2011 Mineral, Virginia, earthquake, there is renewed interest in paleoseismite study in the central Virginia seismic zone. Such work is critical to thoroughly understand regional seismicity and recurrence intervals of the largest magnitude quakes for seismic hazard analysis. This is particularly true in seismic zones characterized by historical low-magnitude and relatively low-frequency events, such as central Virginia (Bollinger and Sibol, 1985; Tarr and Wheeler, 2006).

Much focus is along the South Anna River in the epicentral area of the 2011 temblor (Figs. 29, 31). Here, postquake surficial and bedrock framework studies suggest that this is a zone of long-lived seismic activity along faults that have reactivated many times in the geologic past (Burton et al., 2014, 2015a; Berti et al., 2015). The South Anna River crosses the up-dip projection of the Quail fault zone (M. Chapman, Virginia Tech, 2011, personal commun.), and significant riverbank slumping, rock fall, and all of the small liquefaction features produced by the M 5.8 main shock were discovered along this section of the river (Carter et al., 2012; Green et al., 2015). Work in this area follows that of Obermeier and McNulty (1998). Their paleoseismite reconnaissance of the CVSZ led to the discovery of three sites with potential paleoliquefaction features. The site on the South Anna (M. Tuttle and T. Bush, Tuttle and Associates, 2011, personal commun.) and another along the James River near State Farm (S. Schindler and R. Harrison, USGS, 2012, personal commun.) were revisited following the earthquake, and additional possible paleoliquefaction features were found.

It is along this segment of the South Anna River that the U.S. Geological Survey is currently documenting all primary depositional, pedogenic, and possible paleoliquefaction features exposed in alluvium along the riverbanks, for the purpose of creating a first-of-its-kind baseline inventory of alluvial and possible seismogenic features in an upland Piedmont setting. Characteristics of paleoliquefaction features from central and eastern U.S. seismic zones (CEUS) are well recorded (e.g., Obermeier, 1996, 2009), but this knowledge base comes mostly from studies of the New Madrid seismic zone in the Mississippi embayment (e.g., Tuttle et al., 2002) and around Charleston, South Carolina (e.g., Talwani and Schaeffer, 2001)—areas with surficial geology prime for both creating, and preserving liquefaction features in Cenozoic stratigraphy. Until a thorough and well-documented canvas of all alluvial features in this region is conducted and compiled, suspect paleoliquefaction features found along the South Anna and other CVSZ river systems can only be compared to well-preserved examples elsewhere in the CEUS.

At Horseshoe Farm on the South Anna River, evidence for multiple red, buried soil horizons (paleosols7) beneath young entisols and inceptisols developed in modern fluvial deposits are preserved in riverbank alluvium (Fig. 32). Here, the modern channel is incising through what can be interpreted as point bar, overbank, and channel plug deposits. Multiple abandoned river channels attest to several episodes of river avulsion, including cutoff of the namesake horseshoe. Unconsolidated coarsegrained sediments represented by thin (cm to dm thick) beds of sand and gravel are packaged as lenticular point bar sets up to about 1 m thick, within finer-grained sandy silt and clay-rich muds (Fig. 32). Individual beds within these sets are mostly planar bedded and sorted by grain size, but some are cross-bedded or crudely graded-bedded; dark, heavy mineral-laden beds are common. Isolated beds of sand or gravel, less than several cm thick and laterally discontinuous (< 1.5 m), also occur within the silt and clay, and could be levee blowout flood deposits.

Silt- and clay-rich sediment that encase unconsolidated coarser-grained sand and gravel beds preserve evidence for several cycles of paleosol development. Redoximorphic concentrations give these alluvial sediments typical 2.5YR to 5YR hues, and with illuviated clays, impart multiple distinct profile horizons, up to 0.3 m thick, that stand out in relief along the riverbank. Together, these characteristics define this package of finer-grained sediments as what is best described as a highly iron-indurated hardpan. Of interest are oversized clasts from local bedrock sources (mostly Ellisville Granodiorite and Chopawamsic Formation metasandstone) that occur as isolated cobbles and boulders throughout the hardpan; some remain competent rock, but most have been completely saprolitized in situ, suggesting that they are old (Pavich, 1986, 1990). Capping the hardpan is ~1.5 m of a young but developed soil profile within modern alluvium.

These data indicate the modern South Anna River is currently re-incising into alluvium deposited during a former period of aggradation along this stretch of the river. Published OSL ages from Harrison et al. (2015) shed light on the depositional and landscape evolution in the vicinity of the horseshoe: upland gravels, capping hills ~20 m above the modern flood-plain, yield ages of ~25-42 ka, with sediment from highest terraces along the modern floodplain dated at ~17 ka. Unresolved is the time of deposition of this alluvium, currently at low-water level in the bed of the modern river channel. Bed elevation along the channel of the Horseshoe indicates avulsion and abandonment since ~17 k.y. It is curious, however, that another abandoned channel, and the modern river, cut across and incise the hardpan, rather than skirt around it as does the abandoned horseshoe. Could the abandonment of the horseshoe, in favor of a new channel route through the iron-indurated hardpan, be the result of an earthquake?

The hardpan may serve as a natural paleoseismoscope by acting as a rigid pad during past large magnitude events. Features in the hardpan include a network of polygonal, cm-thick, clay-filled seams, cracks, and fractures (Fig. 33). These are not paleo-liquefaction dikes—neither the clay-filled seams, nor the clayey silt through which they propagate, are of the proper grain size to liquefy during a seismic event. The most reasonable explanation is that the polygonal network of fractures is formed from desiccation cracks developed in the paleosol. Several cracks show evidence for pedogenic modification by meteoric and shallow groundwater flow: redoximorphic reduction (gleying) and clay illuviation in the cracks (Fig. 33).

One feature of interest preserved in the claypan is a several cm-thick sand-filled dike that rises from unconsolidated sand and gravel into the overlying claypan (Fig. 33), which serves as an impermeable cap. Planar beds in the unconsolidated sand and gravel rise gently into the neck of the dike at its base; within the claypan, the dike bifurcates into several sills before pinching out into the network of clay-filled seams near the top of the hard-pan. These observations do fit basic criteria for paleoliquefaction (e.g., Obermeier, 2009), such as: (1) a liquefiable source strata (unconsolidated sand and gravel); (2) disturbed bedding in the source zone (beds rise gently into the neck of the dike); (3) intrusion through impermeable capping strata (iron-indurated hard-pan); and (4) pinch-out terminations along sills, or at the vertex of the dike.

This feature holds significant corollary for assessing seismic hazard in the CVSZ, regardless of its seismogenesis. If it is seismogenic, a reasonable model for its formation is that at some time in the past seismic history of the CVSZ, a large magnitude event jarred open properly oriented desiccation cracks in the iron-indurated hardpan, allowing liquefied sand from underlying unconsolidated layers to rise up into the impermeable paleosols. However, significant post-formation modification by pedogenic and redoximorphication processes highlights one very important attribute here: Intense weathering and soil processes may destroy, mask, or mimic characteristics of paleoliq-uefaction features in the humid subtropical upland Piedmont environment of central Virginia. For this reason, it is critical to document all features observed in South Anna River alluvium, for baseline comparison and assessment throughout the CVSZ. If, on the other hand, this feature is not seismogenic, then negative evidence for liquefaction in this potential and highly susceptible source material (i.e., liquefiable unconsolidated sand beneath a rigid and mostly impermeable hardpan) must place limits on the maximum levels of prehistoric ground shaking (Obermeier, 2009) in the CVSZ.

From Stop 8 at Horseshoe Farm, proceed 0.8 mile north to SR-714/Horseshoe Farm Road; turn right (southeast) onto SR-714/Horseshoe Farm Road and proceed 1.1 miles to SR-604/ Roundabout Road; turn right (southwest) onto SR-604/Round-about Road and proceed 0.8 mile to Roundabout Farm entrance on left; turn left (southeast) into Roundabout Farm and proceed 0.2 mile southeast to Stop 9 at Roundabout Farm.

The goal of this stop is to observe the truncated remnants of middle Pleistocene terraces and colluvial wedges in the Piedmont landscape and put forward a model for terrace stratigraphy along the South Anna River. This stratigraphy defines the geomorphic markers that can be used to measure any differential rock uplift associated with the active tectonics responsible for the 2011 Mineral earthquake.

The Roundabout Farm site (Fig. 34) has been studied extensively and in detail by Richard Harrison (USGS) and access to the site has been graciously granted by the Hopkins family. Several trenches have been excavated on the site and these have been instrumental in defining the alluvial and colluvial stratigraphy of this part of the Piedmont (Fig. 29). Here we summarize the stratigraphy and geochronology of three backhoe dug trenches (trenches A, B, and C) and one hand dug trench (trench E) that constrain the South Anna incision rate for the footwall of the fault that ruptured during the Mineral earthquake. Following from Horton et al. (2015), we will call this fault the Quail fault.

Trench A is excavated along the dirt road extending from the ~107 m (~350 ft) upland down the nose of topography oriented south-southeast toward the South Anna River. This trench exposes two or more colluvial(?) packages of ~1-m-thick mostly matrix-supported vein quartz and quartzite fluvial gravel in a sand, silt, and clay matrix (Fig. 35). At the time of the writing of this paper, discussions continue on the precise interpretation of these packages of surficial sediment—specifically, are they colluvial (favored by Pazzaglia) or at least in part alluvial and in place (favored by Harrison)? The soils formed in these colluvial(?) packages are distinct in their colors and clay content. The upper deposits are reddish-brown in color, have weak argillic B horizons, and are relatively clast-poor. The lower deposit is distinctly red in color and contains a well-developed argillic B horizon with subangular blocky structure. The base of the lower deposit is particularly clast rich with rounded, imbricated (?) fluvial cobbles 2-6 cm in diameter. Here and throughout the Piedmont, these soils are mapped usually, but not exclusively as the Turbeville Series. In contrast to the Masada and soil series observed at Stop 7, these soils are much better drained and are more red in color. The basal, red, colluvial(?) deposit at the head of trench A has been OSL dated at 42.8 ± 4.3 ka. The colluvial(?) packages are in turn unconformably overlain by a massive silt ~30-40 cm thick, interpreted as a Holocene aeolian deposit (Fig. 35; Harrison et al., 2015). Samples have been collected from trench A for use in developing a cosmogenic geochronology.

The colluvial(?) units unconformably overlie a residual soil that grades downward into massive and structured saprolite down slope. The buried residual soil exposed at this site is unlike the stratigraphy observed at Stops 6 and 7, where stratified alluvial material is observed to unconformably overlie saprolite. It is unlikely, but not impossible that the carving of a strath in a valley bottom would allow for the preservation of a residual soil below the alluvial material. It is more likely that a residual soil is preserved in a hillslope position, below colluvial wedges. The saprolite in trench A has numerous fractures that continue into the bedrock. These fractures offset textural and compositional markers in the bedrock and saprolite, but their age and connection to CVZS active tectonics remains enigmatic.

Trench B was excavated into a colluvial hillslope at an elevation of ~99-101 m (~325-330 ft). Here, fluvial gravel-rich colluvial wedges in a brownish-red clay-rich matrix were encountered. Two OSL ages of 23.8 ± 1 ka and 27 ± 1 ka were obtained.

Trench C was excavated into a topographic hollow ranging from ~93-95 m (~305-311 ft). As with trench B, brownish-red, matrix-rich colluvium was encountered. The base of the colluvium was OSL dated to 28.3 ± 2.2 ka and 27.5 ± 2.5 ka. An OSL date of 22 ± 0.6 ka was obtained stratigraphically higher in the colluvial package. Further down in the hollow at an elevation of ~94 m (~307 ft), gray, fine-grained sandy-silt alluvium that caps the colluvium and fills the hollow and tributary stream it feeds was OSL dated at 17.2 ± 0.7 ka.

Trench E (Figs. 34 and 36) was hand excavated to a depth of ~1.5 m at an elevation of ~107 m (~350 ft) that is topographically higher, but stratigraphically older than the head of trench A. The upper meter of the pit is a good match to the brown over red colluvial(7) stratigraphy observed in trench A. Below the lower red colluvial(7) unit, the trench intersected a hard, indurated bed ~25 cm thick of clast-supported quartzite and vein quartz cobbles with a sparse sandy matrix. A cobble weathering out of this horizon and exposed along the road nearby was found to contain Skolithos trace fossils, evidence of its extrabasinal source and possible reworking from much older alluvial deposits no longer preserved in the Piedmont upland. The cobbles sit unconformably atop a hard, red, structured saprolite. Although the colluvial nature of this cobble unit cannot be discounted, it offered the best evidence for the remnants of an intact alluvial deposit high in the Piedmont landscape, some 23 m above the South Anna channel. Three samples for OSL and IRSL geochronology were obtained from this site. All of the OSL samples were saturated, returning minimum ages ~50 ka. The IRSL age from the base of the cobble layer returned finite ages of 391.8 ± 44 ka and 407.8 ± 55 ka. The IRSL age from the top of the cobble layer returned a finite age of 212.3 ± 29 ka (Fig. 36). These ages may reflect mixing of materials both below and above the cobble layer, respectively, but they consistently indicate that the cobble layer is middle Pleistocene in age and significantly older than the late Pleistocene colluvium and alluvium dated in the other trenches. Using the older, average age of ~400 ka for the cobble layer results in a long-term river incision rate of ~58 m/m.y., similar, but more rapid than the rate measured at Stop 6, albeit for a younger terrace.

The stratigraphy and geochronology at the Roundabout Farm, additional data from a similar locality downstream at Yanceyville called the Morgan Farm-Thomas property, and a 1:24,000-scale surficial geologic map of the Ferncliff and Pendleton collectively inform a terrace stratigraphy for the South Anna River (Fig. 37). The elevations of the terraces are only relevant for the South Anna River section between Roundabout Farm (Stop 9, VA-604) and Yanceyville (Morgan Farm-Thomas property, VA-605). We propose five major strath terraces where each terrace unit may be a suite of two or more straths of similar elevation, recognizing the uncertainties that arise without clean, vertical exposures exhibited at Stop 6. The key point is that at several locations along the South Anna River (Stops 6, 7, 11, and Morgan Farm-Taylor; Fig. 29), we have mapped and dated a terrace with similar sedimentologic texture, composition, soils, and age roughly coincident with MIS 4. This is our Qt4 terrace and it forms an important geomorphic marker across the hanging wall and footwall of the Quail fault. The terrace and alluvium and associate colluvial aprons at the Roundabout Farm are assigned to terrace Qt3, forming another, more poorly dated geomorphic marker horizon. Our stops downstream from this location later in Day 2 will compare the elevations and incision rates from terraces Qt3 and Qt4 in the hanging wall of the Quail fault.

From Stop 9 at Roundabout Farm, proceed 0.2 mile northwest to SR-604/Roundabout Road; turn right (northeast) onto SR-604/Roundabout Road and proceed 2.0 miles to SR-646/ Yanceyville Road; turn left (north) onto SR-646/Yanceyville Road and proceed 1.8 miles to VA-208/Courthouse Road; turn right (north) onto VA-208/Courthouse Road and proceed 0.5 mile to SR-630 on left; turn left (northwest) onto SR-630 and proceed 0.2 mile to entrance to Weston Winery; proceed northwest for 0.4 mile to Stop 10 at Weston Winery.

The goal of this stop, in addition to lunch, is to observe the installation of one of the Plate Boundary Observatory (PBO) GPS sites that were completed as part of an NSF-RAPID (Rapid Response Research) project following the 2011 earthquake. The receivers, one in the footwall here and one in the hanging wall of the Mineral earthquake fault ~10 km southeast from this location at Yanceyville, were installed in early November, 2011 during a time when there were still numerous and sizable aftershocks from the 23 August main shock. The installation was probably too late to capture immediate post-seismic elastic relaxation, but the idea is to slowly begin to build a network of high-quality receivers that can detect secular trends in elastic strain accumulation and release in eastern North America. It will take many years, if not decades to assemble an interpretable data set. Part of the time series from the two receivers installed in 2011 are shown in Figure 38.

From Stop 10 at Weston Winery, proceed 0.4 mile southeast to SR-630; proceed 0.2 mile east on SR-630 to VA-208/Court-house Road; turn left (northeast) onto VA-208/Courthouse Road and proceed 0.5 mile to East Main Street; turn right (southeast) onto East Main Street and proceed 0.4 mile to US-33 East; yield right (southeast) onto US-33 East and proceed 7.2 miles to US-522 South; turn right (southwest) onto US-522 South and proceed 0.7 mile south on US-522 to SR-699/Indian Creek Road on right; turn right (west) onto SR-699/Indian Creek Road and proceed 3.5 miles to Stop 11 at Cox Farm.

The goal of this stop is to review the terrace lithostratigraphy, geochronology, correlation, and incision rates for South Anna River terraces in the uplifted hanging wall of the Mineral earthquake fault. The stop provides an overview from a vantage point where the terrace treads along this part of the South Anna River can be easily viewed (Fig. 39). In summary, there are at least two more terraces in the hanging wall of the Quail fault that are not present upstream of Yanceyville. Additionally, the Qt4 and Qt3 terraces project into the hanging wall, but rather than falling with the gradient of the river and remaining subparallel to the South Anna long profile, they gently rise in elevation through the reach here at Stop 11, before gently falling downstream toward the town of South Anna. The terraces have not been mapped further downstream, but doing so could strengthen the proposed correlation (Fig. 40).

Terrace Qt4 has been described from a 2-m-deep hand dug pit at Harts Mill Road where the strath unconformably overlies saprolite at an elevation of ~325 ft (99 m). Here the alluviual, colluvial, and soil stratigraphy closely matches what has been described for Qt4 in the footwall at Stop 7. Samples have been collected and analyzed for OSL and IRSL geochronology. The OSL samples are saturated and return minimum ages of ~50 ka. The IRSL sample returns a finite age of 65.5 ± 8.1 ka. The resulting incision rate is a very rapid 395 m/m.y., an eight-fold increase over the rates calculated for the footwall. Similarly, the Qt3 terrace has a base at ~110 m (~360 ft), and the longer-term rate of incision based on this older terrace would be 91 m/m.y., or twice what is measured for the corresponding terrace in the footwall.

It is difficult to imagine incision proceeding so rapidly in the Piedmont landscape; after all, these are rates commensurate with those measured for the deepest canyons in the American west. It is possible that the IRSL geochronology is dating colluvium, rather than alluvium in place, but the texture and soils of Qt4 here in the hanging wall argue against that. It is also possible that we are dating alluvium of a thick fill far above the strath and thus grossly underestimating the strath age. Assuming for the moment that the incision is this rapid, there are probably several contributing factors. First and foremost is the fact that Stop 11 lies downstream of several large knickpoints, stepping up to Yanceyville (Fig. 10). Recent passage of one or more of these knickpoints through the Stop 11 reach would result in a rapid transient incision rate. Second, and of particular interest to this field trip, is the possibility that the hanging wall of the Quail fault is a growing fold, cored by a blind thrust.

The distinct change in sinuosity and valley width of the South Anna River downstream of Yanceyville where it enters the uplifting hanging wall of the Mineral earthquake fault is further evidence for active rock uplift (Berti et al., 2015; compare sinuosity in Fig. 29 and Fig. 39). As described in Berti et al. (2015), the South Anna channel would be expected to become more sinuous to lengthen its path to reduce its gradient where its valley is being steepened by active rock uplift.

A simple, but useful exercise would be to apportion the observed long-term incision to the background rate fixed by longterm base level fall at the Fall Zone, local incision caused by the transient passage of a knickpoint, and local incision caused by the active tectonics of a growing hanging wall anticline. The long-term incision rate that minimizes the effect of local, transient, knickpoint retreat is represented by Qt4 at Stop 6, and Qt3 at Stop 9, both in the footwall of the Quail fault. In both cases, it is ~50 m/m.y. If the long-term incision rate in the hanging wall is roughly twice that rate, then the inference is that the hanging wall anticline is also growing at a rate of ~50 m/m.y. A Mineral earthquake-sized event is modeled to generate ~7 cm of distributed deformation, represented by rock uplifted in the core of the hanging wall anticline, at the land surface (Walsh et al., 2015). In order to generate 50 m of rock uplift in the past 1 m.y., there would need to be 700 Mineral-sized events, with a recurrence interval of 1500 yr. Although crude and carrying huge uncertainties, the rates, scale of deformation, and frequency of earthquakes are all consistent with what we know about the CVSZ and seismicity in the eastern United States.

From Stop 11 at Cox Farm, proceed south 1.6 miles on SR-699/Indian Creek Road to SR-640/East Old Mountain Road; turn left (southeast) on SR-64/East Old Mountain Road and proceed 3.0 miles to SR-629; turn right (east) onto SR-629 and proceed 0.1 mile to U.S.-522; turn right (south) on U.S.-522 and proceed 7.5 miles to I-64 East; turn left (southeast) at entrance ramp to merge onto I-64 East; proceed southeast on I-64 East for 14.9 miles to Exit 175 for VA-288 South toward Chesterfield; yield right (south) onto VA-288 South and proceed for 8.1 miles to VA-711 Huguenot Trail; at top of exit ramp, turn left (east) onto VA-711/Huguenot Trail (Huguenot Trail becomes Robious Road) and proceed 5.4 miles to VA-147/Huguenot Road; turn left (northeast) onto VA-147/Huguenot Road and proceed 1.0 mile to Woodmont Drive on right; turn right (southeast) onto Woodmont Drive and proceed 0.2 mile to Medina Road; turn left (east) onto Medina Road and proceed 0.4 mile to Traymore Road; turn right (south) onto Traymore Road and proceed 0.1 mile to entrance to Woodmont Recreation Center on right; turn right (west) into parking area for Stop 12 at Woodmont Recreation Center.

Exposed in the creek bank adjacent to Woodmont Recreation Center (Fig. 41) is a well-preserved section of middle Miocene marine clay, nonconformably deposited above Piedmont saprolite (Permian Petersburg Granite), and disconformably buried by high-level Bon Air gravel (Fig. 42). This exposure serves as an opportunity to observe innermost Coastal Plain stratigraphy west of the Fall Line and discuss regional correlations, with implications for significant neotectonic uplift. The contact between gravel and clay is exposed in the creek bank across from the parking area; the Petersburg-clay contact is exposed ~100 m downstream opposite the abandoned tennis courts.

Less than 2 m of high-level Bon Air gravel is preserved in the stream bank next to the parking lot, but what is exposed is very characteristic of the unit (Goodwin, 1980; Carter et al., 2007b). Gravel consists of unsorted, clast-supported, well-rounded pebbles and cobbles, in a yellowish-brown to reddish-brown sandy to clayey sand, grus-like matrix. Clast composition is dominantly vein quartz and quartzite (very few containing the trace fossil Skolithos), with few igneous and metamorphic clasts (most are identified as from local sources—Petersburg Granite and other easternmost Piedmont lithologies). Clasts of Mesozoic rocks are very rare. Some of the quartzite clasts are deeply weathered, with thick weathering rinds, to nearly decomposed, but most are competent. Oddly, in many outcrops, some well-rounded clasts are broken, with sharp, angular edges, indicating minimal transport and rapid deposition after breakage.

Gravel at this outcrop is crudely bedded, with rust lines and thin (<0.5-cm-thick) ferricrete layers bounding several dm-thick beds with slightly contrasting grain size and matrix composition. Lenses of gravel-poor sand, less than several dm-thick and laterally discontinuous (< 1 m), occur in many exposures of Bon Air gravel, but these are difficult to locate in this outcrop. Here, the upper 0.5 m of dull grayish-yellow sand and very unsorted gravel is likely reworked colluvium.

The nature of the contact between Bon Air gravel and underlying clayey silt is debatable. Goodwin and Johnson (1970) believed that the contact is conformable; Johnson et al. (1987) reported it to be scoured and unconformable. Here, the contact between clast-supported gravel with sandy grus matrix and underlying clay is sharp, and marked by both significantly larger cobbles and a layer of ferricrete. Close inspection shows that the uppermost clay bed (28 cm thick) is laminated, but contains small, mostly pea-size pebbles throughout (Fig. 42). Below is light-gray laminated clay with no pebbles. These data can support arguments for conformity with overlying gravel, but reworking of this uppermost clay bed during regression is unlikely, as the pebbly clay is laminated. The distinct lithologic change at the sharp contact suggests a disconformity between overlying sandy gravel and underlying pebbly clay beds. Bon Air gravel demonstrably overlies eastern Piedmont rocks along a regional nonconformity that also planes the unconformable contact between clayey silt and Piedmont basement (Carter et al., 2007b, 2010).

Clayey silt below the gravel is dark greenish-gray to medium bluish-gray, but weathers yellowish-brown, with iron-oxide mottling and staining from disseminated iron sulfides. Clayey silt is massive to finely laminated, and locally interbedded with thin (several cm to dm thick), discontinuous feldspathic sand and clast-supported gravel layers. Clasts within these gravel interbeds are typically well-rounded, quartz pebbles, but other lithologies like granite, metamorphic rocks, sandstone, and chert have also been identified. Within clay-rich beds, subangular to subrounded quartz granules, and abundant flakes of mica are common. Beds dip gently ~6° to the south (Carter et al., 2007b). A basal gravel lag deposit, with a maximum thickness of ~0.3 m, occurs locally at the unconformable contact with underlying Piedmont rocks, and consists of clast- and matrix-supported subangular to well-rounded pebbles, cobbles, and boulders of mostly vein quartz and granite. About 2 m of clayey silt is exposed in this outcrop, but the unit attains a thickness of ~5.5 m in this area.

Petersburg Granite at this exposure is completely saprolitized, but texture and structure within it are still clearly recognizable. The variety of granite here is porphyritic, with potassium feldspar phenocrysts ~2.5 cm in length, which despite saprolitization still retain their pale-pink color. The K-feldspar pheno-crysts are weakly aligned subparallel to biotite foliation in the matrix, which also consists of subidiomorphic quartz and plagio-clase. Northwest-oriented joints and mm-thick chalcedony-filled fractures complete the list of structures within the granite. Of particular note is the absence of a reworked zone at the scoured contact with the overlying estuarine to marine clayey silt—these rocks were not saprolite at the time of marine transgression!

Johnson et al. (1987) considered the clayey silt to be middle Miocene (equivalent to the Choptank and/or Calvert Formations) or older in age, as it underlies high-level gravels they interpreted to be upper Miocene (equivalent to the Eastover Formation). Carter et al. (2007b) compared whole rock geochemistry of clayey silt west of the Fall Line and lower Chesapeake Group sediments of similar lithology east of the Fall Line in this region, and found closest correlation with strata from ~3 m above Paleogene sediments, likely equivalent to the Choptank and/or Calvert formations in agreement with Johnson et al. (1987). Most recently, Weems et al. (2012) confirmed both a middle Miocene age for this unit using dinoflagellates from samples collected at this outcrop, and correlation with the lower Chesapeake Group Choptank Formation. These new data not only corroborate the Johnson et al. (1987) interpretation for the age of the clayey silt, but also lend credence to the Carter et al. (2007b) use of whole rock geochemistry in lieu of detailed paleontologic studies for reconnaissance-level regional correlation of major units across the Fall Line. Most importantly, these data provide crucial evidence for nearly 60 m of offset between middle Miocene clayey silt west of the Fall Line and equivalent strata to the east, though some of this difference in elevation can be attributed to regional base-level slope. To date, direct evidence for significant faults between these outcrop areas is lacking.

At the downstream outcrop near the tennis courts, however, small faults do offset clayey silt beds by several decimeters, with a normal sense of motion. Millimeter-thick clay seams along some of these structures are slickenlined. It is unknown if these small displacement faults are part of a regional system, or simply landslide features in the cut bank of this deep ravine.

In the absence of paleontologic data, most past workers cited regional sea-level curves (Fig. 43) to suggest that high-level Bon Air gravels are also Miocene in age (Johnson et al., 1987; Weems and Edwards, 2007), and likely equivalent to the upper Miocene Eastover Formation. These works regarded Bon Air gravels as regressive fluvial deposits following the middle Miocene sea-level highstand that deposited the underlying clayey silt unit (e.g., Carter et al., 2007a), as sea level was highest during Miocene time. For example, sea-level data indicate that at no time since the Miocene has sea level been high enough to flood the continental margin, assuming tectonic quiescence and a static margin (Fig. 43).

Miocene sediments of the lower Chesapeake Group (Ward and Blackwelder, 1980), including the Calvert, Choptank, St. Marys, and Eastover formations, underlie a large portion of central-eastern Virginia just east of the Fall Line (Mixon et al., 1989), but not all may be present in the Richmond area. At the Fall Line (which parallels I-95 in central-eastern Virginia), ~15 m of lower Chesapeake Group clayey silt sediments abruptly abuts Petersburg Granite (Carter et al., 2007a). These sediments are locally fossiliferous, and include various species of Chesapectin, Mercenaria, and other marine mollusks (Daniels and Onuschak, 1974). These sediments thicken to more than 30 m just east of Richmond. Very thin pebble lines, typically less than a couple of centimeters thick (i.e., the thickness of one layer of pebbles), occur within lower Chesapeake Group clayey silt and likely separate constituent members, including the Eastover Formation (Marr and Ward, 1987; de Verteuil and Norris, 1996; C.R. Berquist, VDGMR, 2005, personal commun.). However, it is difficult to reconcile that as much as 15 m of Bon Air gravel immediately west of the Fall Line at Richmond reduces to a feather-edge pebble line in equivalent sediments less than 10 km to the east. Weems and Edwards (2007), however, counter by pointing out that the upper portions of these units that once may have been present were eroded and planed before deposition of the succeeding member. In other words, the thin gravel lines may be all that is left of thicker still-stand progradation sand and gravel deposits that may have existed.

Regional faulting, however, may allow for correlation of Bon Air gravel with younger upper Chesapeake Group units east of the Fall Line. Approximately 18 km to the east, Carter et al. (2007a) provided a measured section of upper Chesapeake Group sediments (nearshore equivalent of the Yorktown Formation) at Chickahominy Bluff. They demonstrated a triumvirate partition of the 10-m-thick unit, unconformable above lower Chesapeake Group clayey silt: brownish-yellow, silty quartz sand to quartzite pebble gravel at the base, overlain by flaser and lenticular bedded, sandy clay and clayey fine-grained sand, overlain by gravel and upward fining, clayey sand (Fig. 44). These three units represent the Pliocene transgressive-regressive sequence: basal gravel (transgression), sandy clay and clayey sand (high-stand), and upper gravel and sand (regression). These lithologies extend westward to the base of the Chippenham scarp, where fine (pea-size) gravel of the upper gravel and sand package mantles the flat topographic surface (Richmond Plain) below the scarp. The westward extent of the lower package of transgressive gravels has not been thoroughly established in light of the recently established regional faulting that must exist in this area.

We suggest the possibility that prior to faulting, the lower transgressive package may have extended farther west than the Chippenham scarp; i.e., the lower Pliocene Yorktown gravel package at Chickahominy Bluff east of the Fall Line is equivalent to the Bon Air gravel west of the Fall Line. Syndepositional faulting, however, raised the coastline shortly after the onset of transgressive deposition, stranding the Bon Air gravels and reestablishing the westernmost extent of the Yorktown depositional basin at the base of the scarp. Syndepositional faults, well-documented by Berquist and Bailey (1999) in equivalent Pliocene sediments 60 km to the south along the Fall Line, provide supporting evidence. In this model, mid-level terrace gravels west of the Fall Line are equivalent to uppermost Yorktown regressive sediments east of the Chippenham scarp and at Chickahominy Bluff. Lower terraces at 130 to 200 feet above present sea level are equivalent to the Pliocene Bacons Castle Formation to the east. These gravels are easily distinguished from higher Bon Air gravels using clast size, composition, and relative competency. Bacons Castle gravels range up to boulders in size, include abundant Skolithos-bearing quartzite clasts, and numerous and easily recognized Mesozoic clasts (red to maroon siltstone and reddish-gray sandstone). Most of the clasts are also unweathered, with thin to nonexistent weathering rinds.

From Stop 12 at Woodmont Recreation Center, exit left (north) from parking area onto Traymore Road and proceed 0.1 mile north to Medina Road; turn left (west) onto Medina Road and proceed 0.4 mile to Woodmont Drive; turn right (north) onto Woodmont Drive and proceed 0.2 mile to VA-147/Huguenot Road; turn left (southwest) onto VA-147/Huguenot Road and proceed 1.0 mile to VA-711/Huguenot Trail; yield right (west) onto VA-711/Huguenot Trail and proceed 5.1 miles to VA-288 North; yield right (north) onto VA-288 North and proceed 7.8 miles to 1-64 East; yield right (east) onto 1-64 East and proceed 1.9 miles to Exit 177 for 1-295; yield right onto 1-295 and proceed 7.8 miles to Exit 43D-C-B-A for 1-95 North; take the righthand exit onto 1-95 North and proceed 48.9 miles to Exit 133 for U.S.-17/Warrenton Road; take the righthand exit onto U.S.-17 North/Warrenton Road and proceed 0.4 mile to Sanford Drive on left; turn left (southwest) onto Sanford Drive and proceed 0.1 mile to entrance to Comfort Inn & Suites Fredericksburg on right.

2015 Trip Ends—Join Kirk Bryan Field Trip at Great Falls the Next Morning.