Evidence of Cold Climate Slope Processes from the New Jersey Coastal Plain:

Debris Flow Stratigraphy at Haines Corner, Camden County, New Jersey

Abstract

Excavations through surficial deposits across the New Jersey Coastal Plain commonly reveal homogenized surficial sediments, deformed sedimentary structures, chaotically rearranged bed-forms, and wedge-shaped cracks filled with sand from the top-most layers of extant soil profiles. As a whole, these abundant, broadly distributed phenomena are best explained as artifacts of an era of frozen ground during the last Pleistocene glacial maximum. Vigorous freeze-thaw processes and abundant seasonal rainfall created a landscape of low relief covered by highly mobile surficial deposits. The surficial deposits are at grade into broad, flat bottomed valleys now drained by small, tightly meandering, under-fit streams. Modern fluvial, aeolian, and slope processes are ineffectual in either creating or modifying these landscapes.

One particularly brief exposure of complex slope deposits was documented at Haines Corner, Camden County, during the field work (1986) for the Surficial Geologic Map of southern and central New Jersey. The exposure, now presented and interpreted here, provides previously unavailable details of a system of freeze-thaw driven processes that unfolded upon a frozen, impermeable substrate 80 miles south of the southern margin of the Wisconsinan glacial advance to Long Island, N.Y. At the time of these extreme processes, the presently sub-aerial New Jersey Coastal Plain was not proximal to moderating effects of the Atlantic Ocean, being about 100 miles inland and 300 feet above the lowered sea level. Current studies of analogous deposits across the mid-Atlantic Coastal Plain now benefit from dating techniques that were not available during the geologic mapping field work (1985-'92). During the mapping in New Jersey, hundreds of exposures failed to produce datable carbon remains within the stratigraphy of the surficial deposits. Recently reported TL dates from wind-blown sand filling frost wedges, exposed elsewhere in New Jersey, indicate that the widely distributed surficial deposits of the New Jersey Coastal Plain were active during the maximum cold period of the late Pleistocene (around 18,000 years ago).

Introduction

The role of cold climate processes in shaping the surficial geology and geomorphology of the mid-Atlantic Coastal Plain across New Jersey and the Delmarva Peninsula has been a topic of heated speculation since the middle of the twentieth century. Following World War II, engineering geologists with experience in Alaskan permafrost began to recognize land forms, surficial deposits, and soft sediment deformation structures at mid-latitude sites in the lower 48 states. (Pewe, 1973, 1983, French, 1976, Washburn, 1980, Ferrians, 1965, Black and Barksdale,1949, Judson, 1965). Wolfe (1953, 1956, 1977) noted widely distributed fields of closed depressions reminiscent of patterned ground, dormant sand dunes, and near surface deformation of Tertiary and Quaternary deposits across New Jersey; he speculated that permafrost during the Wisconsinan was behind the anomalous landforms, deposits, and structures. Mapping in Delmarva, Rasmussen and Slaughter (1955) noted similar geomorphology, surficial deposits and structures and, in particular, described a distinctive stratigraphic unit, the Parsonsburg sand, as wind blown sand and dunes. These features and deposits were more extensively described and more accurately interpreted by Denny and Owens (1979). In particular they determined that the Parsonsburg sand overlies earlier, peaty deposits and is a relict of the last Pleistocene cold climate period. Related work by Conant and others (1976) in northern Delmarva also presented evidence similar to Wolfe's. Contemporaneous soil scientists, including Tedrow (1986), Ciolkosz (1971), and Nikiforoff (1955) have described evidence of changes in soil forming processes related to cold climates and the superposition of new soil horizons on older, pre-existing soils. In New Jersey, Walters (1978) described a patterned ground signature in thin surficial deposits on Triassic red beds of the Newark basin. However, concerned with a lack of apparent critical scholarship, Black(1983) initiated an effort to discredit the wide interpretation of frost wedges, deformed sediments, and patterned ground as valid evidence of a relict periglacial landscape. Such was the status of observations and interpretations at the commencement of field work in 1984 for compiling the Surficial Geologic Map of Central and Southern New Jersey (Newell and others, 2000). By the end of field work in 1992, it had become apparent that, inspite of his opposition, all of Black's criteria for a valid interpretation of a periglacial environment (Black, 1983) could be satisfied in the surficial deposits of the New Jersey Coastal Plain. A geomorphic system that includes cold climate processes is the best argument for describing and interpreting the surficial deposits and geomorphology (Newell, et al, 1989; Newell and Wyckoff, 1992). A cold climate paradigm was followed in setting up the explanation of the terrestrial, surficial deposits; the processes that formed them are no longer functioning on the present landscape. In fact, modern processes unfolding on the New Jersey Coastal Plain landscape are of insufficient magnitude and frequency to create the Pleistocene deposits in question. Recently, renewed interest in cold climate surficial deposits in New Jersey (French and Demitroff, 2001, French, et al, 2003, 2005) has produced definitive work that confirms the cold climate interpretation of the surficial geology and deposits of the mid-Atlantic region. In the absence of woody materials, analytical techniques for dating surficial deposits were largely untested 20 years ago. The TL dates of French, et al, (2003, 2005), show that wind blown sand and interpreted frost wedge fillings were, in fact, deposited during the last glacial maximum. Our recent studies in progress on Delmarva, on cores from the deep channel of the Chesapeake Bay, and on the stratigraphy of slope deposits on the Blue Ridge Mountains to the west (Joe Smoot, 2004, Eaton, et al, 2003a,2003b) are discovering evidence of periods of rigorous cold climate during the late stages of the Pleistocene. Therefore, it is appropriate to re-examine some of the old, previously unreported data that supported the Surficial Geologic Map of Central and Southern New Jersey (Newell et al 2000) and present it to support new studies on the dynamics of climate change and surficial processes in the mid-Atlantic region. Other new spatial mapping techniques including LIDAR-based DEMs, subsurface profiling with GPR, and the recovery of deep, undisturbed vibrocores, make it possible to revisit the surficial geologic framework and geomorphic history of response to the effects of late Pleistocene global climate change in low altitude, mid-latitude environments.

Location and Setting

Across the New Jersey Coastal Plain ephemeral exposures of all the Pleistocene surficial deposits, as mapped by Newell and others (2000) except for the Holocene and man made deposits, have presented a variety of geomorphic signatures, superficial sediments, and deformed structures that, in the aggregate, indicate a different set of geomorphic processes that are more vigorous than the modern geomorphic system. Subtle, low relief, badlands landforms were massively eroded and large volumes of sediment were transported down slopes through broad flood plains out onto the continental shelf that was sub-aerially exposed during a period of extreme low sea level (330 feet lower than present). Homogenized surface layers, commonly 2-3 feet thick, formerly imbricated pebbles in fluvial gravels are re-oriented with long axes in near vertical planes, frost wedges, involutions, flame structures, closed, elliptical to circular and sometimes rimmed basins, broad terraces and debris flow deposits that are now out of scale with modern fluvial processes are all attributable to repeated freezing and thawing of a shallow substrate above an impermeable layer that no longer exists; perennially frozen ground (permafrost) is a likely candidate for the no longer impermeable horizon. Presently, it is impossible to generate debris flow, solifluction, sheet flow and overland flow because the Cretaceous-Tertiary substrate is commonly highly permeable; all precipitation infiltrates and contributes to groundwater discharge and the base flow of the streams. The streams, largely driven by shallow groundwater discharge, are now under fit within broad, low gradient valleys filled with relict fans and terraces of a formerly active, sediment producing landscape. In October, 1986 we found, photographed, sketched to scale, and described a high-wall exposure in an excavation for a large building complex (originally a hotel). The exposure was over 100 feet long and ranged between 5 and 8 feet high. It was available for study only briefly before disappearing behind a concrete wall.

The excavation and building site is near Hutton Hill at Haines Corner in Voorhees Township of Camden County, New Jersey (see Figure 1). The area had formerly been fallow farm land, and the domain of a local beagle club. The site was midway down a north facing slope ranging from about 165 feet to about 50 feet in elevation. As shown on the generalized cross section (see Figure 2), the top of the hill is underlain by weathered fluvial gravel of the Miocene Bridgeton Formation (Newell, et al, 2000). The underlying Coastal Plain stratigraphy that had been eroded and had contributed to surficial deposits on the hill slope included the Miocene marine Cohansey Formation and the Miocene marine Kirkwood Formation (Owens, et al 1999). Deeper in the substrate and exposed at lower elevations in the landscape is the underlying Vincentown Formation that has contributed distinctive glauconite sand ("greensand") to the lithology of the nearby surficial deposits.

The exposure presented a complex stratigraphy of slope deposits accumulated from numerous debris flow, sheet flow, and solifluction events. Several lithologic sources are distinctive in the aggregate slope deposits. Buried beneath the topmost assemblage of debris flow deposits is a sand dune that subsequently also contributed to the sequence of slope deposits. Laminar shearing, possible dewatering structures and stacked thrust sheets contribute to the interpretation and scale of the processes that moved and stored sediments on the north facing slope. Cutting through the stratigraphy of slope deposits is a small but characteristic frost wedge filled with wind blown sand. The frost wedge indicates that cold climate prevailed for a time following periods of thawing and sediment transport in the active layer. The opportunity to document such details where exposure permits, offers hope for a more comprehensive study of the cold climate processes and the transition to the warmer climate that has followed through the Holocene.

Debris Flow Stratigraphy

The Profile, drawn to scale and without vertical exaggeration in Figure 3, presents a coarsening upward sequence of slope deposits derived from distinctive Tertiary sediments in the substrate. The packages of slope deposits shown in Figure 3, include, from oldest to youngest: weathered Kirkwood Formation, fine grained slope deposits, wind blown sand, sandy slope deposits, gravelly slope deposits, and frost wedge sand fill.

Weathered Kirkwood Formation is the substrate below the slope deposits shown in Figure 3. The exposed, weathered Kirkwood is very fine-grained, silty sand to very fine-grained sand; it is distinctively micaceous and may include feldspar where fresh. Weathered colors are pale-orange to pale-reddish gray. Planar bedded, 4 to 8 inch fining upward beds grade to massive beds that may have originally been sorted but were subsequently disrupted by burrowing organisms; the Kirkwood Formation is an early middle Miocene marine shelf deposit.

Fine-grained slope deposits are derived from, and are unconformable on the weathered Kirkwood Formation. The deposits are very fine- to fine-grained sand, with silty-clayey partings and 2 to 4 inch beds. Locally, reverse graded bedding suggests fine-grained, leading edge of a debris flow overrun by following, coarser material. Internal, discontinuous bed forms include scour and fill possibly from sheet wash events; fine grained laminations locally indicate shearing and realignment of platy clay minerals during down-slope creep that either followed fluid flow as the sediment lost water and became more viscous or occurred as the overburden increased with weight of additional debris flow deposits.

Wind blown dune sand includes fine- to medium-grained, spherical, frosted quartz sand with very fine- to fine-grained, resistant, heavy mineral sand grains; very well sorted in planar to convex upward sets of aeolian cross-beds in randomly stacked packs. Down dip cross-bed vectors indicate variable transport directions (generally sourced from down slope and from the northwest) in cross-bedded packs 0.5 to 1.5 feet thick. Upper bedding surfaces exposed within the dune are marked by 0.5 inch thick soil lamellae of yellow brown to reddish brown iron oxide stained grains and colloidal material. Bedding and the lamellae are perforated by 0.5 inch diameter, 0.5 to 2.5 inch long, insect burrows that are filled with translocated fragments of the lamellae horizons. The insect burrows are also distinguished by the upward/downward translocation of heavy mineral beds in the dune sand (see Figure 4). The sand dune was exposed at the southern (upslope) end of the excavation (see Figue 3). Additional sand dunes, 4 to 5 feet thick at the measured exposure were continuous up slope at least 20 to 30 feet and reached thicknesses of 5 to 8 feet. The sand dune of Figures 3 and 4 is unconformable on the fine grained slope deposits derived from the Kirkwood. The wind blown sand was transported upslope from nearby valley bottoms where it had been derived from earlier slope erosion of the Kirkwood, Cohansey, and Bridgeton Formations. Ultimately, some of the wind blown sand may have traveled from as far as 10 miles from the Delaware River Outwash floodplain to the north and west. However, a more local source is favored by the resistant mineralogy derived from the earlier, extant regolith rather than a labile-rich suite of minerals from glacial outwash in the Delaware Valley. The presence of soil lamellae in the upper beds of the dune suggests that the dune was in place and sporadically active for at least many decades. Lamellae are formed by illuviation of aerosol dust, and fine grained weathered products that are translocated downward in permeable materials; the translocated materials are concentrated at the interface of the saturated-unsaturated boundary of groundwater in the permeable substrate. The insect burrows indicate that the dune was unfrozen during and after lamaellae formation. Surficial wind blown sand and sand derived from the dune field was incorporated in sheared, layered, slope deposits possibly during and after dune formation. The over-riding coarse debris flows suggest that the dune may have been frozen and resistant to erosion at the time of the later, coarse gravelly debris flows.

Sandy slope deposits include fine- to medium-grained quartz-rich sand with very fine- to fine-grained conspicuous, resistant, heavy mineral sand; sandy beds separated by very fine-grained clayey silt partings and laminations. The deposit is characterized by regular 3 to 6 inch beds and sequences of reverse graded beds separated by laminated, sheared, clay silt partings. Discontinuous wavy beds and truncated, convex upward lenses of coarse sand (see Figure 5) may indicate deformation during serial events. Locally, discontinuous, coarser, medium-grained sand beds exhibit deformation of fluvial bed forms into convoluted, convex traces of heavy mineral layers; these structures appear as micro-flame structures and suggest turbulent expulsion of fluid escape following increased pore pressure from loading by subsequent, overlying debris flows (see Figure 6). Alternatively, these features could be very small, imbricate bedding plane thrusts confined to the coarse sandy beds. A third alternative would be disruption of primary bedding by intense, grazing insect burrows. This seems unlikely since other penetrating burrows are not seen in these sediments and the micro structures are confined to the coarser beds. The sandy slope deposits are unconformable on the fine-grained slope deposits. Internal structures, including the reverse-graded beds, deformation of laminations and dewatering structures, indicate a continuing sequence of debris flow events. These sediments are locally derived from the sand dune upslope to the south.

Gravelly slope deposits include Medium to coarse, angular to rounded quartz sand with gray to brown chert, and yellow brown to pink quartz pebbles and sparse cobbles; bedding grades laterally from distinct to massive and the deposit is moderately sorted to diamict; it is as much as 4 feet thick. The unit is unconformable on the fine grained slope deposits, the sandy slope deposits, and the dune sand deposits. The gravelly slope deposits are derived from the fluvial gravel of the Miocene Bridgeton Formation that caps the top of Hutton Hill and other nearby uplands. From the upland source, the sediments were transported several hundred feet from about 150 feet in elevation to about 100 feet in elevation at the site of the excavation. The gravel deposits continue down slope to the valley bottom. The gravel deposits advanced down slope as a series of debris flows that eroded portions of the underlying slope deposits, incorporating the finer grained sediments into the base of the debris flow. Small scale load structures from the fluid pressure of the overburden have been observed along the basal contact and within the gravel deposits. However, the most dramatic deformation is a series of thrust faults triggered by the load of the gravelly debris flow sediments (see Figure 7). Some are internal within the deposit, but the largest incorporate allochthonous fragments from the underlying finer grained slope deposits. Sandy slope deposits are thrust into the diamict sand and gravel sediments and repeated several times, culminating in recumbent folds that surround the allochthonous blocks of sandy slope deposits thrust up from the substrate. The gravelly slope deposits are as much as 4 feet thick but the total thickness is not known because the original surface had been disturbed during excavation.

Frost wedge sand fill includes fine- to medium-grained quartz and heavy mineral sand; grains are rounded and frosted; sediments are similar to the dune sand. Sand is generally massive and fills a wedge shaped crack that cuts across all the slope deposits shown in Figure 3. The crack is 6 to 8 inches wide at the top and diminishes in width near the contact of the gravel debris flow with the underlying sandy slope deposits (see Figure 8). At depth, the cracks converge and are marked with vertically oriented silt-clay partings and a few vertical oriented clasts. The silt and clay filled fractures attenuate with depth. The structure is interpreted as a frost wedge that was relatively short lived but indicates a return to frozen ground conditions after a period of surficial melting and subsequent debris flows. This indicates that cold climate processes were in part seasonal and contemporaneous with the emplacement of the debris flow deposits.

Active Layer Slope Processes

Laminated sequences within each lithostratigraphic unit texturally and structurally indicate a spectrum of slope processes on and within the thawed zone of the surficial deposit. Thin partings and sparse, discontinuous beds and channel fills of sorted sand suggest seasonal overland flow and sheet wash. Thicker beds grade from basal very fine sand and silt upward several inches to medium sand. These reverse graded beds suggest accumulation from the fluid leading edge of a debris flow followed by coarser sediments traveling on top of or at the rear of the flow. Locally, sandy channel fills present deformed structures. Flame structures or small clastic dikes also occur where a subsequent debris flow has over-pressured an underlying saturated horizon. Locally, partings between coarser and finer laminations are suggestive of shearing that may have occurred along the base of a flow that became more viscous as it dewatered but continued to glide down slope. Some apparent glide plane surfaces transition into small bedding plane thrust faults. Friction at the terminus of down slope motion at several locations is indicated by stacked series of small thrust faults that die out in recumbent folds.

Interpretation of Slope Deposit History

Today, moderate but distinctive volumes of loess and wind blown sand are actively transported in the Delaware Valley and surrounding uplands during periods of cultivation and excavation for development. However, none of the slope deposits, especially debris flow deposits, are actively forming on the modern New Jersey Coastal Plain landscape under the present climate. The sequence of weathered Kirkwood Formation, slope deposits, and wind blown deposits indicate a long history of climate forced processes that modified the pre-existing weathered and eroded upland coastal plain landscape of southern New Jersey. The sequence of deposits suggests: Prior to the latest Wisconsinan cold period, during prolonged, warm humid climate of the Sangamon period, the slopes underlain by the Kirkwood, Cohansey, and Bridgeton Formations had been oxidized, leached, and overprinted with a soil profile several feet deep. As the climate became more rigorous and sea level fell, stream gradients were steepened, and the regolith on the various exposed Coastal Plain formations was stripped and transported as colluvium and alluvium down slopes and into the valley bottoms. Seasonal freezing and thawing probably drove the thin, sheet wash and debris flow deposits described as the fine grained slope deposits. A period of cold, possibly dry, air descended from the advancing continental ice cap and swept the landscape putting numerous dune fields and a thin blanket of loess in motion out of the Delaware Valley and across the south New Jersey landscape (Hugh French, personal com. French et al, 2001, 2003, 2005; see also distribution of wind-blown sand and loess as mapped by Newell, et al, 2000). Some of the dunes were periodically stable (decades to centuries) as indicated by the formation of incipient soils and soil lamellae in their topmost horizons. Summer thaws and rain storms reactivated the surface layer of the frozen landscape and slope deposits and debris flows were transported across the sand dunes and pre-existing Kirkwood-derived surficial deposits to form a new sequence of slope deposits derived from the Cohansey Formation up dip, and from the extensive cover of wind blown sand. These deposits were overlaid by debris flow deposits reworked from the Bridgeton Formation outcropping at the top of the hill slope. The cycle of frozen ground and thawed surfaces during summers was repeated possibly for millennia producing complex sequences of diamict slope deposits, frost wedges, and other deformed structures. Following the melt out of the interstitial ice at the end of the Pleistocene, the extreme, rapid drainage of these sandy deposits became re-established and the slope and fluvial processes driven by freezing, thawing and high volume seasonal discharges became dormant. Forest vegetation on the sandy soils was re-established and the nearby streams now derive from base flow discharge. The groundwater discharge driven stream flow is further supported by the accumulation of iron oxide precipitates (bog ore) unburied by minimal sediment transport and deposition, even during extreme storm events.

Using new LIDAR DEM maps, GPR profiling, drilling and excavating, additional stratigraphic detail of the New Jersey Coastal Plain surficial deposits could become routine rather than serendipitous. The details afforded by the Haines Corner site may, at first inspection, appear incongruent with the generalized interpretation of slope deposits presented at the regional scale (1:100,000) on the New Jersey Surficial map (Newell et al, 2000). The "badlands" erosional paradigm used for the regional map is conceptually correct for interpreting the evolution of the landscape with old erosional remnants of surficial deposits perched on noses, eroded remnants, and higher slopes, and with younger deposits reposing on lower slopes and in hollows. However, the details of the Haines Corner site indicate that slope deposits on any position in the landscape, especially in the active layer were not static, but moving, accumulating, eroding, and deforming during the rigors of the last cold climate period. Recognizing active layer stratigraphy and the addition of wind blown cover sand and sand dunes presents a difficult stage for working out the underlying details of both the Cretaceous/Tertiary bedrock geology and the three dimensional framework of the Plio-Pleistocene surficial cover.

Acknowledgments

I thank my colleagues on the U.S. Geological Survey for many ad hoc discussions over the years of the role of a periglacial climate in shaping the surficial geologic framework in New Jersey; these include: Byron Stone, Janet Stone, Carl Kotef (retired), Phil Shafer(deceased), Charles Denny (deceased), and Jim Owens (deceased). John Farnsworth, then of the New Jersey Geological Survey assisted in the field observations. I particularly thank Hugh French (University of Ottawa, retired; University of Delaware, research associate) and Mark Demitroff (Buckhorn Garden Services, Vineland, N.J., University of Delaware) for rekindling present interest in the frozen ground landscapes of New Jersey, and for publishing TL dates that give new credibility to the concept (French, et al 2003, 2005). Adam Benthem assisted with the graphic presentation of information and text.

References