Stratigraphy of Back-Barrier Coastal Dunes, …people.uwec.edu/jolhm/Research_Pub/pdfs of...

Journal of Coastal Research 20 4 980–999 West Palm Beach, Florida Fall 2004

Stratigraphy of Back-Barrier Coastal Dunes,Northern North Carolina and Southern VirginiaK.G. Havholm†, D.V. Ames‡, G.R. Whittecar§, B.A. Wenell†1, S.R. Riggs‡, H.M. Jol†, G.W. Berger*,and M.A. Holmes†2

†Departments of Geology andGeography

University of Wisconsin-EauClaire

Eau Claire, WI 54702, U.S.A.

‡Department of GeologyEast Carolina UniversityGreenville, NC 27858, U.S.A.

§Department of Ocean, Earthand Atmospheric Sciences

Old Dominion University4600 Elkhorn Ave.Norfolk, VA 23529, U.S.A.

*Quaternary Sciences CenterDesert Research Institute2215 Raggio ParkwayReno, NV 89512, U.S.A.

ABSTRACT

HAVHOLM, K.G.; AMES, D.V.; WHITTECAR, G.R.; WENELL, B.A.; RIGGS, S.R.; JOL, H.M.; BERGER, G.W., andHOLMES, M.A., 2004. Stratigraphy of back-barrier coastal dunes, northern North Carolina and southern Virginia.Journal of Coastal Research, 20(4), 980–999. West Palm Beach (Florida). ISSN 0749-0208.

Ground penetrating radar studies of four representative active back-barrier dunes, combined with radiocarbon andphoton-stimulated-luminescence dating techniques and soils analysis, reveal phases of alternating dune activity andstabilization along the North Carolina–Virginia coast. Two smaller dunes represent only the current phase of duneactivity. Two larger dunes preserve evidence of three phases of dune development (ca. 740, 1260 and 1810 AD) andintervening phases of soil development. Climate, particularly moisture conditions, played a part in the timing of duneactivity and stabilization events. All three dune phases are associated with drier conditions whereas soils formationis associated with humid conditions. Modern (phase 3) dunes are more widespread along the coast and their formationis attributed to a combination of dry conditions, increased storminess associated with the Little Ice Age, and risingsea level. Tidal inlet closing and storm overwash processes likely provided sediment point sources for individual dunemasses. The longer history and much greater volume of dune sand in the area of the two larger dunes suggests agreater sediment supply in this locality.

ADDITIONAL INDEX WORDS: Ground penetrating radar, coastal dune, dune reactivation, dune stabilization, paleo-sol, luminescence dating, radiocarbon dating, Little Ice Age.

INTRODUCTION

Coastal dune formation is episodic, and variable in loca-tion, style and extent of development due to the interactionof terrestrial, atmospheric, and near-shore ocean systems. Inhumid coastal regions, vegetation plays a significant role indune development (e.g., GOLDSMITH, 1989). In the last twodecades researchers have begun to develop descriptive andpredictive models to explain patterns of coastal dune devel-opment and preservation (e.g., PSUTY, 1988). Foredunes,which form in situ as a result of trapping of sand by vege-tation and do not migrate (impeded dunes, PYE, 1983), havereceived the most attention because of their role as frontlineprotection from wave attack during storms. Models have beendeveloped to describe conditions favoring foredune formationand stages of foredune development (e.g., HESP, 1988; PSUTY,1988, 1993; KNIGHT et al., 2002). Dunes that actively migrateinland or along the coast, typically over vegetated terrain(‘‘transgressive dunes,’’ GARDNER, 1955; PYE, 1983), have re-ceived less attention. General conceptual models to explainor predict where, when, why and in what form dunes trans-gress inland, especially in non-arid regions, are still in a na-

03503A2 received and accepted in revision 10 April 2003.1 Current address: 2524 W. Twin Lakes Rd., Mausou, IA 50563.2Current address: Hydrology Division, Arizona Dept. of Water Re-

sources, 500 N. 3rd St., Phoenix, AZ 85004.

scent stage (ORME, 1988; HESP and THOM, 1990). (Becausethe term ‘‘transgressive’’ can be confused with marine trans-gression in a coastal setting, the term ‘‘back-barrier dunes’’is used here, although it is an imperfect substitute becausesuch dunes also form on coasts without barrier islands orspits.)

If models of back-barrier dune development in humid coast-al regions are to be enhanced, more data are needed on howsuch dunes develop in time and space. A growing body ofliterature includes earlier, primarily descriptive studies (e.g.,BROTHERS, 1954; COOPER, 1958) and a second generation ofstudies using radiocarbon and, later, luminescence datingtechniques to develop dune event chronologies (e.g., PYE andRHODES, 1985; SHULMEISTER and LEES, 1992; THOM et al.,1994). Ground penetrating radar (GPR) now allows non-de-structive imaging of internal dune stratigraphy that was pre-viously inaccessible, beginning a new generation of dunestudies in both desert dunes (BRISTOW et al., 1996; BRISTOW

et al., 2000a), and coastal dunes (foredunes: e.g., JOL et al.,1998; BRISTOW et al., 2000b; back-barrier dunes: e.g., CLEM-MENSEN et al., 1996).

The main contribution of this study is to integrate GPR andoptical luminescence techniques with trenching, radiocarbondating, and soil analysis to develop a chronology of back-bar-rier dune formation and stabilization in a humid coastal re-gion where no such study has previously been done (Figure

981Coastal Dune Stratigraphy

Journal of Coastal Research, Vol. 20, No. 4, 2004

Table 1. Ground penetrating radar systems and methods.

System

Trans-mitter(volts)

Antennas(MHz)

Antennaspacing

(m)Shot point spacing

(mm)Topographic

correction Post-processing

aPulseEKKObSIR System-2

10001200

100100

1.01.45

0.5 (hand-held);8 to 9 (vehicle-transported)

Total survey station2-ft. contour topo. map

AGCFiltering, AGC, Trace stack

a Sensors and Software, Inc.b Geophysical Survey System Inc.

1). Defining a dune chronology constrains the field of possiblecontrols on dune activity. As such studies continue on variouscoasts, emerging patterns should lead to development of moregeneral models of coastal dune development and stabiliza-tion.

The study supports two additional objectives. The chronol-ogy of eolian processes, integrated with ongoing studies ofregional coastal development through time (e.g., SAGER andRIGGS, 1998), provides a more complete understanding ofcoastal sedimentary processes. Finally, the number and sizeof active dunes in the study area has been steadily decreasingsince the 1930’s (e.g., ALBERTSON, 1936; BIRKEMEIER et al.,1984) and the value of saving the remaining dunes for eco-nomic, ecologic and aesthetic reasons is now recognized. Atthe same time, continued migration of dunes causes problemsin urban areas. An understanding of dune history can informmanagement decisions for coastal dune lands that must bemade by national and state park systems, municipalities, andprivate landowners.

SETTING

The dunes examined in this study are within the Albemarleembayment, one of a series of broad structural embaymentsalong the coastal plain of the eastern U.S. (TRAPP, 1992). Theaxis of the Albemarle embayment parallels Albemarle Sound(Figure 1a). This embayment preserves a complex package ofPleistocene and Holocene marine, estuarine and barrier sed-iments (RIGGS et al., 1992, 1995). The Quaternary sedimentsoverlie a passive margin shelf sequence composed primarilyof Cretaceous and younger sedimentary units overlying olderbasement rocks (e.g., GOHN, 1988).

During glacial sea-level lowstand coastal rivers such as theRoanoke, now a trunk stream flowing into Albemarle Sound,incised valleys and delivered fluvial sediment across the ex-posed shelf (Figure 1c). As sea level rose in the Holocene,channels were flooded and back-filled with sediment (SAGER

and RIGGS, 1998). Limited sand supply in the coastal system,in combination with micro-tidal conditions, have developed aclassic wave-dominated barrier from Cape Henry to CapeHatteros with one modern inlet (Davis, 1994). Historicallyand prehistorically, a number of inlets have opened andclosed in this area (FISHER, 1962; Figure 1d). The sand bar-rier is narrow (0.5–2 km) except where flood deltas formedat sites of past inlets or in areas where large sediment sup-plies allowed for the formation of extensive beach ridge se-quences such as in the Kitty Hawk Woods, Nags Head Woods,and Colington Island (FISHER, 1967). The local abundance of

sand that supplied shoreline progradation in this area maybe related to the presence of paleo-Roanoke riverine sedi-ments on the inner shelf. However, for the past century,much of this part of the Atlantic coastline has been generallyerosional, judging from the lack of foredunes and the ongoingerosion of foredunes constructed in the 1930’s (BIRKEMEIER

et al., 1984).This region has sufficient rainfall (110 cm/year at Duck,

1978–1996 avg., BARON, 1999) to support lush vegetation inthe oak-hickory-pine ecological zone (BURK, 1962; GODFREY,1976). Precipitation from mid-latitude cyclones in winter andconvection and tropical depressions in summer distributesthe rainfall fairly evenly throughout the year (BARON, 1999;LEFFLER et al., 1992). The harsh conditions associated withbarrier islands, such as salt spray, blowing sand, and fre-quent overwash, limit development of mature vegetation suc-cessions (e.g., OOSTING and BILLINGS, 1942; GODFREY andGODFREY, 1976). Active dune surfaces on the Outer Bankstypically support salt- and sand-tolerant species such as seaoats (Uniola paniculata) and panic grass (Panicum amarum).The typical succession for surfaces more than 1.5 m. abovethe water table and where salt spray is not a factor, is toshrubs and trees including live oaks (Quercus virginiana), ju-niper (Juniperus virginiana), and loblolly pines (Pinus taeda;BURK, 1962; FROST, 1999; STALTER, 1993).

Winds blow from all directions, but sand-moving windsfrom the north and northeast predominate in the wintermonths. From May through August, sand-moving winds fromthe southwest quadrant equal and sometimes exceed thosefrom the northeastern quadrant (BARON, 1999; LEFFLER etal., 1992). However, yearly averages suggest that strongwinds from the north and northeast are more frequent, re-sulting in greater sand transport to the south and southwest(Figure 1b). Most of the strongest winds result from extra-tropical cyclones, or ‘‘Nor’easters’’, which are most commonfrom October to March (DOLAN et al., 1988). Less frequently,tropical depressions or hurricanes affect the area during sum-mer and fall; wind patterns from these depend on the trackof the system relative to the coast.

METHODS

Dune Interior Imaging

Ground penetrating radar (GPR) was used to image theinternal stratigraphy of the dunes. Surveys were performedwith two different systems and techniques (Table 1). Two-way travel-time was translated into depth below the surface

982 Havholm et al.


Table 2. Radiocarbon ages.

Sample numberSample locality

and type

Conventionalradiocarbon age

(yrs. BP)a

Calibratedradiocarbon age

(yr. AD)b

Mean calibrated age(yr. AD)c Buried soil

Run HillRH 97-1 In situ stump

Inner wood of tree40 6 60 1690–1730

1810–192017101865

PS2

RH 97-2 In situ stumpOuter wood of tree

0 Modern

RH 98-1 In situ stump 100 6 50 1670–17801795–1945

17251870

PS2

RH 98-2 In situ stump 30 6 50 1695–17251815–1920

17101868

PS2

RH 98-3 Wood fragment 180 6 40 1655–18901905–1950

17731928

PS2

Jockeys RidgeJR 96-1 Soil organics with charcoal 220 6 60 1520–1570

1630–18901905–1950

154517601928

PS2

JR 97-1 In situ stump 170 6 50 1650–18901910–1950

17701930

PS2

JR 98-2 In situ stump 160 6 60 1650–1950 1880 PS2JR 10-U (AMS) Soil organics 220 6 40 1655–1950 1803 PS2JR 98-4 Wood 130 6 70 1650–1950 1800 PS2JR 98-5 (ext. ct.) Charred wood 950 6 110 885–1285 1085 PS1JR 98-3 Charred wood 670 6 60 1260–1410 1335 PS1JR 10-L (AMS) Soil organics 1030 6 50 980–1025 1003 PS1

a Analytical error of 1 sigma.b Calibration error of 2 sigma.c Calibration: STUIVER and LONG, 1993.

Table 3. Photon-stimulated-luminescence ages from quartz.

Sample localityAgea

(yrs. BP)Age

(yr. AD)DunePhase

Penney Hill PEH97-1Run Hill 1 (upper) RNH97-5Run Hill 2 (lower) RNH97-1Jockeys Ridge 1 (upper) JRDG95-1Jockeys Ridge 2 (middle) JKR97-1Nags Head Woods (Tr. 8) NHW95-12Nags Head Woods (Tr. 2) NHW95-4

143 6 16108 6 11

1232 6 65187 6 15742 6 51

1037 6 60826 6 69

ca. 1855ca. 1890ca. 765ca. 1810ca. 1260ca. 960ca. 1175

DP3DP3DP1DP3DP2

a Analytical errors are 1 sigma. Analysis year is 1999.

by measuring near-surface velocity of the electromagneticpulses through the substrate with a ‘‘common mid-point’’ sur-vey (JOL and SMITH, 1991). At Jockeys Ridge the same value(0.14m/ns) was used everywhere. A global positioning system(GPS) was used to locate all transects accurately. Most majorreflections imaged by GPR were ground-truthed on each duneby sampling in outcrop, in a trench, or with an auger.

Dating

Two methods of dating, radiocarbon and luminescence,were used in the project (Tables 2, 3). Ages of in situ treestump remains (typically outer tree rings) and/or loose woodfragments from each buried soil exposure were obtained us-ing standard radiocarbon methods (Beta Analytic Inc.). Soilorganic material from two buried soils was analyzed with the

AMS radiocarbon method. Tree-death age provides a mini-mum age for buried soil development and an approximate agefor the reactivation of the dune that buried the tree.

Photon-stimulated-luminescence (PSL, AITKEN, 1998) dat-ing of 125–212 mm diameter quartz and potassium feldsparsand grains was applied to determine the age of last exposureto sunlight, representing dune activity. Quartz and feldspargrains behave as paleodosimeters. Luminescence intensity,measured on sand collected in light-tight tins, is proportionalto the total absorbed ionizing radiation dose since burial. Di-vided by the radiation dose rate, this luminescence-derivedpost-burial dose yields a PSL age. Samples were collected intrenches from buried dune strata at depths of 0.5 to 1.5 m(see BERGER et al., 2003, Figure 2 for details of collectionsites). The radiation dose rate for the environment of eachsample was calculated from measurements of concentrationsof potassium, uranium and thorium in the sand, combinedwith estimates of local cosmic-ray dose rates. Because of lowfeldspar concentrations in the dune sand (BERGER et al.,1996), quartz ages are reported here. The PSL from quartzwas stimulated by intense, restricted-wavelength blue (470nm) diode light. Single-aliquot-regeneration procedures(MURRAY and WINTLE, 2000) were employed for quartz. Fur-ther details of methods and results are given in BERGER etal. (2003).

Soils AnalysisBuried soils were described in shallow pits and outcrops

using terminology and procedures outlined in SCHOENEBER-



Table 4. Buried soil descriptions for Run Hill and Jockeys Ridge.

Horizon Description

a. Location: Buried soil (PS2) Site RH01, Run Hill, Dare County, N.C.(UTM 18 S 0439525, 3984325) Paleoslope position: shoulder

C dune sand several meters thick; 10YR7/1 light gray fine-medium sand, cross-bedded; clear, wavy boundary

Ab 0–8 cm; 10YR3/1 very dark gray sand with 10YR7/1 lightgray sand in irregular lenses; loose moist, slightly harddry; greatest concentration of humus in center of hori-zon; clear wavy boundary

Eb 8–13 cm; 10YR7/2 light gray sand; loose moist, loose dry;clear to gradual wavy boundary

Bbw 13–21 cm; 10YR7/4 very pale brown sand; loose moist,weakly coherent dry; numerous burrows, 10YR7/2 lightgray sand; gradual to diffuse boundary

C 211 cm; 10YR8/3 very pale brown sand; loose moist, loosedry; cross-bedded dune sand

b. Location: Upper buried soil (PS2) Site JR04, Jockeys Ridge, DareCounty, N.C. (UTM 18 S 0442552, 3980019) Paleoslope position: foot-slope

C2 dune sand several meters thick 10YR7/1 light gray fine-medium sand, cross-bedded; abrupt wavy boundary

Ab 0–18 cm; 10YR3/1 very dark gray sand; loose moist, weak-ly coherent dry; gradual wavy boundary

Bb 18–36 cm; 10YR5/4 yellowish brown sand; loose moist;weakly coherent dry; clear wavy boundary

C1b 36–60 cm; 10YR8/3 very pale brown sand; loose moist,loose dry; gradual wavy boundary

C2 601 cm; 10YR8/2 white sand; loose moist, loose dry; cross-bedded dune sand

c. Location: Upper buried soil (PS2) Site JR05, Jockeys Ridge, DareCounty, N.C. (UTM 18 S 0442550, 3980102) Paleoslope position: con-cave lower side slope

C dune sand several meters thick; 10YR8/1 white fine-medi-um sand, cross-bedded; abrupt smooth boundary

Oib/Ab 0–3 cm; 10YR4/1 dark gray sand with irregular lenses of10YR3/1 very dark gray fibrous peat, 0–3 cm thick; bur-rows filled with 10YR7/1 light gray sand; loose moist,weakly coherent dry; clear wavy boundary

Eb 3–10 cm; 10YR7/1 light gray sand; many fine weak lamel-lae (,1 mm thick) 10YR4/6 spaced 1–3 cm; clear wavyboundary

Bb 10–18 cm; 10YR6/4 light yellowish brown sand; loosemoist, weakly coherent dry; clear irregular boundary

C1b 18–50 cm; 10YR7/3 very pale brown sand with 10YR8/3krotovina and root traces; loose moist, loose dry; clearwavy boundary

C2 501 cm; 10YR8/2 very pale brown sand; loose moist, loosedry; cross-bedded dune sand

d. Location: Lower buried soil (PS1) Site JR05, Jockeys Ridge, DareCounty, N.C. (UTM 18 S 0442545, 3980125) Paleoslope position: toe-slope

C dune sand several meters thick; 10YR8/1 white fine-medi-um sand, cross-bedded; common abrupt wavy dark yel-lowish brown lamellae 10YR4/6 (1–3 mm thick) in lower10 cm; gradual wavy boundary

Ab 0–10 cm; 10YR4/2 dark grayish brown sand with irregularpatches 10YR6/2 light brownish gray; loose moist, loosedry; common abrupt wavy 10YR4/6 dark yellowishbrown lamellae (1–3 mm thick); gradual wavy boundary

A/Eb 10–50 cm; 10YR5/2 dark grayish brown sand with largeirregular patches of 10YR7/3 very pale brown and10YR8/1 white sand; loose moist, loose dry; commonabrupt wavy 10YR4/6 dark yellowish brown lamellae(1–3 mm thick) in upper 20 cm, 1–2 mm thick in lower20 cm; diffuse wavy boundary

Bb 50–80 cm; 10YR8/3 very pale brown sand; loose moist,weakly coherent dry; diffuse wavy boundary

Table 4. Continued.

Horizon Description

C1b 80–100 cm; 10YR7/3 very pale brown sand with 10YR8/3krotovina and root traces; loose moist, loose dry; clearwavy boundary

C2 1001 cm; 10YR8/2 very pale brown sand; loose moist,loose dry; cross-bedded dune sand

GER et al. (1998; Table 4). They can be distinguished fromless well developed dune surface soils that contain A–C pro-files, with very weak color change due to iron stripping justbelow the litter (O) layer. By using outcrop patterns andhand-auger holes to physically trace soil surfaces from an ex-posure to the nearest GPR transect, and from there through-out the area, buried soil units were determined to be identi-fiable and laterally continuous and therefore useful as strati-graphic markers.

DUNE MORPHOLOGY AND HISTORICAL DATA

The active back-barrier dunes of this study (Figure 1d) lie1 km or less from the Atlantic coast and are simple crescenticforms with the exception of Run Hill (Figure 2a), which is acompound crescentic dune (MCKEE, 1979). Jockeys Ridge isa group of 4 dunes (Figure 2b). The two study dunes to thenorth (Penney and Snow Hills) are significantly smaller over-all than the two southern dunes and maximum dune heightincreases for all dunes from north to south (13 to 27 m). SnowHill, Penney Hill and the highest Jockeys Ridge peak havedecreased in height since the mid-1970s (GUTMAN, 1977b;GOLDSMITH, 1977; JUDGE et al., 2000). The dunes are trans-verse (HUNTER et al., 1983), with major slipfaces on theirsouthern side and a southerly migration direction, approxi-mately parallel to the net sand transport direction (Figure1b).

Limited migration data are available for these dunes. SnowHill moved southward 1.5 m/yr in the 1970’s with temporaryslipfaces moving eastward (GUTMAN, 1977a,b). Comparisonof an 1851 map and a modern map (1982) suggest a long-term migration rate for Run Hill of about 7 m/yr (Figure 3).A time series of aerial photos from 1952 to 1996 shows asignificant decrease in migration rate between 1970 and 1982as vegetation and urbanization of the area to the north andeast of the dune progressed (HOLMES et al., 1997; Figure 4a).During three years of stake monitoring (1995–1997) variousparts of the main slipface of Run Hill migrated from 0 to 0.75m/yr. while parts of superimposed dunes moved eastward atrates up to 1 m/yr (WEAVER et al., 1997). Long-term rate ofmovement of Jockeys Ridge dunes can’t be determined fromthe map comparison because specific dune masses can not beidentified on the 1851 map (Figure 3). The southern-mostJockeys Ridge dune was monitored after sand removal and a4-month healing period in early 1995. The slipface migratedat an average rate of 0.3 m/mo (3.6 m/yr) until a series ofstorms in 1999 caused rapid movement, giving a ;5-year av-erage of nearly 0.5 m/mo (6 m/yr; Figure 4b). All of the study

984 Havholm et al.


Figure 1. Location and setting of the study dunes. a. Regional map. b. Sand transport rose, calculated from wind data collected at Duck, N.C. from 1980to 1991 using the method of FRYBERGER (1979). Length of rose ‘‘petals’’ show proportion of potential sand drift in each direction in percent of total sanddrift in all directions. c. Location of Roanoke River paleo-river system (modified from RIGGS et al., 1992). d. Map showing locations of four study dunes,other dunes in the region, and locations mentioned in text. Ovals indicate active dunes, squares indicate stabilized dunes. Black arrows point to locationsof relict tidal inlets (FISHER, 1962). 1. Unnamed prehistoric inlet 2. New Currituck Inlet.

dunes are migrating over variably vegetated landscapes in-cluding grasses and herbs, shrubs and scrub trees. Portionsof Run Hill dune are overrunning a mature maritime forestdeveloped on the stabilized Nags Head Woods dune field (Fig-ure 2a).

Penney Hill and Jockeys Ridge are essentially unvegetatedand have simple cuescentic shapes (Figures 2b, 5c). BothSnow Hill and Run Hill are partially vegetated with grassesand shrubs that are slowing the migration of their upwindstoss components and beginning their transformation fromcrescentic to parabolic forms (Figures 4a, 6c). Dunes just tothe south of Snow Hill have already converted to parabolicforms as a result of the reduction in sediment availabilitysince the initial construction in the 1930’s of artificial fore-dunes along the coast (HENNIGAR, 1977; GUTMAN, 1977a).

Earlier claims that the area around Snow Hill was a ‘‘vege-tated paradise’’ prior to deforestation after the civil war (e.g.,COBB, 1906, quoted in HENNIGAR, 1977) are disputed in morerecent vegetation studies (FROST, 1999; DOLAN and LINS,1986; GODFREY and GODFREY, 1976; DUNBAR, 1958).

Level of traffic on the dunes varies. Snow Hill has virtu-ally no foot traffic, whereas Jockeys Ridge is trampled byover a million visitors per year. Penney Hill experiences sig-nificant vehicular traffic. Although driving on Run Hill hasbeen officially prohibited since the early 1990s, continuedvehicular activity has maintained roads cut through vege-tated portions of the dune and sand movement occurs alongthese trackways. A portion of Run Hill was mined from 1980to 1989 forming a scarp that is now migrating eastward(Figure 4a).



Figure 2. a. Aerial photo of Run Hill dune prior to significant urbandevelopment. It is migrating over parabolic dune ridges of the northernedge of Nags Head Woods dune field, which is stabilized and covered withmature maritime forest. Dashed line indicates position of precipitationridge (leading dune) of Nags Head Woods dune field where it continuesunder Run Hill. (Original scale 1:28,000, March 6, 1952, Frame 179,source unknown) b. Aerial photo-mosaic of Jockeys Ridge dunes. Numbers1–4 are crests of the 4 dunes that comprise this dune mass. Profiles inFigure 4b were measured along line labeled S. (Photos taken Nov. 6, 1995at 10 5 8009 by N.C. Department of Transportation.)

DUNE STRATIGRAPHY

Snow Hill and Penney Hill

GPR reveals that the two smaller northern dunes eachcomprise one package of predominantly high-angle reflec-tions (up to 10 m thick, Snow Hill; 12 m thick, Penney Hill),interpreted as dune strata (Figures 5, 6). The high-angle re-flections overlie a package of sub-horizontal reflections, in-terpreted as the pre-dune barrier stratigraphy. Dune strataare predominantly foresets formed by migration of the dunelee-slope to the SSW. Coarse variation in reflection dip anglerepresents GPR sampling of strata formed by different lee-slope components (slipface vs. non-slipface, different lee-faceorientations). More subtle changes in dip orientation and mi-nor truncations of reflections represent minor modificationsof lee-slope shape resulting from shifts in wind direction (e.g.,bounding surface, position 140–168, Figure 5a). Near thedune crests, dune topsets that formed on the gentle slopebetween the dune crest and brink are preserved (e.g., position140, Figure 6a; position 160, Figure 5a). The water table isimaged as a prominent horizontal reflection that cuts acrossdune strata at Penney Hill (Figure 5a,b) and follows the pre-dune surface at Snow Hill (Figure 6a,b).

A sediment sample taken from dune foresets on the upwindside of Penney Hill (Figure 5c) gives a PSL burial date of 1436 16 years (PEH97-1, Table 3). This corresponds to about1855 AD. Although the dune strata sampled are nearly theoldest preserved strata within the dune, this date representsa minimum age of dune development because older dunestrata may have been entirely eroded during dune migration.

Run Hill

The internal stratigraphy of Run Hill is more complex (Fig-ure 7). A prominent sub-horizontal reflection at a depth of270 ns (18 m) on transect RH1r1 (Figure 7a) was determinedby auger to be the water table. A prominent hummocky re-flection with up to 20 m relief, identified in outcrop as a bur-ied soil (PS2), separates two major packages of high-anglereflections (DP1, DP3, Figure 7a,b). The buried soil, a TypicQuartzipsamment, is thin and is characterized by a dark grayA horizon that is weakly cemented by humus, forming small(10 cm) scarps where exposed at the surface. A subsoil (B)horizon of variable thickness is weakly stained by iron and athin leached (E) horizon may be present (Table 4a, Figure8a). The widespread extent of the buried soil and its char-acteristic level of development suggest it formed under a welldeveloped forest that generated a continuous litter layer fora period of many decades to a few hundred years. By com-parison, soils on coastal dunes in the mid-Atlantic states thatform over a longer time period (a few thousand years) developinto Spodic Quartzipsamments. These soils develop irregulardark brown patches and pendants in their B horizons whereiron, aluminum and organic coatings accumulate (HATCH etal., 1985; WHITTECAR et al., 1982; Whittecar and Baker,1984). Tree stumps rooted in the buried soil (e.g., Figure 8a)have radiocarbon ages corresponding to calendric dates of ca.1650 AD or younger (samples from 4 sites, Table 2, Figure7c).

986 Havholm et al.


Figure 3. Maps of the vicinity of Run Hill and Jockeys Ridge. 1851 map, Kill Devil Hills to Nags Head, N.C., U.S. Coast Survey, Register #351, originalat 1:20,000 courtesy of Outer Banks History Center. Modern 7.59 U.S.G.S. topographic maps, originals at 1:24,000, portions of Manteo (1953) and KittyHawk (1982) sheets spliced at 368N latitude.

Reflections in the unit below the buried soil (DP1) dipsteeply to the SE (up to 168) and SW (up to 298) and areinterpreted to be strata formed by dunes during an earlierphase of activity. Thickness of the lower unit is variable. Amound in this unit that rises to 13 m above the water table(PR in Figure 7b), corresponds in shape and location to the

ridge on the western (leading) edge of the Nags Head Woodsdune field to the south that is being buried by Run Hill. Thisprecipitation ridge (leading dune migrating over vegetatedcoastal area, COOPER, 1958; also termed long-walled trans-gressive ridge, THOM, 1978) of the older dune complex ex-tends southward along the edge of the field of stabilized, tree-



Figure 4. a. Time series of land use maps of eastern Run Hill, originally drawn from aerial photographs scaled at 1:6000. 1952 photo shown in Figure2a. 1970 photo: taken Nov. 24 at 1:12,000, Frame 98, source unknown. 1982 photo: taken Nov. 30 at 1:6000, frame 3-63, by Kimball and Assoc. for DareCo. 1996 photo: taken April 3 at 1:6000, frame 36-17 by Kimball and Assoc. for Dare Co. b. Successive profiles of the leading edge of the southernmostdune of the Jockeys Ridge group (position S, dune 4, Figure 2b) beginning in Jan. 1995 when 15,000 cubic yards (11,500 m3) of sand were removed toprotect the road.

covered parabolic dunes (Figure 2a). GPR shows three bun-dles of predominantly SW dipping reflections separated bynortheast dipping reflections within this ridge. The geometryindicates that the ridge grew by successive dunes stacking upto form the high precipitation ridge (Figure 7b, position 40NEto 70SW). A very limited number of trench exposures of thedune strata of unit DP1 reveal a wide variety of dune foresetorientations (true dip azimuths of N628W, N258E, S118W,S648E), and foreset dip angles ranging from near-horizontalto 358. Disruptions of low-angle strata suggest the presenceof vegetation during deposition. Overall, the characteristicsof the strata in the lower sediment package (DP1) are con-sistent with formation by parabolic dunes. Sediment from

this unit gives a PSL date of deposition of about 765 AD (12326 65 yrs, RNH97-1, Table 3, Figure 7c).

The upper unit (DP3) comprises two sub-units. The mainsubunit (a) is composed of steeply dipping reflections that dippredominantly to the southern quadrant, and represent duneforesets generated by migration of the main dune bedform(Figure 7a). These were observed in several places in surfaceexposures and trenches. The less extensive subunit (b), lo-cated atop subunit a, has reflections dipping generally east-ward. These are confirmed by trench exposures to be strataformed by the eastward migration of superimposed dunes onRun Hill (S898E, WEAVER et al., 1997), and are inferred to beseparated from the main dune strata by a bounding surface

988 Havholm et al.


Figure 5. Penney Hill dune. a,b. Representative GPR transects showing interpretation on lower transect. GPR reveals Penney Hill to be a relativelysimple dune with no buried soils. Reflection identified as water table confirmed by auger, pre-dune surface confirmed by auger and natural exposure atsouth end of dune. Dune foreset strata confirmed by trenches. c. Outline map of Penney Hill barchan showing locations of GPR transects, trenches andauger holes, and sample site for dune age determination.

(e.g., Figure 7a, position 250–310). A sample taken from duneforesets in unit DP3 above the buried soil has a PSL depo-sition age of about 1890 AD (108 6 11 yrs; RNH97-5, Table3, Figure 7c).

Jockeys Ridge

Jockeys Ridge stratigraphy displays even greater complex-ity (Figures 9–11). One major extensive reflection can betraced throughout much of the sampled area at depths of upto about 300 ns (25 m) below the surface (PS2, Figures 10,11). The undulatory (up to 15 m relief) reflection approachesthe dune surface in several locations where it can be identi-fied in outcrop as a buried soil atop an irregular dune surface(Figure 8b). The upper buried soil (PS2) at Jockeys Ridge isvery similar to the buried soil at Run Hill (Table 4a–c). Agreater range of properties (horizon thickness, color) is seenat Jockeys Ridge because buried soil profiles now exposedformed in a wider range of landscape positions and drainageconditions (swales and toeslopes predominate) than the bur-

ied soil exposures seen at Run Hill (mostly well drained crestand upper slope sites). Darker color in the A horizons reflectsabundant charcoal, which is absent at Run Hill (WHITTECAR

et al., 1998). A stump, wood fragments and soil organics fromburied soil PS2 provide radiocarbon ages of 1650 AD or later(Table 2, Figure 9).

A second prominent undulatory reflection is present insome areas below the upper (PS2) reflection (Figures 10, 11).This reflection has 10 m of relief and where it is exposed atthe surface it is identified as a well developed buried soil(PS1), twice as thick as the upper buried soil. Soil morphologyis also more complex with multiple dark A and pale gray Ehorizons and dark brown soil lamellae up to 3 mm thick (Ta-ble 4d; Figure 8c). Where well preserved and exposed, thelower soil profile has a notable pale yellow ‘‘color B’’ horizon.Although the thicknesses and strength of colors in the E andB horizons vary among outcrops, they are within ranges typ-ical of soils on young Holocene dunes. Carbonized wood frag-ments from buried soil PS1 provide radiocarbon ages corre-



Figure 6. Snow Hill dune. a,b. Representative GPR transects. GPR reveals Snow Hill to be a relatively simple dune with no buried soils. Reflectionsidentified as water table and pre-dune surface confirmed by auger, dune foreset strata confirmed by trenches. True dip angle and azimuth shown attransect intersection is from stereonet manipulation. The sloping water table surface in transect FC1c2 is anomalous and may result from an error intopographic measurement west of position 105. We were unable to return to the site to confirm this. c. Outline map; note that Snow Hill is beginning todevelop the shape of a parabolic dune with trailing arms. Locations of GPR transects, trench and auger hole shown.

sponding to calendric dates of 885–1285 AD and 1260–1410AD; soil organics date to 980–1025 AD (Table 2).

Jockeys Ridge dune strata are divided into three packagesby the two buried soils, each with an irregular geometry. Theupper package (DP3), above the upper buried soil, is up to 20m thick and is the most clearly and extensively imaged. Oneast-west transects reflections dip predominantly westward(e.g., Figure 10b); on north-south transects reflections dip pre-dominantly southward (e.g., Figure 10d,e). Trenches confirmthat these reflections represent dune foreset strata with truedips predominantly towards the southwestern quadrant, rep-resenting migration of dune lee slopes in this direction inresponse to the prevailing wind. Wind reversals are recordedprimarily in multiple bounding surfaces (reactivation surfac-es) that truncate dune strata (e.g., Figure 11c). Sediment fromdune foresets in unit DP3 give a PSL age of about 1810 AD(187 6 15 yrs BP, JRDG95-1, Table 3).

The package of reflections between the two buried soils(unit DP2) is up to 15 m thick and is clearly imaged in a few

areas. Reflections dip predominantly westward on east-westtransects (e.g., Figure 10c, 11b) although eastward-dippingreflections are also present (e.g., Figure 10b). Exposures ofthese strata confirm that they are also dune foreset strata.Sediment from the middle sediment package (DP2) gives aPSL deposition age of about 1260 AD (742 6 51 yrs BP, JKR97-1, Table 3).

Below the lower buried soil is a third package of reflections(unit DP1), which, where imaged on a north-south transect,displays reflections that dip steeply to the south (Figure 11a).On east-west transects reflections dip both to the west andto the east (Figure 10b). A trench confirms that these aredune foreset strata. The base of unit DP3 is shallow and re-flective enough to be imaged on some transects and is inter-preted to be the pre-dune surface (e.g., Figure 11a). In someareas it is imaged as an extremely prominent double reflec-tion (Figures 10b–d), and is tentatively interpreted as ashelly gravel lag by comparison with related GPR work else-where on the North Carolina coastal plain. Locally, packages

990 Havholm et al.


Fig

ure

7.R

unH

ill

dune

.a,

b.R

epre

sent

ativ

eG

PR

tran

sect

s.R

unH

ill

disp

lays

anol

der

dune

pack

age

(DP

1)be

low

abu

ried

soil

(PS

2),

and

anup

per

dune

pack

age

com

pris

edof

two

part

s.T

hem

ain

dune

fore

sets

dip

toth

eS

W(D

P3a

)an

d,ab

ove

abo

undi

ngsu

rfac

e,su

peri

mpo

sed

dune

fore

sets

dip

toth

eE

(DP

3b).

c.O

utli

nem

apof

Run

Hil

l,sh

owin

glo

cati

ons

ofG

PR

tran

sect

s,si

tes

whe

reth

ese

wer

egr

ound

-tru

thed

,and

sam

ple

site

sfo

rdu

nean

dbu

ried

soil

age

dete

rmin

atio

n.



Figure 8. Photos of buried soils. a. Buried soil profile (PS2) on Run Hill (see Table 4a for description). Weakly developed A, E, and B horizons noted.In situ carbonized tree trunk (arrow) buried by dune after soil development; this is the source of the wood for radiocarbon samples RH97-1,2 (Figure 7c,Table 2). The stump appears lower than the soil because the soil slopes very steeply towards the viewer. b. Concave-up exposure of upper buried soil(PS2) formed in interdune swale on Jockeys Ridge. Numerous exposures of such soil-capped swales are visible in the northern half of the dune afterperiods of strong deflation. Shovel handle (arrow) is approximately 10 cm long. c. Distinctive features in the upper horizons of the lower buried soil (PS1)at Jockeys Ridge (see Table 4d for description, location). Note the relatively thick soil lamellae formed in the A and E horizons and the multiple humus-rich (cumulic?) zones within the A horizon.

of sub-horizontal to dipping reflections are imaged below thepre-dune surface, suggesting the nature of barrier islandstratigraphy (Figures 10a,c,d, 11a). A dominant sub-horizon-tal reflection imaged only in the finer-scale GPR transects(Figure 11) was identified by augering as the water table.

Additional Dates from Nags Head Woods Dunes

PSL dates have been determined in two places on high-angle dune foreset strata of stabilized parabolic dunes in theNags Head Woods dune field (Table 3). These sediments, de-posited in the southern half of the dune field, between RunHill and Jockeys Ridge, give PSL depositional dates of about960 AD (1037 6 60 yrs, NHW95-12) and about 1175 AD (8266 69 yrs, NHW95-4).

PHASES OF DUNE ACTIVITY AND STABILITY

Three phases of dune activity punctuated by two phases ofdune stability are recognized in the stratigraphy of the activedunes in this study (Figure 12). Strata from the earliest dunephase (DP1) are preserved at Run Hill and Jockeys Ridge,and in the stabilized dunes of Nags Head Woods. Dune activ-

ity dates to about 765 AD at Run Hill and 960 AD in thesouthern part of Nags Head Woods (PSL method). JockeysRidge experienced stabilization and soil formation (PS1) by1000 AD (radiocarbon method). Based on degree of develop-ment, the soil formed on the dune surface over a period ofseveral hundred years. A corresponding soil is not found onRun Hill. Either Run Hill remained active and eolian sedi-ments deposited during this time were not preserved or notsampled for luminescence dating, or the first buried soil (PS1)developed as at Jockeys Ridge, but was entirely removed dur-ing an ensuing phase of eolian activity. Additionally, at leastone dune ridge in Nags Head Woods dune field was active at1175 AD (PSL method), indicating that while a soil developedin the Jockeys Ridge area, partially vegetated parabolicdunes remained active nearby to the north.

A second phase of dune activity (DP2) resumed by 1260 ADat Jockeys Ridge (PSL method), partially burying and par-tially deflating the older soil. Sediment of this depositionalage was not found on the other dunes. A second phase ofstabilization occurred by 1700 AD (radiocarbon method). Soilformed during this time is represented at both Run Hill andJockeys Ridge (PS2). Being less well developed than buried

992 Havholm et al.


soil PS1, it likely formed over a somewhat shorter time pe-riod. By 1810 AD (PSL method) the modern active dunes ofdune phase 3 (DP3) were burying this younger soil at JockeysRidge. Samples from strata of this phase of activity from Pen-ney Hill and Run Hill date to ca. 1855 and 1890, respectively.Although not dated, it is assumed that Snow Hill, with noburied soils, also formed during this latest phase of dune ac-tivity.

On all four study dunes, dune strata formed during dunephase 3 are composed primarily of steep lee-slope strata thatdip in a southerly to southwesterly direction, and can be in-ferred to have formed as a result of typical transverse dunemigration. Evidence of wind reversals is primarily preservedin reactivation surfaces formed by erosion of the upper leeslope. These surfaces are subsequently buried as prevailingwinds resume and the southwest-facing slipface rebuilds andcontinues to migrate. The reactivation surfaces were imagedwith GPR on all 4 dunes, but are best displayed at JockeysRidge (Figure 11c). Similar strata are sketched by HUNTER

(1981, Figure 10c) and have been documented, for example,at White Sands where they were formed by simple transversecrescentic dunes in a bimodal wind regime (CRABAUGH,1994).

CAUSES OF DUNE DEVELOPMENT ANDSTABILIZATION

Back-barrier dunes develop where there is an adequatesupply of sand available to wind strong enough to move thesand (e.g., KOCUREK and LANCASTER, 1999). In humid coast-al regions, vegetation is an important control on sedimentavailability (HESP and THOM, 1990). For dunes to migrateinto the vegetated back-barrier area, the combination of sed-iment supply and wind velocity must be sufficient to over-come the tendency of vegetation to stabilize the wind-blownsand. A number of causes of humid coastal back-barrier duneinitiation have been described (e.g., PYE, 1983; HESP andTHOM, 1990; KLIJN, 1990). Some initiate localized dune de-velopment, such as fire or sediment supplied at a streammouth. Others potentially affect much a broader region, suchas sea level changes or climatic shifts. Causes of dune devel-opment and stabilization in the study area are discussed be-low, based on the stratigraphy and resulting chronology ofdune activity and stabilization (Figure 12).

Sediment Supply

The Atlantic coast in the study area comprises a relativelystraight, continuous barrier spit with only one active inlet(Figure 1d). The only potential modern sources of sand areriver sediments from Chesapeake Bay to the north, and ero-sion of sediment from the inner shelf or sedimentary unitsthat occur in the barrier shoreface. The greatest amount ofdune development over time is concentrated in the JockeysRidge/Run Hill/Nags Head Woods area. All three dune phasesare represented here, as are the largest modern dunes. High-resolution seismic and vibracore studies have shown that theRun Hill/Nags Head Woods/Jockeys Ridge dune mass is lo-cated where the Roanoke paleochannel traverses the moderncoast (Figure 1c; RIGGS et al., 1992). This fluvial channel,

incised into the shelf during the last sea level lowstand, filledwith fluvial and estuarine sediments. These sediments arethe largest potential source of unconsolidated quartz sand be-tween Cape Hatteras and Cape Henry (RIGGS et al., 1995).The sand would have become available at the Atlantic coastas the shoreface ravinement surface progressively eroded theinner shelf during the Holocene sea level rise (RIGGS et al.,1992).

Sediment sources for the smaller modern dunes in thestudy area north of Run Hill (such as Snow and Penney Hills,Figure 1d) are more difficult to identify. Sediment could havebeen made available by closing of former inlets. Snow Hill islocated south of an unnamed prehistoric inlet and PenneyHill rests at the southern edge of the relict tidal delta of theNew Currituck Inlet (Figure 1d). Another potential source ofsand is washover events (ROSEN, 1979; LEATHERMAN, 1976;KOCHEL and WAMPFLER, 1989).

Forcing Factors

Coastal back-barrier dunes have formed in a number ofhumid coastal areas in the last millennium, particularly inthe last 400 years. Formation of these dunefields has beenattributed to a variety of factors including fire damage (e.g.,FILION, 1984), overgrazing (e.g., MCCANN and BYRNE, 1994),coastal progradation (e.g., CHRISTIANSEN and BOWMAN,1986), sea-level (or lake-level) rise (e.g., LOOPE and ARBO-GAST, 2000), climate deterioration during the Little Ice Age(GROVE, 1988) from the 1200’s to 1850 AD (e.g., PYE andNEAL, 1993) and combinations of these (e.g., CLEMMENSEN etal., 1996; JELGERSMA and VAN REGTEREN, 1969). For thestudy area the first three of these factors can be ruled out ascauses for development of modern dunes (DP3). The buriedsoil at Run Hill (PS2) that underlies the modern dune showsno evidence of fire; vegetation studies suggest that overgraz-ing has not had significant impact in the area (FROST, 1999);absence of natural foredunes or modern beach ridge plainsindicates this shoreline was likely erosional during formationof modern dunes. Climatic deterioration (increased stormi-ness, cooler temperatures) associated with the later stages ofthe Little Ice Age, and rising sea level since 1600 AD (SAGER

and RIGGS, 1998) are more likely contributing causes of mod-ern dune development in the study area. Combined withdrought conditions that persisted in this region from about1650 to the late 1800’s (STAHLE et al., 1988), cooler temper-atures, increased storminess and gradually rising sea levelcould have provided the right combination of conditions fordune activation. Rising sea level and more severe wave cli-mate probably increased coastal erosion, opening inlets andincreasing overwash, moving more sediment onto land. In-creased wind velocities and decreased vegetation in droughtconditions likely enhanced eolian transport and dune sandmovement.

It may be possible to attribute timing of the two earlierdune phases and the ensuing periods of dune stabilization toclimatic variation, as well. The middle dune phase at JockeysRidge (DP2 at ca. 1260 AD) occurred near the beginning ofthe climatic deterioration leading to the Little Ice Age (POR-TER, 1986). It also occurred near the end of nearly a century



Figure 9. Map of Jockeys Ridge showing locations of GPR transects andsample sites for dune and buried soil age determination. Area of mapshown on Figure 2b.

of persistent drought in this region that accompanied the Me-dieval Climatic Anomaly (STAHLE et al., 1988). Although theM.C.A. was a period of more temperate climate in Europe (ca.1000–1300 AD, LAMB, 1985; GROVE, 1988) and few docu-mented glacial advances in the northern hemisphere (1090–1230 AD, PORTER, 1986), it brought aridity in the south-western U.S. (ca. 900–1350 AD, STINE, 1994) and to the mid-Atlantic coast (;1050–1280 AD, STAHLE et al., 1988). Theyounger soil (PS1) formed sometime after 1300 AD and beforethe modern dune activity. The 14th through 16th centurieswas a relatively wet period in this area (STAHLE et al., 1988).The older, more well-developed soil (PS1) at Jockeys Ridgedeveloped by around 1000 AD. The century from about 950to 1050 AD was on average a rather wet period in this region(STAHLE et al., 1988). The older phase of dune formation(DP1, ca. 765 AD) occurred at about the time of the mostsevere drought indicated in the 1600-year tree-ring climateproxy record for the region, and near the end of a century

dominated on average by drier conditions (STAHLE et al.,1988). This time also corresponds to the period of early Me-dieval glacial advances, which may have had a climate moresimilar to the Little Ice Age (PORTER, 1986).

External forcing factors, such as a climate change, arelikely not necessary to explain every event of dune stabili-zation in a humid coastal region. As dunes migrate inlandthey move away from their sediment source near the shore,where surf and salt spray limit vegetation growth, and intoan ecological zone where vegetation can begin to take holdon the dune. A typical sequence for a coastal transversedune is to be progressively vegetated, turning into an activeparabolic dune, and finally becoming fully stabilized, a pro-cess that has been documented in aerial photographs for theFalse Cape dunes in Virginia (HENNIGAR, 1977) and thesouth coast of Israel (TSOAR and BLUMBERG, 2002), and intopographic maps for the Rabjerg Mile in Denmark (AN-THONSEN et al., 1996). Such partly or fully stabilized para-bolic forms are common on humid coasts around the world(e.g., Australia, Europe and NW U.S.A, review in PYE andTSOAR, 1990, p. 200; also New Zealand, BROTHERS, 1954;South Africa, TINLEY, 1985; Sri Lanka, SWAN, 1979). Be-cause the past hundred years was a period of warming andmoist conditions, it is difficult to determine whether suchtwentieth-century coastal dune stabilization by vegetationis driven primarily by climatic effects, or whether it is thenatural sequence of events for such dunes in any climaticconditions. Various human impacts on the dunes also com-plicate the interpretation. However, the overall results ofthis study suggest that climate contributes to phases of de-velopment and stabilization of coastal dunes on humidcoasts. Dune formation is associated with drought periodsand dune stabilization and soil formation is associated withmoister conditions. Increased storminess and may also con-tribute to dune activation.

CONCLUSIONS

1. Of four active modern dunes studied along the Atlanticcoast of northeastern North Carolina and southeastern Vir-ginia, the two smaller, more northern dunes show only onephase of eolian activity. The two larger, more southern dunesrecord more than one phase of dune development, punctuatedby periods of stabilization with vegetation and soil develop-ment.

2. The sequence of phases of dune activity and stabiliza-tion of these four dunes is as follows, listed youngest to oldestto parallel Figure 12.

Dune Phase 3 All dunes active since at least 1810 ADBuried Soil 2 Run Hill

Jockeys Ridgedeveloped by 1700 AD

Dune Phase 2 Jockeys Ridge(Run Hill?)

active by 1260 AD

Buried Soil 1 Jockeys Ridge(Run Hill?)

developed by 1000 AD

Dune Phase 1 Run HillJockeys Ridge

active by 765 AD

The Nags Head Woods dune field, active by 765 AD at Run

994 Havholm et al.


Figure 10. Representative GPR transects of Jockeys Ridge dunes imaged with GSSI equipment (Table 1). See Figure 9 for transect locations. a–cTransects with west-east orientation. d,e Transects with north-south orientation.

Hill, remained active in areas between Jockeys Ridge andRun Hill through at least 1175 AD. This suggests that thestabilizing conditions under which the first soil (PS1) devel-oped at Jockeys Ridge were represented, in the southernNags Head Woods dunes, by development of parabolic dunesthat remained active.

3. The larger, longer-lived dune systems formed only in thevicinity of Run Hill and Jockeys Ridge, which suggests a lo-calized source of sediment for the large volume of eolian sand.The paleochannel of the Roanoke River, the major trunkstream draining this portion of the coastal plain/piedmontduring glacial times, passes under the barrier island in thevicinity of these dunes. The fluvial and estuarine sedimentfill of this channel likely provided the source of abundantsand-size sediment for later reworking by littoral and finallyeolian processes.

4. The temporal distribution of dune and soil development

in the study area suggests that dune activation and stabili-zation may be related to climatic fluctuations. All three phas-es of dune development are associated with extended and/orsevere drought conditions. Soils develop during periods of ex-tended humid conditions.

5. Widespread development of back-barrier dunes at inter-vals along the Atlantic coast in the study area since 1700 ADis tentatively attributed to a combination of drought, in-creased storminess associated with the Little Ice Age, and agradual sea level rise. Increased storm frequency and risingsea level probably caused increased erosion and transfer ofsediment from the foreshore to the coast; increased wind anddecreased vegetation allowed that sand to be efficiently re-worked by the wind. Storm overwash and ebb tidal deltasands associated with closed inlets may have provided theprimary point sources of unvegetated sediment available fordevelopment into eolian dunes.



Figure 10. Continued.

ACKNOWLEDGMENTS

We greatly appreciate logistical and financial assistancefrom the following: Nags Head Woods Ecological Preserveand The Nature Conservancy (Barb Blonder and Jeff SmithDeBlieu), Jockeys Ridge State Park (George Barnes and Kel-ley Thompson), Friends of Jockeys Ridge State Park (PeggyBirkemeier), North Carolina Department of Parks and Rec-reation (Marshall Ellis and Sue Regier), and the Field Re-

search Facility of the Army Corps of Engineers at Duck (BillBirkemeier). This project was also supported financially bythe Office of University Research and Sponsored Projects atthe University of Wisconsin-Eau Claire and East CarolinaUniversity Geology Department. False Cape State Park pro-vided access to Snow Hill. Invaluable assistance in the fieldwas provided by Dan Holloway, Sarah Weaver, Jan DeBlieu,Chad Bartz, Ann and Dick Wood, Con Weltman, and StacinMartin.

996 Havholm et al.


Fig

ure

11.

Rep

rese

ntat

ive

finer

-sca

legr

ound

pene

trat

ing

rada

rtr

anse

cts

ofth

eno

rthe

rnen

dof

Jock

eys

Rid

ge,i

mag

edw

ith

puls

eEK

KO

equi

pmen

t(T

able

1).S

eeF

igur

e9

for

tran

sect

loca

tion

s.



Figure 12. Schematic depiction of phases of dune development, stabilization, and soil development of Run Hill and Jockeys Ridge. Snow Hill and PenneyHill, not shown, have no buried soils and have likely been active only during Dune Phase 3.

LITERATURE CITED

AITKEN, M.J., 1998. An Introduction to Optical Dating. New York:Oxford University Press, 262p.

ALBERTSON, C., 1936. The Dunes of Dare Raleigh, N.C.: CapitalPrinting Co., 25p.

ANTHONSEN, K.L.; CLEMMENSEN, L.B., and JENSEN, J.H., 1996.Evolution of a dune from crescentic to parabolic form in responseto short-term climatic changes: Rabjerg Mile, Skagen Odde, Den-mark. Geomorphology, 17, 63–77.

BARON, C., 1999. Climate data from Duck, N.C., collected at theField Research Facility of the U. S. Army Corps of Engineers.Posted at http://www.frf.usace.army.mil/

BERGER, G.W.; SADOVSKY, T.F., and DAVIS, D.A., 1996. A simplestaining method for K-feldspars before sample preparation for IR-OSL dating. Ancient TL, 14, 9–11.

BERGER, G.W.; MURRAY, A.S., and HAVHOLM, K.G., 2003. Photonic

dating of Holocene back-barrier coastal dunes, northern NorthCarolina, U.S.A.

BIRKEMEIER, W.; DOLAN, R., and FISHER, N., 1984. The evolutionof a barrier island: 1930–1980. Shore and Beach, 52, 2–12.

BRISTOW, C.; PUGH, J., and GOODALL, T., 1996. Internal structureof aeolian dunes in Abu Dhabi determined using ground-pene-trating radar. Sedimentology, 43, 995–1003.

BRISTOW, C.S.; CHROSTON, P.N., and BAILEY, S.D., 2000a. Thestructure and development of foredunes on a locally progradingcoast: insights from ground-penetrating radar surveys, Norfolk,U.K. Sedimentology, 47, in press.

BRISTOW, C.S.; BAILEY, S.D., and LANCASTER, N., 2000b. The sedi-mentary structure of linear sand dunes. Nature, 406, 56–59.

BROTHERS, R.N., 1954. A physiographic study of Recent sand duneson the Auckland west coast. New Zealand Geographer, 10, 47–59.

BURK, J.C., 1962. The North Carolina Outer Banks: a floristic in-

998 Havholm et al.


terpretation. Journal of the Elisha Mitchell Scientific Society, 78,21–28.

CHRISTIANSEN, C. and BOWMAN, D., 1986. Sea-level changes, coast-al dune building and sand drift, North-Western Jutland, Den-mark. Geografisk-Tidsskrift, 86, 28–31.

CLEMMENSEN, L.B.; ANDREASEN, F.; NIELSEN, S.T., and STEN, E.,1996. The late Holocene coastal dunefield at Vejers, Denmark:characteristics, sand budget and depositional dynamics. Geomor-phology, 17, 79–98.

COBB, C., 1906. Where the wind does the work. National GeographicMagazine, 37.

COOPER, W.S., 1958. Coastal sand dunes of Oregon and Washington.Geological Society of America Memoir 72, p. 169.

CRABAUGH, M.C., 1994. Controls on accumulation in modern andancient wet eolian systems. Unpublished Ph.D. dissertation, Uni-versity of Texas at Austin, 162p.

DAVIS, R.A., 1994. Barrier island systems—a geologic overview. In:DAVIS, R.A., JR., (ed.), Geology of Holocene barrier island systems.Berlin: Springer-Verlag, pp. 1–46.

DOLAN, R. and LINS, H., 1986. The Outer Banks of North Carolina.United States Geological Survey Professional Paper 1177B, 47p.

DOLAN, R.; LINS, H., and HAYDEN, B.P., 1988. Mid-Atlantic CoastalStorms. Journal of Coastal Research, 4, 417–433.

DUNBAR, G.S., 1958. Historical geography of the North Carolina Out-er Banks. Baton Rouge: Louisiana State University Press, 234p.

FILION, L., 1984. A relationship between dunes, fire and climate re-corded in the Holocene deposits of Quebec. Nature, 309, 543–546.

FISHER, J.J., 1962. Geomorphic expression of former inlets along theOuter Banks of North Carolina. Unpublished Masters thesis, Uni-versity of North Carolina, Chapel Hill, 120p.

FISHER, J.J., 1967. Development pattern of relict beach ridges, OuterBanks barrier chain. Unpublished Ph.D. thesis, University ofNorth Carolina, Chapel Hill, N.C., 325p.

FROST, C., 1999. Vegetation History of the Outer Banks. Unpub-lished manuscript, 116pp.

FRYBERGER, S.G., 1979. Dune forms and wind regime. In: MCKEE,E.D., (ed.) A Study of Global Sand Seas. U.S. Geological SurveyProfessional Paper 1052, pp. 137–169.

GARDNER, D.E., 1955. Beach-sand heavy-mineral deposits of easternAustralia. Bureau of Mineral Resources Bulletin, 28, 103.

GODFREY, P.J., 1976. Comparative ecology of east coast barrier is-lands: hydrology, soil, vegetation. In: Technical Proceedings of the1976 Barrier Island Workshop. Washington, D.C.: The Conserva-tion Foundation, pp. 5–34.

GODFREY, P.J. and GODFREY, M.M., 1976. Barrier Island Ecology ofCape Lookout National Seashore and Vicinity, North Carolina.Washington D.C.: National Park Service Scientific Monograph Se-ries 9, 160p.

GOHN, G.S., 1988. Late Mesozoic and early Cenozoic geology of theAtlantic Coastal Plain: North Carolina to Florida. In: SHERIDAN,R.E. and GROW, J.A., (eds.), The Atlantic Continental Margin: U.S.,The Geology of North America, v. I-2, pp. 107–130.

GOLDSMITH, V., 1989. Coastal sand dunes as geomorphic systems.In: Proceedings of the Royal Society of Edinburgh, vol. 96B. pp. 3–15.

GOLDSMITH, V., 1977. Stop descriptions. In: Coastal Processes andResulting Forms of Sediment Accumulation, Currituck Spit, Vir-ginia/North Carolina. Virginia Institute of Marine Science SRAM-SOE No. 143, pp. 1–46.

GROVE, J.M., 1988. The Little Ice Age. London: Methuen, 498p.GUTMAN, A.L., 1977a. Orientation of coastal parabolic dunes and

relation to wind vector analysis. In: Coastal Processes and Result-ing Forms of Sediment Accumulation, Currituck Spit, Virginia/North Carolina. Virginia Institute of Marine Science SRAMSOENo. 143, pp. 423–438.

GUTMAN, A.L., 1977b. Movement of large sand hills: Currituck Spit,Virginia–North Carolina. In: Coastal Processes and ResultingForms of Sediment Accumulation, Currituck Spit, Virginia/NorthCarolina. Virginia Institute of Marine Science SRAMSOE No. 143,pp. 439–457.

HATCH, D.R.; BELSHAN, J.E.; LANTZ, S.M.; SWECKER, G.R., and

STARNER, D.E., 1985. Soil Survey of City of Virginia Beach, Vir-ginia: U.S. Department of Agriculture, 131 p. plus maps.

HENNIGAR, H.F., 1977. Evolution of coastal sand dunes: Currituckspit, VA/NC. In: Coastal Processes and Resulting Forms of Sedi-ment Accumulation, Currituck Spit, Virginia/North Carolina. Vir-ginia Institute of Marine Science SRAMSOE No. 143, pp. 403–422.

HESP, P.A., 1988. Morphology, dynamics and internal stratificationof some established foredunes in Southeast Australia. Sedimen-tary Geology, 55, 17–41.

HESP, P.A. and THOM, B.G., 1990. Geomorphology and evolution ofactive transgressive dunefields. In: NORDSTROM, K.F.; PSUTY,N.P., and CARTER, R.W.G. (eds.), Coastal Dunes: Form and Pro-cess. John Wiley and Sons, Ltd., pp. 253–288.

HOLMES, M.; HAVHOLM, K.G., and JOL, H.M., 1997. Eolian dunemorphology, Run Hill, N.C. Proceedings of the 5th Annual Univer-sity of Wisconsin-Eau Claire Student Research Day, p. 28.

HUNTER, R.E., 1981. Stratification styles in eolian sandstones: somePennsylvanian to Jurassic examples from the western interiorU.S.A. SEPM Special Publication 31, pp. 315–329.

HUNTER, R.E.; RICHMOND, B.M., and ALPHA, T.R., 1983. Storm-con-trolled oblique dunes of the Oregon coast. Geological Society ofAmerica Bulletin, 94, 1450–65.

JELGERSMA, S. and VAN REGTEREN, J.F., 1969. An outline of thegeological history of the coastal dunes in the western Netherlands.Geologie en Mijnbouw, 48, 335–342.

JOL, H.M. and SMITH, D.G., 1991. Ground penetrating radar ofnorthern lacustrine deltas. Canadian Journal of Earth Sciences,28, 1939–1947.

JOL, H.M.; VANDERBURGH, S., and HAVHOLM, K.G., 1998. GPRstudies of coastal aeolian (foredune and crescentic) environments:examples from Oregon and North Carolina, U.S.A. Proceedings ofthe Seventh International Conference on Ground Penetrating Radar(GPR ’98), May 27–30, University of Kansas, Lawrence, KS, USA,Volume 2, pp. 681–686.

JUDGE, E.K.; GARRIGA, C.M., and OVERTON, M.F., 2000. Topograph-ic analysis of dune volume and position at Jockey’s Ridge StatePark, North Carolina. Shore and Beach, 68, 19–24.

KLIJN, J.A., 1990. Dune forming factors in a geographical context.In: BAKKER, W.M. (ed.). Dunes of the European Coasts: geomor-phology, hydrogeology, soils. Catena Supplement 18, pp. 1–14.

KNIGHT, J.; ORFORD, J.D.; WILSON, P., and BRALEY, S.M., 2002.Assessment of temporal changes in coastal sand dune environ-ments using the log-hyperbolic grain-size method. Sedimentology,49, 1229–1252.

KOCHEL, R.C. and WAMPFLER, L.A., 1989. Relative role of overwashand aeolian processes on washover fans, Assateague Island, Vir-ginia-Maryland. Journal of Coastal Research, 5, 453–475.

KOCUREK, G. and LANCASTER, N., 1999. Aeolian system sedimentstate: theory and Mojave Desert Kelso dune field example. Sedi-mentology, 46, 505–515.

LAMB, H.H., 1985. An approach to the study of the development ofclimate and its impact on human affairs. In: WIGLEY, T.M.L.; IN-GRAM, M.J., and FARMER, G. (eds.), Climate and History. Cam-bridge: Cambridge University Press, pp. 291–309.

LEATHERMAN, S.P., 1976. Barrier island sediment dynamics: over-wash processes and eolian transport. Proceedings of the 15th Con-ference on Coastal Engineering, American Society of Civil Engi-neers, pp. 1958–1973.

LEFFLER, M.W.; BARON, C.F.; SCARBOROUGH, B.L.M.; HATHAWAY,K.K., and HAYES, R.T., 1992. Annual Data Summary for 1990CERC Field Research Facility, Technical Report CERC-92-3.Vicksburg: U.S. Army Corps of Engineers, 129p.

LOOPE, W.L. and ARBOGAST, A.F., 2000. Dominance of an ;150-yearcycle of sand-supply change in the late Holocene dune-buildingalong the eastern shore of Lake Michigan. Quaternary Research,54, 4141–422.

MCCANN, S.B. and BYRNE, M.L., 1994. Dune morphology and theevolution of Sable Island, Nova Scotia, in historic times. PhysicalGeography, 15, 342–357.

MCKEE, E.D., 1979. An introduction to a study of global sand seas.In: MCKEE, E.D. (ed.), A Study of Global Sand Seas. U.S.G.S. Pro-fessional Paper 1052, pp. 1–19.



MURRAY, A.S. and WINTLE, A.G., 2000. Luminescence dating ofquartz using an improved single aliquot regenerative dose proto-col. Radiation Measurements, 32, 57–73.

OOSTING, H.J. and BILLINGS, W.D., 1942. Factors affecting vegeta-tional zonation on the coastal dunes. Ecology, 23, 131–142.

ORME, ANTONY R., 1988. Coastal dunes, changing sea level, and sed-iment budgets. Journal of Coastal Research, Special Issue 3, 127–129.

PORTER, S.C., 1986. Pattern and forcing of northern hemisphere gla-cier variations during the last millennium. Quaternary Research,26, 27–48.

PSUTY, N.P., 1988. Sediment budget and dune/beach interaction.Journal of Coastal Research, Special Issue 3, 1–4.

PSUTY, N.P., 1993. Foredune morphology and sediment budget, Per-dido Key, Florida, U.S.A. In: PYE, K. (ed.), The Dynamics and En-vironmental Context of Aeolian Sedimentary Systems, GeologicalSociety Special Publication 72, 145–157.

PYE, K., 1983. Coastal dunes. Progress in Physical Geography, 7,531–557.

PYE, K. and NEAL, A., 1993. Late Holocene dune formation on theSefton coast, northwest England. In: PYE, K., (ed.), The Dynamicsand Environmental Context of Aeolian Sedimentary Systems. Geo-logical Society Special Publication, v. 72, pp. 201–217.

PYE, K. and RHODES, E.G., 1985. Holocene dvelopment of an episodictransgressive dune barrier, Ramsay Bay, North Queensland, Aus-tralia. Marine Geology, 64, 189–202.

PYE, K. and TSOAR, H., 1990. Aeolian Sand and Sand Dunes. Lon-don: Unwin Hyman, 196p.

RIGGS, S.R.; CLEARY, W.J., and SNYDER, S.W., 1995. Influence ofinherited geologic framework on barrier shoreface morphology anddynamics. Marine Geology, 126, 213–234.

RIGGS, S.R.; YORK, L.L.; WEHMILLER, J.F., and SNYDER, S.W., 1992.Depositional patterns resulting from high-frequency Quaternarysea-level fluctuations in northeastern North Carolina. In: FLETCH-ER, III, C.H. and WEHMILLER, J.F. (eds.), Quaternary Coasts of theUnited States: Marine and Lacustrine Systems. SEPM Special Pub-lication No. 48, pp. 141–153.

ROSEN, P.S., 1979. Aeolian dynamics of a barrier island system. In:LEATHERMAN, S.P., Barrier Islands. New York: Academic Press,pp. 81–98.

SAGER, E.D. and RIGGS, S.R., 1998. Complex Holocene infill historyof a mid-Atlantic incised valley-fill system: an integrated litho-,seismic, and chronostratigraphic synthesis. In: ALEXANDER, C.R.;DAVIS, R.A., and HENRY, V.J. (eds.), Tidalites: Processes and Pro-duts. SEPM Special Publication 61, pp. 119–127.

SCHOENEBERGER, P.J.; WYSOCKI, D.A.; BENHAM, E.C., and BROD-ERSON, W.D., 1998. Field book for describing and sampling soils.

Lincoln, NE: Natural Resources Conservation Survey, USDA, Na-tional Soil Survey Center, 182 p.

SHULMEISTER, J. and LEES, B.G., 1992. Morphology and chronostra-tigraphy of a coastal dunefield; Groote Eylandt, northern Austra-lia. Geomorphology, 5, 521–534.

STAHLE, D.W.; CLEAVELAND, M.K., and HEHR, J.G., 1988. NorthCarolina Climate Changes Reconstructed from Tree Rings: AD 372to 1985. Science, 240, 1517–1519.

STALTER, R., 1993. Barrier Island Botany. Dubuque: Wm. C. Brown,164p.

STINE, S., 1994. Extreme and persistent drought in California andPatagonia during mediaeval time. Nature, 369, 546–549.

STUIVER, M. and LONG, A., 1993. Calibration 1993. Radiocarbon, 35(1), 230p.

SWAN, B., 1979. Sand dunes in the humid tropics: Sri Lanka. Zeit-schrift fur Geomorphologie, 23, 152–171.

THOM, B.G., 1978. Coastal sand deposition in southeast Australiaduring the Holocene. In: DAVIES, J.L. and WILLIAMS, M.A.J.(eds.), Landform Evolution in Australasia. Australian NationalUniversity Press, pp. 197–214.

THOM, B.; HESP, P., and BRYANT, E., 1994. Last glacial ‘‘coastal’’dunes in Eastern Australia and implications for landscape stabil-ity during the Last Glacial Maximum. Palaeogeography, Palaeocli-matology, Palaeoecology, 111, 229–248.

TINLEY, K.L., 1985. Coastal dunes of South Africa. South AfricanNational Scientific Programmes, Report 109, 298p.

TRAPP, H., JR., 1992. Hydrogeologic framework of the northern At-lantic coastal plain in parts of North Carolina, Virginia, Maryland,Delaware, New Jersey, and New York. U.S. Geological Survey Pro-fessional Paper 1404-G, 44p.

TSOAR, H. and BLUMBERG, D.G., 2002. Formation of parabolic dunesfrom barchan and transverse dunes along Israel’s Mediterraneancoast. Earth Surface Processes and Landforms, 27, 1147–1161.

WEAVER, S.J.; WEAVER, K.D., and HAVHOLM, K.G., 1997. Surficialdune processes and migration of Run Hill dune, Bodie Island,North Carolina. The Compass, 73, 110–123.

WHITTECAR, G.R.; HATCH, D.R., and BAKER, J.C., 1982. Soil devel-opment and geomorphic history, Cape Henry, Virginia. GeologicalSociety of America, Abstracts with Programs 14 (1 & 2), p. 95.

WHITTECAR, G.R. and BAKER, J.C., 1984. Soil development on Ho-locene dunes and beach ridges, Cape Henry, Virginia. Washington,D.C.: Program Abstracts, Association of American Geographers, p.47.

WHITTECAR, G.R.; HOLLOWAY, D.M.; BECRAFT, S.; HAVHOLM, K.G.;WENELL, B.A., and JOL, H.M., 1998. Topography and character-istics of paleosols within coastal back-barrier dunes, northeasternNorth Carolina. Geological Society of America Abstracts with Pro-grams, 30, pp. 138.

Stratigraphy of Back-Barrier Coastal Dunes, …people.uwec.edu/jolhm/Research_Pub/pdfs of...

Documents

Transcript of Stratigraphy of Back-Barrier Coastal Dunes, …people.uwec.edu/jolhm/Research_Pub/pdfs of...