EOLIAN SEDIMENT TRANSPORT ON NORTH CAROLINA … · Joaquin Valley (Wilshire et al., 1981) and up to...

Copyright @ 2006 by Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

EOLIAN SEDIMENT TRANSPORT ON NORTH CAROLINA COASTALPLAIN AGRICULTURAL FIELDS

Paul A. Gares1, Michael C. Slattery2, Patrick Pease1, and Jonathan D. Phillips3

The coastal plain of the eastern United States has long been viewed asan area unaffected by wind erosion. However, high winds in the spring,sandy soils, and local farming practices suggest that soil erosion by windmay be a bigger problem than expected. Field studies conducted ineastern North Carolina provide data about eolian sediment transport,quantifying the role of wind processes in agricultural field soil loss.Eolian sediment flux ranged from 5.5 to 139 kg haj1 hj1 in seven eventsmonitored over a 9-year period. The range in sediment flux seems to berelated to a combination of soil types and antecedent moistureconditions on the fields. The study reveals that eolian processes aresufficiently frequent and sufficiently powerful to produce long-termredistribution of sediment on coastal plain fields resulting in thetruncation of the soil profile. The continued long-term loss of soil fromthe fields has potential repercussions on their productivity. (Soil Science2006;171:000–000)

Key words: Eolian sediment transport, North Carolina, soil degradation.

WIND-INDUCED erosion is well docu-mented in regions dominated by soil-

moisture deficits. In arid and semiarid areas ofthe world, the severity and frequency of erosion,marked by large quantities of sediment trans-port, have been equated with low annual rainfalland decreasing vegetation densities before highwind events (Fryrear, 1981; Middleton et al.,1986; Stout, 2001). Eolian soil erosion inagricultural areas has been recognized as aserious problem at least since the 1930s Bdustbowl^ when a single storm was reported to havemoved up to 270 million tons of soil (Kimberlinet al., 1977). More recently, single wind eventshave scoured as much as 0.6 m of sediment fromparts of agricultural fields in California’s SanJoaquin Valley (Wilshire et al., 1981) and up to1 m of sediment during 12 h of high wind innorthwest Texas (McCauley et al., 1981).

The climatic index of wind erosion (C) isused to characterize the eolian erosion potentialfor particular regions of the United States(Chepil et al., 1962; 1963; Chepil and Woodruff,1963; Woodruff and Siddoway, 1965). C valuesrange from greater than 200 for southernArizona to 1 for the Appalachian Mountains(Kimberlin et al., 1977). The coastal plain ofNorth Carolina has C values that range from 3to 10, suggesting that this region has a low riskof wind erosion. This, in combination withgentle slope gradients and a predominance ofsoils with sandy, permeable surface horizons, hasled to a conception of the coastal plain as aregion with limited erosion due to either wateror wind. This conception is further supportedby the fact that the region is geologically stableand has led some to conclude that the region isneither an eroding or degrading landscape overgeologic time scales (Daniels et al., 1984;Markewich et al., 1990).

Despite this impression of relative stability,there is increasing recognition of the importanceof soil erosion on the Atlantic coastal plain. Anumber of studies suggest that coastal plainerosion rates have averaged about AQ19 to 10 Mghaj1 yj1 over the last 200 to 250 years (Lowrence

1

0038-075X/06/17110–000–000 October 2006

Soil Science Vol. 171, No. 10

Copyright * 2006 by Lippincott Williams & Wilkins, Inc. Printed in U.S.A.

200120

1Department of Geography, East Carolina University, Greenville, NC. Dr. Gares

is corresponding author. E-mail: [email protected] of Geology, Texas Christian University, Ft. Worth, TX.3Department of Geography, University of Kentucky, Lexington, KY.

Received Feb. 13, 2006; accepted May 6, 2006.

DOI: 10.1097/01.ss.0000230126.78771.7a


et al., 1986, 1988; Phillips, 1993; Phillips et al.,1993). These rates are similar to measurementsand estimates of contemporary and historic ero-sion rates in the southeastern Piedmont, a regionrecognized as having rapid and problematic soilloss (Trimble, 1974; SCS, 1982; 1983). Studies ofsediment transport and storage in local coastalplain streams confirm the high rates of erosion onthe uplands, but they also show that the amountof sediment transported out of the system byfluvial processes is not equal to the amountremoved at the source (Simmons, 1988; Hubbardet al., 1990; Phillips, 1992a, b). Phillips et al.(1993) suggest that the missing sediment mayeither be in storage somewhere along the convey-ance route or have been removed by wind. Spatialpatterns of soil profile truncation and burial are insome cases consistent with eolian erosion andredeposition (Phillips et al. 1999a, b), and Nanneyet al. (1993) recognize that sandy soils on theAtlantic coastal plain represent a wind high erosion

potential, despite the indications of the winderosion index (C).

There is very little information on coastalplain soil erosion due to eolian processes. Blow-ing soil is common in the region (Fig. 1A),particularly in late winter and early spring whenwind velocities often are high, temperatures (andinsolation) begin to rise, soil-moisture is likely tobe low, and vegetation on agricultural fields isabsent. Eolian features, such as ripples, field-edgedunes, and blowouts can be observed on oraround many agricultural fields (Fig. 1B-D).Pease et al. (2002) demonstrate that significantdust emissions occur during wind events on theNorth Carolina coastal plain. However, the totalvolume removed from a field in the form of dustis minimal and would have limited impact on soiltruncation. Sediment transported from a field ascreep or saltation would be expected to bevolumetrically more significant than thatremoved as suspended load and would, thus,

Fig. 1. Photographs of wind events and eolian landforms seen on the North Carolina coastal plain. A, Dust cloud onField C; B, Eolian deposit and surface ripples at Field C; C, Field-edge dune at Field E; D, Field-edge blowout insouthern Pitt County.

2 GARES, ET AL. SOIL SCIENCE

200120


potentially have a greater impact on the volumeof soil loss. The purpose of this paper is to reporton several short, site-specific studies that quantifythe amount of sediment removed from coastalplain agricultural fields by eolian sediment trans-port in contact with the bed (saltation and creep).These studies provide an initial assessment of thepotential role of eolian processes in the truncationof coastal plain soil profiles.

BACKGROUND

On agricultural fields, eolian sediment trans-port is affected by numerous factors, particularlymoisture content (McKenna-Neuman andNickling, 1989; Jackson and Nordstrom, 1997;Namikas and Sherman, 1995), soil characteristics(Zobeck, 1991; Fryrear et al., 1994), crusting(Potter, 1990; Goossens, 2004), and vegeta-tion (Bilbro and Fryrear, 1994; Hagan andArmbrust, 1994). Studies of sediment transportacross beaches suggest that sediment moisturecontent is the most important factor affecting thequantity of sediment moved because it promotescohesion between particles (Sarre, 1988; 1989;Jackson and Nordstrom, 1997). On agriculturalfields, this may be even more important becausethe greater capillary tension associated withsmaller particles found in soils promotes theretention of water in the pores, maintaining ahigh moisture content. Goossens (2004) reporteda direct linear relationship between crust strengthand horizontal sediment flux and found smallparticles to be the most affected. Crusts dodevelop on the agricultural fields of the NorthCarolina coastal plain as fine particles adhere toeach other during surface drying after rainfall.This may initially inhibit sediment transport,but high wind velocities ultimately put sedi-ment particles into motion, and these maybreak the crusts down, thereby facilitatingsediment transport (Potter, 1990; Goossens,2004). The density of vegetation on agriculturalfields will vary considerably throughout theyear as fields go from bare surfaces to maturecrops. The inverse relationship between vege-tation density and eolian sediment movementhas been demonstrated (Hagan and Armsbrust,1994; Bilbro and Fryrear, 1994). The exacttiming of vegetation changes on agriculturalfields may ultimately depend on the farmer’scrop choices.

Field observations of sediment removal fromagricultural fields reflect the effects of combina-tions of these variables. In Argentina,

Buschiazzo et al. (1999) report that on twofields with different soils 3 to 18 kg haj1 hj1

(0.3–1.8 Mg haj1 for the entire event) wereremoved during a storm with an average windspeed at 3-m height of 3.9 m secj1 and aduration of 103 h. A second storm of 25-hduration and average wind speed of 6 m secj1

produced a sediment flux of 30 to 40 kg haj1

hj1 (0.75 to 1 Mg haj1). The differences intransport between two events are attributed tothe duration of the events and to the variabilityof wind speeds during each event. Monitoringof several fields in the central plains of theUnited States shows that 40 to 3300 kg haj1

hj1 (0.5–70 Mg haj1) were removed duringstorms with durations of 3.5 to 34.7 h in whichmaximum wind speeds at 2-m height rangedfrom 10 to 23.7 m secj1 (Fryrear, 1995). Insemiarid Niger, 7.2 to 283.7 kg mj1 of sedi-ment were transported from an experimentalfield during wind events whose velocities at aheight of 2 m ranged between 7.6 and 10.3m secj1 (Sterk and Stein, 1997). In humidIndiana, Nanney et al. (1993) report an annualerosion amount of 193 Mg haj1 from a specificfield. The potential significance of wind erosionin humid environments is also supported by datapresented by Robinson (1968) who reports5-cm topsoil loss during a 5-day event inLincolnshire, England.

STUDY AREA

Since 1993, studies of wind conditions andsediment transport have been conducted onagricultural fields at seven locations in southernPitt County, on the North Carolina coastal plain(Fig. 2). Eastern North Carolina has a humidsubtropical climate with an average rainfall ofabout 120 cm falling year-round, with summerprecipitation exceeding winter by about 27%(Clay et al., 1975). We conducted a monthlywater balance analysis for this region usingaverage temperature and precipitation data forthe period 1993 to 2002 for Kinston, NorthCarolina, located in the center of the coastalplain (Fig. 2). Potential evapotranspirationranges from 7 mm in January to 163 mm inJuly and equals actual evapotranspiration inevery month except June in which the deficitis only 1 mm. The monthly analysis masks dailyconditions that are more significant to eoliansediment transport. Although average monthlytemperatures range from 8.5 -C in January to28.2 -C in July, the 10-year range of mean daily

AQ2VOL. 171 ~ NO. 10 NORTH CAROLINA COASTAL PLAIN AGRICULTURAL FIELDS 3

200120


January temperatures is 4.7 to 11.6 -C, whereasfor July, the range is 26.4 to 29.8 -C. Themaximum temperatures in this period were27.2 in January and 37.7 -C in July. The

maximum temperature values would yieldpotential evapotranspiration values in excess of100 mm even in January. If these conditionsoccur in combination with a dry period, the

Fig. 2. Eastern North Carolina, showing the location of the Pitt County study sites.

Fig. 3. The wind climate of eastern North Carolina. A, Directional frequency; B, Average wind speed by direction; C,Seasonal distribution by direction; D, Average monthly wind speed.


200120


result is a rapid drying of the soil surface thatwould increase its susceptibility to erosion.

Coastal plain soils in general have a high sandcontent, suggesting that there is a potential forerosion despite a low climatic index (C) of winderosion. The wind data recorded at Kinston at aheight of 2 m for 1993–2003 (Fig. 3) show that thecommonly (Bagnold, 1941AQ3 ) accepted thresholdvelocity (5.4 m secj1) for saltating particles in themedium sand range is exceeded about 30% of thetime. Wind directional patterns are associated withthe passage of fronts with southwest winds preced-ing the fronts and northwest winds after them.Southwest winds dominate in spring and summerand the northerly winds dominating in the fall andwinter. Average wind speeds tend to be higher inthe spring than any other time of year.

The higher wind speeds in the spring,combined with the activities of farmers whoare preparing their fields for planting, makes soilerosion a frequent occurrence during this season(Pease et al., 2002). In North Carolina, fields aretraditionally plowed under between Novemberand February. Field preparation begins in Feb-ruary, and planting starts in late March withtobacco, followed by corn in early April, cottonin late April, and soy in early May. Harvestbegins with tobacco in late July and August,continuing until early November for soy. Thisschedule leaves many of the fields bare duringthe winter and early spring, the period with the

highest wind velocities. This is the period whensoil erosion could be expected to occur.

The surface of the fields examined isgenerally smooth and flat with slight undula-tions, some of which are possibly due to eoliandeposition (Fig. 1). The Pitt County Soil Survey(USDA, 1974) provides information about thecombinations of soil types found on each field,along with their general size, drainage, andinfiltration characteristics (Table 1). Phillipset al. (1999a, b) demonstrated that local varia-tions on these fields are much greater than areaccounted for by the USDA soil surveys. Thesewithin-field variations undoubtedly affect themovement of sediment by wind at the localscale, but, to maintain methodological consis-tency, the published soil survey data are usedhere for all sites. The USDA information can beused to characterize the erosional susceptibilityof the soils. Several soil types seem more proneto eolian erosion because of a combination ofhigh sand content, high infiltration, and rapiddrainage. These erodible soils are Alaga, Craven,Lakeland, Ocilla, Olustee, Pactolus, andWagram. In general, these soils contain at least70% sand, except the Craven series. The lowersand content of the Craven soils is offset by thegood drainage/infiltration characteristics, andthe combination makes these soils susceptibleto wind erosion. The Ocilla and Olustee soilshave a very high proportion of sand-sized

TABLE 1

Soil types on study fields (by percent of field area) and drainage/infiltration characteristics (source: USDA, 1974)

Soil Type Drain� Infil.. Sand (%)- A B C D E F G

Alaga‘ 1 1 70–85 44 39 14

Aycock 2 2 40–50 27 32

Bibb 4 2 50–60 2 27

Craven‘ 2 2 50–60 12

Lakeland‘ 1 1 88–95 9

Lenoir 4 2 20–40 25

Norfolk 2 2 30–40 63

Ocilla� 4 1 70–85 5 4

Olustee‘ 5 1 70–85 36

Pactolus‘ 3 1 70–85 13 8

Portsmouth 5 2 25–40 16 57 55 56

Rains 4 2 25–40 16

Tuckerman 4 2 50–64 13 31

Wagram‘ 2 1 70–85 55 41

�Drainage 1indicates excessive; 2, good; 3, moderate; 4, poor; 5, very poor..Infiltration 1 indicates rapid; 2, moderate.-Percentage of sand in top horizon‘Soil type has drainage/infiltration characteristics that would promote rapid drying of sediment and produce susceptibility to

removal by wind.

VOL. 171 ~ NO. 10 NORTH CAROLINA COASTAL PLAIN AGRICULTURAL FIELDS 5

200120


particles, but they have very poor drainage dueto a high concentration of clay in the subsurfacelayers (USDA, 1974). The high proportion ofsand-sized particles in they surface layer andrapid infiltration of both soil types, however,makes them susceptible to wind erosion.

A detailed analysis of sediment size wasconducted for thirty surface samples collected atSite C (Fig. 2) where Alaga and Wagram soilsdominate (Table 1). The average sedimentdistribution data show a predominance of sand-sized particles (83.9%) over silt (15.9%) with avery small proportion of clay (0.2%). Thisconfirms the USDA (1974) information forWagram and Alaga soils (Table 1). The dataalso show that the nonsand materials are con-centrated in the silt range. Silt-sized particles aresusceptible to wind erosion, unlike clay-sizedsediments that are more resistant due to theircohesive characteristics. Within the sand frac-tion, 97.4% of the particles in the samples werein the fine- to very fine–sand range (0.063–0.25 mm), and the average sand particle for thefield is 0.223 mm. These small sand sizes arevery conducive to eolian transport in a saltationlayer. The silts are subject to transport in suspen-sion, as attested to by the dust clouds that come offthe fields during high winds (Pease et al., 2002AQ4 ).

METHODS

Two different field approaches were used inthis study. One approach focused on measuringeolian sediment transport rates during a windevent. The second approach measured total fluxfrom a field as a result of a single wind event.The transport rate approach used in situ windmeasurements to obtain a wind velocity profile.A tower equipped with 5 R.M. Young cupanemometers spaced 0.25, 0.5, 1, 2, and 4 mabove the surface was installed in the middle ofthe field. Data from the sensors were recorded at1-sec intervals over 30-min periods with a datalogger. Average wind speeds are determined foreach sampling height, and the data are plotted toobtain the vertical profiles. Sediment movementwas monitored over the sampling period withcylindrical PVC pipe traps, designed afterLeatherman (1978). The trapped data representtotal weight of sediment trapped during the30-min sampling period. Traps were deployedalong a line perpendicular to the wind direction.The cylindrical traps record only saltation load.The trap data are converted to a transport rateexpressed in kg mj1 hj1.

Total sediment flux was determined after aspecific high wind speed event in which thewind blew from a particular direction for aconsiderable period. These conditions resultedin the formation of an obvious depositionalfeature downwind of the field in question. Thearea of the depositional feature was computedfrom data obtained with normal surveyingmethods. Depth of deposition was sampled atregular intervals along transects established acrossthe feature. Deposit depth was measured fromsoil samples collected with a soil augur. Thesample clearly showed the newly depositedmaterial as a distinct light-colored layer abovethe older darker organic material. It was a simpletask to use a ruler to measure the thickness ofthe light-colored layer. Total volume depositedwas computed from the area and depth data.This approach yields a total volume of sedimentmoved by the wind event and is expressed asvolume per unit of field area (m3 haj1). Thevolume data were converted to a weight using abulk density of 1.6 g cm3, an average of samplescollected from the deposited sediment at thefield sites. The total flux data are reported inmegagrams per hectare. These data are alsoconverted to a transport rate (kg mj1 hj1)using the known dimensions of the field and theduration of the event. This conversion allowsthe data obtained from the two differentsampling approaches to be compared.

RESULTS

The presentation of the results focuses oneach individual field. The study methodologyused at each site is described along with its soilcharacteristics. The sites where sediment trans-port was measured are presented first (Sites A, B,and C), followed by those where total fluxmeasurements were made (Sites D, E, F, and G).

Field A

In the spring of 1993, we conducted ourfirst sediment transport study at Site A, a 16.8-hafield located in the southeast corner of PittCounty (Fig. 2). This field consists mainly ofAlaga loamy sand soils with small areas ofLakeland, Ocilla, Portsmouth, Pactolus, andTuckerman soils (Fig. 4). The traps were placedat 20-m intervals in a T-shaped layout just northof the center of the field. The tower was located10 m west of the traps. Data were collectedon April 25, 1993, between 10:00 a.m. and12:00 midnight AQ5when wind speeds exceeded the


200120


sediment transport threshold. Data recorded atKinston indicated that this event lasted 30 h,wind speeds were 4 to 8 m secj1,AQ6 and winddirection was south-southwest. Precipitationduring the 10 days before sampling was very

limited, but 0.8 cm of rain fell the afternoon ofApril 25th, after wind samples were collected.At the time at which the field data werecollected, the wind speed at Kinston wasreported at 5.4 to 6.7 m secj1, and the direction

Fig. 4. Field A, showing the location of the traps and anemometer mast, the distribution of soils, the wind speed/direction hourly data for the event monitored and the precipitation record for the 10 days before the wind event.


200120


was predominantly from the southwest. Themean wind speed at the field site during the datacollection was 6.6 m secj1 at the 4 m height.The wind velocity data collected show aconsistent linear form when plotted (Fig. 5A).The shear velocity (u*), calculated after Baueret al. (1992AQ3 ) averaged 0.334 m secj1 for the tworuns. The amount of sediment trapped recordedin the traps ranged from 3.0 to 8.9 kg mj1 hj1

with an average of 5.7 kg mj1 hj1 (Table 2).The mean grain sizes of sand trapped rangedfrom 0.243 to 0.296 mm, and the average of alltraps was 0.263 mm, very similar to the meansize of parent material at Field C.

Field B

Study area B, a field of 14.4 ha, is locatedabout 2.1 km northwest of site A. Its soils aremainly Portsmouth loam, with a small area atthe northern end of the field consisting of Alagasoils, both loamy sands (Table 1). At this site, thetraps were deployed 10 m apart along an east-west line at the northern end of the field (Fig. 6).Wind data were measured during an event thatbegan on April 5, 1994, that lasted 17 h, withsouth-southwest winds whose speeds rangedfrom 3.6 to 9.4 m secj1. The 3 days beforedata collection were dry, although periods ofprecipitation occurred 4 to 10 days before theevent. Average wind speed at the top of themast (4 m) during the data collection run inthe late afternoon of April 5 was 8.4 m secj1.The wind velocity profile for the samplecollected at field B has a linear form (Fig. 5B)with an average shear velocity (u*) of 0.392

m secj1. Data from three of the eight cylindricalsediment traps deployed at Site B were judgedunusable because holes were discovered in thereceiving bags and sediment escaped from thetrap. The remaining five traps recorded trans-port values of 11.1 to 12.2 kg mj1 hj1, with anaverage of 11.7 kg mj1 hj1 (Table 2). Sedimentsize analysis of the trapped sand shows that themean particle size of the sample within each traphad a mean of 0.234 mm, with values rangingfrom 0.231 to 0.241 mm.

Field C

Field C (Fig. 7), with an area of 5.7 ha,consists mainly of Wagram loamy sand andAlaga loamy sand soils (Table 1). Wind andsediment transport data were collected onMarch 3, 1999, after a dry period of 3 days.Using the protocol used on fields A and B, theanemometer tower was placed approximately inthe center of the field, and the traps were

Fig. 5. Wind velocity profiles. A and B, Profiles for event on 4/25/93 at Field A; C, Profile for event on 4/5/94 onField B; D, Profile for event on 3/3/99 on Field C.

TABLE 2

Amount of sediment trapped at Field A, B, and C

SiteSediment flux

(kg mj1 hj1)

Average grain size

(mm)

Site A max 8.9 0.296

Site A min 3.0 0.243

Average Site A 5.7 0.263

Site B max 12.2 0.241

Site B min 11.1 0.231

Average Site B 11.7 0.234

Site C max 36.7

Site C min 11.9

Average Site C 20.4


200120


deployed in a single line along the north-south axisof the field with individual traps located approx-imately 20 m apart. This event lasted 30 h, with amean speed of 8.1 m secj1 and a range of 5.4 to13 m secj1. The winds were mainly from thewest-southwest during the data collection periodwhen the mean wind speed recorded at the topof the tower (4 m) was 11.8 m secj1. The profilefor this wind event was linear (Fig. 5C). Theshear velocity (u*) for the site C velocity profile is

0.608 m secj1. The amount of sediment trans-ported on Field C varied from 11.9 to 36.7 kgmj1 hj1, with an average of 20.4 kg mj1 hj1

(Table 2). Although the trapped sand sampleswere not analyzed to determine their particlesizes, this was the field where the surface sampleswere analyzed and the mean grain size for thesource material here was 0.223 mm. The 10-dayperiod before March 3 had three small precip-itation events that contributed 7 cm of rainfall.

Fig. 6. Field B, showing the location of the traps and anemometer mast, the distribution of soils, the wind speed/direction hourly data for the event monitored and the precipitation record for the 10 days before the wind event.


200120


Fig. 7. Fields C and D, showing the location of the traps and anemometer mast, the distribution of soils, the windspeed/direction hourly data for the event monitored and the precipitation record for the 10 days before the windevent.


200120


Field D

Field D, located adjacent to Field C (Fig. 7),has an area of 3.8 ha. This field consists ofWagram, Alaga, and Bibb fine sandy loams(Table 1). The poorer draining Bibb soils arelocated at the western end of the field. Wind-storms occurred on February 4 and 18, 2002,that transported large amounts of sediment tothe eastern end of field D where it was depositedin a linear dune along the road (Fig. 1C). Thespeeds at Kinston during the two northwesterlywind events averaged 7.0 m secj1, with themaximum reaching 9.8 m secj1. The winddirections were 302- and 332-, and the eventslasted 26 h combined. The 10-day period beforethe wind event was dry, with a single very smallprecipitation event on February 11. The volumeof the dune was determined using data collectedon February 19, 2002. The length of the featurewas measured with a tape, and widths anddepths were measured along sample transectsestablished 10 m apart. The thickness of the sanddeposit was used along with the horizontaldistance data to obtain a cross-sectional area forthe sampling transect. The cross-sectional areaswere then integrated along the length of thedune to determine a total volume of sand in thefeature. The windstorm removed 48.4 Mg haj1

from this field. Using the standard bulk densityfor sand size particles (1.6 g cmj3), this valuecan be converted to a transport rate of 59.7 kgmj1 hj1 (Table 3).

Field E

March 19–23, 1996, was a period of highwind activity in eastern North Carolina withwind speeds at Kinston averaging 8.4 m secj1

and reaching 17.9 m secj1. Approximately threequarters of a centimeter of rain fell in the 3 daysimmediately before the wind event. The longaxis of the 2.2-ha field at Clayroot, North

Carolina (Fig. 8), runs in a west-northwest toeast-southeast direction. The average winddirection for this event was 236-, slightly southof the field orientation. The soils are predom-inantly Norfolk loamy sands with a highsusceptibility to wind erosion. During the windevents in question, sediment was removed fromthe field and carried eastward where a deposi-tional fan was formed in the grass on an adjacentuncultivated path wide enough to allow passageof farm equipment. No wind recording equip-ment was deployed at the site at the time of theevent. The volume of sediment deposited onthe path was determined using the same tech-nique that was applied at Field C. The totalvolume of sand accumulated along the margin ofthe field amounted to 9.5 Mg haj1 or 5.9 kgmj1 hj1 (Table 3). This volume does notinclude any sediment transported across the pathand deposited on the adjacent field, so theamount of deposition measured underrepresentsthe total flux from the field.

Field F

A significant wind event occurred on March31 to April 1, 1997, that resulted in obvioussediment redistribution on the fields of southernPitt County. The northwest wind had anaverage wind speed at Kinston of 8.1 m secj1

and a maximum of 10.3 m secj1, lasting 25 h.Rainfall during the period before the event waslimited to single significant occurrence onMarch 26 when 1.2 cm of precipitation wasrecorded at Kinston. Sediment accumulation onthe downwind margins of many fields wasclearly visible after the event where there hadbeen none previously. The field at location F is480 m long by 180 m wide, giving an area of 8.9ha (Fig. 8). The long axis of the field is orientedin a north-northwest to south-southeast direc-tion. The soils are composed primarily of

TABLE 3

Wind event data and wind potential index

DateField

(ha)

Area

(m secj1)Direction Speed

Duration

(h)

Probability

(%)

Flux Mg

haj1

Transport

kg mj2 hj1

4/25/93 A 16.8 205 5.9 30 11.0 5.4 5.7

4/5/94 B 14.4 208 6.52 17 19.4 7.8 11.7

3/3/99 C 5.7 249 8.10 30 1.2 25.7 20.4

2/3–11/02 D 3.8 331 6.6 27 2.4 48.4 59.7

3/19–21/96 E 2.2 253 8.05 43 1.0 9.5 5.9

3/31/97 F 8.9 335 8.14 18 2.0 10.4 31.1

3/31/97 G 2.7 335 8.14 18 2.0 150.8 139.2

Probability of occurrence determined from wind Kinston wind data for 1993–2003.


200120


Portsmouth loamy soil. There is a small area ofAlaga loamy sand along the south side of thefield. A road runs along the southeastern end ofthe field. The road surface is 0.25 to 0.5 mbelow the field surface and so represents a naturalsand trap for sediment removed from the fieldduring westerly winds. A significant depositionalfeature accumulated along the edge of the field,at the road margin, consisting of a dune ridge ofsand with an obvious stoss slope into the road(Fig. 1C). After the wind event, the area of thedepositional dune ridge was measured withsurveying equipment, and a map was producedof the feature. The depth of deposition wasdifficult to determine in this instance because thesediment was packed so loosely that cores couldnot be obtained with the augur. Estimates of

deposition depths were obtained by trenchingthrough the dune at several locations. The slopeof the underlying original roadside ditch wasestimated with these data and was assumed toextend along the entire length of the dunefeature. This ditch slope was used as the bottomof the newly deposited dune and volume wascomputed accordingly. It was determined that10.4 Mg haj1 (Table 3) of sediment weredeposited in the dune as a result of the wind-storm. This is the equivalent of 31.1 kg mj1 hj1.

Field G

The March 31, 1997, wind storm alsoeroded significant amounts of sediment fromField G that has an east-west orientation. Thefield measures 300 m in length and 90 m in

Fig. 8. Field E, showing the location of the traps and anemometer mast, the distribution of soils, the wind speed/direction hourly data for the event monitored and the precipitation record for the 10 days before the wind event.


200120


width, giving it an area of 2.7 ha (Fig. 9). Thesoil on this field is primarily Portsmouth loam,along with a large area of Olustee loamy sand(Table 1). Sediment removed from the fieldduring this event produced a depositional fan onan adjacent field located immediately to the east.The extent of the fan was clearly visible as thedeposited sediment consisted of light-coloredsandy material overlaying the darker soil on thereceiving field. The fan volume was determinedby using the survey procedure described pre-viously. The amount of sediment deposited inthe fan amounted to 150.8 Mg haj1 or 139.2 kgmj1 hj1 (Table 3).

DISCUSSION

Large amounts of sediment moved during

the wind events monitored in this study despite

relatively low average wind speeds. The

amounts of sediment moved in North Carolina

(Table 4) compare with those reported for other

humid and even semiarid sites (Buschiazzo et al.,

1999; Fryrear, 1995; Sterk and Stein, 1997),

although these values fall below those measured

for extreme conditions in other semiarid to arid

areas (Stout, 2001). Most of these studies used

sampling systems that captured dust emissions as

well as some saltation load. Thus, the numbers

Fig. 9. Fields F and G, showing the location of the traps and anemometer mast, the distribution of soils, the windspeed/direction hourly data for the event monitored and the precipitation record for the 10 days before the windevent.


200120


reported here for North Carolina fields under-represent the true sediment flux from thesefields. Pease et al., (2002) sampled at a variety ofheights above the surface and report totalsaltation and suspension fluxes of 1.14 kg haj1

hj1 and 762 kg haj1 hj1 for two events inFebruary 2002. Both events, the proportion ofsediment trapped above 0.25-m heightamounted to less than 10% of the total flux.Thus, one can conclude that the data reportedhere underrepresent the total flux from thesefields by approximately 10%. However, it alsoshould be recognized that the relative propor-tion of suspended and saltation loads on thefields can be expected to vary in association withthe source material on the field. Pease et al.(2002) report that there were large variationswithin the study field that they associate withthe presence of different soil types on the field.Fields with a high proportion of sand in thesource sediment can be expected to have alower amount of suspended load. On the otherhand, a high proportion of silty material in thesource sediment does not necessarily mean thatthere will be a large amount of suspended loadbecause the increased cohesion between par-ticles inhibits the release of sediment from thesource (Goossens, 2004). All our data show thatthat, at least on an event basis, eolian sedimentredistribution on eastern North Carolina coastalplain agricultural fields is comparable with otherlocations around the world. This suggests thateolian geomorphological processes should notbe ignored in what are apparently stable (in aneolian sense) environments.

The volumes of sediment transport measuredin Pitt County range from 5.5 to 140 kg mj2 hj1,a range of two orders of magnitude. Although

there should be a general correspondencebetween average event wind speed and theamount of sediment moved, local field conditionsexert significant influence on the amount ofeolian erosion, confusing the relationship. Weattribute the variations in sediment flux to thecharacteristics of the soils on each field. Moisturecontent of the sediment has been shown to be asignificant determining factor in explaining thevariations in sediment transport (Sarre, 1988,1989; Namikas and Sherman, 1995). Althoughwe did not measure moisture content in the soilsof our study sites, we did identify the precipitationpattern and amounts in the days before the eventsin question. The amount of sediment movedduring the event seems to be affected by therecency/magnitude of the rainfall. Thus, sedimentflux at Field E was quite small during a wind eventthat had a high average wind speed. We believethat eolian erosion was limited by the occurrenceof rainfall just 48 h before the wind event.

Another factor influencing the wind eventmagnitude/transport relationship is the soil ateach site. This is important because antecedentsoil-moisture is a controlling factor in sedimentflux from these agricultural fields particularlybecause the mixture of silts and clays with sandsin the soil help to retain moisture. Soils withgreater silt/clay fractions would remain wetterand would yield less sediment during eolianevents. There were parts of several study fieldsthat consisted of USDA soil types that aredescribed as containing significant proportionsof silt-sized particles, as well as smaller amountsof clays. Field F measuring 8.6 ha, yielded on10.4 Mg haj1, whereas Field G, measuring 2.7ha, yielded 15 times more sediment during thesame wind event. We attribute the difference tothe fact that Field F consisted mostly of fine-grained Portsmouth and Tuckerman soils,whereas 48% of Field G consists of less fineOlustee and Pactolus soils.

Eolian processes are inconspicuous factors inmodifying the landscape in a wide range ofenvironmental systems worldwide. However,the quantities of sediment redistributed havethe potential repercussions on agricultural pro-ductivity. This study shows that significantsediment deposits can accumulate along thefield margins after a wind event. Although thissediment is sandy in nature and might be viewedas having little importance to the productivity ofthe field, it represents a loss from the productiveA horizon. When considered in combinationwith the removal of finer particles during the

TABLE 4

Comparative flux and sediment transport values in North

Carolina and other selected areas of the world

FieldFlux

Mg haj1

Transport

kg mj1 hj1

A 5.4 5.7

B 7.8 11.7

C 25.7 20.4

D 48.4 59.7

E 9.5 5.9

F 10.4 31.1

G 150.8 139.2

Buschiazzo et al. (1999) 0.3–1.8

Buschiazzo et al. (1999) 0.75–1.0

Fryrear (1995) 0.5–70

Sterk and Stein (1997) 42–250


200120


same events (Pease et al., 2002), the long-termconsequences should be of concern to farmers.

Our observation that local field conditionsexert significant influence over the magnitude ofeolian erosion has management implications. Onemight conclude that individual field managementought to involve the adoption of a worse casemanagement approach in which the maximumdeflation would be identified for a field and thatwould be adopted as the standard for erosioncontrol. In North Carolina, it is clear that somefields are very susceptible to sediment removal bywind, whereas others are not. Differentiating afield’s susceptibility to eolian processes wouldallow farmers to focus their field managementefforts. The biggest problem with this approach isthat it requires identifying the in situ sedimenttransport and soil characteristics. As Phillips et al.(1999a, b) demonstrate, there can be largevariations from the accepted distribution of soiltypes (USDA, 1974), and this will influence localsediment transport conditions. There is theopportunity to incorporate remote sensing andGIS to map out areas highly susceptible todeflation on individual fields. This may representa fruitful area for future research.

CONCLUSIONS

The results of these small field studies showthat eolian sediment transport can be significanton agricultural fields in eastern North Carolina,a region where climate and soil conditionsindicate a low potential for soil erosion by wind.Although the results presented here do notprovide direct cause and effect relationshipsbetween the propensity for sediment transport/field erosion and potential factors that mightexplain both between and within field variations,there are indications that soil conditions areprimarily responsible for those variations. How-ever, the relatively high concentration of sand-sized particles in most of the soils of eastern NorthCarolina provides a ready source of material forsaltation load. The observed amounts of sedimentflux from the fields studied here also providesupport for the suggestion made by Phillips(1993) and Phillips et al. (1993) that soil horizontruncation may be partly attributable to sedimentremoval by wind. The importance of windprocesses on these fields needs to be confirmedthrough more detailed field studies that incorpo-rate a variety of methodologies designed tomeasured wind velocity characteristics, sedimenttransport, soil-moisture variations attributable to

drying rates of surface particles. In addition, soiltilling in the spring may have an important effecton loosening the soil particles in during a seasonwhen insolation is increasing, and this in turnwill affect surface evaporation rates, drying outthe soil and making it more susceptible towind erosion.

ACKNOWLEDGMENTS

The authors would like to acknowledge thecontributions of graduate students at EastCarolina University who over the years haveassisted the authors by helping to set up fieldequipment and analyze samples. In particular,the authors would like to recognize the contri-butions of Matt Flynn, Mark Lampe, and MarkLange. The authors would like to thank twoanonymous reviewers for comments made on anoriginal draft that led to a significant improve-ment in the manuscript. The research was in largepart funded by a grant from the USDA.

REFERENCES

Bilbro, J. D., and D. W. Fryrear. 1994. Wind Erosionlosses as related to plant silhouette and soil cover.Agron. J. 86(3):550–553.

Buschiazzo, D. E., T. M. Zobeck, and S. B. Aimar.1999. Wind erosion in loess soils of the semiaridAregentinian Pampas. Soil Sci. 164(2):133–138.

Chepil, W., F. H. Siddoway, and D. V. Armhurst.1962. Climate factor for estimating wind erodibilityof farm fields. J. Soil Water Conserv. 17:162–165.

Chepil, W., F. H. Siddoway, and D. V. Armhurst.1963. Climatic index of wind erosion conditionsin the Great Plains. Soil Sci. Soc. Am. Proc.27:449–451.

Chepil, W., and N. P. Woodruff. 1963. The physicsof wind erosion and its control. Adv. in Agron.,15:211–302.

Clay, J. W., D. M. Orr, Jr., and J. E. Stuart. 1975.North Carolina Atlas. The University of NorthCarolina Press, Chapel Hill, NC, 331 pp.

Daniels, R. B., H. J. Kleiss, S. W. Buol, H. J. Byrd,and J. A. Phillips. 1984. Soil systems of NorthCarolina.N.C. Agric.Res. Bull. (RaleighNC) 467:77.

Fryrear, D. W. 1981. Long-term effect of erosion andcropping on soil productivity. Geol. Soc. Am.Spec. Pap. 186:253–260.

Fryrear, D. W. 1995. Soil losses by wind erosion. SoilSci. Soc. Am. J. 59:668–672.

Fryrear, D. W., C. A. Krammes, D. L. Williamson, andT. M. Zobeck. 1994. Computing the wind erodiblefraction of soils. J. SoilWater Conserv. 49(2):183–188.

Goossens, D. 2004. Effect of soil crusting on theemission and transport of wind-eroded sediment:


200120


field measurements on loamy sandy soil. Geo-morphology. 58:145–160.

Hagan, L. J., and D. V. Armbrust. 1994. Plant canopyeffects of wind erosion saltation. Trans. Am. Soc.Agric. Eng. 37:461–465.

Hubbard, R. K., J. M Sheridan, and L. R. Marti.1990. Dissolved and suspended solids transportfrom coastal plain watersheds. J. Environ. Qual.19:413–420.

Jackson, N. L., and K. F. Nordstrom. 1997. Effects oftime-dependent moisture content of surface sedi-ments on aeolian transport rates across a beach.Earth Surf. Processes Landf. 22:611–621.

Kimberlin, L. W., A. L. Hidlebaugh, and A. R.Grunewald. 1977. The potential wind erosionproblem in the United States. Trans. Am. Soc.Agric. Eng. 20:873–879.

Leatherman, S. P. 1978. A new eolian sand trapdesign. Sedimentology. 25:303–306.

Lowrence, R., J. K. Sharpe, and J. M. Sheridan.1986. Long-term sediment deposition in theriparian zone of a coastal plain watershed. J. SoilWater Conserv. 41:266–271.

Lowrence, R., S.McIntire, andC. Lance. 1988. Erosionand deposition in a field estimated using sesium-137activity. J. Soil Water Conserv. 43:195–199.

Markewich, H. W., Pavich, M. J. & Buell, G. R.1990. Contrasing soils and landscapes of thepiedmont and coastal plain, eastern United States.Geomorphology. 3:417–448.

McCauley, J. F., M. J. Grolier, and D. J MacKinnon.1981. The US dust storm of February, 1977.Geological Soc. Am. Spec. Pap. 186:123–148.

McKenna-Neuman, C, and W. G. Nickling. 1989. Atheoretical and wind tunnel investigation of theeffect of capillary water on the entrainment ofsediment by wind. Can. J. Soil Sci. 69:79–96.

Middleton, N. J., A. S. Goudie and G. L. Wells.1986. The frequency and source areas of duststorms. In: Aeolian Geomorphology. W. G.Nickling (ed). Allen and Unwin, London, pp.237–259.

Namikas, S. L., and D. J. Sherman. 1995. A review ofthe effects of surface moisture content on aeoliansand transport. In: Desert Aeolian Processes. V. P.Tchakerian (ed.). Chapman-Hall, London, pp.269–293.

Nanney, R. D., D. W. Fryrear, and T. M. Zobeck.1993. Wind erosion prediction and control. WaterSci. Technol. 28:519–527.

Pease, P., P. A. Gares, and S. Lecce. 2002. Eoliandust erosion for an agricultural field on theNorth Carolina coastal plain. Phys. Geogr. 23(5):381–400.

Phillips, J. D. 1992a. The source of alluvium in largerivers of the lower coastal plain of North Carolina.Catena. 19:59–75.

Phillips, J. D. 1992b. Delivery of upper-basin sedi-

ment to the lower Neuse River, North Carolina.Earth Surf. Processes Landf. 17:699–709.

Phillips, J. D. 1993. Pre- and post-colonial sedimentsources and storage in the lower Neuse Riverbasin, North Carolina. Phys. Geogr. 14:272–284.

Phillips, J. D., M. J. Wyrick, J. G. Robbins, andM. Flynn. 1993. Accelerated erosion on the NorthCarolina coastal plain. Phys. Geogr. 14:114–130.

Phillips, J. D., M. C. Slattery, and P. A. Gares. 1999a.Truncation and accretion of soil profiles on coastalplain croplands: implications for sediment redis-tribution. Geomorphology. 28:119–140.

Phillips, J. D., H. Golden, K. Cappiella, B. Andrews,T. Middleton, D. Downer, D. Kelli, and L.Padrick. 1999b. Soil redistribution and pedologictransformations in coastal plain croplands. EarthSurf. Processes Landf. 24:23–39.

Potter, K. N. 1990. Estimating wind-erodible materialson newly crusted soils. Soil Sci. 150:771–776.

Robinson, D. N. 1968. Soil erosion by wind inLincolnshire, England. EastMidl. Geogr. 4:351–362.

Sarre, R. D. 1988. Evaluation of aeolian sand trans-port equations using intertidal zone measure-ments, Saunton Sands, England. Sedimentology.35:671–679.

Sarre, D. R. 1989. Aeolian drift from the intertidalzone on a temperate beach: potential and actualrates. Earth Surf. Processes Landf. 14:247–258.

SCS (Soil Conservation Service). 1982. Upper TarRiver Erosion Survey. Raleigh. U.S. Departmentof Agriculture, pp. 57.

SCS, (Soil Conservation Service). 1983. Upper NeuseRiver Erosion Survey. Raleigh. U.S. Departmentof Agriculture, pp. 54.

Simmons, C. E. 1988. Sediment Characteristics ofNorth Carolina Streams, 1970–1979. USGS OpenFile Report 97-701. US Geological Survey,Washington, DC, pp. 130.

Sterk, G., and A. Stein. 1997. Mapping wind-blownmass transport by modeling variability in space andtime. Soil Sci. Soc. Am. J. 61:232–239.

Stout, J. E. 2001. Dust and environment in theSouthern High Plains of North America. J. AridEnviron. 47(4):425–441.

Trimble, S. W. 1974. Man-induced soil erosion in thesouthern Piedmont, 1741–1970. Soil ConservationSociety of America, Ankeny, IA, pp. 180.

USDA. 1974. Soil Survey, Pitt County, NorthCarolina. U.S. Department of Agriculture, pp. 73.

Woodruff, N. P., and F. H. Siddoway. 1965. A winderosion equation. Proc. Soil Sci. Soc. Am. 29:602–608.

Wilshire, H. G., J. K. Nakata, and B. Hallet. 1981.Field observations of the Decemeber 1977 windstorm, San Juaquin Valley, California. Geol. Soc.Am. Spec. Pap. 186:233–252.

Zobeck, T. M. 1991. Soil properties affecting winderosion. J. Soil Water Conserv. 46:112–118.


200120

Date post:	31-Oct-2020
Category:	Documents
Upload:	others
View:	2 times
Download:	0 times

EOLIAN SEDIMENT TRANSPORT ON NORTH CAROLINA … · Joaquin Valley (Wilshire et al., 1981) and up to...

Documents