LETTERSPUBLISHED ONLINE: 21 OCTOBER 2012 | DOI: 10.1038/NGEO1615

Linking the historic 2011 Mississippi River flood tocoastal wetland sedimentationFederico Falcini1,2,3, Nicole S. Khan1, Leonardo Macelloni4, Benjamin P. Horton1, Carol B. Lutken4,Karen L. McKee5, Rosalia Santoleri2, Simone Colella2, Chunyan Li6, Gianluca Volpe2,Marco D’Emidio4, Alessandro Salusti1,7 and Douglas J. Jerolmack1*Wetlands in the Mississippi River deltaic plain aredeteriorating1 in part because levees and control structuresstarve them of sediment2–4. In spring 2011 a record-breakingflood brought discharge on the lower Mississippi River todangerous levels, forcing managers to divert up to 3,500 m3 s−1

of water to the Atchafalaya River Basin5. Here we use field-calibrated satellite data to quantify differences in inundationand sediment-plume patterns between the Mississippi andAtchafalaya River. We assess the impact of these extremeoutflows on wetland sedimentation, and use in situ data col-lected during the historic flood to characterize the Mississippiplume’s hydrodynamics and suspended sediment. We showthat a focused, high-momentum jet emerged from the leveedMississippi, and delivered sediment far offshore. In contrast,the plume from the Atchafalaya was more diffuse; divertedwater inundated a large area, and sediment was trapped withinthe coastal current. The largest sedimentation, of up to severalcentimetres, occurred in the Atchafalaya Basin despite thelarger sediment load carried by the Mississippi. Sedimentaccumulation was lowest along the shoreline between the tworiver sources. We conclude that river-mouth hydrodynamicsand wetland sedimentation patterns are mechanistically linked,providing results that are relevant for plans to restore deltaicwetlands using artificial diversions2–4,6–8.

Protecting and expanding coastal wetlands is vital for ecosystemservices of the Mississippi River Delta9–12, and harnessing naturalprocesses of wetland building using the Mississippi River and itssediments is an essential component of restoration plans2–4. Theonly portion of the delta experiencing significant expansion ofcoastal wetland at present is at the mouth of the Atchafalaya River(Fig. 1a), where a higher mineral (that is, non-organic) sedimentconcentration3 and hydrodynamic factors8 allow sufficient sedi-ment deposition6 to outpace subsidence and sea-level rise13. Therecently released 2012 Coastal Master Plan14 proposes river diver-sions and channel realignment to divert sediment and fresh waterfrom the Mississippi River and Atchafalaya River into adjacentbasins, to reconnect the river to delta wetlands. Successful designand implementation of such measures require an understandingof diverted sediment movement and deposition, especially duringhigh-water events when the potential sediment load is greatest.

The Mississippi River flood of spring 2011 was one of the largeston record5,15. Both the Mississippi River and Atchafalaya River

1Department of Earth and Environmental Science, University of Pennsylvania, Philadelphia, Pennsylvania 19104, USA, 2Istituto di Scienze dell’ Atmosfera edel Clima, Consiglio Nazionale delle Ricerche, Rome 00133, Italy, 3St Anthony Falls Laboratory, and National Center for Earth-surface Dynamics, Universityof Minnesota, Minneapolis, Minnesota 55414, USA, 4Mississippi Mineral Resources Institute, University of Mississippi, University, Mississippi 38677, USA,5US Geological Survey, National Wetlands Research Center, Lafayette, Louisiana 70506, USA, 6Department of Oceanography and Coastal Sciences,School of the Coast and Environment, Louisiana State University, Baton Rouge, Louisiana 70803, USA, 7Dipartimento Scienze Geologiche, Roma Tre,00146 Rome, Italy. *e-mail: sediment@sas.upenn.edu.

exhibited elevated suspended sediment loads (Fig. 1d). Floodwaterdischarge (Qw) at the Old River Control Structure—the locationwhere theMississippi River bifurcates into the Atchafalaya andMis-sissippi channels—crested at a value of ∼20,000m3 s−1 for the pe-riod of 14–31May 2011 (ref. 15). To relieve pressure on levees alongthe downstream portion of the Mississippi River in Baton Rougeand New Orleans, the Morganza Spillway (Fig. 1a) was opened on14 May 2011 for the first time in almost 40 years. At peak flooding,3,500m3 s−1 of water was being diverted to the west into theAtchafalaya Basin15, flooding the swamps andmarshes along the en-tire length of theAtchafalayaRiver (Supplementary Information).

Although both the Mississippi River and Atchafalaya Riverchannels had obvious sediment-laden plumes emanating fromtheir mouths, the differences in plume patterns and extent ofinundationwere striking (Fig. 1 and Supplementary Figs S2 and S3).We performed time-series analysis of suspended sediment con-centration (SSC) from Moderate Resolution Imaging Spectrora-diometer (MODIS) Aqua satellite data, calibrated using field mea-surements (Supplementary Information). The Mississippi River—whose floodwaters upstream were completely contained within ar-tificial levees—exhibited narrow and focused jets of sediment-ladenwater, especially from Southwest Pass, which penetrated the coastalcurrent with limitedmixing16–20 for the duration of the flood (Fig. 1and Supplementary Fig. S8). The intentionally flooded AtchafalayaBasin inundated a∼100-km-wide coastal zone, and sediment fromits broad plume seemed to be trapped in the nearshore zone for fourweeks, where it thoroughlymixedwithmarinewaters (Fig. 1a–c andSupplementary Fig. S8). The diffuse nature of the Atchafalaya Riverplume may also have been enhanced by the shallow depth of thereceiving bay, which would increase bottom friction compared withthe Mississippi River. We expected greater wetland sedimentationover a broad area in the Atchafalaya Basin, from both directdeposition by floodwaters and indirect deposition through coastalreworking of the plume (Fig. 1a–c; Supplementary Figs S2 and S8).

To test these ideas we conducted a sedimentation survey of45 sites by helicopter across the Mississippi Birdsfoot, Barataria,Terrebonne and Atchafalaya basin wetlands during 21–27 June2011 (Fig. 2). Shallow sediment cores (5 cores per site) wereextracted from the marsh surface at a consistent distance (5m)from waterways. The surface sediment layer was presumed to be arecent deposit on the basis of distinguishing features such as lack ofplant roots and different colour and consistency from underlying

NATURE GEOSCIENCE | VOL 5 | NOVEMBER 2012 | www.nature.com/naturegeoscience 803

© 2012 Macmillan Publishers Limited. All rights reserved.

LETTERS NATURE GEOSCIENCE DOI: 10.1038/NGEO1615

continuous current meter transects

NE1NE2

NE3 SE1

SE2

SE3

SE4

SE5

SW

Pass 1

Pass 4

P2-3P3-3

50 km

30° N92° W

Morganza Spillway

Baton Rouge

Bonnet Carre

Belle Chasse

New Orleans

Lake Pontchartrain

Mississippi RiverBirdsfoot DeltaAtchafalaya Bay

Atchafalaya River

Mississippi River

Hydrographic/current meter stations

Latit

ude

(° N

)

Latit

ude

Longitude Longitude

0

10

20

30

40

50

0

20

40

60

80

Long

itude

(° W

) Distance (km

)

92° W 91° W

900 2,200 3,500

SSC (mg l¬1)

QW

SSC MR AR

120

160

200

240

280

320

360

400

Continuous current meter transects

29.1

29.2

29.3

29.4

29.5

29.6

Atchafalaya River

90° W 89° W

Mississippi River

+

89.2

89.4

89.6

89.8

5 75 150

29° N

30° N

0.4

0.8

1.2

1.6

2.0

2.4

2.8

3.2

3.6

SSC (m

g l ¬1)

Apr. 15 Apr. 30 May 15

Qw

(m

3 s¬

1 × 1

04)

May 30 Jun. 14 Jun. 29

a

b

c d

30 Apr. 16 May 31 May 16 Jun. 01 Jul 30 Apr. 16 May 31 May 16 Jun. 01 Jul.

Atchafalaya River Mississippi River

Louisiana Mississippi

Distance (km

)

Figure 1 |Mississippi and Atchafalaya river plume patterns during the 2011 flood. a, MODIS ocean colour image on 1 June 2011 showing the suspendedsediment pattern along the coast; boat survey locations and points of interest are indicated. b, Hovmöller plots of SSC (mg l−1) from field-calibratedMODIS Aqua (Supplementary Information) along cross-plume transects of the Atchafalaya (left) and Mississippi (right) rivers, indicated on 1 May–31 July,2011 (c). c, Cumulative SSC over the same period as in b. Values on the top (bottom) side of the colour bar refer to b (c). d, US Geological Survey gaugedata of daily discharge (Qw (m3 s−1), lines) and SSC (mg l−1; symbols) for the Mississippi River at Belle Chasse, Louisiana and the Atchafalaya River atMorgan City, Louisiana. SSC data for the Atchafalaya River were collected at Simmesport, Louisiana. Note that sediment load, Qs=Q∗wSSC.

804 NATURE GEOSCIENCE | VOL 5 | NOVEMBER 2012 | www.nature.com/naturegeoscience

© 2012 Macmillan Publishers Limited. All rights reserved.

NATURE GEOSCIENCE DOI: 10.1038/NGEO1615 LETTERS

Terrebonne

Barataria

Atchafalaya

Mississippi River (Birdsfoot) Delta

Longitude (decimal degrees)

Ave

rage

accu

mul

atio

n(g

cm

¬2 )

0 25 50 75 10012.5 km

Average depth (cm)< 0.50.5 - 2.5> 2.5

Longitude

Latit

ude

89° W90° W91° W

SSC (m

g l ¬1)

200

150

100

50

0

29° 40' N

29° 20' N

29° N

28° 40' N

91.6 91.1 90.6 90.1 89.6 89.10

5

10b

a

Figure 2 | Spatial distribution of sediment during the 2011 flood. a, Locations, and measured recent sediment accumulation, from shallow cores along thedelta shoreline (circles), merged with a map of SSC on 1 June 2011 derived from field-calibrated MODIS Aqua data. b, Recent sediment accumulation(±s.d.) at each sampling site, calculated using the average thickness (n= 5) and bulk density (n= 2) of the flood sediment layer.

sediments. The thickness of the surface sediment layer varied from0.0 to 8.3 cm, with a coast-wide average of 1.5 cm. These sedimentswere composed of moderately sorted fine silt (median value fromall basins, 13.4mm) with a low organic content (10± 1%) andmoderate bulk density (0.60± 0.05 g cm−3), features that did notvary significantly across basins (Supplementary Table S2). Usingsediment thickness and bulk density, we calculated that recentsediment accumulation was greatest in the Atchafalaya (1.61±0.96 g cm−2, n= 14), supporting expectations of greater potentialcontribution of sediment to wetland accretion (Fig. 2). Recentaccumulation in the Birdsfoot Delta was smaller but still substantial(1.14±0.78 g cm−2, n=9), showing thatMississippi River sedimentreached some marsh areas through small channels or depositionfrom its spatially restricted coastal plume (Fig. 2). Much lesssediment accumulated in the Terrebonne (0.42±0.18 g cm−2, n=14) and Barataria (0.34±0.22 g cm−2, n=8) basinwetlands, locatedfarthest from the rivers. The correspondence between zones ofhigh shoreline deposition, and coastal SSC patterns identified fromsatellite data (Fig. 2), suggests plume-derivedwetland deposition.

Recent sediments in the Atchafalaya and Birdsfoot basinswere comparable to the underlying deposits in terms of bulkdensity and grain size (Supplementary Table S2), but contained agreater abundance of centric diatoms, which are planktonic formscommonly found in the water column21. Underlying sedimentwas dominated by pennate taxa typical of benthic environments21(Supplementary Fig. S9). The ratio of centric/pennate diatoms inthe surface sediment was 127% higher than in the underlyinglayer in the Atchafalaya and Birdsfoot basin sites, but wasonly 23% higher in the Terrebonne and Barataria basins. Thisfinding supports the interpretation that recent accumulation in theAtchafalaya River and Mississippi River resulted from depositionof suspended sediment, either from overbank flooding or anocean-mixed plume. The former is more likely the case for sitesfarther inland, whereas the latter probably dominated for coastallocations (Fig. 2). We also observed these sediments to be relativelyenriched in diatom species indicative of low-salinity environments(Supplementary Information), supporting riverine influence. Thesimilarity in physical characteristics of recent flood sediments andunderlying deposits, and the low organic content of all samples,indicate that mineral sediment of river origin has been a dominantcontributor to building andmaintaining these coastal wetlands.

To better understand the Mississippi River plume dynamicsduring this extraordinary event, and to calibrate satellite data, we

carried out oceanographic transects off the Birdsfoot Delta duringthe peak of the flood (Fig. 1 and Supplementary Figs S4 and S5).Although studies have used satellites to track Mississippi plume dy-namics during floods22, in situ hydrographic measurements duringsuch a large flood have not been conducted. Our transect surveyscaptured vertical profiles of flow velocity, temperature, salinity,SSC and grain-size distributions, allowing us to characterize thehydrodynamics of the effluent plume and to construct a sedimentbudget for the Mississippi River flood outflow (Fig. 3). Resultsfrom the perimeter of the Birdsfoot Delta showed three mainfreshwater outflows corresponding to the Southwest, Southeastand Northeast passes, each distinguished by salinity < 15 PSU,temperatures cooler than ambient (Supplementary Information)and SSC > 40mg l−1. We estimated a total sediment discharge ofQs =Q∗wSSC∼ 3.9×103 kg s−1, in agreement with values reportedfrom an upstream gauge (Fig. 1d).

A Southwest Pass survey showed that its jet was easilyrecognized, even 50 km offshore, as a persistent core of high currentvelocity and SSC (Fig. 3). To relate jet dynamics to sedimentsuspension, we computed the Rouse number—essentially the ratioof particle settling velocity (vs) to fluid shear velocity (u∗)—for eachmeasurement station. We expected sediment to be well suspendedfor vs/(ku∗)<2.5 (ref. 23), where k is von Kármán’s constant. Shearvelocity and particle settling velocity were estimated from currentmeter and grain-size data, respectively, collected off the threestudy passes. The typical range of mineral suspended sediment was0.02–0.2mm in diameter, corresponding to representative values ofvs/(ku∗) of 0.016–2.262, respectively. This ratio was constant fromthe mouth of Southwest Pass up to ∼40 km offshore, confirmingthat all sediments were suspended within this narrow river jet(Supplementary Information).

The exceptionally coherent plume generated by SouthwestPass—the largest of the Mississippi River outflows—penetrated a∼1m s−1 northwest-trending coastal current (Fig. 1 and Supple-mentary Fig. S8). The absence of a littoral sediment plume and asso-ciated sedimentation along the Terrebonne and Barataria shoreline(Fig. 2 and Supplementary Fig. S8) corroborates that the SouthwestPass flood plume was insensitive to coastal processes. To examinethe physical basis for these dynamics, we calculate the potentialvorticity for this plume using the bulk parameter derived in ref. 8:5c =UC/(hW ), where U (m s−1) and C (mg l−1) are scale quan-tities for jet velocity and SSC, respectively, and h (m) and W (m)are channel outlet depth and width, respectively (Supplementary

NATURE GEOSCIENCE | VOL 5 | NOVEMBER 2012 | www.nature.com/naturegeoscience 805

© 2012 Macmillan Publishers Limited. All rights reserved.

LETTERS NATURE GEOSCIENCE DOI: 10.1038/NGEO1615

A

A’

B B’

C

C’

C’

BB’

Velocity contours: 1.2, 0.8, 0.4 m s¬1

Velocity contours: 1.2, 0.8, 0.4 m s¬1

SW

SE

NE

80

60

40

20

0

89° W89.5° W

Longitude

Latit

ude

89° W89.5° W 89° W89.5° W

SSC (mg l¬1)

SSC (mg l¬1) SSC (mg l¬1)

A A’

h (m

)

h (

m)

Section distance (km)

Section distance (km)

40

40

40

40

8020

20

60

Ocean data view

Ocean data view

Ocean data view

Ocean data view

0

5

10

15

200 50 100 150

80

60

40

20

0

0

5

10

15

20

h (m

)

0 5 10 15 20Section distance (km)

0 5 10 15 20

80

60

40

20

0

0

5

10

15

20

a b

c d

SW SE NE

29° N

28.8° N

28.6° N

29.4° N

29.2° N

C

Ocean data view

29° N

28.8° N

28.6° N

29.4° N

29.2° N

Ocean data view

29° N

28.8° N

28.6° N

29.4° N

29.2° N

Figure 3 | SSC (mg l−1) and velocity (m s−1) profiles around the Mississippi River Birdsfoot Delta (1 June 2011). a, Map indicating a large-scale SSCvertical transect A–A’ around the Birdsfoot Delta; Southwest, Southeast and Northeast passes are highlighted. b, SSC vertical profile along transect A–A’;three main sediment outflows are indicated. c, SSC and velocity profiles along transect B–B’ offshore of Southwest Pass; inset shows location. d, SSC andvelocity profiles along transect C–C’.

Information). High potential vorticity inputs to the ocean—suchas cold-wind-generated filaments24 or bottom-influenced jets25,26—tend to conserve their potential vorticity along the direction offlow, producing coherent self-sharpening jets24–28 and, in the caseof sediment plumes, localized deposition at jet margins8. We foundthat offshore potential vorticity values were comparable to thatat the channel mouth (that is, potential vorticity was conserved):5c ∼ 2.49 × 10−5, 2.24 × 10−5, 2.14 × 10−5 kgm−4 s−1, for theSouthwest Pass, Pass 1 and Pass 4 transects, respectively (Figs 1a and3 and Supplementary Table S1). Data indicate that the SouthwestPass plume indeed behaved as a self-sharpening jet, a class of flowsknown to experience limitedmixingwith ambient waters27,28 (Figs 1and 3). The large momentum flux through a relatively narrowchannel outlet, and the absence of strong frictional effects, seemto be responsible for the lack of interaction between the SouthwestPass plume and the coastal current (Supplementary Fig. S8). Thispattern is in contrast to the Atchafalaya sediment plume that shifteddirection following ocean currents (Supplementary Fig. S8).

Our analysis suggests that river-mouth hydrodynamics influ-enced sediment deposition patterns during the spring 2011 flood.The historic Morganza Spillway opening simulated a more naturalflooding scenario in the Atchafalaya River: this diffuse plume—influenced by coastal currents and winds—delivered substantialsediment over a broad area, both directly to wetlands throughinundation and to the nearshore zone where tides and currentscould potentially carry it onshore. Although the Mississippi Rivercarried a larger sediment load than the Atchafalaya River, it pro-duced less sedimentation. Flow confinement promotes delivery ofvast quantities of sediment far offshore, where it cannot build aland platform to support wetlands. If the Mississippi River plume

was diffuse, its sediments would probably have been carried shore-ward with the coastal current to produce substantial depositionat Barataria and Terrebonne. To address this problem, MississippiRiver diversions are proposed upstream of the Birdsfoot wheresediment would be delivered into shallower receiving basins3,4,29not currently fed by the Mississippi River at present. Althoughthe ultimate success of such diversions will depend on a varietyof factors7,30, our work shows how fine sediments carried in aflood and diverted to shallow marine settings could contributesubstantially to marsh sedimentation. This finding complementsa recent study demonstrating significant sand deposition in theBonnet Carre Spillway resulting from diversion during the sameflood29. Engineered diversions could harness the full spectrum ofriver sediment tomitigate wetland loss in key areas.

MethodsWater discharge and SSC data were obtained from US Geological Surveysurface-water time series (National Water Information System; http://la.water.usgs.gov/MississippiRiverFlood2011.html/, accessed 12 July 2012) over the rangefrom 1 April to 30 June 2011. These data were collected by automatic recorders andmanual measurements at field installations. Satellite ocean true-colour images wereobtained from theMODIS.We employedMODIS images processed by the Instituteof Marine Remote Sensing of the University of South Florida (http://imars.marine.usf.edu/, accessed 5October 2011). Satellite SSC data were obtained using processedMODIS Level-1A products, by following a procedure for estimating suspendedload from remote-sensing reflectance high-resolution band 1 at 645 nm. MODISimages were downloaded through the NASA (National Aeronautics and SpaceAdministration) Internet servers OceanColor Web (http://oceandata.sci.gsfc.nasa.gov/, accessed 16 July 2012). Coastal currents were examined using nowcast resultsfrom the South Atlantic Bight and Gulf of Mexico Circulation Model, which isdeveloped, operated and maintained by the Ocean Observing and Modeling Groupof the Department of Marine, Earth and Atmospheric Sciences at North CarolinaState University (http://omglnx6.meas.ncsu.edu/sabgom_nfcast/, accessed 13 July

806 NATURE GEOSCIENCE | VOL 5 | NOVEMBER 2012 | www.nature.com/naturegeoscience

© 2012 Macmillan Publishers Limited. All rights reserved.

NATURE GEOSCIENCE DOI: 10.1038/NGEO1615 LETTERS2012). Satellite sea surface temperature data were recorded by the NOAA (NationalOceanic and Atmospheric Administration) Advanced Very High ResolutionRadiometer sensor. We used sea surface temperature maps provided by the EarthScan Laboratory, Coastal Studies Institute of the Louisiana State University (http://www.esl.lsu.edu/imagery/AVHRR/, accessed 5 January 2012). For a full descriptionof the methods, see Supplementary Information.

Received 13 August 2012; accepted 24 September 2012;published online 21 October 2012

References1. Couvillion, B. R. et al. US Geological Survey Scientific Investigations Map 3164,

scale 1:265,000, 12 p. pamphlet (2011).2. Day, J. W. et al. Restoration of the Mississippi Delta: Lessons from Hurricanes

Katrina and Rita. Science 315, 1679–1684 (2007).3. Kim, W. et al. Is it feasible to build new land in the Mississippi River Delta?

Eos Trans. AGU 90, 373–374 (2009).4. Paola, C. et al. Natural processes in delta restoration: Application to the

Mississippi Delta. Annu. Rev. Mar. Sci. 3, 67–91 (2011).5. http://www.mvn.usace.army.mil/bcarre/floodfight.asp.6. Blum, M. D. & Roberts, H. H. Drowning of the Mississippi Delta due to

insufficient sediment supply and global sea-level rise. Nature Geosci. 2,488–491 (2009).

7. Kearney, M. S., Riter, J. C. A. & Turner, R. E. Freshwater river diversions formarsh restoration in Louisiana: Twenty-six years of changing vegetative coverand marsh area. Geophys. Res. Lett. 38, L16405 (2011).

8. Falcini, F. & Jerolmack, D. J. A potential vorticity theory for the formation ofelongate channels in river deltas and lakes. J. Geophys. Res. 115, F04038 (2010).

9. Batker, D. et al. Gaining Ground—Wetlands, Hurricanes and the Economy:The Value of Restoring the Mississippi River Delta (Earth Economics Tacoma,WA, 2010). Available at http://www.eartheconomics.org/FileLibrary/file/Reports/Louisiana/Earth_Economics_Report_on_the_Mississippi_River_Delta_compressed.pdf.

10. Törnqvist, T. E. et al. Mississippi Delta subsidence primarily caused bycompaction of Holocene strata. Nature Geosci. 1, 173–176 (2008).

11. Kirwan, M. L. & Murray, A. B. A coupled geomorphic and ecological model oftidal marsh evolution. Proc. Natl Acad. Sci. USA 104, 6118–6122 (2007).

12. Kirwan, M. L. et al. Limits on adaptability of coastal marshes to rising sea level.Geophys. Res. Lett. 37, 1–5 (2010).

13. Donoghue, J. Sea level history of the northern Gulf of Mexico coast and sealevel rise scenarios for the near future. Climatic Change 107, 17–33 (2011).

14. Coastal Protection and Restoration Authority of Louisiana Louisiana’sComprehensive Master Plan for a Sustainable Coast. Coastal Protection andRestoration Authority of Louisiana. Baton Rouge, LA (2012); available at http://www.coastalmasterplan.louisiana.gov/2012-master-plan/final-master-plan/.

15. http://la.water.usgs.gov/MississippiRiverFlood2011.html#.16. Hu, C. et al. Mississippi River water in the Florida Straits and in the Gulf

Stream off Georgia in summer 2004. Geophys. Res. Lett. 32, L14606 (2005).17. Shi, W. & Wang, M. Satellite observations of flood-driven Mississippi River

plume in the spring of 2008. Geophys. Res. Lett 36, L07607 (2009).18. Peckham, S. D. A new method for estimating suspended sediment

concentrations and deposition rates from satellite imagery based on thephysics of plumes. Comput. Geosci. 34, 1198–1222 (2008).

19. Walker, N. D., Rouse, L. J. Jr, Fargion, G. S. & Biggs, D. C. The Great Floodof summer 1993: Mississippi River discharge studied. Eos Trans. AGU 75,414–415 (1994).

20. Schiller, R. V., Kourafalou, V. H., Hogan, P. J. & Walker, N. D. The dynamicsof the Mississippi River plume: impact of topography, wind and offshoreforcing on the fate of plume waters. J. Geophys. Res. 116, C06029 (2011).

21. Round, F. E., Crawford, R. M. & Mann, D. G. The Diatoms, Biology AndMorphology of the Genera 747 (Cambridge Univ. Press, 1990).

22. White, J. R. et al. Mississippi River flood of 2008: Observations of alarge freshwater diversion on physical, chemical, and biologicalcharacteristics of a shallow estuarine lake. Environ. Sci. Technol. 43,5599–5604 (2009).

23. Bagnold, R. An approach to the sediment transport problem from generalphysics. US Geol. Surv. Prof. Paper 422-I (1966).

24. Bignami, F. et al. On the dynamics of surface cold filaments in theMediterranean Sea. J. Mar. Res. 74, 429–442 (2008).

25. Narimousa, S., Maxworthy, T. & Spedding, G. R. Experiments on the structureand dynamics of forced, quasi-two-dimensional turbulence. J. Fluid Mech. 223,113–133 (1991).

26. Stern, M. E. & Austin, J. Entrainment of shelf water by a bifurcating continentalboundary current. J. Phys. Oceanogr. 25, 3118–3131 (1995).

27. Wood, R. B. & McIntyre, M. E. A general theorem on angular-momentumchanges due to potential vorticity mixing and on potential-energy changes dueto buoyancy mixing. J. Atmos. Sci. 67, 1261–1274 (2010).

28. Dritschel, D. G. & Scott, R. K. Jet sharpening by turbulent mixing. Phil. Trans.R. Soc. A 369, 754–770 (2011).

29. Nittrouer, J. A. et al. Mitigating land loss in coastal Louisiana by controlleddiversion of Mississippi River sand. Nature Geosci. 5, 534–537 (2012).

30. Day, J. W. et al. Vegetation death and rapid loss of surface elevation intwo contrasting Mississippi delta salt marshes: The role of sedimentation,autocompaction and sea-level rise. Ecol. Eng. 37, 229–240 (2011).

AcknowledgementsThis work was supported by NSF-RAPID awards (EAR-1140269; OCE-1140307), aNOAA grant (NA11OAR4310101) and the University of Pennsylvania’s BenjaminFranklin Fellowship, and received further logistical support from the Luquillo CriticalZone Observatory (EAR-0722476). The CNRWORK was partially financially supportedby the European Commission MyOcean-2 Project grant agreement (283367). We thankC. Vervaeke and A. Constantin for assistance with the helicopter survey and T. Touchet,M. Enache, M. Mills and T. Dura for assistance with sediment and diatom analyses. Fieldsupport from the crew of R/V Acadiana, C. Zhang, P. Dash, A. and A. Kolker is verymuch appreciated. We also thank V. Forneris for the collection of L-1A data products.Any use of trade, product or firm names is for descriptive purposes alone and does notimply endorsement by the US Government. This paper is a contribution to IGCP project588 (Preparing for coastal change) and PALSEA.

Author contributionsF.F. performed the satellite analysis, developed the potential vorticity theory forsuspended sediment and coordinated the overall study. N.S.K. contributed to collection,analysis and presentation of the wetland sediment and diatom data. L.M. and C.B.L.contributed to the overall study and led the boat survey. B.P.H. contributed to the overallstudy approach and participated in data interpretation. M.D. and A.S. performed thehydrographic and suspended sediment data acquisition and analysis. K.L.M. contributedto the overall study approach and led the wetland sediment survey. R.S., S.C. and G.V.contributed to the processing of satellite data, subsequent analysis and presentation.C.L. contributed to the collection of river-mouth flow velocity profiles, subsequentdata analysis and presentation. D.J.J. supervised the research, participated in datainterpretation and led the writing of the main text. All authors contributed to thewriting of this manuscript.

Additional informationSupplementary information is available in the online version of the paper. Reprints andpermissions information is available online at www.nature.com/reprints. Correspondenceand requests for materials should be addressed to D.J.J.

Competing financial interestsThe authors declare no competing financial interests.

NATURE GEOSCIENCE | VOL 5 | NOVEMBER 2012 | www.nature.com/naturegeoscience 807

© 2012 Macmillan Publishers Limited. All rights reserved.