DSpace at VNU: Recent sedimentation and sediment accumulation rates of the Ba Lat prodelta (Red River, Vietnam)

A steep frontal prodelta slope is characterized by very rapid sedimentation tens of cm per year of muddy sediments under inXuence of the turbid river plume.. Analysis of the downcore 210

1367-9120/$ - see front matter © 2006 Elsevier Ltd All rights reserved.

doi:10.1016/j.jseaes.2006.03.006

Recent sedimentation and sediment accumulation rates

of the Ba Lat prodelta (Red River, Vietnam) G.D van den Bergha,¤, W Boera, M.A.S Schaapveldb, D.M Ducc, Tj.C.E van Weeringa

Received 8 September 2003; received in revised form 21 February 2005; accepted 2 March 2006

Abstract

The Ba Lat River is the major distributary of the Red River system in North Vietnam To assess the recent to subrecent depositional processes in the Ba Lat prodelta, a detailed sediment analysis was conducted Bottom samples were collected during two Weld surveys, one

in the dry season (winter) and one in the wet season (summer) A steep frontal prodelta slope is characterized by very rapid sedimentation (tens of cm per year) of muddy sediments under inXuence of the turbid river plume Beyond direct inXuence of the river plume the bottom slope decreases and bottom transport by the prevailing southward directed currents becomes important Coarse-grained tempestites alter-nate with the dominating muddy sediments Downcore changes in the 234Th activities indicate that the subaqueous delta progrades to the southwest, with erosion and reworking of older sediments occurring north of the present outlet The southwestward progradation is also encountered in the trend of 210Pb activities indicating that this process has continued for at least 100 years Avulsion of the Ba Lat outlet

in 1973 has led to a decrease in sedimentation rates north of the Ba Lat outlet

© 2006 Elsevier Ltd All rights reserved

1 Introduction

The Ba Lat is the main distributary of the Red River

sys-tem in North Vietnam (Fig 1) The coastal plain near the

Ba Lat outlet has accreted over a distance of 23 km during

the last 500 years, and is build up by an alternation of fossil

beach-spit systems with back barrier swamp deposits in

between (Thanh et al., 1997) The river discharge follows a

clear seasonal pattern reXecting the variation in rainfall

under the constraint of a monsoonal climate Besides

sea-sonal diVerences, the inter-annual variation in suspended

sediment transport varies between 30 and 120 million tons

per year passing Son Tay Environmental changes, both

anthropogenic as well as natural induced changes, have a

high potential to aVect the coastal zone of the Ba Lat Delta The main Ba Lat channel debauched at its present position before 1938 and after 1973 From 1938 until 1971 it entered the sea 10 km more to the north, when during a severe Xood

in August 1971 its location shifted to a position south of the present outlet During the typhoon Kate in 1973 the frontal sand barrier broke through and the main outlet started to occupy its present position (Thanh et al., 1997)

There are indications that the frequency of typhoons

aVecting the coast of Vietnam has increased during the sec-ond half of the 20th century (Thanh et al., 1997) Further-more, the construction between 1979 and 1994 of the Hoa Binh Dam in one of the three major tributaries of the Red River (Fig 1A), has nearly halved the average annual sus-pended sediment concentration at Son Tay gauging station (van Maren, 2004) These are some of the factors that are expected to inXuence the densely populated coastal zone of the Red River in the coming decades, and the development

* Corresponding author Tel.: +31 222 369 394; fax: +31 222 319 674.

of sustainable coastal zone management becomes

increas-ingly necessary Integrated coastal zone studies provide

use-ful information not only in documenting modern terrestrial

and marine environments, but also in understanding

ero-sion patterns within geological and historical contexts (

Mil-liman et al., 1987)

Our main objective within the framework of the Red

River Delta Research Program was to gain an

understand-ing of the sedimentological processes that govern the

devel-opment of the present Ba Lat prodelta The study is based

on the analysis of bottom samples recovered from the Ba

Lat prodelta The sedimentological and geochemical

imprints in the sedimentary record were analyzed in order

to reconstruct the processes that resulted in the present

conWguration of the prodelta on a 100-years time scale

Analysis of the downcore 210Pb activity has been

per-formed on a number of gravity cores, to assess spatial

vari-ability in recent accumulation rates on a »100 years time

scale Analysis of excess 234Th has been applied on box core

samples, in order to determine seasonal diVerences in

depo-sition, re-suspension, and mixing-processes of the surface

sediments Other parameters that were analyzed are the

composition, grain-size distributions and organic carbon

and nitrogen contents of the sediments A study on acoustic

facies analysis and prodelta geometry is presented

else-where in this volume (van den Bergh et al., 2006) For the

hydrodynamical and climatic conditions governing the

delta development the reader is referred to van Maren and

Hoekstra (2004, 2005) and van Maren et al (2004)

2 Delta setting

Based on the acoustic study carried out by van den Bergh

et al (2006), a morphogenetic subdivision of the study area was made (Fig 1B) This subdivision consists of: (1) delta front, (2) prodelta, (3) Gulf of Tonkin Shelf, and (4) erosional shoreface zones The delta front fringes the delta plain and is marked by a slope break between 5 and 7 m water depth at the transition with the prodelta The delta front has not been sampled in the course of this study The prodelta forms a rel-atively steep muddy slope that merges in to the shelf of the Gulf of Tonkin around the 30 m isobath The recent prodelta deposits are recognizable on the acoustic proWles as a dark band of high reXectivity with multiple sub-bottom reXectors, which become weaker and gradually converge with the bot-tom reXector in oVshore direction Based on bottom gradi-ents, bottom proWle, and the relative position with respect to the Ba Lat River mouth, the prodelta can be subdivided into: (A) the Northern Prodelta, (B) the Frontal Prodelta, and (C) the Southern Prodelta (Fig 1) The Frontal Prodelta, located adjacent to the Ba Lat mouth, is only 6 km wide and has the steepest bottom gradient of »6.5 m/km between 12 and 23 m water depth Towards the South the prodelta rapidly widens

to more than 20 km across and bottom gradients decrease to

»1.5 m/km The Southern Prodelta is detached from the coast Towards the North the prodelta widens as well, but near coastal bottom gradients remain relatively steep (»4.5 m/km) The Northern Prodelta is also detached from the coast by a 2–3 km wide erosional shoreface

Fig 1 (A) Map of Vietnam with location of the study area and (B) map of the study area showing a sub-division of the Ba Lat prodelta (various shaded zones), based on the acoustic study of van den Bergh et al (2006) Also shown are the positions of bottom sampling stations Hatched area represents the zone of maximum accumulation, where the most recent acoustic unit has a thickness in excess of 2 m.

3 Methods

In 2000 Weld campaigns where conducted in February–

March and July–August, during the dry and wet monsoon,

respectively Bottom sampling stations were selected based

on a preliminary study of the shallow penetrating acoustic

proWles (van den Bergh et al., 2006) A gravity corer with a

length of 2 m and a diameter of 9 cm was used Gravity

cor-ing stations are shown in Fig 1, and a list of all cores with

their coordinates is presented in Table 1 At the Weld station

cores were split, photographed and described

macroscopi-cally Magnetic susceptibility was measured at 1 cm

inter-vals with a handheld Bartington MS2E1 surface sensor

Sub-samples with known volume were taken at 5-cm

inter-vals for standard Dry Bulk Density (DBD) measurements

The top part of each core was more intensely sampled for

210Pb analysis Then the cores were covered with plastic foil

and sealed in plastic and stored horizontally at a

tempera-ture of 5–7 °C The intact core halves were shipped to the

NIOZ in the Netherlands for X-ray photograpy and further

analysis (granulometry, XRF, Corg and C/N ratios)

The XRF Cortex-corescanner, is a non-destructive,

semi-quantitative logging instrument for major element

determination (Jansen et al., 1998) The elements Fe and Ca

were analyzed at 1-cm intervals Stereo X-ray photographs

were made of selected intervals Based on a study of the

photographs, additional sample levels for grain-size and,

for cores 2 and 15, Corg and C/N ratio analyses, were

selected besides standard intervals of 5 cm

Grain-size analyses were performed on 15 cores using

a Coulter LS 230 analyser Approximately 0.1 g of sedi-ment was weighted in glass beakers and mixed with 15 ml

of tap water The samples were put in an ultrasonic bath for 5 minutes Then the sample was passed through a

2 mm sieve and measured under continued ultrasonic treatment

Organic carbon and nitrogen contents were determined

on samples from cores 2 and 15, using a Carlo Erba

NA-1500 series 2 Nitrogen Carbon Sulphur Analyser Sample treatment and analysis was according to the method of Ver-ardo et al (1990) The data are presented as weight percent-ages of organic carbon and nitrogen versus depth and C/N versus depth

Measurements for 210Pb analysis were made following the methods outlined in Boer et al (2006) Coarse-grained intervals, recognized by the dry bulk densities or macro-scopic descriptions, were omitted for 210Pb analysis The best model Wts through the data points were calculated using the Constant Flux and Constant Sedimentation (CF–CS) model (Appleby and OldWeld, 1992; Boer et al.,

2006) For gravity cores 6, 10 and 15 the supported 210Pb activity, as deWned by 226Ra, was determined by analyzing the 226Ra activity using gamma spectrometry, according

to the method outlined in van den Bergh et al., 2003) In several cores with very high accumulation rates, sup-ported 210Pb activities were not reached at the base of the core If no 226Ra-based supported activities were avail-able, model Wts were obtained with pre-deWned supported Table 1

List of Red River Delta gravity cores sampled in 2000

a Core not located along acoustic transect: water depth is estimated.

activities of 36.4 and 41.3 Bq kg¡1, based on the minimum

and maximum 226Ra-based supported activities measured

in other cores The development of a surface mixed layer

(SML) was not always evident in core proWles In case of

doubt solution Wts were forced both with and without

SML, and the calculated variables resulting from the

vari-ous solutions are presented as a range in tables and

Wgures

In four cores (2, 10, 12 and 18) vertical grain-size

Xuctua-tions appeared to result in highly irregular 210Pb activity

plots, due to rather variable speciWc surface areas in these

samples In these cores 210Pb activities were measured on

the <25m fractions

Boxcores for 234Th analysis were retrieved (Fig 1B)

using a cylindrical boxcorer with an inner diameter of 12

and 38 cm length Some stations were sampled during

both seasons in order to assess seasonal variability Only

cores were used that showed no signs of disturbance

upon retrieval On ship deck an 11 cm-diameter plastic

pipe was pushed into the sediment and sealed at both

ends after the water on the surface of the sediment had

been removed Boxcores were transported vertically to

the Weld station, where they were sliced The freeze–dried

samples were analysed for 234Th using -spectrometry

following the procedure as mentioned in Schmidt et al

(2002)

4 Results

The most recent depositional unit that is recognizable on

the acoustics has been mapped separately as Unit Z1 (van

den Bergh et al., 2006: Fig 6) The area of major recent

accumulation, where Unit Z1 has its thickest development

of >2 m, is indicated hatched in Fig 1B This area of major

accumulation covers the steep frontal delta slope and a

large proximal area of the Southern Prodelta Gravity cores

3, 15, 16 and 23 are located in this area of major

accumula-tion Stations 1, 2, and 12 cover the more distal areas of the

Frontal Prodelta Stations 6–8, 13–14, and 24–27 are

located in the distal parts of the Southern Prodelta At

sta-tions 8 and 8B stiV, Wne-grained sandy sediment was

retrieved in the core catcher

Cores 10–11, 21 and 29–31 penetrate the Northern

Prodelta Several short cores (stations 22, 18, 32 and 33)

sampled the surface sediments of the Gulf of Tonkin

Shelf At station 22 coring failed but some coarse sandy

material was recovered in the core catcher Station 18 is

located in an oVshore depression, which is bounded to the

west by a SSW-NNE trending fault with surface

expres-sion at the bottom and to the east by a convex ridge

(Fig 1B)

The erosional coastal zones were not sampled, but are

clearly revealed on the acoustic proWles (van den Bergh

et al., 2006: Fig 3A) At station 9, located between the Hai

Hau coast and the Southern Prodelta, multiple coring

attempts resulted in the retrieval of only 15 cm of stiV,

well-consolidated mud The acoustic proWle that runs along this

station shows inclined sub-bottom reXectors that are trun-cated by the bottom reXector, indicating erosion of older prodelta deposits

At 12 stations boxcores were retrieved, either during the dry season, the wet season, or during both seasons (Fig 1B) The boxcore retrieved at station 3 during the dry season at 15.5 m water depth showed fresh elongated paral-lel scours of several mm deep at the surface, indicating ero-sion by strong bottom currents

4.1 Sediment characterization

Data compilations of lithology and various analyzed parameters are shown in Figs 2A–C and 7 Muddy sedi-ments from the Ba Lat prodelta can be easily distinguished from the contrasting sandy deposits covering the Gulf of Tonkin Shelf The modern prodelta deposits have a reddish brown color, whereas the sandy shelf de posits have a greenish gray color, are coarser-grained, and contain abun-dant shell debris

The prodelta sediments contain quartz, feldspar and mica as major grain types in the coarse silt and Wne sandy fractions Silt-sized detrital grains are mostly covered with a reddish brown coating of Fe-oxides XRD analysis

on bulk samples indicated the presence of 14 Å clay minerals, mostly chlorite Kaolonite is probably also present in the clay fraction, besides hematite Calcite shows very weak intensities on the XRD diagrams The contrasting color diVerence between the prodelta muds and the sandy shelf deposits is caused by relatively low amounts of Fe-oxides in the latter The XRF measure-ments demonstrate the relatively low amount of Fe miner-als in these sandy shelf deposits (e.g Fig 2B: base of core 18), and the low Fe contents also correlate well with very low magnetic susceptibility values No hematite could be demonstrated to be present in the sandy shelf samples Instead, pyrite is more prominently present in these sedi-ments At stations along the distal margin of the prodelta, where cores penetrated the muddy prodelta deposits into the underlying shelf sands, the transition between both sed-iment types was always found gradual and marked by a 10–

20 cm thick interval of heavily bioturbated and mottled sediments

The prodelta deposits consist predominantly of muddy sediments with a few thin (<5 cm) coarser-grained interca-lations Median grain-sizes are mostly between 5 and 10m, whereas the thin-bedded coarser-grained intercalations have 90% of their volume in the range smaller than 100m (very Wne sandy silt) Notable are the upwards-coarsening trends at the top of cores from the Northern Prodelta (cores 10, 21, 31, 11, 21 and 12) Cores located on the Fron-tal prodelta (core 2) and in shallower water (cores 10 and 21) show the most pronounced jumps in grain-size Cores located more distally on the Southern Prodelta (cores 6, 7 and 25) have the most homogeneous grain-size distribu-tions of very Wne silt Carbonate building organisms were overall quite rare

The thin silt and sandy silt layers frequently exhibit

par-allel lamination Occasionally cross-lamination (current

ripples), erosional bases, and Wning-upwards trends are

developed In the more oVshore cores coarser layers usually

constitute less than 4% of the total core length The

coarser-grained layers tend to become more frequent in cores

located in shallower water, varying between 5 and 21% of

the total core length The sandy layers may consist of

rela-tively well-sorted sand or silt, and occasionally contains

abundant plant remains (e.g core 2, 143–145 cm depth)

4.2 210 Pb analysis

210Pb dating is an important tool widely used to assess spatial variability in recent accumulation rates on a »100 years time scale (Krishnawami et al., 1980; Benninger et al., 1997; Fuller et al., 1999; Chague-GoV et al., 2000; van den Bergh et al., 2003; Boer et al., 2006) 210Pb (with a half-life

of 22.3 years) is produced by the decay of atmospheric

222Rn and is removed from the atmosphere as fallout In the marine environment it rapidly adheres to the surface of

Fig 2 Data compilations of gravity cores from the Ba Lat prodelta, showing lithology, downcore values of DBD, magnetic susceptibility, grainsize, and for some cores Ca and Fe contents and Corg and C/N ratios (A) Cores from the Northern Prodelta (B) Cores from the Frontal Prodelta (C) Cores from the Southern Prodelta (next page) For the grain size columns dots represent median values and squares represent the grain size of the 90% percentile Gray bands in the right columns of core 15 correspond with tempestites.

A

B

sediment particles, to become incorporated into

accumulat-ing sediments

Grain-size Xuctuations have an eVect on the 210Pb

activi-ties Smaller grains result in higher speciWc surface area and

thus in a higher potential to capture 210Pb (Eisma et al.,

1989) Variable grain-sizes aVected the outcome of the 210Pb

activity proWles of cores 2, 10, 12 and 18 negatively

Examina-tion of grain-size distribuExamina-tions showed that coarser-grained

samples usually showed major peaks in the coarse silt or Wne

sand fractions (Fig 3A) Using the <25m fraction for

mea-suring 210Pb activities, negative eVects from grain-size

Xuctu-ations were largely reduced (Fig 3B) Even when only the

<25m fractions was used, some marked shifts in activity

remain (e.g cores 2, 10, Fig 4) These Xuctuations

presum-ably reXect highly variable activities of the accumulating

sed-iment, in particular for cores located relatively close to the

river mouth Erosional events may also account for some of

the discontinuities in the proWles This is probably the case in

core 2, where a sharp decrease in activity occurs between 15

and 20 cm core depth at the top of a silty interval

The proWles of total downcore 210Pb activities for the 12

analyzed cores are shown in Fig 4 The model results of

supported activities (Csupp), inventories, extrapolated

activi-ties at the surface (C0), accumulation rates (), mixing rates

(Db) and depths of the Surface Mixed Layer (Zmix) are

sum-marized in Table 2 The accumulation rates are presented in

cm yr¡1

Surface Mixed Layers (SML) only occur at stations

located along the outer margin of the prodelta (stations 12,

25 and 26), reaching a thickness of up to 30 cm These two

stations have the lowest recorded sedimentation rates of

<1 cm yr¡1 Station 25 is the furthest away from the Ba Lat

River mouth An accumulation rate of between 0.7 and

1.6 cm yr¡1 was obtained, depending on an interpretation

with or without SML

Highest accumulation rates of more than 3 cm yr¡1 were found at the Frontal Prodelta (stations 1–3) and the proxi-mal part of the Southern Prodelta (station 15) Station 3, located on the steep slope of the Frontal Prodelta, has an extremely high sedimentation rate Values of between 33 and 94 cm yr¡1 are obtained when extrapolating excess activities downward beyond the length of the core and assuming supported activities of between 36.4 and 41.3 Bq kg¡1 as measured in other cores The extremely high sedimentation rate at station 3 reXects direct inXuence of settling from the river plume

Cores from the Northern Prodelta (10 and 11) have intermediate sedimentation rates of between 1.5 and 2.1 cm yr¡1 The low accumulation rate (1.0 cm yr¡1) is com-parable to stations along the distal edge of the prodelta

4.3 234 Th analysis

Downcore analysis of excess 234Th is commonly used for the assessment of short-term and seasonal diVerences in deposition, re-suspension, and mixing-processes in various marine environments (Aller et al., 1980; Aller and DeMas-ter, 1984; Fuller et al., 1999) 234Th activities were analyzed

on 14 boxcore samples from the Ba Lat prodelta (Fig 5) Surface 234Th excess activities Xuctuated between negligible values and 200 Bq kg¡1 Maximum penetration depths ranged between 1 and 2.5 cm During the dry season in March, highest inventories were recorded at stations 14–16, located in the zone of maximum deposition of the Southern Prodelta (Fig 6) Unfortunately, no wet season measure-ments are available for these stations

The highest inventory during the wet season was recorded at station 18, followed by stations 7 and 26 Sta-tions 13 and 27, located on the oVshore margin of the Southern Prodelta, both have intermediate wet season

Fig 2 (continued )

C

1 10 100 1000

Core 15

Fe% of total counts

DBD & magsus

0 0.5 1 1.5

Grainsize

(µm)

XRF

Ca% (of total counts - Fe)

Lithology

80 85 90

0

20

40

60

80

100

120

140

CGS units

DBD (g.cm ) -3

C/N ratio

%C org

%N org

0.0 0.1 0.2

Core 6

DBD & magsus

CGS units

Grain size Lithology

0.0 0.5 1.0 1.5

0

20

40

60

80

100

120

140

160

(µm)

DBD (g.cm )-3

inventories Notably, station 26, located along the oVshore

edge of the Southern Prodelta, has a much higher wet

sea-son inventory than stations 13 and 27, which are closer to

the Ba Lat In the Northern Prodelta inventories for the dry

and wet season are close to zero at station 10, reXecting

consistent non-sedimentation or erosion during both

sea-sons At station 11 inventories during both seasons are of

intermediate magnitude, but the dry season inventory is

slightly higher than that of the wet season

4.4 Organic carbon and nitrogen

Color banding is well-developed in cores from the

Frontal Prodelta (cores 1–3) and to a lesser degree in the

more distal cores from the Southern Prodelta (cores 6–7,

15–16 and 23) In core 2 the dark bands are 1–3 cm thick,

show Munsell colors 7.5YR3/2 or ¡3/3 (brownish black

or dark brown) The dark bands tend to have a sharp base

with the dark colors gradually fading upwards (Fig 7)

The lighter colored bands show Munsell color 5YR4/3

(dull reddish brown) and may be thicker or thinner than

the dark colored bands In the cores from the Southern

Prodelta dark/light bands are weaker developed, show

more mottling and are on average thicker (10–15 cm) In

order to investigate weather these color bands were gener-ated by possible seasonal variability in the supply of organic material, cores 2 and 15 were selected for analysis

of organic carbon (Corg) and nitrogen (Ntotal) Both cores have high 210Pb accumulation rates and therefore poten-tially a good resolution

Core 2 is composed of clayey silts with a few pro-nounced coarser-grained, silty levels, containing benthic foraminifera and ostracods besides siliciclastic grains The

Corg in core 2 Xuctuates around an average of 0.7%, with a single peak of over 3% coinciding with the coarse-grained layer at 143 cm core depth, which contains macroscopically visible plant remains (Fig 7) The C/N ratio shows a single peak of 19.2 at this sandy layer, in accordance with land-derived plant material (Hedges et al., 1999) For the remain-ing part of the core the Corg and Ntotal contents are posi-tively linked to each other, the C/N ratio Xuctuating between 4.2 and 9.1, indicating a common marine source These values are comparable to bottom sediments from the Mekong prodelta (Landmann et al., 1998) They do not correlate with volume percentages of the size fractions smaller than 8m, 25–63 m or larger than 63 m Neither

is there a strong correlation between darker color and higher C

Fig 3 Graphs showing the grain-size eVect on 210 Pb activity in core 10 (A) Grain-size distribution plots at 5 cm intervals (B) Mass percentages of the

>25 m fraction of the bulk samples (black dots), 210 Pb activity of the bulk samples (closed squares) and 210Pb activity of the <25m fraction (open squares).

Particle size (µm)

25 µm

0

10

20

30

40

50

60

70

80

Activity (mBq/g)

mass/mass % (fraction >25 µm)

210Pb(tot) 210Pb(tot) - fraction < 25 µm m/m% > 25 µm

The granulometric curve of core 15 (Fig 2C) shows

sim-ilar Xuctuations as in gravity core 2 Organic carbon

per-centages are Xuctuating around an average of 0.9%, slightly

higher than in core 2 The organic carbon and nitrogen curves show two large peaks at 10.5 and at 116.5 cm core depth, both levels with macroscopically visible plant

Fig 4 210 Pb total activity proWles of core sediments from the Ba Lat prodelta Open squares represent data points that were omitted during Wtting, because they were coarser-grained than the remaining part of the core For many cores multiple Wtting solutions are plotted, depending on the model choice (with

or without SML) or on the choice of the supported 210 Pb activity (based on minimum and maximum values obtained in other cores).

Southern prodelta

0

20

40

60

80

100

120

140

160

180

200

210Pb (<25mu fraction)

fit incl lower point

fit excl lower points

226Ra

210Pb (bulk fraction) fit with SML fit without SML 226Ra 226Ra (average)

210Pb (<25 mu fraction) best fit Csupp set at value core 6 Csupp core 6 Csupp best fit

210Pb (bulk fraction) Csupp set at value core 10 Csupp Set at value core 6 Csupp core 10 Csupp core 6

0

20

40

60

80

100

120

140

160

180

200

210Pb (bulk fraction)

fit with SML

fit without SML

226Ra

226Ra (average)

210Pb (bulk fraction) Csupp Set at value core 6 Csupp Set at value core 10 Csupp core 10 Csupp core 6

210Pb (bulk fraction) best fit Csupp set at value cores 12 & 15 Csupp cores 12 &

15 Csupp best fit

210Pb (bulk fraction) Csupp set at value core 6 best fit Csupp core 6 Csupp best fit

210Pb (<25 mu fraction) best fit

Csupp best fit

210Pb (bulk fraction) Csupp set at value core 10 Csupp set at value core 12 Csupp core 10 Csupp cores 12 &

15

210Pb (bulk fraction) Csupp set at value core 10 Csupp set at value core 6 Csupp core 10 Csupp core 6

210Pb (bulk fraction) Csupp set at Csupp set at value core 6 Csupp core 6 Csupp core 10

Table 2

Parameters derived from model Wtting of 210 Pb activity proWles

a For most proWles supported activity was not reached at depth Csupp values obtained by measuring 226 Ra activities are printed bold For cores where

Csupp was not directly measured and was not reached at the core base, values from other cores were substituted.

b Depending on the choice of the supported activity, various best Wt solutions were found To the left values leading to a minimum accumulation rate ( min), to the right values leading to a maximum accumulation rate ( max) are given.

c In these stations the model choice (with or without SML) lead to slightly varying solutions Only the sieved fraction <25m was used for 210 Pb analysis.

Station C0 (Bq kg ¡1 ) Csuppa (Bq kg ¡1 ) Zmix (cm)  min.b (cm yr ¡ 1) C0 (Bq kg ¡1 ) Csupp (Bq kg ¡1 ) Zmix (cm)  max.b (cm yr ¡1 )  average (cm yr¡1 )

remains The C/N ratios are considerable higher at these

level (17.3 and 16.1, respectively) than in the remaining part

of the core

5 Discussion

5.1 Spatial accumulation patterns

The diVerence in color and mineral content between the

gray shelf sands and the reddish brown delta deposits

reXects a diVerence in depositional conditions Iron

miner-als are concentrated in the Wne-grained sediments that settle

out from the modern river plume, whereas the sandy shelf

sediments were most likely deposited under high energy

conditions The gray shelf sands are generally well

consoli-dated and the echosounder proWles did not show active

bedforms, so they are assumed to be relicts from the

post-glacial transgression

Maximum grain-size of suspended sediments from the

Ba Lat plume is up to 60m, with a median grain-size vary-ing between 4 and 8m (van Maren and Hoekstra, 2004) This corresponds well with the median grain-sizes in the muddy prodelta deposits, indicating that the major sedi-ment source is provided by suspended sedisedi-ment from the river The turbid plume extends some 5 km oVshore from the river mouth during the wet season The steepest part of the Frontal Prodelta down to a depth of 22 m lies directly under the inXuence of this wet-season turbid plume Fur-ther oVshore surface waters are still characterized by rela-tively low salinities (<12 ppt at least 10 km in oVshore direction), but these older generations of river plumes appear to have lost most of their suspended sediment load due to Xocculation processes (van Maren and Hoekstra,

2004) Direct settling from the river plume leads to an extremely high 210Pb accumulation rate on the Frontal Prodelta slope at station 3 (30–90 cm yr¡1) The rapid Fig 5 234 Th activity proWles of boxcores from the Ba Lat prodelta.

Station 10, waterdepth 22 m Station 11, waterdepth 27.5 m Station 18, water depth 30.5 m

Station 14, water depth 21 m: Station 16, waterdepth 16.5 m Station 15, waterdepth 23 m:

Dry season: diamonds Wet season: triangles

238

U-based supported activities indicated by dashed line

Station 7, waterdepth 22 m: Station 13, water depth 26.5 m Station 27, waterdepth 27.5 m: Station 26, waterdepth 29 m

0 0.5 1 1.5 2 2.5 3

0 50 100 150 200 250 300

Bq/kg

0

0.5

1

1.5

2

2.5

3

0 50 100 150 200 250 300

Bq/kg

0

0.5

1

1.5

2

2.5

3

0 50 100 150 200 250 300

Bq/kg

0 0.5 1 1.5 2 2.5 3

0 50 100 150 200 250 300

Bq/kg

0.0 0.5 1.0 1.5 2.0 2.5 3.0

0 50 100 150 200 250 300

Bq/kg

0

0.5

1

1.5

2

2.5

3

0 50 100 150 200 250 300

Bq/kg

0 0.5 1 1.5 2 2.5 3

0 50 100 150 200 250 300

Bq/kg

0 0.5 1 1.5 2 2.5 3

0 50 100 150 200 250 300

Bq/kg

0 0.5 1 1.5 2 2.5 3

0 50 100 150 200 250 300

Bq/kg

0 0.5 1 1.5 2 2.5 3

0 50 100 150 200 250 300

Bq/kg

accumulation explains the lowest DBD values at this

sta-tion (average DBD D 0.66 g cm¡3) The high accumulation

by means of Xocculation and settling maintains the steep

slope of the Frontal Prodelta slope At stations 1 and 2,

located oVshore from the slope break at 22 m water depth

(in between stations 2 and 3), the accumulation rate has

decreased drastically down to values of between 3 and

4 cm yr¡1 The distal slope break of the prodelta marks the

point beyond which direct settlement from the turbid

plume rapidly decreases

The large thickness of the youngest acoustic unit (Unit

Z1) in the area southwest of the river mouth (see Fig 6 in

van den Bergh et al., 2006) suggests that after settling,

bot-tom transport redirects a large amount of sediment in

southwestern direction This is again conWrmed by the

rela-tively high accumulation rate at station 15 (3.2–4.5 cm yr¡1)

Towards the eastern margin of the Southern Prodelta and

further south the 210Pb accumulation rates decrease with

decreasing thickness of the youngest acoustic unit In the

Northern Prodelta on the other hand (stations 10 and 11)

accumulation rates are low, varying between 1.5 and

2 cm yr¡1 These observations are in line with the prevailing

current patterns Southward directed residual Xow

domi-nates near the Red River Delta coastline during the NE

Monsoon (dry season) During the wet season residual Xow

is probably weaker and more variable, though primarily

directed southward The Coriolis force seems to be a

domi-nant factor in the southward Xow of surface currents,

strengthened by the prevailing wind patterns during the dry

season (van Maren and Hoekstra, 2005) However, of

greater importance for sediment bed-load transport is the

tidal asymmetry, which leads to a net southwards transport

throughout the year (van Maren et al., 2004) It is likely

that the stations in the Northern Prodelta receive most of

their sediment by means of longshore transport from the

more northern river mouths The 234Th measurements

indi-cate that non-deposition or erosion during both seasons

occurs at station 10 At station 11 the 234Th inventory of the

dry season appeared larger than during the wet season, sug-gesting that southward bottom transport and focusing of re-suspended sediments occurs mostly during the dry sea-son Cores 10, 11, 21 and 29, all from the Northern prod-elta, are capped by a 5 to 10 cm thick interval that is markedly coarser-grained than the sediment underneath (Fig 2A) These coarse-grained bottom sediments seem to represent lag-deposits that result from winnowing of the silt-clay fraction by bottom currents

The observed long-term accumulation at station 10, as indicated by the 210Pb activity proWle, could be explained by attributing the accumulation at this station as dating back

to the period before 1971, when the main river outlet entered the sea opposite station 10

5.2 Tempestites

The occurrence of coarser-grained layers in the prodelta deposits indicates that occasionally high energetic condi-tions occur near the sea-bed, that are able to transport sand-sized particles as bed-load The sandy silt layers and silt layers developed in most cores except at the most distal deeper water stations, are likely formed during stormy con-ditions In the Ba Lat area, landward-directed storms are generated due to the intense summer lows and winter highs that characterize the local climate Table 3 presents a list of typhoons that made landfall within 100 km from the Ba Lat between 1964 and 2000 On average the delta is aVected by

a heavy storm once every 4 years Heavy storms directly hitting the Ba Lat area occurred in 1973 (Kate) and 1983 (Georgia)

Storms frequently produce sheet-like sands of consider-able lateral extent, which may extend to depths several tens

of meters below the fair-weather wave base These tempes-tites are individual graded sandy beds produced in waters where sandy transport is not an important factor during fair weather conditions Tempestites show a sharp erosional base, followed by a coarse-grained (lag) deposit overlain by Wning upward sediments They are typically 5–10 cm thick and may exhibit cross-bedding structures at the base or intermediate levels (Einsele, 1992) Parallel lamination replaces the cross-stratiWcation at greater water depths When a storm is accompanied by heavy Xooding in the near-by land areas, large volumes of sediment and land-derived organic material may be supplied in addition to sediment eroded from the shoreface

The laminated sandy layers in cores 2 and 15 meet these criteria in most cases An exception is the layer at 113–

120 cm in core 15, which shows a coarsening upward trend (Fig 2C) However, the larger particles in this layer repre-sent mostly plant remains Typhoons in North Vietnam occur mostly towards the end of the wet season and thus do not necessarily co-occur with the major supply of sediment rich in organic material On the other hand, storm surges Xooding the coastal areas may transport large amounts of terrestrial organic material seawards Events where large amounts of terrestrial organic material was transported Fig 6 Comparison of 234Th inventories for the various stations during

March (dry season) and August (wet season).