DSpace at VNU: An analysis of coastal erosion in the tropical rapid accretion delta of the Red River, Vietnam

DSpace at VNU: An analysis of coastal erosion in the tropical rapid accretion delta of the Red River, Vietnam tài liệu,...

An analysis of coastal erosion in the tropical rapid accretion delta of the Red

River, Vietnam

Do Minh Duca,⇑, Mai Trong Nhuanb, Chu Van Ngoia

a

Faculty of Geology, College of Science, Vietnam National University, Hanoi, 334 Nguyen Trai, Thanh Xuan, Hanoi, Viet Nam

b

Vietnam National University, Hanoi, 144 Xuan Thuy, Cau Giay, Hanoi, Viet Nam

a r t i c l e i n f o

Article history:

Received 9 November 2010

Received in revised form 8 August 2011

Accepted 10 August 2011

Available online 13 September 2011

Keywords:

Red River delta

Shoreline

Accretion

Erosion

Sediment transport

a b s t r a c t

The largest plain in the North Vietnam has formed by the redundant sediment of the Red River system Sediment supply is not equally distributed, causing erosion in some places The paper analyzes the evolvement and physical mechanism of the erosion The overlay of five recent topographical maps (1930, 1965, 1985, 1995, and 2001) shows that sediment redundantly deposits at some big river mouths (Ba Lat, Lach, and Day), leading to rapid accretion (up to 100 m/y) Typical mechanism of delta propaga-tion is forming and connecting sand bars in front of the mouths Erosion coasts are distributed either between the river mouths (Hai Hau) or nearby them (Giao Long, Giao Phong, and Nghia Phuc) The evolvement of erosion is caused by wave-induced longshore southwestward sediment transport Mean-while sediment from the river mouths is not directed to deposit nearshore The development of sand bars can intensively reduce the erosion rate nearby river mouths Erosion in Hai Hau is accelerated by sea level rise and upstream dams Sea dike stability is seriously threatened by erosion-induced lowering of beach profiles, sea level rise, typhoon, and storm surge

Ó 2011 Elsevier Ltd All rights reserved

1 Introduction

The Red River begins from the mountains of Yunnan province

(China) The Red River delta in Vietnam territory is formed by

the Red and Thai Binh river systems, which is commonly called

the Red River system (Fig 1) The delta is about 15,000 km2 It is

a rich agricultural area and densely populated Along the coastline,

the interaction between the sea and big rivers has created a typical

tropical natural condition which is suitable for tourism,

agricul-tural and aquaculture development

The Red River delta develops in a very dynamic fluvial and

mar-ine environment The river basin is characterized by an alternation

of wet and dry seasons producing a huge total annual suspended

sediment load (Hoekstra and Van Weering, 2007) The delta is river

dominated (Fig 2) The annual amount of sediment transported by

the Red River system into the East Vietnam Sea is about

82  106m3 In the wet season (from April to September), about

90% of the annual sediment supply is transported through the

var-ious distributaries (Nhuan et al., 1996) Of the total amount of

sed-iment supplied, 11.7% passes through the Van Uc and Thai Binh

river mouths, 11.8% through the Tra Ly river mouth, 37.8% through

the Red River (Ba Lat) mouth and 23.7% through the Day river

mouth (Duc et al., 2007)

The northern part of the coast (from Ba Lat to Hai Phong) has a diurnal tidal regime with average amplitude of 2.5–3.5 m In the southern part, from Ba Lat to Day mouth, the tide is mixed with

a diurnal dominance The average tidal amplitude is 2–3 m (Nhuan

et al., 1996) Waves usually have a dominant direction from the east, northeast during the dry season (October–March) and from east, southeast during the wet season (April–September) The aver-age and maximum wave heights are 0.7–1.3 m Wave heights in se-vere typhoons can reach over 5 m (Nhuan et al., 1996)

The large amount of sediments has made the delta a rapid con-tinuous advancing to the sea Old shorelines of the delta are recog-nized through series of old sand bars, historical and anthropogenic proofs (Hoan and Phai, 1995) The delta was enlarged 20–30 km from the 10th to 15th century and 10 km from 15th to 19th cen-tury (Fig 1) Sediments supplied by the big mouths (Tra Ly, Ba Lat, Lach, and Day) are mainly deposited at shallow sea and form sand bars in front of the mouths They protect shorelines behind against wave and current attacks making a suitable condition for rapid accretion

Most of sediments discharged from rivers deposit in front of the river mouths and causes rapid accretion As a consequence, severe shoreline retreat occurs at some other places due to sediment def-icit Erosion coast in the Red River delta has length and area much less than those of accretion coast However, it has seriously damaged the coastal villages and made an obstacle for economic development in the region The distribution of erosion shoreline

1367-9120/$ - see front matter Ó 2011 Elsevier Ltd All rights reserved.

⇑Corresponding author Tel.: +84 4912042804; fax: +84 4 38583061.

E-mail address: ducdm@vnu.edu.vn (D.M Duc).

Contents lists available atSciVerse ScienceDirect Journal of Asian Earth Sciences

j o u r n a l h o m e p a g e : w w w e l s e v i e r c o m / l o c a t e / j s e a e s

can be a typical characteristic of the Red River delta They are

either in the middle of the big river mouths (Hai Hau) or very close

to them (Giao Long, Giao Phong, and Nghia Phuc)

The paper has objective to outline shoreline change in the

re-cent time and find out the physical mechanism of erosion in the

tropical rapid accretion delta of the Red River Recent evolvement

of shoreline and the reasons are elucidated by the analysis of

topographical maps and nearshore sediment transport Factors

affecting shoreline retreat such as typhoon, sea level rise, and

up-stream dams are studied to assess potential acceleration of erosion

and its impacts to coastal structures

2 Materials and methods

2.1 Topographical maps

A series of topographical maps are used to investigate recent

changes of shoreline (Table 1) The maps were established using

different co-ordinate systems and scales WGS-84 stands for World

Geodetic System which is currently the reference system being

used by the Global positioning system WGS-60 is one of the

Fig 1 Red River system and locations of old shorelines.

Fig 2 The Red River in the chart of delta classification of Coleman and Wright

former systems of WGS-84 HN-72 or Hanoi-72 is the system of

Vietnam which was used before 2000 It used Krasovski ellipsoid

and the datum (origin co-ordination) was transferred from

Mos-cow to Hanoi It is not a spatial unified system and requires

differ-ent formulas to convert to other systems VN-2000 is the currdiffer-ent

national geodetic system in Vietnam The origin co-ordination

was installed in Hanoi VN-2000 is a unified system for the whole

country

Maps are scanned and then the shoreline in each map is

digi-tized in MapInfo software The vector maps are then transformed

to the same scale and datum which is used here as WGS-84 The

converting procedures follow the instruction of the circular on

‘‘Guidelines for the Application of VN-2000 System’’, established on

20 June 2001 by theGeneral Department for Land Survey(Ministry

of Natural Resources and Environment) The 1930 map is

sepa-rately analyzed Some national roads are assumed not to be

chan-ged and can be reference to overlay this map with 1965 map

Because of uncertainty for 1930 map the change of shoreline from

1930 to 1965 has not high reality

Shoreline in topographical maps is considered as the line

be-tween seawater and land when water level is at longterm mean

tide The tide range at Hondau station (Quang Ninh province) is

used for the North Vietnam where the longterm spring tide is

4.0 m Obviously the longterm neap tide is 0 m Therefore shoreline

of the Red River delta in topographical maps is defined at the mean sea level of 2.0 m above the neap tide

2.2 Sediment sampling and testing

A rectangle net of survey along the coast of Hai Hau at the depths of 0–30 m was set up The distances between investigation points are 2.5 km and 5 km in the depths of shallower and deeper than 10 m water deep, respectively (Fig 3) During the fieldwork small ships were used The position of sampling stations was deter-mined using a GPS with an accuracy of 5–100 m A total of 52 sed-iment samples were taken by grab sampler This is a part of investigation in the Red River coast in 1996 and 2000 (referred

toDuc et al (2007))

Grain size distributions of sediment samples were analyzed by means of sieve for the sandy fractions (sieve sizes: 2, 1, 0.5, 0.25, 0.125 and 0.063 mm), and by means of pipette analysis for samples containing particles smaller than 63lm

A thin-walled tube (ASTM, 2001) was manually inserted to sur-face sediment in tidal flat to take undisturbed geotechnical sam-ples Six samples were retrieved along the coast in 2008 (Fig 3) Water content (W), bulk density (c), and grain density (D) of sed-iments are defined in the laboratory Porosity (n) of sediment is then defined as:

Table 1

List of recent topographical maps.

Datum Ellipsoid

1965 WGS-60 WGS-60 1:50,000 Published 1978

1985 HN-72 Krasovski 1:50,000 Published 1991

1995 VN-2000 WGS-84 1:50,000 Published 2001

2001 VN-2000 WGS-84 1:50,000 Published 2005 (Ba Lat mouth only)

n ¼ 1  c

2.3 Net sediment transport

McLaren and Bowles (1985)proposed a hypothesis that relates

two cases of grain-size trends to net transport paths According to

this model, along the direction of net transport sediments can be

either better sorted, finer and more negative skewed (measured

in / units) or better sorted, coarser and more positively skewed

The model has been re-examined byGao and Collins, 1990and

1992 They proposed a procedure to define two dimensional net

sediment transport pathways, including some steps as follows:

(1) comparisons of grain size parameters at a station with the ones

at adjoining stations (the distance between them is not longer than

a characteristic distance It represents the space-scale of sampling)

to define unit vectors, i.e if there is one or not a net sediment

transport from the station to another; trend vector at a station is

defined by sum of unit vectors; (2) averaging trend vector at the

station and other ones of adjoining stations to remove noise and

define transport vectors and (3) significance test on the transport

vectors The parameters of sorting, mean diameter of grain sizes,

and skewness are considered to be equal importance in defining

net sediment transport pathways

The grab sampler takes sediment samples from the bottom

sur-face to the depth of 10–15 cm It may represent different time

peri-ods (e.g., a longer or shorter periperi-ods are taken into accounts at sites

of higher and lower sedimentation rates, respectively) The

charac-teristic distance in this area is assumed to be 5 km, which is the

longest distance between two adjacent sample stations The

differ-ences of sedimentation rates between stations of shorter than 5 km

apart are supposed to be small At about assumed characteristic

distance, the difference in recent sedimentation rates achieved

by210Pb analysis in 2 gravity cores (i.e cores 6 and 7 off the

south-west coast of Hai Hau) is only 0.5 cm/y (vanden Bergh et al., 2007)

Therefore the assumption is acceptable

2.4 Longshore sediment transport

Waves change the propagating direction when they reach to the

shallow water it due to bottom fiction To a certain depth waves

break and induce currents The currents then cause longshore

sed-iment transport which is the main reason of coastal erosion

Vol-ume of sediment transport is estimated by CERC formula (US

Army Corps of Engineers, 2002: Manual of Coastal Engineering)

The potential longshore sediment transport rate, dependent on

an available quantity of littoral material, is most commonly

corre-lated with the so-called longshore component of wave energy flux

or power:

Eb¼qgH2b

cgb¼c

2khb

sinh2khb

!

where Ebis the wave energy evaluated at the breaker line,abthe

wave angle relative to the shoreline (°), Hbthe wave height at

break-ing (m), cgbthe wave group speed at the breaker line, Lsthe wave

length (m), c the wave velocity (m/s),qthe density of water (kg/

m3), g the gravitational acceleration (g = 9.82 m/s2), and hb is the

depth of wave break (m)

The amount of longshore sediment transport is expressed as the volume transport rate (Ql) which is estimated by the formula:

ðqsqÞgð1  nÞPl ðm

where K is the experimental coefficient, equal to 0.39, qs the density of sediment grains (kg/m3), and n is the porosity of sediments

3 Results 3.1 Shoreline change The overlaying of maps shows quantitative figures of shoreline change at the coast (Fig 4and Table 2) The Red River delta is intensively moved seaward at the big river mouths such as Ba Lat, Day, and Lach The distribution of erosion comes between accretion segments

The average velocity of accretion is 65 m/y (1930–1965), 84 m/y (1965–1985), and 60 m/y (1985–1995) at the Ba Lat mouth (Table

2) The propagation of shoreline has close relation to the formation and enlargement of sand bars (Fig 5) A small bar (Vanh sand bar) was formed during the period from 1930 to 1965 The main direc-tion of development is to the NE (left bank of the river) The mouth was then rapidly moved toward the sea (1965–1985) The Vanh bar was intensively extended, and two other bars (Ngan and Lu) were formed The accretion at the right bank was dominant The seaside of Lu bar in 1985 was 7.1 km away from 1965 shoreline, i.e an advancing rate of 350 m/y on average This period is consid-ered as the strongest development of the Ba Lat mouth Following that mechanism a new series of sand bars was formed during the period 1985–1995, which was then enlarging and connected to each other in 2001 The mouth developed symmetrically The Lach and Day river mouths have not typical mechanism of propagation as the Ba Lat mouth Beside the sediment budget transported from the Ninh Co and Day rivers, longshore sediments from erosion at Hai Hau is intercepted by river currents and depos-ited in between the mouths and sand bars However there are still some small creeks between bars Their channels change frequently and are easy to be filled up Therefore a large continuous area of ti-dal flat is formed at the Day and Lach mouths The average accre-tion rates were 95–110 m/y and 27–35 m/y, respectively Sediment is mainly deposited at the big river mouths, causing erosion in other places The erosion occurs either near the big river mouths or in the middle of them It has caused land loss of several villages (Fig 6) Nearby the Ba Lat mouth, the shoreline of 22 km

in Giao Long and Giao Phong was eroded during the period 1930–

1965 The maximum retreat rate at Giao Phong was 24 m/y The ero-sion was even more severe in the period 1965–1985 with the aver-age velocity of 1.5 times larger than it was in 1930–1965 However the erosion was interrupted in 1985–1995 along with the south-ward enlargement of the Lu bar A short segment of 2.5 km was re-corded as weak erosion in 1995 The remaining 18 km of shoreline turned to be very strong accretion The shoreline moved seaward 100–430 m during the period 1985–1995 Another eroding coast

is Nghia Phuc which situates nearby the Lach mouth Erosion has ta-ken place since 1965 in the length of about 0.5 km The retreat rate was 8–10 m/y The shoreline is now at the trough of sea dikes The most severe erosion is the coast of six coastal communes (Hai Dong, Hai Ly, Hai Chinh, Hai Trieu, Hai Hoa, and Hai Thinh)

in Hai Hau district The erosion is considered to start from the beginning of the 20th century (1905) (Pruszak et al., 2002) It has

a close relation to the degradation of the Ha Lan river mouth (the former main river mouth of the Red River system at that time) The clear evidence of Ha Lan mouth degradation can be found at Giao Long and Giao Phong shorelines where were continuously

accreted with rapid rates (reaching to 100 m/year in some

seg-ments during the period of 1905–1930) (Fig 4) However the main

river mouth was then shifted to the Ba Lat mouth and shorelines

changed to erosion During the period 1930–1965 the maximum

retreat rate was 22 m/y in Hai Ly and Hai Chinh communes The

Hai Ly coast was then significantly eroded in 1965–1985 The

aver-age rate was 21 m/y At the same period the rates were 5 m/y at

Hai Dong coast and 11 m/y at Hai Chinh–Hai Thinh coast The

south part of Hai Thinh commune was accreted Upto 1995, major

parts of shoreline reached to the trough of sea dikes, i.e the water

level at the mean tide touched the dikes Lateral movement is

stopped Shoreline retreat is only realized at some segments in

Hai Ly and Hai Chinh where the former dikes were broken and

the locations of new dikes shifted landward Shoreline change at

other parts of the Hai Hau coast cannot be estimated by

topograph-ical maps The evolvement of erosion is then recognized by the

change of bottom topography and landscapes on the beach during low tide (see Section 3.2)

Nowadays, the shoreline in Hai Dong has been changed to accretion (by eye-seeing and personal conversation with local authority for some recent years) However the erosion continues

to increase in other segments The most severe erosion segment

is now shifting to Hai Thinh commune It is very significant by a series of three photos taken at the same place at the coast of Hai Thinh commune from 2003 to 2005.Figs 7 and 8show that a small tent for mineral exploitation was almost disappeared during

10 months (from 02 September 2003 to 25 July 2004) The shore-line retreated about 30 m Nine months later all the pine trees were destroyed The shoreline reached to the sea dike with a lateral movement of about 40–50 m (Fig 9) The result proves an actual situation of increasing erosion that is opposite to a remark of re-cent decrease of erosion (Pruszak et al., 2002)

Fig 4 Shoreline change in Hai Hau coast and adjacent areas.

Table 2

Accretion and erosion at the Red River delta coast.

Length (km) Average rate (m/y) Area (ha/y) Length (km) Average rate (m/y) Area (ha/y) 1930–1965

1965–1985

1985–1995

Day mouth a

a

The shoreline of the Day mouth in Ninh Binh province was not taken into account.

3.2 Nearshore topography change

The erosion has caused a remarkable change of bottom

topogra-phy along the coast The depth contour of 3 m in 1985 is

approxi-mately matched with the one of 5 m in 1965 (Fig 10) The 2 m

contour (if is considered as the middle between 1 and 3 m tours) was moved landward 1–2 km from 1965 to 1985 The con-tour then continued moving 1.5–3 km from 1985 to 1995 The maximum movement occurred at the south of Giao Phong com-mune An opposite sign of erosion is also recognized at the north part of Giao Phong commune where the 2 m contour of 1995 inter-cepts the one of 1985

Fig 5 Recent progress of the Ba Lat mouth.

Fig 6 Land loss due to erosion in Hai Ly commune.

Fig 7 Hai Thinh, 02 September 2003.

Fig 8 Hai Thinh, 25 July 2004.

3.3 Sediment properties and net transport

Laboratory testing of undisturbed soil samples (Table 3) shows

that nearshore sand is medium sand with the porosity of 0.42–

0.49 Grain size parameters of surface sediments are shown in

Table 8 Based on grain sizes, two main types of sediments are

de-fined such as sand and silt Sand is distributed along the shoreline

in water depths of 3–5 m, except to the southeast of the Red River

mouth, where sand extends down to the water depth of 15 m

(Fig 11) The recent sand is very well sorted and consists on

average for 98.5% of sandy and 1.5% of silt particles Silt is widely

distributed along the coast stretching from northeast to southwest Most of the silt is poorly sorted The composition is dominated on average by 70% silt, 22% clay and 8% sand Besides, sandy silt dis-tributes at the eastern and southeastern margin of the study area

at the depth is over 25–30 m It is the old sediment units (Duc

et al., 2007)

A set of 52 sediment is used to define the net transport accord-ing to the method ofGao and Collins, 1992 The results inFig 11 shows that the sediment from the Ba Lat mouth is not deposited nearshore, but moves seaward up to the water depth of 25 m It

is very significant along the Giao Long – Ha Lan coast at the depth

of 5–25 m In Hai Thinh shoreline, the sediment is transported along the coast southwestward In Giao Long–Giao Phong shore-line, the sediment is transported along coast northeastward The reason may be the northeast waves do not have strong effect on the coast because of the sand bars in front of the Red River mouth 3.4 Longshore sediment transport

The volume of longshore sediment transport is calculated by the formula (6), with the wave monitoring data at Hai Ly from 1976 to

1994 (Table 4) The result shows that the sediment is dominantly transported southwestward by the northeast and east waves The

Fig 9 Hai Thinh, 17 April 2005.

Fig 10 Nearshore topography change.

Table 3

Physical properties of nearshore sand.

No Sample Percentage of grain sizes (mm) Water content

(%)

Bulk density (g/cm 3 )

Grain density (g/cm 3 )

Porosity 1.0- 0.50- 0.25- 0.125- >

0.50 0.25 0.125 0.063 0.063

Table 4 Wave parameters at the Hai Hau station (01 January 1976–31 December 1994) ( Pruszak et al., 2002 ).

No H s (m) T p (s) h (°) ab (°) Duration (days)

1 0.57 2.93 87.5 44.5 36.5

2 1.22 5.14 81.7 50.3 35.1

3 1.67 6.49 81.6 50.4 1.4

4 2.04 7.00 77.0 55.0 0.2

5 0.51 3.00 65.0 67.0 10.9

6 1.13 5.13 63.6 68.4 2.2

7 1.79 5.94 73.2 58.8 9.9

8 1.99 6.82 68.8 63.2 4.6

9 2.48 7.18 67.6 64.4 4.2

10 3.20 8.20 63.5 68.5 0.7

11 0.80 2.65 43.0 89.0 19.1

12 1.16 5.00 43.0 89.0 3.8

13 1.91 6.00 43.0 89.0 0.3

14 2.50 7.00 43.0 89.0 0.1

15 4.25 8.25 40.0 92.0 0.1

16 0.59 3.27 20.0 112.0 3.4

17 1.32 5.12 18.4 113.6 1.8

18 1.60 6.00 26.0 106.0 0.2

19 2.03 7.47 16.7 115.3 0.7

20 2.52 7.79 18.4 113.6 0.2

21 3.25 9.05 24.0 108.0 0.1

22 0.50 3.11 3.2 3.2 11.0

23 1.19 5.44 1.2 130.8 7.0

24 1.74 6.79 11.8 120.2 0.4

25 0.54 3.20 22.6 154.6 7.4

26 0.47 5.80 47.0 179.0 33.1

27 1.28 9.36 63.4 195.4 0.1

28 1.76 11.00 47.0 179.0 0.1

29 0.39 5.00 70.0 202.0 3.2

volume is 654,078–801,078 m3/year in Hai Dong and Hai Ly

sec-tion (Table 5) The figure of Nghia Phuc section is 440,979 m3/y

Volume of southwestward transport at Giao Long section (1965)

was 741,335 It gives an evidence of strong erosion in this area

dur-ing the period 1965–1985 Sediment transport changed to

north-eastward at this section in the context of 1995 topography The

volume is 80,798 m3/y

4 Discussion

4.1 Impacts of sea level rise

IPCC (2007)indicated a clear trend of sea level rise (SLR)

world-wide with the average rate of 1.8 mm/y over 1961–2003 A

compar-ative study on impacts of SLR has confirmed that Vietnam is the

most vulnerable country to sea level rise in Southeast Asia and

one of top five most vulnerable countries in the world (Susmita

et al., 2007) Relative SLR in Vietnam is mainly calculated from

tide-gauge data collected at the four chief stations: Hon Dau (Quang

Ninh province – North Vietnam), Da Nang, Qui Nhon (Center Part)

and Vung Tau (South Vietnam) The longest tide data is achieved

at Hon Dau station from 1960 to 2000 The SLR of 1.9 mm a year

has been observed in this period (Hanh and Furukawa, 2007).Thuy

(1995)analyzed two tidal gauges in the North coast, one is at Hon

Dau and another is at Hai Hau The result shows that from 1950s

to 1990s the average rate of SLR is 2.24 mm/y The recorded data

of four chief stations shows that the increments in sea level varying

from 1.75 to 2.56 mm/y along the coast of Vietnam in 50 recent

years It is 3 mm/y over 1993–2008 (MONRE, 2009)

To estimate the increase of shoreline erosion the formula of the

so-calledBrunn’srule (1962) is used The formula shows the

rela-tion between SLR and the increase of shoreline erosion as

following:



where S is the SLR (mm/y); R1the exceeding rate of erosion due to SLR (m/y); L⁄

and (h⁄

+ B) are the width and vertical extent of the active beach profile

The results (Table 6) show that the increase of erosion rate can reach to 0.14–0.31 m/y along the coast of the Red River delta How-ever the erosion rate depends on many factors such as human activity, change of direction of sediment flow, waves, and currents (Duc et al., 2007) It is hard to define the accurate contribution of SLR on the increase of erosion rate To have a raw estimation of SLR effect, the erosion rate at the south Hai Thinh commune is ta-ken into account The rates were approximately 0 and 11 m/y dur-ing the period 1965–1985 and 1985–1995, respectively It is about

40 m/y in 2005 Therefore SLR contributes 34% to the increase of erosion rate during the period 1965–1995 and 12% from 1995 to 2005

4.2 Impacts of tropical cyclones Tropical cyclone is a typical climatic event in the North Viet-nam The so-called storm season often starts in June and ends in October About 13% of the total tropical cyclones attacked the country landed on the North coast Tropical cyclones, especially ty-phoons have caused many severe lost of properties and lives For instant, the typhoon PAT (23 October 1998) made 500,000 home-less and 90 death in the North coast The imprints of typhoons are recognized at the 22 m water deep by laminated sand layers between silty clay layers in a gravity core (van den Bergh et al.,

2007) Storm surge due strong winds and heavy rainfall in a ty-phoon can reach to a height of 2.6 m (Table 7) This phenomenon always leads to serious losses The most recent Damrey typhoon landed in the high spring tide caused very disastrous damages on

Fig 11 Net sediment transport pathways at the Hai Hau coast (1985 topography, map of surface sediment is extracted from Duc et al (2007) ).

sea dikes, mangrove, shrimp ponds, and infrastructure Hundreds

thousand people had to emigrate

According to the formula ofKriebel and Dean (1993)the retreat

distance caused by extreme wave heights can be estimated as

following:

R1¼HsðWb hb=moÞ

Wb¼ hb

A

Ts¼ 320 H

3=2

b

g1=2A3 1 þ

hb

moWb

hb

where Hsis the significant wave height (m), Hbthe wave height at

breaking (m), hb the depth of wave break (m), Wb the width of

breaking wave zone (m), B the height of berm (m), mothe beach

slope, t the duration of extreme wave heights (h), A the sediment

scale or equilibrium profile parameter (m1/3), R(t) is the retreat

distance caused by extreme wave heights (m)

The recorded wave heights during typhoons at Hai Hau tide

sta-tion (1976–1994) were 3.2–4.25 m (Table 4).Table 8indicates that

the erosion rate can reach to 7.1 m when the wave height is 4.25 m

and the duration is 2.4 h

The research of NCDC (1996)emphasized the increase in the

number of tropical cyclones attacked Vietnamese coast during

the period 1920–1994 The most recent statistical data of the

annual number of tropical cyclones shows that the number of

cy-clones does not have any clear trend during the period 1960–

1990s It had a significant reduction in number from 2000 to

2004 and then has been increasing very rapidly from 2005 up to present (Duc et al., 2009) This matter cannot all be claimed on cli-mate change However it is evidence showing that the variability of extreme events at the coast occurring more complicated Kleinen (2007) suggested an increase in occurrence and intensity of ty-phoons in the Western part of the Pacific, especially the ones that hit Vietnam The threat of typhoons on coastal zone in general and

on sea dike stability in particularly is expected to be more serious

in the near future

The analysis of statistical data on typhoons of the National Centre for Meteorology and Hydrology( http://www.thoitietnguy-hiem.net) shows that there were 86 typhoons directly hit the coast

of the Red River delta over 1962–2010, i.e an average of two ty-phoons annually The return periods of tyty-phoons with the intensity

of equal or greater than 10, 11, 12, and 13 (Beaufort scale) are about 3, 5, 10, and 21.5 years, respectively (Fig 12) As recognized from the Damrey typhoon (September 2005), storm surge and wave-run up were the main reasons for sea dike destruction and coastal flooding The correlation between typhoon intensity and

Table 5

Volumes of longshore sediment transport.

Location Beach slope Shoreline orientation (°) SW NE Total (m 3

/y) a

Giao Long (1965) 0.00800 44 1,000,002 258,667 741,335 Giao Long (1995) 0.00400 76 275,402 356,200 80,798

a

‘‘Plus’’ and ‘‘minus’’ are sediment transport to the SW and NE, respectively.

Table 6

Increase of erosion rate due to SLR.

Section SLR (mm/y) h ⁄

(m) Increase of erosion rate (m/y)

Table 7

Heights of storm surge in severe typhoons.

No Typhoon Date of formation Landing place Storm surge height (m)

1 PHILLIS 02 July 1966 Nam Dinh, Ninh Binh 1.10

5 WARREN 16 August 1981 Thai Binh, Nam Dinh 1.15

8 DAMREY 19 September 2005 Nam Dinh, Hai Phong 2.50

storm surge (Fig 13) indicates a unit increase of typhoon intensity

causes an increase of about 30 cm of storm surge height

To assess combined impacts of typhoon and SLR on storm surge

height, medium emission scenario (B2) is taken into account The

B2 scenario expects that sea level in Vietnam will rise 30 and

75 cm in 2050 and 2100, respectively (MONRE, 2009) As shown

inFig 14, the return period of a 2.6 m storm surge height reduces

from 20 years at the present to 9 years in 2050 and only 4.5 years

in 2100

4.3 Upstream dams

The Hoa Binh dam (fully operated in 1989) does not change the

amount of water but reduced 56% of suspended sediment budget

in the downstream flow as recorded at Son Tay and Hanoi

hydrolog-ical monitoring stations (Table 9) Sediment transported to the coast

can also derive from lateral and bottom erosion along the channels,

but no monitoring data about suspended matters in river flows near

the coast is recorded Meanwhile the accretion rate decreased from

84 m/year (1965–1985) to 60 m/year (1985–1995) at the Ba Lat

mouth It shows a tendency of reduction in coastal accretion due

to dam construction In the near future, as Tuyen Quang, Son La and other hydropower plants will be operating the sediment supply for the coast is going to decrease significantly The reduction of accretion at the river mouths and more severe erosion in the Hai Hau coast are irreversible

4.4 Suggestion of coastal protection The protection of the costal zone in the Red River delta is very important because of high population density and economic bene-fits Sea dike is the most important measure to protect the coast Simple dikes of compacted soils were common in the 1980s The dikes are easy to be eroded and severely damaged in a typhoon Such type of dikes is still at some parts of the coast in the Hai Chinh and Hai Dong communes To reinforce the dikes groynes were used They were constructed by a chain of concrete tubes with diameter of 1 m, thickness of 10 cm, and height of 1.5 m The tubes were filled up with sand bags and placed continuously at the depth

of 0.5 m under beach surface The distance between groynes is

80 m Mangrove forest is another measure against coastal erosion

A hundred meters of mature mangrove can reduce 0.1 m of wave height (Mazda et al., 1997;Quartel et al., 2007) However it cannot

be used in areas of severe erosion The sea dikes at erosion coasts, i.e Hai Hau and Nghia Phuc have been intensively reinforced since

1998, especially after the Damrey typhoon in September 2005 Dikes have mild slope of 1:2.2-3 with the height extended to +4.5 to 5.5 m and The dike footing was placed at the depth of

Table 8

Erosion rate caused by an extreme wave height.

Section S (m) h b (m) H b (m) B (m) m o D 50 (mm) A (m 1/3

) T (h) R(t) (m) Giao Phong 4.25 6.96 3.15 2.00 0.0040 0.143 0.0798 2.4 6.6 Hai Dong 4.25 9.23 4.10 2.00 0.0150 0.143 0.0798 2.4 7.1 Hai Hoa-Hai Thinh 4.25 8.18 3.78 2.00 0.0100 0.147 0.0840 2.4 3.1 Nghia Phuc 4.25 8.83 3.23 2.00 0.0045 0.157 0.0872 2.4 3.6

Table 9

Average water, sediment discharge before and after Hoa Binh dam.

Parameter Station

1956–1988 1989–1994 1995–1998 1956–1988 1989–1994 1995–1998 Average water discharge (bill m 3

Average sediment budget (10 6

Fig 13 Correlation between storm surge height and intensity of typhoon.

Fig 14 Estimated storm surge height and the return period (storm surge height is

estimated by the regression equation in Fig 13 ).

Fig 15 Current concrete seadike system.