Assessment of salinity intrusion in the red river delta vietnam

To bring a reasonable operation for both electric production and saltwater prevention is urgent duty of Hoabinh reservoir in the future, finding a numerical model for simulation and pred

ASSESSMENT OF SALINITY INTRUSION IN THE RED RIVER DELTA

VIETNAM

by

3B

Le Thi Thu Hien

A thesis submitted in partial fulfillment of the requirements

for the degree of Master of Engineering

4B

Dr Sutat Weesakul (Co-chairman)

Water Resources Engineering Water Resources University Hanoi, Vietnam

Fellowship Donor: The Government of Denmark

Asian Institute of Technology School of Civil Engineering

Thailand May 2005

1.2.3 Hydraulic Constructions in Study Area

1 Hoabinh Hydropower Plant

2 Son la Hydropower Plant: On-Going construction

3.2.1 Finite Difference Tidal Hydraulics Equations 22 3.2.2 Finite Difference Salt Balance Equations 22

iv

Chapter I INTRODUCTION 1.1 Problem Identification

Intrusion of salt-water in dry season is a well-known phenomenon in the Thaibinh estuaries In the rainy season from June to November the discharge of freshwater from upstream is high, the saltwater is pushed to the sea and the problem of salinity intrusion is not present But in the dry season from December to May of the next year the discharge of freshwater from upstream is small and the salinity intrusion problem becomes serious In some branches of the Red river system, the distance of salinity intrusion may be up to 40 km As the increasing of the freshwater intake for irrigation, the salinity intrusion is causes a lot of problems for irrigation, aquaculture and other economic activities due to lack of freshwater The knowledge of the characteristics of salinity intrusion therefore is very necessary for solving the problems and utilizing the river

Red-Saltwater intrusion in Red river delta has been studied for several years ago Many institutions involve research works for controlling and predicting Red-river salinity intrusion by various methods with mathematical models VRSAP (Water Resources Planning Institute, Hanoi Water Resources University), TL1, TL2 (Institute

of Mechanics), the hydrological and meteorological models (Hydrological and Meteorological Services), the multivariable relational model (Institute for water resources research) However there is a lack of efficient numerical hydrodynamic models that consider effect of Hoabinh reservoir as well as calculation and prediction salinity intrusion in Red river delta Sonla Hydropower Plant is going to built in upstream of Da River to reduce flood damage and improve irrigation in the Red River Delta Increasing inflow for irrigation in dry season can cause a change of salinity concentration for planning aquaculture area in Red River Delta’s coastal zone How to supply sufficient freshwater for paddy crops while controlling salinity concentration for aquaculture area? This is an important issue to assess the entire effect of Sonla Hydropower Plant to downstream area

Moreover, global climate changes in some recent years have deep effect to hydrology condition of Red river delta “Global Warning” could cause sea level rise 0.5

to 1 meter by the current century due to the “Greenhouse Effect” A rise in sea level enables salt water penetrates upstream and inland, and would threaten human uses of water particularly during droughts

To bring a reasonable operation for both electric production and saltwater prevention is urgent duty of Hoabinh reservoir in the future, finding a numerical model for simulation and prediction salinity intrusion in Red river delta for future is also very important

1.2 Study Area Introduction

1.2.1 Geographical Condition

Red - Thai Binh River System is the second largest river system in Vietnam, after Mekong River It originates from Nguy Son Mountain in Yunnan province of China

Fig 1.1 Red River System in Vietnam Territory The whole basin areas occupy 169,020kmP

2

P with a population of 18.56 millions in the year 2000

As a large river basin with a complex topography including mountains and hills (covering 90% of the area), delta and coastal areas, the Red - Thai Binh river basin, hosting a diverse and more and more developed socio-economy, makes a significant contribution to the national economy

Figure 1.2 River Network in Red River Delta

S«ng Hµn

S«ng Hång

S«ng Th¸i B×nh

S«ng Luéc S«ng §×nh §µo S«ng Kinh Trai S«ng DiÔn Väng S«ng §uèng

§Çm V©n Tr×

Hå

S«ng B«i S«ng §¸y

Fig 1.3 Study Area

1.2.2 Hydrological Condition

The Red River Delta is in reality the delta of two river systems: the Red River System and Thai binh River System The Red River System consists of 3 major river branches namely the Da, Lo and the Thao Rivers The Thaibinh River System is also comprised of 3 river branches, which are the Cau River, Thuong and the Luc Nam River as shown in Figure 1.4 The two river systems are connected through the Duong and Luoc rivers forming the Red and Thaibinh River Basin

Fig 1.4 Schema of River Network System

Duong River Viettri

SCHEMA OF RIVER SYSTEM

Da River

Thao River

Lo River

Hanoi

Hoabinh Reservoir

Cau River

Thuong

Sea East

Phalai Thaibinh River Luoc River Red

River Sontay

Catchments

Catchment’s Area (kmP

2

P)

Area Percentage in Red River Delta (%)

3

P (equal to 7.2% ones of Red river at Sontay station) with annual discharge is 318mP

3

P/s

Apart from inflow from Cau, Thuong and Luc Nam River, one numerous inflow is passed from Red River at downstream of Phalai through Duong River This flow is nearly triple are compared with Thaibinh’s (25kmP

3

P compare with 8.26kmP

3

P) In addition, Thaibinh River also gets supplementary volume from Red river through Luoc River with total volume is 13 kmP

3

P per year before flowing to the sea

In the dry season, water level in Red River fall down very low; in somewhere freshwater altitude of river is less than altitude of field’s surface inside the dyke However, water resources of Red River keep in plentiful state so the lowest monthly average inflow

at Sontay is 691mP

3

P/s

1.2.3 Hydraulic Constructions in Study Area

1 Hoabinh Hydropower Plant

Hoabinh Hydropower Plant was built in 1980 in the northern mountainous province

of Hoabinh with assistance from the former Soviet Union

Some characteristics of Hoabinh reservoir

• Surface of the reservoir F=200 km2

Hoabinh Hydropower Plant has been completely constructed in 1979 with 8 electricity generation units It has raised the discharge of flow of Da (Black river) and Red rivers in dry season up to 400-600 mP

3

P/s The flow regulation also facilitates to put saltwater into river mouth in dry season

2 Sonla Hydropower Plant: On-Going Construction

Sonla Hydropower Project to be constructed on the Da River, it is far from Hoabinh Hydropower Plant nearly 250 km towards upstream and about 320 km of Hanoi

The proposed Sonla Dam would be the largest dam in Vietnam The Sonla Hydropower Station Project will be the largest of its kind in south East Asia

Sonla Hydropower together with Hoabinh Hydropower Plant will improve Vietnam's electricity fuel mix, reduce flood damage and improve irrigation in the Red River Delta Sonla reservoir will hold a total of 25 billion m3 of water Together with the Hoabinh reservoir, the water volume will total 36 billion mP

3

P With the Sonla reservoir, safety discharge to Hoabinh in the dry season is 759 mP

3

P/s, raising 115 mP

3

P/s if has only

Hoabinh reservoir (Source: Proceedings of the Workshop on Methodologies for EIA of

Development Projects, Hanoi, July, 1999)

Electricity of Vietnam (EVN) plans begins construction on Sonla Hydropower Plant late 2005 First turbine expected operable 2012, the entire of construction expected compliable in 2015

U

Major objectives

• Energy production: 14.16 billion KWh/year

• Regulation flood stream: very important for Hoabinh Dam and downstream areas, including Hanoi (ensuring water level in Hanoi during flood season not to exceed 13 m)

• Water supply: providing to the Red River Delta about 6 billion m3; during dry season will ensuring a sanitary run-off of 300-600m3/sec

• Creating new opportunities for regional socio-economic development

U

Some characteristics of construction

• Normal water level: 265 m

• Dam height: 177 m

• Volume of reservoir: 25.4 billion m3

• Surface of reservoir: 440 km2

• Installment capacity: 3.600 MW

Fig 1.5 Location of Sonla and Hoabinh Reservoirs

1.2.4 Tidal Regime and Salinity Intrusion

The mixing of fresh and marine waters also is accelerated by tidal action The tidal regime in this area is irregularly diurnal, but is more regularly diurnal upstream The maximum tidal range along the coast of the Delta is approximately 4 m The tidal transfer speed in the river mouth approaches 95-150 cm/sec and the tidal influence extend 150-180

km from the river mouths (Source: Nguyen Ngoc Thuy, 1982)

Due to low terrain and improved river mouths so much, seawater and salinity are easy to go Red River Delta in almost of annual In Thaibinh River, low river bottom datum, large estuary and upstream inflow create a good condition for severe saltwater intrusion up far from the sea to Lucnam, Cau and Thuong River In the Red River, distance of saltwater intrusion was recorded at location which is 10 km far from Hanoi station above and 185

km far from the sea

Salinities increase from about 0.5 ppt in the rivers to 30.0 ppt Fluctuation widely of salinity depends on the flow in the river and state of the tide Salinity concentration 1 ppt can intrude about 30 – 40 km in average in the main branches with complicated characteristic

1.2.5 Existing Land Use

Almost the entire delta has been reclaimed for agricultural land, aquaculture ponds, forestry and urban development Approximately 53% of the delta is agricultural land, 6.4%

is forestry land and there are only some 3.8% of permanent lakes and ponds for aquaculture as shown in Fig 1.4 and Table 1.2

1 In general

The principal land use throughout the delta is the annual cultivation of rice, in

whole region produces about three million tons of rice per year (an average yield of 2,835 kg/ha in 1995)

To facilitate rice production, some 1,080 km of embankments, 34,400 km of canals, 1,310 drains, 217 reservoirs and 1,300 pumping stations have been constructed

In spite of the low salinity of estuarine water, the production of table salt by traditional measures in estuarine waters has been developed Each year the salt fields of this area have provided North Vietnam a table salt production of 20,000 – 30,000 tons

Table 1.2 Existing Land Use in 1998 (Unit: 1000 ha) Total Area Agricultural Land Forestry Land Aquaculture Land

Table 1.3 Agricultural Crop Land (Unit: 1000 ha)

Annual Crop Land

Perennial Crop Land

2 In Coastal Zone Area

Almost coastal zone area in Red River Delta no has agricultural land and has traditionally depended on fishing and salt production Production of catching fish is getting decreased Life of many stakeholders in the area is below poverty line

Coastal zone has 3 different sorts of water, including fresh water, brackish water and brine

Brine surface: set for the exploitation of sea products Some main sea products are

bream, Chinese herring, Khoai fish, grey mullet, Vuoc fish (perch), Van shrimps, Bop shrimps, and pawns At present, the seafood catching activities are natural and being carried out on small-scale A majority of aquatic products are used in processing traditional lines such as fish sauce, shrimp paste and seafood

Area of brackish water surface: Being mainly available in the Red, Thaibinh and

Traly river mouths thanks to an abundant source of short-lived creature, algas and aquatic botany as natural food used in process of breeding aquatic products Thaibinh province has about 20,705ha (Tienhai district has 9,949ha and Thaithuy district 10,756ha), of which

15,839ha is able to breed brackish water products (Tienhai 7,179ha and Thaithuy 8,660ha), including 10,386ha of tide-water region and 5,453ha of low productivity rice-transplantation salt-land likely being used for breeding brackish water sea-products At present, about 3,629ha is tapped for breeding shrimps, crab, arca, mussel and gracilaria

Fresh water region: The total area of aquatic products breeding is about 9,256ha, of

which 6,020ha has been exploited for breeding Besides, more than 3,000ha of low productivity hollowed

In brief, the estuaries of Red River Delta offer good conditions for aquaculture as follows:

- Water available for aquaculture development is large, estimated at over 1,000 ha

- Natural food sources are abundant, natural seed stock, particularly shrimp, is diverse in species composition

- High tide level assists in the supply and drainage of water and so the reception of natural food and seed from the sea and to the sanitation of the rearing ponds

- The mangroves in costal zone help protect aquaculture ponds and contribute to the supply of aquaculture seed (crabs, shrimp and certain species of fish) and feed (molluscs, trash fish, small mangrove crabs etc.)

Since the early 1980s, the aquaculture farming for export in Red River Delta has been encouraged and promoted by the government Furthermore, a high economic return leads to the widespread practice of this lucrative activity

There are many districts in coastal zone convert of salt fields, intertidal areas and mangrove forests into aquaculture ponds with highly profitable, at least in the short term

Fig 1.7 Districts along the Coast Having Aquaculture Production

Basing on different conditions about topography area, tidal regime, salinity concentration and so on, the sort of aquaculture species and pond size in each regional area

is different

Table 1.4 Aquaculture Productions in Districts along The Coast (Source data: 2002)

Nam dinh

Nghiahung 19P

o

P56-20P

o

P00N 106P

o

P07-106P

o

P12E

Shrimp

1040 Crab

Venus Clams Haihau 20P

o

P00-20P

o

P15N 106P

o

P12-106P

o

P20N 106P

o

P22-106P

o

P27N 106P

o

P27-106P

o

P37N 106P

o

P24-106P

o

P

Mud Crab Hai phong Tienlang 20P

o

P30-20P

o

P55N 106P

o

P28-106P

at present and future condition

In order to archive the above requirements, the mathematical model of MIKE11 is used to evaluate the characteristics of the freshwater flow and salinity intrusion based on the recent observed data The result will be estimated under the conditions of sea level rise due to the Greenhouse Effect

The objectives of this study are as follows:

- To estimate the longitudinal dispersion coefficients at different braches in the Red river delta at present

- To assess the effect of Hoabinh reservoir to salinity intrusion in condition with or without the reservoir

- To assess the future-effect of Sonla reservoir to salinity intrusion

- To forecast characteristics of flow and salinity intrusion in the future

Chapter II LITERATURE REVIEW

2.1.1 Mathematical Formulas

For many years, a number of systematic attempts have been made with more or less success to correlate the intrusion of saline water with tidal characteristics on the basic of actual observations of salinity condition in the estuaries

TAYLOR (1935) developed the turbulence theory and used the statistical approach

to formulate the dispersion coefficients for the case of two-dimensional motion as follow:

where:

DR x R, DR y R : the dispersion coefficients in x and y directions

U’R L R, V’R L R : the velocity fluctuation in x and y directions

RR u R, RR v R : the auto-correlation of velocity in x and y directions

A requirement is that the velocities be measured according to the Lagrangian standpoint However, actual data for velocity are normally obtained by measurements taken at fixed points, that is, they are expressed in the Eulerian point of view Therefore, the TAYLOR theorem cannot be applicable to the data available in most cases A transformation between the Eulerian and Lagrangian description of velocities was made by HAN and PASQUILL (1957), WADA et al (1975); they suggested that the dispersion coefficient can be expressed as follows:

Eu E

Ev E

where:

u’R E R, v’R E R : the Eulerian velocities fluctuation in x and y direction

β : a dimensionless parameter depending upon the scale of turbulence

TR E R : the Eulerian time scale

KETCHUM (1951) has presented an approach to the steady state salinity intrusion

problem based on dividing an estuary into segments whose lengths are equal to the average excursion of a particle of water during the flood tide Complete mixing is assumed within each segment at high ride, and exchange coefficients are based on this assumption As a result of the complete mixing assumption this method is limited to steady-state studies of estuaries where the well mixed condition is approached Estuaries of this type are characterized by very large rations of tidal prism to freshwater discharge and are a rather limited class as compared to the partially mixed estuary

ARON and STOMMEL (1951) have proposed a mixing-length theory of tidal

mixing as a means of treating the time average (over a tidal cycle) salinity distribution in a rectangular estuary The one-dimensional conservation-of-salt equation was employed with

a convective term for the river flow and horizontal eddy diffusivity The latter is assumed

to be equal to the product of the maximum tidal velocity at the estuary entrance, the tidal excursion length, and a constant of proportionality By integrating the conservation

as a classification of estuaries by means of a “flushing number” obtained by a best fit of field salinity measurements with one of the family of curves

They used the steady state model to study the problem of salinity intrusion equation is:

d dx

S d

From the above equation, DR x R can be estimated if the salinity distribution along the estuary is known

TAYLOR (1954) established that the longitudinal dispersion in a long straight pipe

may be characterized by a one-dimensional dispersion equation, in which the diffusive and convective process occurring throughout the cross section interact to produce a longitudinal dispersion coefficient:

*

0

u : the shear velocity

This result was probably the best known as well as the simplest of all equations describing turbulent dispersion

FURUMOTO and AWAYA (1955) proposed a numerical model to calculate

salinity intrusion in tidal estuaries by mean of transforming the independent variable x in the advective-dispersion equation into the storage volume V They obtained the longitudinal distribution of the dispersion coefficient in the estuary based on the quasi-steady transformed dispersion coefficient equation with the aid of the observed S-V relationship and fresh water inflow

THOMAS (1958) applied TAYLOR’s concept to flow in an infinitely wide two

dimensional open channel in which the flow is described by a power-law distribution He obtained a complicated functional relationship between dispersion and Reynolds number

ELDER (1959) duplicated THOMAS’s effort, for assuming a logarithmic velocity

profile, obtained a remarked simple result:

*

93

5 hu

in which h is the depth of flow

PRITCHARD (1959) presented a mathematical model representing the variation

of salinity concentration from tidal cycle to tidal cycle:

x

S A D x

S Q t

DR x’ R: time-averaged over a tidal-cycle of dispersion coefficient

DR x’ R was obtained by integrating the steady state equation corresponding to Eq (2.9)

By fitting data to Eq (2.15) D’R x R could be obtained

IPPEN and HARLEMAN (1961) used the steady state model and analyzed the

results of salinity intrusion experiment in the tidal plume of the Waterways Experiment Station (WES) to show that:

DR x R : the dispersion coefficient at station x at the low tide

DR o R : the dispersion coefficient at x=0 at low tide

x : 0 at river mouth

B : the distance seaward from x=0 to the point where S=So at low tide

The parameter Do is found to be correlated with a stratification parameter G/J, where:

fluid unitmassof ygainedper

ntialenerg rateofpote

d massofflui ionperunit

gydissipat rateofener

J

HARLEMAN and ABRAHAM (1966) re-analyzed the WES data and found that

the stratification parameter G/J was related to another parameter called “estuary number”

ED They formulated the following correlations:

2 1 1 2

055

70.0

o

E T

F P E

f

D t D

2

in which

PR t R : tidal prism, defined as the volume of water

FR D R : densimetric Froude number

uR o R : maximum tidal velocity

h : depth at the ocean velocity

ρ

∆ : change of density over the entire length of the estuary

study the salinity intrusion in a constant width representation of the Rotterdam Waterway The unsteady state diffusion equation:

∂

∂

x

s A D x

QS t

AS

They applied boundary conditions repeating from tidal cycle to tidal cycle, thus creating also a repeating time-varying salinity distribution The dispersion coefficient was assume to be in form:

The value of DR o R at any instant of time was determined by using the ocean boundary condition for salinity

FISCHER (1966-1968) made an important step in the development of methods for

predicting longitudinal dispersion coefficient in natural stream based on Taylor’s theory

He presented two ways of predicting a dispersion coefficient for a natural stream: the Method moment and the Routing method His methods required field measurement of channel geometry, concentration and cross-sectional distribution of velocity

The method of moment is based on the equation:

1 2

2 1 2 2 2

2

12

1

t t dt

d

x x

2 1 2 2 22

1

t t u

σ : the variance of the concentration distribution with respect to time, measured at

a fixed point in the stream

u : the mean velocity of the flow

t : the time of passage of centroid of concentration

Subscripts 1 and 2 refer to the two measuring stations

In the Routing method, a value of DR x R is assumed The validity of DR x R may be tested

by the beginning with a measured concentration curve at a particular time, applying the theory to predict a concentration curve at the same later time, at which one was actually measured The comparison between the observed data and routed results demonstrates the validity of the predict dispersion coefficient

BOICOURT (1969) used the approach of Prichard to study the salinity of Upper

Chesapeake Bay He obtained the dispersion coefficient by integrating equation

x S

dx t

S A S Q A D

x

f TA

A K x

∂

∂

(2.22) where

K’ is a constant during the time period of (t2-t1) and could be computed from measured data

A is the cross-section area

They derived the longitudinal dispersion coefficient, which was assumes constant during a time period:

x

o dt A K

t t QS t M t M

where:

M : total mass of salt

The value of Q, A, S are measured at station

THARCHER and HARLEMAN (1972) improved the model used by Stigter and

Siemons (1967) They extended the problem to transient boundary condition and proposed

a formula in which the dispersion coefficient varied with time and space:

S S K

x

S S K

L : length of estuary from sea entrance to head of tide

SR o R : salinity of sea water

S : local salinity

DR t R : dispersion coefficient die to the shear flow

6 / 5

002

o

E L

u

K

(2.27)

FISCHER (1973) showed that a quantitative estimate of the dispersion coefficient

in a real steam could be obtained by neglecting the vertical profile entirely and applying TAYLOR’s analysis to the traverse velocity profile:

t x

u w I D

ε

2 2

LIU (1977) also suggested a similar equation as Equation (2.28):

3

*

2

R u

Q

where:

β: a coefficient

Q : the discharge of the river

R : the hydraulic radius

LIU deduced an expression to estimate the coefficient

5 1

*18

VONGVISESSOMJAI, ARBHABHIRAMA and APICHATVULOP (1978)

formulated a mathematical model to investigate the effect of upstream fresh water discharge and tidal conditions on the salinity concentration and intrusion length along the Chao Phaya and the Mae Klong rivers The dispersion coefficient expression suggested by Thatcher and Harleman was used in this model in the following form:

=

x

S K nuR K

3

P/s)/(ppt/km) are appropriate for the Chao Phaya river

• K1 is equal to 100(mP

3

P/s)/(ppt/km); K2 is equal to 200(mP

3

P/s)/(ppt/km) for the Mae Klong river

PRANDLE (1981) analyzed the measured data from eight estuaries and shows that

these data could be fitted reasonable well with each of three expressions the dispersion coefficient:

o x

FWQA is usually called as ORLOB following the name of Dr Geral T Orlob This model was to be used in actual cases Both set of Saint-Vanant equations and dispersion equation are solve with a consideration of tidal effects The first application of FWQA was Sacramento-San Joaquin, California

b) SALFLOW of Delf Hydraulics

SALFLOW (1987) is production of cooperation between Hydraulics Institute of Netherlands and Mekong Committee It is one of the newest achievements in numerical salinity intrusion model

Test model in Netherlands achieved good results and doing apply in Mekong delta

In addition, there are modules of salinity intrusion in some hydrodynamic model in recent year as ISIS (English), MIKE11 (Danish) and HEC-RAS (US) but have not applied in Vietnam

2.1.3 Salinity Intrusion Study in Vietnam

Salinity Intrusion in Mekong Delta Project (Southern of Vietnam) in 1980 under the Mekong Committee assistance promoted the research of salinity intrusion in Vietnam Within the framework of this project, some of saltwater and salinity intrusion models were found by Mekong Committee and Institute of Water Resources Planning and Institute of Mechanics These models are used in research of Mekong delta planning, in estimate effect

of anti-salinity-intrusion constructions to enlarge crop area in the dry season as well as prediction salinity intrusion These models have important contribution to study of salinity

intrusion in Vietnam

On contrary, research of salinity intrusion in Red-Thaibinh delta is mentioned less than The following are some previous study

VI (1980) by analyzing the data recorded at the stations in the Red river system

states that in dry season the intrusion length of salinity at some branches of the Red river system may be longer than 30 km; also the freshwater discharge and the slope of salinity intrusion was not present due to the large amount of freshwater discharge from upstream

THUY (1985) studied the characteristics of tide in the Red River estuary He found

that the tidal properties vary greatly from the rainy season to dry season and the predominant components of tidal waves are diurnal

THUY (1987) applied a numerical model to study the flow in the river system

during flood and dry season He found that in dry season, tidal waves could propagate more than 100 km upstream along main branch of the river system

PHUC (1990) used 1D numerical model with much success However in the

model, the effects of density differences were not considered The data used for calibration were limited and the verification of the model was not possible Moreover, data were used

in the model such as datum of all station was not possible to bring to standard altitude; cross sectional areas of river system were not measured at the same years Thus the results were very limited

NGO (1991) based on the recorded data of salinity concentration at stations along

estuaries of the Red River System has drawn some primary remarks on the characteristics

of salinity intrusion there Details of salinity intrusion in each tributary of the river network were not investigated

dispersion coefficient varies in the same manner as those of salinity concentration

CA (1996) based mainly on two previous publications by Vu (1990) and Vu et al

(1991) Using many year recorded data of salinity concentration at stations along the estuaries, monthly-average salinity concentration at each estuary is computed The salinity intrusion length in each estuary was also estimated Details of salinity concentration distributions along the estuaries were studied using a numerical model of the transport and dispersion of salinity He found that in the dry season, the salinity intrusion length is as long as 20 km in the main river and mire than 20 km for some tributaries In the main river and tributaries with high freshwater discharge, the maximum salinity concentration is observes in January while for the tributaries with low freshwater discharge, the maximum salinity concentration is observed in March

HUNG‘s study (1998) of saline intrusion in the Red river delta has been also

limited by data used for calibration of the model and the dispersion coefficients were not accounted for the saline gradient along to estuaries also the verification was not carried out

AN NIEN, NGUYEN (1999) has summarized studies relating to saline intrusion in

Vietnam and has pointed out that at estuaries the salinity is in the range of 22-28ppt The saline intrusion length in the Red river delta is not so long The distributaries connected to the open sea is at acute angle, thus bands affected by salinity are narrow with the width of

12 km

The above studies are the first studies in some rivers without consideration of whole river system

Control of salinity concentration is primary importance in development of aquaculture in coastal zone as well as water intake to irrigate for crop fields in the dry season

According to Water Quality Standards (TCVN5943-1995) and Quality Criteria of Water for Aquatic Life (28TCN171-2001); (28TCN191-2004), salinity concentration is required for water intake into paddy fields and aquaculture ponds as followings:

- Gate of weirs under the dykes can be opened to intake for rice seeds fields while salinity concentration is 1g/l With growing-paddy, maximum salinity concentration is allowed in 4g/l

- The procedure for intensive culture of Tiger shrimp assign that salinity concentration of shrimp ponds as well as for nursery of shrimp from post-larvae 15 to post-larvae 45 is from 10 to 30 (past per thousand) (the best range: 15ppt-25ppt)

- River water can be used for men and livestock with salinity concentration is 0.4g/l

Chapter III THEORETICAL CONSIDERATIONS

The characteristics of flow and salinity intrusion in the rivers are governed by three predominant factors, which control the magnitude and direction of current at different depth and at different distances from the estuary month The three factors are:

i) The effect of tide in generating the tidal current and turbulence;

ii) The effect of upstream discharge of freshwater in producing a net seaward transport; an

iii) The effect of gravitational forces due to density differences between the upstream freshwater and downstream seawater or the sediment

The rise and fall of the tide at the mouth and the associated exchange of water masses through the entrance result in the temporary storage of large amounts of sea water

in the estuary during high tide and the drainage of this water seaward during low tide The total volume so exchanged is known as the tidal prism, which, for a given estuary, is variable only with tidal amplitude

Of significance in relation to this tidal prism is the continual inflow of fresh water from upland sources which results in a volume of fresh water equal to discharge rate totaled over the tidal period While this discharge rate may vary slowly with time, the ratio

of fresh-water volume to the tidal prism has proved of value in the general classification of estuaries

The type of an estuary depends on the ratio of volume of seaward freshwater flow

in a tidal cycle and volume of the tidal prism which govern the mixing of the estuarine water The volume of seaward freshwater flow in a tidal cycle is:

T AU

while

the tidal prism is:

T Au

T u A

u

U P

= (3.3) where

A: cross-sectional area at the estuarine mouth

UR f: R: seaward freshwater flow velocity

πo

u

u=2 : mean tidal velocity in a half tidal cycle

uR o R: maximum tidal velocity

3.1.2 Partially Mixed Estuary

When the ration of volume of seaward fresh water flow in a tidal cycle and volume of the tidal prism, Eq 3.3, is in the range of 0.2 to 0.5, the estuary type is partially The tidal velocity is strong enough to produce sufficient turbulence to induce mixing horizontally and vertically between fresh and salt water, therefore, there exists no distinct interface of fresh and salt water However there exists a gradient of salinity in the vertical direction due

to the partial mixing

3.1.3 Well Mixed Estuary

When the seaward fresh water flow is small, during the dry season, with respect to tidal flow,Q T /P≤0.1, the fresh and salt water are well mixed due to strong tidal action Under this condition, there exists only a little gradient of salinity in the vertical direction and salinities decrease progressively from the sea water at the mouth of the estuary

In these four estuaries, stratification is strongly season-dependent resulting in partly stratified conditions or well-mixed in the dry season when the dominant stratifying is tidal straining, tidal advection, and bed-generated turbulent mixing The rainy season is characterized by stratified conditions when estuarine circulation and advection of stratification by tidal currents and river flow are the main stratifying mechanisms

In the dry season, almost of estuaries in Red river delta is partially mixing with the ration between volume of freshwater in a tidal cycle and tidal prism as follows:

Table 3.1 Classification of Mixing Type of Main Braches of Red River Delta

(Source of data: Center of Estuary and Coastal Engineering)

Therefore, applying one-dimension model to evaluate the salinity intrusion in rivers

of Red river delta is appropriate

0.60 1.05

Partilly mixed Partilly mixed

4

Day

0.19 1.13 0.52

0.6 0.63 0.70

0.95 0.65 2.3

Partilly mixed Partilly mixed Stratified

5 Thai Binh 0.59 0.65 0.91 Partilly mixed

6 Van Uc 1.00 0.7 1.43 Stratified

7 Cam 0.66 0.75 0.88 Partilly mixed

8 Da Bac 0.92 0.75 1.23 Stratified

q x

Q t

z

∂

∂+

2 2

=+

∂

∂+

∂

∂

R A

Q Q gA n x

z gA A

Q x t

S u t

B: width at the water surface of the river cross section, including specified storage for each segment [m]

A: cross section area [mP

2

P]

n: Manning’s roughness coefficient

S: salinity concentration at time t [mP

SR l R: lateral salinity concentration

Both the tidal dynamics and the salt balance equations are solved using implicit finite difference method The domain, the x-t plane, is discretized into rectangular grids of size ∆x by ∆t Value of water level (H), discharge (Q), and salinity (S) at all grid points are to be determined

Firstly, the water level and discharge along the river are determined from the tidal dynamics Secondly, using these water level and discharge values, the salt balance equation

is solved to give the salinity along the river

The general outline of the MIKE 11 is used to compute the tidal hydraulics andsalinity intrusion in the Red river delta is as follows:

• Use the finite difference formulae with Abboth and lonescu 6-point implicit scheme

• Use the interpolate formulae to determine the intermediate grid points

3.2.1 Finite Difference Tidal Hydraulics Equations

- Continuity equation – h centered

- Momentum equation – Q centered

j

n j n j n j n j

x

Q Q Q Q x

+ +

2

22

1 1 1 1 1 1

(3.7)

3.2.2 Finite Difference Salt Balance Equations

To be simple, lateral flow is eliminated An arbitrary six point scheme constructed

by means of weighting factors and applicable to

t

S

∂

∂ and

x

S

∂

∂term of salt balance equation Fig 3.6 show how the weighting coefficients are assigned The weighting coefficients

m g

12

=++

=+++

m g

d b a

12

2

1

2

21

2

1 1 1 1

1 2

1 1

1 1

1 1 1

1 1 1

1 1 1 1

1 1 1

− +

+ + +

−

+

−

+ +

+ +

x

S S S

x

S S S

D

S S S

S d S S a S S b

x

u

S S S

S g S S m

t

n j n j n

j n

j n j n

j x

n j n j n

j n j n

j n j n

j n j

n j n j n

j n j n

j n j

Q Q

Center point

Fig 3.1 Staggered Meshes in Space and Time

3.3 Model characteristics

One-dimensional, unsteady salinity intrusion in a tidal estuary is solved using a salt balance equation while the tidal dynamics are solved using a system of one-dimensional, unsteady continuity and momentum equations based on Manning’s roughness coefficient

Both tidal dynamics and salt balance equations are solved using finite difference methods for implicit schemes The tidal dynamics are calibrated by varying Manning’s roughness coefficient until the model produces the water level and discharge relationships observed in prototype

2 Vertically well-mixed salinity condition exists

3 The seasonal lateral out/inflow from groundwater is negligible

MIKE 11 is a professional engineering software package for the simulation of flows, water quality and sediment transport in estuaries, rivers, irrigation systems, channels and other water bodies MIKE11 is based on an integrated modular structure with a variety

of basic modules and add-on modules, each simulating certain phenomena in river systems MIKE 11 includes basic modules for the following:

• Rainfall - Runoff

• Hydrodynamics

• Advection-dispersion and cohesive sediments

• Water quality

• Non-cohesive sediment transport

The modules used for this study consists of hydrodynamic (HD) and dispersion (AD) modules The application the MIKE 11 model in this study takes place in two steps:

advection-• Computation of the river flows and water level by MIKE 11-HD

• Computation of the river salinity concentration using MIKE11-AD

Chapter IV DATA COLLECTION

The collected data involves geometry data of cross-sections, hydrology, hydraulic and salinity data at stations in the Red-Thaibinh River System in 1993, 2002 and 2003

Two sets of geometry data in 1993 and in 2000 along main branches of whole Red River Delta are collected The accuracy of both of topographical data is very high and reliable In this study, geometry data of cross-section in 2000 is used

Fig 3.2 Arbitrarily Weighted Six-point Computational Molecule

to 5km

Fig 4.1 Spatial Distribution of Available Cross-Section in Red-Thaibinh River System

Data of water level, discharge and salinity concentration in the dry season mainly from December to April of the next year at the stations along the main branches of the Red river and Thaibinh River in 1993, 2002, 2003 are collected Table 4.1, 4.2, 4.3 shows the location of stations, the measured components of each set of data, respectively Station river networks of each year above are displayed in Figure 4.2, 4.3, 4.4

Measured data in 1993 for Red river basin are very good for hydrodynamic simulation However, the measurement of salinity concentration was focus on Balat estuary which is the main branch of Red river basin Salinity concentration along the other branches as Day, Ninhco, Traly was not measured or measured at only one station Therefore, these data are not enough to calibrate and verify for model

Measured data in 2003 is the newest data However, the measured campaign was limited in Red river basin from only Hanoi station to the sea Hence the change of upstream inflow of Red River Delta cannot study

Measured data in 2002 in all of Red River delta is the most adequate to requirement

of this study The data are both homogenous and long enough in order to be used in numerical computation of salinity intrusion For this reason, hydrological and salinity data from 5th to 20th March, 2002 is used in this study

Measured campaign was carried out in the same time at all rivers from upstream to Gulf of Tonkin Salinity was observed at every hour during high tide and every bihourly during low tide At every station, salinity is measured at three depths (surface, middle and bottom) Fig.4.9 illustrates the salinity concentration at three depths at Balat(km0) station

Salinity concentration tends to increase from surface to bottom of water depth Water

levels are measured at every hour Fig 4.10 presents measured water level and salinity concentration at Balat(km0) station in the Red River The maximum salinity concentration happens when water level is also the maximum

Table 4.1 Collected Data in 2003 for Numerical Model (Time of measurement: Feb/1/2003 to March/27/2003)

No River Station Code

Geographical Position Measured

Components Eastern

Fig 4.2 Locations of Gauging-Stations in Measured Campaign by The Year 1993

23

7 10

5 1

Table 4.2 Collected Data in 2002 for Numerical Model

No River Length

No

of Station Code Geographical Position

Measured Components

10 Luoc 68050 33 Trieu Duong 5 106.07 20.39 * *

11 Hoa 18000 19 Quan Khai 29 106.28 20.35 * *

12 Thai Binh 92100 50

Dong Xuyen 28 106.33 20.4 * * Cau Phao han 27 106.20 20.57 * *

13 Gua 2000 5 Ba Nha (Trung Trang) 34 106.27 20.52 *

15 Van Uc 35600 21 Kenh Khe(Quang Phuc) 36 106.32 20.45 * *

Trung Trang 35 106.29 20.5 * * *

18 Lach Tray 42100 24 Kien An 37 106.37 20.49 * *

20 Kinh Thay 37200 25 Ben Binh 30 106.21 21.03 * * *

Fig 4.3 Locations of Gauging-Stations in Measured Campaign by the Year 2002

Ninh

H~t

Q~t

12 11 10 9 8

14

15

16

5 22

32 34

38 35

30

26 7

25 4

6 1

Table 4.3 Collected Data in 1996 for Numerical Model (Time of measurement 12 to 11/23/1993 and 11 to 3/17/1996)

(Source of Data: Center for Estuary and Coastal Engineering Vietnam Institute of Water Resources Research)

No River Code Station

Geographical Position Measured

Components Eastern

Figure 4.4 Locations of Gauging-Stations in Measured Campaign by the Year 2003

Upstream Boundery Downstream Boundery

LEGEND

Ninh

3 2

4 10 11

4.3 Based Maps

1 Based map of Red River - Thaibinh River in ArcView of scale 1:250000 with

database management system on topography, hydrology, transport system, communes and land use for Red-Thaibinh River are collected to use in application of this study

Fig 4.5 Based Maps in ARCVIEW of scale 1:250000

2 Based map of Vietnam in MAPINFO environment including several basic map layers: topographic, geologic, administrative boundaries, basin boundaries, river and stream network, transportation network, hydraulic works systems, etc They are used for determination of gauging-stations are showed in Fig 4.6, 4.7, and 4.8

Fig 4.6 Map of Discharge Observation Stations

Fig 4.7 Map of Water Level Observation Stations

Hai hau Dis trict

Quyetchien

Docbo Trucphuong

Phalai

Catkhe Trieuduong Thuongcat

Namdinh

Vuquang

Hoabinh Yenbai

Phuly

Hanoi Sontay

Dongxuyen Trungtrang Benbinh

Trucphuong

Vuthuan

Connam

Balat Ngodong Dinhcu

Phule Binhhai Quyetchien

Phalai

Catkhe Trieuduong Thuongcat

Namdinh

Tienhoang Docbo

Nhutan Ninhbinh

Fig 4.8 Map of Salinity Intrusion Observation Stations

3/5/02 3/7/02 3/9/02 3/11/02 3/13/02 3/15/02 3/17/02 3/19/02 3/21/02 0

Salinity Concentration at three depth of water: Surface, Middle and Bottom

at Balat(km0) Station

Fig 4.9 Salinity Concentration at Three Depth of Water at Balat (km0) Station

Hai hau Dis trict

Tienhoang

Nhutan

Kienan

XM Hoangthach Caokenh

Dinhcu

Balat(km0) Balat

Kenhkhe Trungtrang Quankhai

Benbinh

Cauphaohan Tientien

kilometers'

100

3/5/02 3/7/02 3/9/02 3/11/02 3/13/02 3/15/02 3/17/02 3/19/02 3/21/02 0

1.4 LEGEND

S(psu) H(m)

Peak of Water Level

Fig 4.10 Salinity Concentration and Water Elevation at Balat (km0) Station

3.3.1 Hydrodynamic Module

The HD module contains an implicit, finite difference computation of unsteady flow in rivers and estuaries Within the standard HD module, advanced computational formulations enable flow over a variety of structures to be simulated:

by considering the cross sections The flood plain depression within the model area is defined as flood cells or storage cells, which have been connected to the main river through the link channel The runoffs from catchments of the region are directed to the hydrodynamic model either as lateral inflows distributed over a reach or as point inflows The area to be attached at each point or reach has been determines examining the detailed drainage pattern, established with the help of topo maps

3.3.2 Advection-Dispersion Module

The AD module is based on a one-dimensional equation of conservation of mass of

a dissolved or suspended material, i.e the dispersion equation The dispersion equation is solved numerically using an implicit finite difference scheme, which has negligible numerical dispersion An AD simulation is carried out on the basis of results from the hydrodynamic model

advection-The movement of salt in the model depends on flow velocities and the amount of mixing, which occurs with water of differing salinity Water level and discharge boundaries of hydrodynamic model must be specified as either open or closed for salinity

Times series of salinity concentrations needs to be specified in open boundaries, where salt can enter or leave the model In closed boundaries, it is assumed that no net transport of salt occurs Usually downstream water level boundaries are open concentration boundaries where as upstream discharge boundaries are closed

The program for salinity intrusion in an estuary is describe in Fig 3.4

READ, INPUTS

1 GEOMETRY &

ROUGHNESS

2 DISPERSION COEFFICIENTS

CALCULATION OF WATER LEVEL AND DISCHARGE BY HD-

PRINT OUTPUTS

CALCULATION OF SALINITY BY AD-MODEL

STOP STAR

Fig 3.3 Flow Dia gra m of Program for Sa linity Intrusion in Estua r y

3.4 Statistical Criterion

For the evaluation of the performance of the mathematical model compared

preferably against measurements, some statistical parameters are often used in order to

obtain some quantitative measure for the deviations

3.4.1 Efficiency Index or Coefficient of Efficiency

Efficiency index can be used for the measuring the efficiency or accuracy of the

model

2 1

2 1

2 1

)(

)(

)(

X X

Y X X

X

i i

i n

i i n

i i

(3.11) where,

SST is the total variation of the value in calibration or validation stage

1

2

)(

SSE is the sum of square errors (sum of squares of the differences between

observed and calculated value)

1

2

)(

where,

Xi is observed (measured) data at time i

X

11

Yi computed (predicted) data at time i

N number of data points

3.4.2 Standard Deviation (s)

1

)(

n

i i

1

)(

n

i i

Y

11

3.4.3 Root Mean Square Error (RMSE)

Root mean square error is used to compare the performance of two or more models when used for the same data set

RMSE

1

2)(

1

(3.14)

product of their standard deviations The correlation coefficient must have -1 R 1 The value R = 1, implies a perfect linear relationship with a positive slope, while a value of R = -1 results from a perfect linear relationship with a negative slope As the covariance gets smaller relative to the variance, so does the correlation coefficient, approaching zero If the variances are very small, then the covariance can be small and there will be a strong relationship The coefficient of determination is also defined as RP

2

P Both the correlation coefficient, R and the coefficient of determination are not well suited for model performance evaluation

y

x s s

XY

)1(

))(

*)((

n

i

i i

(3.16)

3.5.1 General Introduction

A worldwide rise of sea level is among the predicted consequences of the

“Greenhouse Effect”, the global warming expected as a result of the accumulation in the Earth’s atmosphere of carbon dioxide and other greenhouse gases generated by industrial and agricultural activities It has been suggested that increasing concentrations of these gases will lead to a rise of between 1.5P

o

P and 4.5P

There have been various scenarios for the scale of this sea level rise Analyses carried out by the Environmental Protection Agency in Washington led Hoffman to predict that global sea level would rise a meter during the next 50 to 150 years, with the most likely scenario of an accelerating sea level rise attaining a meter a century

Sea level rise scenario based on the prediction from the Climatic Research Unit, University of East Anglia, that global sea water level will stand to 18 cm higher by the year

2030 and the prediction from the Environmental Protection Agency, Washington, that it will rise 1m by the year 2090 as shown in Figure 4.2.5 (BIRD, 1989)

3.5.2 General Effects of A Sea Level Rise

A global sea level rise of one meter would greatly, modify coastal environments, producing erosion and submergence, especially on low-lying sectors Such a rise will enable the highest tides to reach levels of at last 1m above their present limits, allowing for

a possible increase in tide range as nearshore water deepen

A rising sea level will result in submergence and widening of the mouths of rivers with increased penetration by salt water, which may invade underground aquifers A sea level rise will also enlarge and deepen lagoons, causing erosion of fringing swamp areas,