study on inundation due to storm surge for phu quoc islands

Storm surge at 8 points around islands in scenario 3 71 Figure 24.Vegetation cover layer on the Phu Quoc island 76 Figure 26.Hydrological data layer around Phu Quoc island 77Figure 27.Si

MINISTRY OF EDUCATION AND

TRAINING

MINISTRY OF ARGICULTURE AND RURAL DEVELOPMENT

THUY LOI UNIVERSITY -

MINISTRY OF EDUCATION AND

TRAINING

MINISTRY OF ARGICULTURE AND RURAL DEVELOPMENT

THUY LOI UNIVERSITY -

Declaration

I hereby certify the work which is being presented in this thesis entitled, “Study on storm surge due to inundated for Phu Quoc island” in partial fulfillment of the requirement for the award of the Master of Coastal Engineering Management, is an authentic record of

my own work carried out under supervision of Ass Prof PhD Vu Minh Cat The matter embodied in this thesis has not been submitted by me for the award of any other degree

or diploma

Date: May 30, 2016

Vu Van Lan

I would like to thank faculty of Marine and Coastal Engineering of Thuy Loi University and Faculty Marine Science & Island, Ha Noi University for Natural Resources and Environment for enabling me thesis before the deadline

Finally, I would like to express my special appreciation to my friends and colleagues for their support, encourage and advices The deepest thanks are expressed to my family member for their unconditional loves

TABLE OF CONTENS

LIST OF FIGURES 7

LIST OF TABLES 9

INTRODUCTION 10

1 The necessity of the study 10

2 Objectives 11

3 Objects and scope of the study 11

4 Study approaches and methodology 11

5 Structure of the thesis 13

CHAPTER 1: OVERVIEWS ON STORM SURGE STUDY AND STUDY AREA 14

1.1 Literature reviews 16

1.1.1 International researches on storm surges 16

1.1.2 Storm surge researches in Viet Nam 20

1.2 Brief description on study area 21

1.2.1.Natural conditions 21

1.2.2 Climatic and oceanographic characteristics 24

1.2.3 Hydrological and oceanographic characteristics 25

1.2.4 Social and economic features 27

CHAPTER 2: APPLICATION OF DELFT3D TO STUDY STORM SURGE 29

2.1 Data used for the simulation 29

2.1.1 Statistical typhoon data 29

2.1.2 Water level 36

2.1.3 On land and seabed topography 36

2.2 Model description 36

2.2.1 Hydrodynamic Model 36

2.2.2 Typhoon model 43

2.3 Set up the hydrodynamic model 50

2.3.1 Computational grid 50

2.3.2 Topography 51

2.3.3 Boundary conditions 52

2.3.4 Other parameters 53

2.4 Model calibration 53

CHAPTER 3: SIMULATION OF STORM SURGE IN PHU QUOC ISLANDS 59

3.1 Typhoon zoning along the coastlines of Viet Nam 59

3.2 Generation of typhoon scenarios for Phu Quoc areas 62

3.3 Extraction of water level around Phu Quoc islands 62

3.3 Simulated results 63

3.3.1 Scenario 1: Simulation of typhoon namely Linda (later it is called Linda typhoon) that approached to the study area in November, 1997 63

3.3.2 Scenario 2: Scenario 1 in case of the typhoon Linda coming at the same time of flood tide at the study area 66

3.3.3 Scenario 3: Simulation of typhoons according scenarios approved by Ministry of Natural Resources and Environment (MoNRE), but wind velocity changes with the time V= f (t) 69

CHAPTER IV: BUILDING INUNDATED MAPS CAUSED BY STORM SURGE FOR PHU QUOC ISLANDS 72

4.1 Introduction on application of GIS 72

4.2 Application of ArcGIS software to build up inundated map 74

4.2.1.Topographic data 75

4.2.2 Hydrologic data 77

4.3 Building up inundation maps 78

4.3.1 The inundation map of scenario 1 78

4.3.1 The inundation map of scenario 2 82

4.3.2 Inundation of Phu Quoc island according scenario 03 85

CONCLUSION AND RECOMMENDATION 89

Conclusion 89

Recommendation 90

References 92

LIST OF FIGURES

Figure 2 Comparison of the calculated results top, legs and wave height of Boussinesq

Figure 3 Location of Phu Quoc islands on satellite image 22Figure 4 Statistical storm paths approaching to Viet Nam coasts 29

Figure 9 Observed and computed water level at Phu Quoc (C=55) 55Figure 10 Observed and computed water level at Phu Quoc (C=60) 56Figure 11 Observed and computed water level at Phu Quoc (C=65) 56Figure 12 Observed and computed water levels for model verification 57Figure 13 Observed and computed water levels in verification step 57

Figure 15.Basic characteristics and storm risk in Viet Nam Coasts 61

Figure 18.Water level field around Phu Quoc islands in scenario 1 65Figure 19 Storm surge at 8 points around islands in scenario 1 66Figure 20.Tidal series at Phu Quoc during Linda typhoon 67Figure 21 Storm surge at 8 points around islands in scenario 2 68Figure 22 Storm surge at 8 points around islands in scenario 3 71

Figure 24.Vegetation cover layer on the Phu Quoc island 76

Figure 26.Hydrological data layer around Phu Quoc island 77Figure 27.Simulated water level and topography in coast of Duong Dong and Ham

Figure 29 Inundation maps of the Cua Duong, Duong Dong, Duong To communes 81

Figure 32 Inundated mapping of Duong To, Cua Duong and Duong ông communes

84Figure 33 Inundated mapping of Duong To, Cua Duong and Duong ông communes

84Figure 34.Spatial disribution of flooded areas in the Phu Quoc island in scenario 3 86Figure 35 Inundated mapping of Duong Dong and Cua Duong commune 87

Figure 37 Inundated mapping of D ng To commune 88

LIST OF TABLES

Table 1 Coordinate of Phu Quoc area shown in map scale of 1/ 50.000 23

Table 2 Monthly and yearly average temperature in Phu Quoc and Rach Gia (oC) 24

Table 3.Monthly and yearly average and minimum humidity (%) at Phu Quoc 24

Table 4.Monthly and yearly average and minimum humidity (%) at Phu Quoc 24

Table 5 Monthly wind velocity and main direction at Phu Quoc 25

Table 6.The statistical result of wave height and its period at Phu Quoc 26

Table 7.Water level at Phu Quoc (103058 E – 10013 N) station (1990-2008) 27

Table 8 Statistics on typhoon hitting to Phu Quoc and surrounding areas 30

Table 9 Characteristics of typical typhoons approaching to the southern coasts 31

Table 10.Tidal constituents at 3 boundaries 53

Table 11.Coefficient RMSE 56

Table 12.The locations where water level is extracted 62

Table 13.Linda typhoon’s parameters 63

Table 14.Highest storm surge and appearance time at the extracted points 65

Table 15.Adjustment of Linda typhoon time to fit to spring tide 67

Table 16.Highest water level and appearance time at 8 points in scenario 2 68

Table 17.Typhoon parameters used to simulate in scenario 3 69

Table 18 Highest water level and appearance time at 8 points in scenario 3 70

Table 19 The maximum storm surge at the point around Phu Quoc islands 71

Table 20.Clasification of inundated depth 78

Table 21 Flooded area for Phu Quoc island in scenario 1 81

Table 22 Flooded area for Phu Quoc island in scenario 2 85

Table 23 Flooded area of Phu Quoc island in scenario 3 87

INTRODUCTION

1 The necessity of the study

In recent years due to the impact of global climate change, natural disasters become more complex, especially storms, accompanied by rising sea levels caused flooding of coastal estuaries The sea level rise due to storm caused flooding of coastal areas and break dike, especially storm occur during high tides So the study, calculated and forecasting extreme storm surge in coastal area and flooding risk due to storm are positive tasks to find appropriate solutions for prevention and reduction of damages in coastal areas The components cause extreme water level during storm including tides, storm surge, and wave surge, in which the storm surge is an important one

Storm surge is a dangerous natural phenomenon which causes lost lives, destruction of socio-economic infrastructures and valuable resources when typhoon attacking to coastal areas Worldwide, storm surge has caused major damages such as the typhoon

in 1970 and 1990 with water surge more than 7 m, generated large wave, inundated to delta of Bangladesh and over 400,000 people were killed On the Caribbean, highest water surge of typhoon Flora is 8 meter, it had cause flood and over 5000 people were killed Coastal of the United States had been affected by historic storm surge of up to 7.4 m The countries on the Northern coasts of Europe had been affected serious of storm surge in 1916, 1953, 1962, 1976, in which the storm occurred in 1953 in Netherland caused large inundation and over 1400 people killed

Storm surges may be defined as high sea water level above mean sea level which is caused by strong winds and low atmospheric pressures of a storm Winds which blow towards land exert a shearing stress on the surface, causes an increase in the sea water level near the coastlines Low atmospheric pressure also produces high elevation due to the so-called inverted barometer effect The highest surges have generated by strong tropical cyclones The surge belongs to the same class of phenomena as tide waves and tsunamis Its horizontal scale depends on the parameters of the storm In general, the storm surge occurs in duration of several hours, but it can sometimes last for a few days

It is obvious that prediction of surges is a very urgent issue to be addressed, especially

attempts to develop methods for forecasting storm surges in Vietnam At present, all kinds of activities are increased in number and almost marine constructions need sea level data for designing Sometimes they need the values of sea level rise which happen

at rare frequencies, while the observed stations located along coasts and islands are scarce That’s why in this study numerical models are used to simulate storm surge based

on the data of bathymetry, figure of coastlines, climate characteristics of typhoons and oceanic data at sea and coastal stations

2 Objectives

The general objective of the present work, therefore, is to develop a method based on hydrodynamic models to determine maximum surface water elevation generated by typhoons The computed results of model produce an atlas of pre-computed surges and

a collection of several possible typhoon conditions from many potential surges It is straightforward to determine the highest possible surge at all vulnerable coastal locations from a particular family of tracks and simulated storm surges as input data for preparation of potential map of inundation in Phu Quoc island

3 Objects and scope of the study

+ Objects of the study: The extreme water level during storm at the shorelines and potential inundation caused by storm surge

+ Scope of the study: Phu Quoc islands and surrounding areas

4 Study approaches and methodology

The simulation of storm surge and land inundation by using Delft3d is shown in figure

1, of which the following steps are conducted

Figure 1 Flowchart to illustrate the study approach

Step 1: Collection of data

These data are used as input for calibration and verification of models and simulation of storm surge These are included:

+ Hydro-meteorological data such as water level series at the sea and observed stations, atmospheric pressure, water temperature, wind, wave, currents etc

+ Topography including sea bathymetry and land elevation surrounding the study islands

+ Socio-economic data: including infrastructures around the coasts, lands, forests, ecosystem and all socio-economic activities that will be damaged if coastal strip of island is inundated by typhoon water

Step 2: Set up computational model

It includes computation network, meshes, defined boundaries such as tide series, seabed topography and typhoon information including central typhoon pressure, typhoon

data, the calibration and verification are conducted to find model parameters that are fit between simulated and real observed data both for hydrodynamic as well as typhoon models

Step 3: Create scenarios for simulation of storm surge in the study area

The scenarios proposed for study of storm surge are based on the real typhoon which had been occurred in the past and potential typhoon that can occur under the climate change conditions These are presented in detail in chapter 3

Step 4: Simulation of storm surge around the study area according to scenarios proposed

in step 3

Results of this step are resultant water level fields (including tide plus storm surge) at the study areas The real storm surge can be taken by subtracting total water level and astronomical tide at the same time The computation of resultant water level, potential land inundation around the island at many points is taken for each scenario

Step 5: Inundated mapping for each scenario

By overlapping simulated water level map on to topographical map with the support of GIS software, the potential inundated map can be produced for each scenario

Step 6: Conclusions and recommendations

In this content, author will summary the results conducted in the research and also propose the future works that should be continued to serve socio-economic development

in Phu Quoc islands

5 Structure of the thesis

Besides the introduction, conclusion, recommendation and annexes, the study is consisted 4 chapters as following:

Chapter 1: Overviews on storm surge study and study area

Chapter 2: Application of Delft 3D to study storm surge

Chapter 3: Simulation of storm surge in Phu Quoc islands

Chapter 4: Building inundated maps caused by storm surge

CHAPTER 1: OVERVIEWS ON STORM SURGE STUDY AND STUDY AREA

Storm surges may be defined as high sea water level caused by strong winds and low atmospheric pressures at the center of a storm Winds which blow towards land exert a shear stress on the surface, causing an increase in the sea surface elevation near the coastline Low atmospheric pressure also produces high elevation due to the so-called inverted barometer effect The highest surges are generated by strong tropical cyclones and it is considered the same category of phenomena as tide waves and tsunamis Its horizontal scale depends on the parameters of the storm In general, the storm surge can occurs in the duration of several hours, but it can sometimes last for a few days It is obvious that prediction of surges is a very urgent issue to be addressed, especially in coastal regions which are affected by tropical cyclones

There have also been many attempts to develop methods for forecasting storm surges in Vietnam At present, on the coastal areas, all kinds of activities are increased in number and almost marine structures need sea level data for designing Sometimes they need the values of sea level rise which happen at rare frequencies, while the sea level stations located along coasts and islands are scarce That’s why we need to find another way to define the maximum values of sea level rise that can happen in the chosen areas The way mentioned here is numerical model that is used to simulate sea level rise with the data of bathymetry, coastal topography and climate characteristics of typhoons, tide and wave conditions occurred at the chosen places

Worldwide, storm surge has caused serious damages in the coastal areas For example the typhoons attacked the Bangladesh in 1970 and 1990 created a storm surge of more than 7 m, generating high waves, and inundated large area of Bangladesh delta and more than 400,000 people were killed in these events On the Caribbean Sea, the highest water surge of typhoon Flora was 8 m causing serious flood and over 5,000 dead people; Coastlines of the United States had been affected by big storm with storm surge up to 7.4 meters The countries on the Northern coasts of Europe had also been affected by serious storm surge in 1916, 1953, 1962, 1976 in which the typhoon in 1953 hitting the

coastlines of Netherlands caused sea dike breaches, resulting large inundation and over

1400 dead people

Storm surge is very dangerous natural phenomenon which causes to destroy valuable properties, lost lives and all socio-economic infrastructures when typhoons attack to coastal areas Friction of wind on water surface and decreasing pressure at typhoon center are main reasons to create high storm surge The bathymetry and parameters of typhoon including center typhoon pressure, maximum wind radius, wind velocity, storm track, river flow, and tidal regime are factors which affected to storm surge That’s why the problem is very urgent and important to study

There have also been many attempts to develop methods for forecasting storm surges in Vietnam At present on the coastal areas, all kinds of activities are increased in number and almost marine constructions need data for designing in which sea water level is very important because it happens at rare frequencies, meanwhile the observed stations located along coasts and islands are scarce That’s reasons why we need to find other ways to define maximum values of sea level rise occurred in a certain areas The way mentioned here is an application of numerical models for simulation For doing this data

of bathymetry, coastal topography, hydrodynamic parameter such as tide, waves, sea currents and typhoon characteristics in the interested areas are needed

According to Le Van Thao et al (2000), storms occur in Viet Nam unevenly The most affected areas are the northern and the central coasts The southern coast is less affected both in number and intensity, but damages was more serious because less awareness on the typhoons of local people Typhoon Linda in November 1997 was an example and considered as a very uncommonly strong storm in the past 100 years to the southern coasts area There was about 778 people killed, 1142 and 2541 injured and missing,

2789 boats sank etc The total economic loss was estimated about 480 million USD (Le Van Thao et al, 2000)

Phu Quoc and Tho Chu islands belonging to Kien Giang province are located in the western sea of Thailand gulf They are considered as strategic locations in socio-economic development as well as defense and security in the south of Viet Nam due to

there terrain and resources For sustainably economic development and environmental protection, the study all on natural disasters, specially typhoon and storm surges is a priority tasks For which we can assess the flood inundation and damages due to typhoon and storm surges to serve marine spatial planning as well as to make strategy for mitigation of natural disasters for Phu Quoc islands

1.1 Literature reviews

1.1.1 International researches on storm surges

Because of the direness of the storm surge disaster, the studies of theory and scene from which construct methods, technological modeling to calculate and forecast storm surge, which have been conducted for so long According to Brestschneider (1959), different factors can cause change of the water level in coastal areas during a hurricane are: the parameter of storm (atmospheric pressure, wind speed ), the rotary motion of the earth, wave, and rain Later, Pore (1965) has added factors: tide, shape of shoreline and water depth

Currently there are several methods of calculation and forecast storm surge such as method uses semi-empirical formula, diagram method, artificial neural systems method and numerical model methods

In the method using semi-empirical formula (Ippen and Hallerman, 1966), surge magnitude is calculated based on ground level wind speed, wind fetch length, the angle between the wind direction and the axis perpendicular to the shoreline and the water depth This method is very simple but precision is not high because it does not describe all the factors which impact on storm surges

Diagram method (Yang et al, 1970, Horikawa, 1985) is often used to forecast storm surges for some ports, where have many monitoring data on hurricanes and storm surges The content of the method is to construct the monogram based on the relationship between monitoring data of water level with parameters of hurricane storm (the largest wind speed, wind direction, reduce of pressure in the center) Therefore, the method is very limited when data series is not long enough (usually around 100 years if require result is high precision) and often only true for the nearest observation station

Numerical models method was created to overcome the deficiencies of empirical measurement data The advantage of this method is reduction of cost compared with experimental measurement methods In addition, this method also allows calculation, forecast the evolution of the phenomenon based on a lot of assumed scenarios, which does is not yet exist in reality present but likely to happen in the future

In studies by numerical models, storm surge phenomenon is modeled based on the shallow water equations (2 or 3 dimensions) Depending on the purpose, in forecast about storm surges, 2-dimensional model does not take much time to calculate but it can achieve full accuracy When the need for simulation and calculations in more detail, eg distribution according to the flow rate of the water layer, 3-D model is needed With more detail simulation and calculations, such distribution flow rate under water layers

is revealed in 3-dimensional model At the beginning, the numerical models were built

to simulate storm surge, which are limited by several reasons: (1) Usually only simulation, calculation of individual phenomena such as tide, wave, storm surges; (2) The grid is very course, which does not cover about the detailed topography of coastal area; (3) In addition, many effects that affect to storm surges in the equations system is ignored Therefore, accuracy of the results of the model varies among areas or simulation is very good for a storm but limited with other storms For example, Jelesnianski’s model (1965), which has ignored friction component and nonlinear component, so calculated results were reasonable in spatial distribution of water rise and the time storm surges is the highest, however tends to overestimate the water height in some l SPLASH model (Special Program to List Amplitude of Surge from Huricanes) was built in 1972 by Jelesnianski and then SLOSH model (Sea, Lake, and Overland surges from Hurricanes) was developed to simulate the storm surge in coastal areas, sea and lake, which NOAA (National Oceanic and Atmospheric Administration) used to simulate coastal flooding caused by storm surge in the United States (Jelesnianski et al,

1984, 1992) but there are many restrictions such as the use a grid with fixed structure, which cannot simulate the coastal areas with complex topography and shoreline The Concept of storm surge in the previous calculations usually understood as the water level rises due to the impact of the wind stress and reduction of pressure in center of the

storm However, in fact wave stress generated surge wave, which occupies a very significant part of the storm in the shallow waters Therefore, recently wave setup has been interested and considered as an important part in the warning news, forecasts in countries like the US, Japan, UK Due to the complexity of the wave setup phenomenon, calculation has just followed by the analytic formula Longuet- Higgins and Stewart (1963)

This study has shown that the magnitude of wave set up depends on the horizontal gradient variation of wave radiation stresses Longuet-Higgins and Stewart’s theory has explained the mechanism of phenomenon of rising water and ebb water around break water area in shallow water zones Longuet-Higgins and Stewart’s theory proved quite suitable, when verified with experimental data of Bowen et al (1968) about phenomenon

of rising water and ebb water in points around break water area Bowen et al's experiments (1968) on the phenomenon of wave surges and ebb water, which are caused

by waves have been used to verify numerical models, which simulate the phenomenon

of wave propagation in coastal zones as shown in figure 2

Figure 2 Comparison of the calculated results top, legs and wave height of

Boussinesq 1D model with experimental data of Bowen (1986)

Recently wave setup was considered in calculating total of storm surges by combining numerical models in many studies Funakoshi et al (2008) have combined ADCIRC model which is used to simulate about storm surge and SWAN model, which is used to simulate wave This study indicates that, wave setup may contribute 10-15% in extreme water levels in storms Another study combined storm surge model and wave model as

Chen et al (2008) in 2005 Hurricane Katrina in the US, which concluded that storm surge by affection of coastal wave contribute 80% in extreme water levels while other influences such as tide, surface wave and rising water by wind contributed only 20%

In 2010, Youl Kim Soo et al have developed model to predict storm surges, which integrated tide both waves (Surge Wave and Tide - Suwat) This model was designed with integrated mesh to calculate storm surges in the Tosa Gulf - Japan and the results are consistent with the measured data, while previously many models not interested wave setup give lower results Youl Kim Soo's research also shows that to study wave setup need to perform on the calculated grid which has detailed resolution After Youl Kim Soo model has been used to forecast storm surges in many ports in Japan

In recent years, due to the development of systems of monitoring and transmit water level data in real time, data assimilation techniques of water level in tidal forecast model, from that storm surges has been built and development (Lewis and Derber, 1985; Thacker and Long, 1988)

To accurately predicting storm surge depends on accurately predicting the field of pressure and wind in storms However, if the water level monitoring data is regularly updated in the forecast calculations, the error will be decrease significant When Lionello P (1996) used data assimilation techniques in water level forecasting models showed that the storm surge forecasting results in report in Atlantic Beach has reduced errors up to 50% for forecast from 1 to 3 day

Assess the risks of hurricanes and storm surges also follow the traditional approach of assessing risk method of natural disasters is based on statistical methods In developed countries like the US, Canada, Australia, the European Community (England, Poland, Croatia, Italy, Netherlands, Spain), Asia (Japan, South Korea, the East Asian countries (SY Wang et al, 2007), there have many research programs to develop response methods early To calculate the possibilities of disaster risk, Monte-carlo (PPMC) method is used

a lot in disaster applications: storms, storm surges, waves and waves in storms, floods, landslides, earthquakes Specifically with surges in Australia, scientists have simulated storm for 3,000,000 years from data of historical hurricanes in 30 years, the US used

storm simulated data from 2,000 years from data of historical hurricanes in 100 years, which is used as input of the storm surge model, from that construct frequency line of storm surge with period is from 2-100 years and distributed risk map of large wave

1.1.2 Storm surge researches in Viet Nam

Vietnam is a coastal country with high potential risk of storm surges That’s why the study storm surges is paid much attention long time ago with many methods from experiences to mathematical models

According to statistical researches, the first study of Vu Nhu Hoan (1988) was presented,

in which storm surges are estimated according to statistical methods and charts Recently, Hoang Trung Thanh (2010) used observed data of water levels in the oceanographic and estuary stations to assess water surge generated by wind and thus gave overview about time and rise and set down trends at the monitoring stations Although there the advantage of being simple and easy to use but limited the application

of statistical methods Because of very sparsely observed stations, the accuracy of this method is not so high Therefore this method is only suitable in some monitoring stations where data series are long enough and with this reason it is rarely used in Vietnam

In research methods using numerical models, there are three main directions being used These are included self-built models, research and development of open source models from abroad; and using commerce models from abroad These information can be seen

in the researches of Vu Nhu Hoan, Do Ngoc Quynh, Le Trong Dao, Bui Xuan Thong, Dinh Van Manh, Nguyen Thi Viet Lien, Nguyen Vu Thang and Nguyen Xuan Hien When researching storm surges in coastal areas of Tokin Gulf, Le Trong Dao (1998) have used finite element method to calculate tide and storm surges and had conclusion that due to large tidal difference up to 4.0 m, so storm surge was more impacted by tidal regimes Also by using this method, Nguyen Vu Thang (1999) got result about prediction storm surges at Hai Phong coast using finite element methods

Also by using finite element method, in the governmental projects namely KT.03.03,

Do Ngoc Quynh and Pham Van Ninh (1999) used the bi-directional shallow water

equations to calculate the tide and storm surges for all Vietnam coastal areas Accordingly, the current situation and the risk of storm surges was calculated and partitioned by latitude The results of the study have served for disaster prevention and build coastal constructions Also according to the finite difference method, Bui Xuan Thong (2000) has developed cage mesh to increase the details the points need calculated

as well as reducing the time to calculate when the calculate storms surges In 2001, the model predicted storm surge had calculate to tidal and design on the cage mesh by the Institute of Mechanics have been applied on calculated the storm surge with detailed resolution to 1.0 km serving the coastal constructions such as dikes, jetties, after that, this model has been applied in many subjects, different projects related to storm surge

in Vietnam Author Phung Dang Hieu (2013) have built models that predict storm surges that taking into account the influence of the tide on the system nonlinear shallow water equations and according different method of SMAC combined with schemetic CIP there are tertiary accuracy for nonlinear components, The model was applied to simulate surges and flooding coastal areas of Thua Thien Hue very reliable results when compared with observation data

In recent years, due to the development of computational systems and information technologies have had many foreign models are built and developed towards commercialization as well as shape open source to community develop The popular commercial model is being applied in Vietnam as models MIKE Danish Hydraulic Institute (DHI), SMS model of the US Navy, the Delft-3D model of the Hydraulic Institute Delft In the thesis had used Delft 3D- Flow simulation about storm surge in Phu Quoc Island

1.2 Brief description on study area

112 km and 45 km accordingly Administratively, Phu Quoc district consists of Phu Quoc island and 2 other smaller islands namely An Thoi and Tho Chu with total area of 593.05 km2 in which Tho Chu archipelago is farthest from the main land (about 115 km)

Figure 3 Location of Phu Quoc islands on satellite image

The shape of Phu Quoc island is nearly triangular, base side in the north, narrow gradually to the south (figure 1.3) The research site would be covered Phu Quoc water area of 220 km2, from the coastline of island to the water depth up to 20 m The study area is defined by the points from A1 to A26 with coordinates (National Coordinate System VN 2000) showed in table 1

Table 1 Coordinate of Phu Quoc area shown in map scale of 1/ 50.000

Point Northern

latitude

Eastern longitude Point

Northern latitude

Eastern longitude A1 103° 59' 54.55" 9° 56' 23.54" A14 103° 59' 47.59" 10° 28' 6.96"

The typical topography of Phu Quoc island is low hill The coastline is zigzag, divided

by various channels and rocky mountain The sea bed bathymetry of Phu Quoc is clearly distinguished into 2 levels of depths:

+ From 0 – 8 m: generally even and flat At the southern site, bathymetry is more complicated with unstable slope, submerged dunes and deep channels due to the impacts

of submarine canyons The estuarine topography is consisted of submerged dunes, bar between channels and varies with seasons due to river-sea dynamic interaction

+ From 8 -20 m: the bathymetry is deeper from the shorelines to the offshore It is the boundary of modern sediment deposition area and existing kinds of shorelines: This kind

of shoreline was observed in every original rock before Quaternary era in Phu Quoc island and other islands, composition is mostly high stable continental sedimentation

1.2.2 Climatic and oceanographic characteristics

In general, Phu Quoc area is belonging to tropical monsoon climate with dry and rainy seasons annually Main climatic characteristics computed based on data collected during 1992-2003 in the islands are shown as below

Temperature: monthly air temperature is presented in table 2 For that annual average temperature is 27.4oC; hottest in May (28.70C) and coldest in January (26.10C)

Table 2 Monthly and yearly average temperature in Phu Quoc and Rach Gia (oC)

is less than 50% in dry season and around 60% in rainy season

Table 3.Monthly and yearly average and minimum humidity (%) at Phu Quoc

Meanvalue 77 77 77 80 83 85 86 86 87 86 79 74 81 Min value 40 36 39 43 46 60 61 65 60 50 44 39 36

Precipitation: Yearly rainfall at Phu Quoc is 2983 mm The rainy season starts from May to October with about 130 rainy days and total rainfall of 2397 mm, approximately 80% The dry season lasts from November to April next year with total rainfall of only 20% Maximum rainfall is in August and minimum value is in February/January Daily maximum rainfall is also in August with value of about 330mm

Table 4.Monthly and yearly average and minimum humidity (%) at Phu Quoc

Monthly

Daily max 77.9 39.3 103.2 127.1 84.1 126.8 196.5 327.1 162.1 132.9 136.0 71.0

Wind: Phu Quoc is within tropical monsoon zone which appears main wind directions such as North and northeast in winter and west and southwest in summer Calm wind just only takes 8% over the year Monthly wind velocity and main direction at Phu Quoc

is shown in table 5

Table 5 Monthly wind velocity and main direction at Phu Quoc

Storms: In general Thailand gulf including Kien Giang province has less storm vents in comparison to East sea and northern part of Viet Nam coasts As statistics in recent 60 years (1955-2015), there were fewer than 40 storms hitting to these areas in which there were only 8 storms attacked Kien Giang coast Storm season usually occurs in last months of a year Despite of less storm, but damages cause by these typhoons were very serious due to less awareness of local people The Linda storm occurred in 1997 was one example

1.2.3 Hydrological and oceanographic characteristics

a River system: due to being formed on the small catchment areas, geomorphology and

climate conditions, so river system in the Phu Quoc islands is less developed and mostly small with very short in length and steeping and later we call springs The flow in these springs only exist in rainy season with very small discharge The main springs are Duong Dong, Cua Can, Rach Tram, Cua Lap and Ham Ninh

Duong Dong spring: It is the biggest channel in Phu Quoc originated from the center of

a main island and flowing to the sea in the west coast at Duong Dong estuary This is the biggest socio – economic and tourism center of Phu Quoc In environmental point

of view, the waste water due to all activities is discharged through this channel to cause pollution seriously for the coastal and estuarine area of Duong Dong town

Cua Can spring: It is the second rank channel of Phu Quoc island, flowing from the East

to West at Cua Can estuary where saline water is intruded deeply to the land due to very small fresh water from upstream Cua Can is also a fishing port and very important economic centre of Phu Quoc District

Rach Tram is a third rank channel located at the northern part of the main island, but very small in size and short in length and covered by vegetation Fresh water exists only

in upper part in rainy season The mouth of Rach Tram is salted by sea water The density of population at this estuary is too high

Ham Ninh is a spring formed in Ham Ninh island It is also very small channel developed

in west to east direction The population in this location is high with main activities of tourism and aquaculture

b Wave climate: According to long-term meteorological data, Phu Quoc is belonging

to monsoon zone with northeast monsoon from November to April and southwest monsoon from May to October

Based on observed data of wave and analyzed, it shows that about 40% of wave are W; 30% of NE and E directions; 20% of other directions and 10% of calm wave

SW-Table 6.The statistical result of wave height and its period at Phu Quoc

The tidal regime in Thai Land Gulf is considered as mixed irregular tide with duration

of ebb and flood tide is approximately the same (11.3-12.0 hours) Tide periods is 24.3 hours, Tidal range is approximately 1.2m The highest astronomical tide is appeared from September to November and lowest astronomical tide is in April to June

Table 7.Water level at Phu Quoc (103058 E – 10013 N)station (1990-2008)

Parameters

Water level (cm)

I II III IV V VI VII VIII I X X XII XII Year Mean 103 97 91 85 80 76 80 81 86 96 103 103 90 Max 175 159 157 142 151 141 146 158 172 170 176 187 197 year 2000 2000 2005 1999 1999 1999 1999 1997 2006 1999 1999 1999 1999

year 1997 1997 1992 1992 1992 2005 1992 1992 1992 1992 1992 1992 2005

1.2.4 Social and economic features

Population: As statistical data in 2006, the population at Phu Quoc district was 86.908

people, equivalent to density of 147 people per km2 The rate of population growth in the district is about 15% in 2005 due to the migration from different inland provinces to this island The distribution of people is uneven and mostly concentrated at the western coasts at Duong Dong and An Thoi towns where the land is of 20.8%, but the population

of over 65% in comparison to the entire island

Statistically data on socio-economic in Phu Quoc district is as flowing

+ Urban population: 49.000 people

+ Under aged people: 50.102 people of about 54% of the population

+ Labor in industries: 11.934 employees, mostly working in handicraft and local and light industries

+ Labor in the service sector: nearly 8.000 mostly in tourism field such as catering hotels, motels

+ Over 6.000 employees working in agricultural and forestry

Phu Quoc has many favorable conditions for socio-economic development, especially

in eco-tourism With very famous sandy beaches and original forest covered mostly the island and very sunshine and high temperature around the year, Phu Quoc becomes very interested place for ecotourism Also marine resources, land resources, water resources

are great potential resources for economic development Fish-sauce and pepper are famous products in Phu Quoc

CHAPTER 2: APPLICATION OF DELFT3D TO STUDY STORM SURGE

2.1 Data used for the simulation

2.1.1 Statistical typhoon data

The study area is rarely affected by typhoons According to statistics, every 2 to 5 years, there is one storm to attack this area Also typhoon intensity when coming to this area

is not so strong With above two conditions, local people seem to be forgotten on typhoon They are not ready to prevent, so the typhoon is coming, it had caused the extremely damages both property and lives for Mekong provinces due to flood and storm surge both in lands and islands in the areas With above reasons, the study on typhoon and water surge is very urgent and necessary task for the areas

Figure 4 Statistical storm paths approaching to Viet Nam coasts

According to the General Statistics of Southeast Asian coasts, annually Vietnam coasts are subjected about 10 typhoons aquavelent to 33% of total number of typhoons occur

in East sea in which more than 80% hitting to northern and central coasts and only less than 20% going to southern coasts including Phu Quoc

Typhoon occuring is at the same time of rainy season, from May to October, while in the southern coasts, it occurs during last quater of the year (October to December) The typical paths of typhoon approaching to Viet Nam Coasts are illustrated in figure 4

Table 8 Statistics on typhoon hitting to Phu Quoc and surrounding areas

18 TWENTYFIVE Tropical Depression 14-14 NOV, 2012

Sources: http://www.jma.go.jp/jma/jma-eng/jma-center/rsmc-hp-pub-eg/besttrack.html

http://weather.unisys.com/hurricane/

Characteristics of typical typhoons approaching to the southern coasts including Phu Quoc islands are shown in table 9

Table 9 Characteristics of typical typhoons approaching to the southern coasts

8.7 100.9 11/25/12Z 30 1003 1004

2.1.2 Water level

Water level is used as boundary conditions of Delft3D model and observed water level

at Phu Quoc station which are used for calibration and verification of the model before simulating according to proposed scenarios

+ Water level at boundary is taken from global data sources and using one module in Delft3D to generate water level along the open sea boundaries

+ Observed water level at Phu Quoc station is used to calibration and verification the model These data will be elaborated later during setting up and simulation

2.1.3 On land and seabed topography

This data was taken from the project namely KC09.16/11-15 ‘‘Marine spatial planning Phu Quoc – Con Dao served for sustainable development” with a map scale of 1/10,000

on land and seabed nearby the cosatal trip For deep water, seabed elevation is taken from DEM 90m x 90m and general map produced by Viet Nam agencies

2.2 Model description

2.2.1 Hydrodynamic Model

WL/Delft Hydraulic Institute had developed a unique, fully integrated computer software namely Delft3D for a multi- disciplinary approach and these software package are used to compute for coastal, rive and estuary areas It can carry out simulation of flows, sediment transports, wave, water quality, morphological developments and ecology It has been designed for expert and non- experts alike The Delft3D software

is composed of several modules, grounded around a mutual interface, while being capable to interact with another Delft3D- Flow, which this manual is about, is one of these modules

Deflt3D- Flow is a multi- dimensional (2D or 3D) hydrodynamic (and transport)

result from tidal and meteorological forcing on rectilinear or curvilinear, boundary fitted grid In 3D simulation, the vertical grid is defined following the sigma co- ordinate approach

Storm surge is a long gravity wave with a length scale similar to the size of generating tropical storm, and last for several hours depending on the size and speed of movement This produces sustained elevation of the water surface above the levels caused by normal astronomical tides, however, its behavior is different in deep water and in shallow in the deep water, far from a coast, the surface wind stress by a tropical creates a rotating mound, or vortex, of water by diffusing momentum downward The ocean elevation is small, approximately the hydrostatic uplift response to the low central pressure (the inverted barometer effect) and some minor long term Coriolis effect On entering the shallow water of continental shelf, dynamic effects become pronounced, conservation

of the potential vortices of the mound requires development of marker divergence Local bathymetry reflection from the coast also contribute to substantially amplify the surge high To calculate the extreme water level a hydrodynamic model Delft 3D- FLOW for continental shelf is use

Basic equation

The hydraulic of the continent shelf in the storm conditions is simulated by solving the system of two – dimensional of shallow water equation that consists two horizontal momentum equation and on continuity equation

Conservation of momentum in x– direction (depth and density averaged)

w

g U v P

 ;

v y

C cheesy coefficient  water level above a referent level

d Bottom depth u, v depth averaged velocity

f Coriolis parameter w mass density of water

 Diffusion coefficient (eddy viscosity)

U absolute magnitude of total velocity, U (u2v2 1/2)

Cartesian co- ordinates ( , ) and spherical co -ordinates ( , ) Transform between

Cartesian and spherical co- ordinates is done with

(2-With Q representing the contributions per units area due to the discharge or withdrawal

of water, precipitation and evaporation

q in: source of water per unit volume

q out: Sink of water per unit volume

P: Non-local source term of precipitation

E: Non – local sink term due to evaporation

Momentum

The equations of Momentum in  and

direction are written as:

2

2 2

0

( ) ( )

2

2 0

( ) ( )

Delft3D calculates he vertical velocity from the continuity equation Vertical velocity