Evaluation of water stress and water quality under the impact of climate change in the upper thai binh river basin, vietnam

VIETNAM NATIONAL UNIVERSITY, HANOI VIETNAM JAPAN UNIVERSITY HTET THU SOE EVALUATION OF WATER STRESS AND WATER QUALITY UNDER THE IMPACT OF CLIMATE CHANGE IN THE UPPER THAI BINH RIVER B

VIETNAM NATIONAL UNIVERSITY, HANOI

VIETNAM JAPAN UNIVERSITY

HTET THU SOE

EVALUATION OF WATER STRESS AND WATER QUALITY UNDER THE IMPACT

OF CLIMATE CHANGE IN THE UPPER THAI BINH RIVER BASIN, VIETNAM

MASTER’S THESIS

VIETNAM NATIONAL UNIVERSITY, HANOI

VIETNAM JAPAN UNIVERSITY

HTET THU SOE

EVALUATION OF WATER STRESS AND WATER QUALITY UNDER THE IMPACT

OF CLIMATE CHANGE IN THE UPPER THAI BINH RIVER BASIN, VIETNAM

MAJOR: ENVIRONMENTAL ENGINEERING

CODE: 8520320.01

RESEARCH SUPERVISORS:

Associate Prof Dr SATO KEISUKE

Dr PHAM QUY GIANG

My appreciation also extends to Ms Pham Thi Kieu Chinh who assisted me to complete research works I could not have been completed without her kind supports I

am also grateful to Assistant Professor Dr Taishi Yazawa, for his invaluable research advices

I also would like to thank every single member of Master’s Program in Environmental Engineering for their kind supports during these two years, from enrollment to graduation

Finally, yet importantly, my gratitude goes to my parents and soulmate who always respect my decisions, and they had acted as my rock in times of troubles Their encouragement had been the crux of this research journey

This research was financially supported by JICA Research Grant Program, Ritsumeikan University and the Ministry of Education, Culture, Sports, Science and Technology-MEXT/the Japan Society for the Promotion of Science-JSPS KAKENHI Grant Program (JP 18H04153)

Thanks All!!

TABLE OF CONTENTS

CHAPTER 1: INTRODUCTION 1

1.1 General Background 1

1.2 Research Motivation 3

1.3 Target Basin 3

1.4 Problem Statements 4

1.5 Objectives 5

1.6 Thesis Structure 5

1.7 Baseline Information about the Study Basin 6

1.7.1 Hydrological Features 6

1.7.2 Topography and Administrative Boundaries 7

1.7.3 Climatic Condition 7

1.8 Summary 10

CHAPTER 2: LITERATURE REVIEW 11

2.1 Administrative Provinces 11

2.2 Hydrological Modelling 12

2.3 Climate Change 13

2.4 Water Stress Assessment 15

2.5 Water Quality Evaluation 17

2.6 Summary 20

CHAPTER 3: HYDROLOGICAL SIMULATION 21

3.1 SWAT Hydrological Model 21

3.1.1 Preparation of In-put Data for Model Set-up 23

3.2 SWAT Model Set-up 29

3.2.1 Watershed Delineation 29

3.2.2 HRU Analysis 29

3.2.3 Integration with Weather Database 29

3.3 SWAT Model Calibration and Validation 30

3.4 Model Performance Evaluation 30

3.5 Results of Simulation 31

3.5.1 SWAT Model Calibration and Validation Result 31

3.6 Summary 35

CHAPTER 4: CLIMATE PROJECTION 36

4.1 Future Climate Scenario 36

4.2 Performance Analysis of Bias Correction Method 38

4.3 Results of Climate Projection 39

4.3.2 Projected Precipitation Data 39

4.3.3 Projected Maximum and Minimum Temperature 41

4.4 Summary 45

CHAPTER 5: WATER STRESS ASSESSMENT 46

5.1 Water Demand 46

5.2 Water Resources 46

5.3 Water Stress 46

5.4 Future Water Stress 47

5.5 Result of Water Stress Assessment 48

5.5.1 Current Water Demand 48

5.5.2 Current Water Resource Potential 49

5.5.3 Current Water Stress 49

5.5.4 Future Water Demand 50

5.5.5 Future Water Resource Potential 51

5.5.6 Future Water Stress 51

5.6 Summary 53

CHAPTER 6: WATER QUALITY EVALUATION 54

6.1 Water Quality Parameters 55

6.2 Vietnamese National Water Quality Index (VN_WQI) 56

6.2.1 Calculating WQI in this Study 56

6.3 Data Analysis 57

6.4 Result of Water Quality Assessment 58

6.4.1 Water Quality Results 58

6.4.2 Results of Statistical Analysis 62

6.4.3 Results of Water Quality Index 85

6.5 Future Water Quality Status under the Impact of Climate Change 88

6.6 Summary 89

CHAPTER 7: CONCLUSION, LMITATATIONS AND FUTURE TREND 90

REFERENCES 92

APPENDICES 99

Appendix 1 Result of Population Projection 99

Appendix 2.A Detailed Statistics About Current Water Stress 100

Appendix 2.B Detailed Statistics About Future Water Stress 101

Appendix 3 Photo Records of River Water Sampling Point 102

Appendix 4 Scenes During River Water Sampling 105

Appendix 5 List of Survey Team Member 105

Appendix 6 Analytical Methods 106

Appendix 7 Detailed WQI Calculation Method 110

LIST OF TABLES

Table 2.1 Variation of Precipitation (%) during 1958 – 2014 14

Table 3.1 SWAT Land Use Code and Statistics (2015) 26

Table 3.2 Summary for Preparation of Model Input-Data 28

Table 3.3 Model Performance Rating 31

Table 3.4 Goodness-of-fit Statistics for Discharge Simulation 32

Table 3.5 Calibrated Parameters and Fitted Values 32

Table 3.6 Average Monthly Hydrological Components 34

Table 3.7 Annual Water Balance Statistics 34

Table 4.1 Description of RCM 36

Table 4.2 Period of Study and Projected Climatic Variables 36

Table 4.3 Performance Evaluation Results 39

Table 4.4 Changes in Seasonal Precipitation 40

Table 4.5 Changes in Seasonal Maximum Temperature 42

Table 4.6 Changes in Seasonal Minimum Temperature 42

Table 4.7 Changes in Long Term Annual Temperature 43

Table 6.1 Location of River Water Sampling Points 54

Table 6.2 Summary of Water Quality Parameters 56

Table 6.3 VN_WQI Based Classification for Surface Water Quality 57

Table 6.4 In-situ Water Quality Result of Cau River Sub-basin 58

Table 6.5 Ex-situ Water Quality Result of Cau River Sub-basin 59

Table 6.6 In-situ Water Quality Result of Luc Nam River Sub-basin 60

Table 6.7 Ex-situ Water Quality Result of Luc Nam River Sub-basin 60

Table 6.8 In-situ Water Quality Result of Thuong River Sub-basin 61

Table 6.9 Ex-situ Water Quality Result of Thuong River Sub-basin 61

Table 6.10 Mean Values of Water Quality Parameters observed in CRSB 62

Table 6.11 Mean Values of Water Quality Parameters observed in LNRSB 63

Table 6.12 Mean Values of Water Quality Parameters observed in TRSB 63

Table 6.13 Correlation Matrix 66

Table 6.14 Total Variance Explained for Wet Season of CRSB 67

Table 6.15 PCA Loadings (CRSB-Wet Season) 68

Table 6.16 Total Variance Explained for Dry Season of CRSB 70

Table 6.17 PCA Loadings (CRSB-Dry Season) 71

Table 6.18 Total Variance Explained for Wet Season of LNRSB 73

Table 6.19 PCA Loadings (LNSB-Wet Season) 74

Table 6.20 Total Variance Explained for Dry Season of LNRSB 76

Table 6.21 PCA Loadings (LNSB-Dry Season) 77

ii

Table 6.23 PCA Loadings (TRSB-Wet Season) 79 Table 6.24 Total Variance Explained for Dry Season of TRSB 81 Table 6.25 PCA Loadings (TRSB-Dry Season) 82

LIST OF FIGURES

Figure 1.1 Location of Target Basin (Source: TA-7629, VIE, 2012) 4

Figure 1.2 Long Term Annual Maximum Temperature Trend 8

Figure 1.3 Long Term Annual Minimum Temperature Trend 8

Figure 1.4 Long Term Monthly Precipitation Trend 9

Figure 1.5 Average Monthly Discharge at the Gia Bay Hydrological Station (2005-2019) 9

Figure 1.6 Work Flow of the Research 10

Figure 2.1 Annual Changes of Temperature (Ngu et al., 2016) 14

Figure 2.2 Water Stress Levels in Vietnam (2030WRG, 2017) 17

Figure 2.3 River Water Quality in Vietnam (2030WRG, 2017) 20

Figure 3.1 DEM of the UPTBRB 23

Figure 3.2 High Resolution Land Use and Land Cover Map of the UPTBRB (2015) 25

Figure 3.3 Major Land Cover Statistics in the UPTBRB 25

Figure 3.4 Soil Map of the UPTBRB 26

Figure 3.5 The Proportion of Soil Classes observed in the UPTBRB 27

Figure 3.6 Location of Hydro-meteorological Stations 28

Figure 3.7 Formation of HRUs 29

Figure 3.8 Calibration Result at the Gia Bay Hydrological Station 33

Figure 3.9 Validation Result at the Gia Bay Hydrological Station 33

Figure 3.10 Annual Water Balance of the UPTBRB (2008 – 2019) 35

Figure 4.1 Location of Projected Meteorological Stations 38

Figure 4.2 Average Monthly Precipitation Trends (5-stations average) 41

Figure 4.3 Annual Changes of Precipitation Trends (5-stations average) 41

Figure 4.4 Average Monthly Changes of Maximum Temperature (5-stations average) 43

Figure 4.5 Average Monthly Changes of Minimum Temperature (5-stations average) 44

Figure 4.6 Annual Changes of Maximum Temperature (5-stations average) 44

Figure 4.7 Annual Changes of Minimum Temperature (5-stations average) 45

Figure 5.1 Proportion of Sectoral Water Demand 49

Figure 5.2 Distribution of Current Water Stress Levels 50

Figure 5.3 Distribution of Predicted Water Stress Levels 52

Figure 5.4 Comparison of Current and Future Water Statistics (Annual) 52

Figure 5.5 Changes in Water Demand by Each Sector 53

Figure 6.1 Location Map of Sampling Points 55

iv

Figure 6.3 Bi-plot Illustration for PC1 and PC2 in CRSB (Wet Season) 69

Figure 6.4 Bi-plot Illustration for PC2 and PC3 in CRSB (Wet Season) 69

Figure 6.5 Bi-plot Illustration for PC1 and PC3 in CRSB (Wet Season) 70

Figure 6.6 Eigen Values and Proportion of Variances (CRSB-Dry Season) 71

Figure 6.7 Bi-plot Illustration for PC1 and PC2 in CRSB (Dry Season) 72

Figure 6.8 Bi-plot Illustration for PC2 and PC3 in CRSB (Dry Season) 72

Figure 6.9 Bi-plot Illustration for PC1 and PC3 in CRSB (Dry Season) 73

Figure 6.10 Eigen Values and Proportion of Variances (LNSB-Wet Season) 74

Figure 6.11 Bi-plot Illustration for PC1 and PC2 in LNRSB (Wet Season) 75

Figure 6.12 Bi-plot Illustration for PC2 and PC3 in LNRSB (Wet Season) 75

Figure 6.13 Bi-plot Illustration for PC1 and PC3 in LNRSB (Wet Season) 76

Figure 6.14 Eigen Values and Proportion of Variances for LNRSB (Dry Season) 76

Figure 6.15 Bi-plot Illustration for PC1 and PC2 in LNRSB (Dry Season) 77

Figure 6.16 Eigen Values and Proportion of Variances (TRSB-Wet Season) 78

Figure 6.17 Bi-plot Illustration for PC1 and PC2 in TRSB (Wet Season) 79

Figure 6.18 Bi-plot Illustration for PC2 and PC3 in TRSB (Wet Season) 80

Figure 6.19 Bi-plot Illustration for PC1 and PC3 in TRSB (Wet Season) 80

Figure 6.20 Eigen Values and Proportion of Variances (TRSB-Dry Season) 81

Figure 6.21 Bi-plot Illustration for PC1 and PC2 in TRSB (Dry Season) 82

Figure 6.22 Bi-plot Illustration for PC2 and PC3 in TRSB (Dry Season) 83

Figure 6.23 Bi-plot Illustration for PC1 and PC3 in TRSB (Dry Season) 83

Figure 6.24 Cluster Dendrogram for Wet Season in the UPTBRB 84

Figure 6.25 Cluster Dendrogram for Dry Season in UPTBRB 85

Figure 6.26 Seasonal Variation of Observed WQI in the CRSB 86

Figure 6.27 Seasonal Variation of Observed WQI in the LNRSB 87

Figure 6.28 Seasonal Variation of Observed WQI in the TRSB 88

LIST OF ABBREVIATIONS

CEM : Center for Environmental Monitoring

CMIP6 : Coupled Model Inter-comparison Project Phase 6

CRSB : Cau River Sub-basin

DEM : Digital Elevation Model

LNRSB : Luc Nam River Sub-basin

TRSB : Thuong River Sub-basin

PCA : Principal Component Analysis

RCP 4.5 : Representative Concentration Pathway under 4.5 Scenario

RCM : Regional Climate Model

SWAT : Soil and Water Assessment Tool

UPTBRB : Upper Thai Binh River Basin

VEA : Vietnam Environmental Administration

WQI : Water Quality Index

1

CHAPTER 1: INTRODUCTION 1.1 General Background

According to the Burmese proverb, “Rice offers survival for seven days; however, water gives life only a day” Not only human, none of a single living thing cannot survive without water Although water is a renewable resource, sound management and adequate exploitation strategy are essential to ensure a sustainable water supply In the United Nation’s Sustainable Development Goal No (6), target indicator (6.4) clearly highlighted the critical role of water regarding the efficient use of water and water scarcity (UNHCR, 2017)

Vietnam is rich in water resources with dense river networks accompanied by

2360 rivers The river basins in Vietnam can be grouped into three main categories based

on the basin area and share of administrative boundary; multiple provinces, two provinces and a single province (Taylor & Wright, 2001) There are 9 major basins namely Red River, Thai Binh, Bang Giang – Ky Cung, Ma, Ca La, Thu Bon, Ba, Dong Nai and Mekong that all shared together about 80% of the total country area (ADB, 2009) The average annual surface water runoff is approximately 830 – 840 bn m3 in which only 43% is available for sustainable exploitation (2030WRG, 2017) Groundwater potential accounts for 63 bn m3 but only 7% can be sustainably exploited (FAO AQUSTAT, 2011) However, the sustainability of water resource in Vietnam is greatly associated with the transboundary water management as 63% of total water resources was originated in outside of the country; China, Cambodia, Lao PDR and Thailand (MONRE, 2006 & 2030WRG, 2017)

Mekong and Red-Thai Binh are well known basins in Southern and Northern Vietnam since they together contribute about 42% of the country GDP In order to sustain country GDP growth, the long term effective water resource management strategies for these basins are crucial due to having their high trans-boundary water dependency; 95% in Mekong and 40% in Red-Thai Binh Basin (2030WRG, 2017)

Vietnam is famous for agricultural business and that contributes 18% of the country GDP but consumes about 80% of the total water share One third of the country area was occupied by agriculture in which 20.6% is available for farming, 12% is identified as permanent crops and the permanent pasture mounts to 2.1% (2030WRG, 2017) Main crops are paddy, maize, coffee and sugarcane Mekong delta, Red River delta and Northern highland regions are serving as the rice bowls of Vietnam According

to agricultural master plan (2020), the paddy fields will be maintained to not more than 3.8 million ha but the production is targeted to increase 41 to 43 million ton/year in 2020 and 44 ton/year by 2030 (MPI, 2016 & 2030WRG, 2017) It is highlighting that large amount of water resources are required to meet the current and target production line

Annual water demand in Vietnam is increasing year by year; 80.2 bn m3/year in

2009, 80.6 bn m3/year in 2015 and projected to reach 95 bn m3/year by 2030 (ADB,

2009 & IWRP, 2015) Water demand for agricultural sector is 76 bn m3 in 2016 and expected to increase 91 bn m3 by 2030 Aquaculture demand is 10 bn m3/year in 2016 and forecasted to reach 12 bn m3/year in 2030 Annual industrial water demand is 6 bn

m3/year in 2016 and expected to increase 15.6 bn m3/year by 2030 Annual municipal water requirement is 3.1 bn m3/year in 2016 and forecasted to increase 5.7 bn m3/year Hydropower sector demand is 57 bn m3/year in 2016 and projected to increase 63 bn

m3/year (2030WRG, 2017)

Wastewater treatment system is still under development in Vietnam To elaborate, current treatment system has the capacity to treat only 12-13% of municipal wastewater, and 10% of industrial wastewater (Tien, 2015 & 2030WRG, 2017) Consequently, three basins have been identified as the most polluted river basins; Cau River Basin, Nhue-Day River Basin and Dong Nai River Basin (MONRE, 2006) In 2015, with the purpose

to enforce the regulations on water resource management and quality control, “the Ministry of Natural Resources and Environment (MONRE)” amended the technical guidelines for surface water quality (QCVN 08: 2008/BTNMT) Currently, the MONRE has been conducting regular water quality monitoring activity for Cau River Basin four times a year and the results are disclosed to the public through the index score (Tran et

3

Climate change is strongly linked with the watershed hydrology By 2030, it was estimated that the future annual runoff is expected to reach 15 bn m3/year where higher runoff is predicted to increase 25 bn m3/year in wet season and lesser runoff in dry season by 10 bn m3/year (2030WRG, 2017) It is showing the changes of rainfall pattern

on seasonal water availability The major climatic variables such as precipitation and temperature are greatly influencing the water resource sustainability

Water demand and resources should be widely discussed considering potential population growth, industry and agricultural aspects Integrated approaches would be more effective towards quantitative and qualitative development of water resources It

is crucial to manage the existing water resources effectively to ensure safe and reliable water supply required for multi-sectoral development

1.2 Research Motivation

Red-Thai Binh plays a critical role in socio-economic development of Vietnam

as its contribution to the country GDP is approximately 15% The basin possesses a strategic geographical location for doing business ranging from high land mountain in the North to alluvial plain land in North-East of the country About agriculture, 15% of total rice production comes from this basin serving as a promising resource for ensuring national food security and foreign income Numerous industrial clusters, craft villages

(i.e., 65% of the whole country) and intensive agricultural fields are concentrated in the

Red Thai Binh Basin (2030WRG, 2017) For the sake of people life and water resource sustainability, the call for a comprehensive water resource exploitation plan was made

to the scientists and experts by Mr Tran Hong Ha, the Minister for the Ministry of Natural Resources and Environment (MONRE, 2016)

1.3 Target Basin

Target basin of the study is Upper Thai Binh River Basin (UPTBRB) located in the Northern Vietnam The UPTBRB is a component of Red-Thai Binh River Basin that occupied almost the entire area of North East of Vietnam (See figure 1.1) Being a transboundary river basin, the catchment area of Red-Thai Binh River Basin is distributing in three countries; China (48%), Laos PDR (1%) and Vietnam (51%) The basin provides 17% of total annual surface water runoff in Vietnam following after

Mekong River Basin with 59% thus play an essential role to provide the sufficient water resources for the capital Hanoi (2030WRG, 2017)

Figure 1.1 Location of Target Basin (Source: TA-7629, VIE, 2012)

1.4 Problem Statements

The UPTBRB is having at high risk of water stress due to high water demand from various sectors According to the 2030 Water Resources Group, the overall water demand in the Red-Thai Binh Basin is projected to increase 42% by 2030 Regarding the water exploitation index, Red-Thai Binh Basin had been fallen under low water stress category in 2016 and is expected to reach water stressed category by 2030 The competition for a regular access to a steady supply of water share will be more significant in the near future

The UPTBRB is also highly vulnerable to the impact of climate change The projection statistics are highlighting that the average surface temperature under RCP 4.5 scenarios would increase 1.9 to 2.4°C by 2100 Similarly, the precipitation is expected

to increase 5 – 15% The intensity and pattern of precipitation is expected to alter

5

On the other hand, Cau river had long been identified as the most polluted watershed in Northern Vietnam (Tran et al., 2017) Water quality evaluation in the Cau River still remains challenges due to spatial and seasonal variation Basin-wide water quality assessment should be carried out in order to identify the pollution status especially between the main rivers and tributaries Likewise, the lack of water quality database in the remaining two basins; Thuong and Luc Nam River Sub-basin still pose the significant challenges for effective water resource management

1.5 Objectives

The main objectives of this study are as below;

- To evaluate basin water stress through integration of hydrological model and social statistics

- To evaluate basin river quality by multivariate statistical analysis and the Vietnamese Water Quality Index

- To contribute integrated water resource management through quantitative and qualitative approaches

1.6 Thesis Structure

This thesis was prepared in a structure in compliance with the instructions for preparation of Master’s thesis guidelines issued by Vietnam Japan University General research information with the corresponding chapters were summarized below;

 Chapter 1 is the general background about Vietnam relating to the status of water resources It briefs the current water demand by each sector including the efforts

of related bodies in water resource management Significant water related challenges are identified, and research objectives are proposed based on motivation Baseline natural and hydro-climatic information of the target basin

is also provided

 Chapter 2 is mainly about literature review of research scope and content It was designed to provide the review of previous researches and important highlights

in the basin The comparative studies which have a common research approaches

in other basins are widely discussed for referencing purposes

 Chapter 3 describes about hydrological modelling in the basin Detailed step by step method to construct the watershed model, technical considerations and results are provided

 Chapter 4 summarizes climate projection part to predict the future climate condition represented by regional climate model Analysis and discussion are made on the different time frames of historical, current and future scenario in order to provide a comprehensive view of study on climate change in the basin

 Chapter 5 discusses water stress through estimation of sectoral water demand and resources This session applied the outcomes of chapter 3 and 4 to determine current and future water stress through integration of multi-assessment methods

 Chapter 6 provides river water quality evaluation Seasonal river water sampling and laboratory analysis are performed The results are driven by multivariate statistical analysis and the Vietnamese regulations

 Chapter 7 is a conclusion and recommendation session in which all the key findings and necessary solutions are provided including limitations and future research direction

1.7 Baseline Information about the Study Basin

1.7.1 Hydrological Features

Upper Thai Binh River Basin (UPTBRB) is composed of three major rivers; Cau River, Thuong River and Luc Nam River thus forming three sub-basins together The basin has a total number of 7 tributaries namely Cho Chu, Thuong Nghinh, Du, Cong and Calo rivers in Cau River Sub-basin (CRSB), while Luc Nam River Sub-basin (LNRSB) has one upstream tributary, Dinh Dem river and Thuong River Sub-basin (TRSB) also has the only Rang tributary located in the upstream region Cau River is the longest river with 288 km and originated at Bac Kan Province The length of Thuong River is 157 km that is flowing from Lang Son Province to Bac Giang Province The length of Luc Nam River is 200 km and is serving as a blood line of Bac Giang Province All the three main rivers flow together into Thai-Binh River at the Pha Lai The UPTBRB is thus rich in river network and forming as a complex river basin that lies

7

1.7.2 Topography and Administrative Boundaries

Geographic location of the basin extends from 21°48'54.81"N to 106°27'44.98"E The basin has an area of 12,720 km2, equivalent to about 4% of the total country area The upper part of the basin is surrounded by dense high mountain ranges extending North-East direction The highest recorded elevation is 1,592 m (Tam Dao Mountain) while the lowest is 3.8 m at Pha Lai where Thai Binh river is formed The low alluvial plain land occupied the central and downstream for the basin that favors good conditions for cultivation and settlement The basin covers many administrative provinces such as Bac Kan, Thai Nguyen, Hanoi, Vin Phuc, Bac Ninh, Lang Son and Bac Giang As of

2015, the basin population is estimated at approximately 4.95 million The average basin population density is 389 persons per km2 that is higher than the country’s 280 person per km2

1.7.3 Climatic Condition

A tropical monsoon climate mainly dominated in the basin The recorded current long term annual maximum temperature is about 28 °C and 21°C is the minimum thus the gap is about 6°C (See figure 1.2 & figure 1.3) The annual long term mean precipitation account for approximately 1550 mm (See figure 1.4) There are four seasons and each season lasts three months; winter (December to February), Spring (March to May), Summer (June to August) and autumn (September to November) (MONRE, 2006) The heaviest precipitation is mostly received in July and August with the amount over 300 mm/month (Thai et al., 2017) During 2005 – 2019, about 67%, two-third of total runoff, occurred within June to September of the rainy season (See figure 1.5)

Figure 1.2 Long Term Annual Maximum Temperature Trend

Figure 1.3 Long Term Annual Minimum Temperature Trend

9

Figure 1.4 Long Term Monthly Precipitation Trend

    Figure 1.5 Average Monthly Discharge at the Gia Bay Hydrological Station

1.8 Summary

It was clearly observed that the annual increase of water demand while water resource availability is highly dependent on seasonal variation and also associated with the impact of climate change Moreover, limited water treatment facilities are threatening river water quality along with the socio-economic growth It is important to evaluate the water resource in terms of quantitative and qualitative perspectives to let the water stakeholders adopt the necessary practices and measures The following figure 1.6 was designed to provide the corresponding approaches to achieve the specific objectives

Figure 1.6 Work Flow of the Research

in all year round, more than 2294 mm and the average annual temperature is 20.6 °C (Child Fund, 2013)

Thai Nguyen

Thai Nguyen province is located in the center of North East Region The province area is 3526.64 km2 that exists together with Bac Kan Province in the North In 2019, it has a population of about 1.29 million while population density is 365 persons per km2 The average temperature is 25 °C and the recorded average precipitation varies 2000 to

2500 mm In 2010, industrial clusters were rapidly established thus the main economy

of the province had shifted to industry-construction from agriculture (Thai Nguyen Overview, 2015)

Vinh Phuc

Vinh Phuc Province has the area of 1237.52 km2 As of 2013, the average population is 1.03 million and population density is 832 persons per km2 The average temperature is 24 to 25 °C while the average precipitation is 1650 mm Agriculture is the major business that shared about 70% of the total land area (Vinh Phu Statistics Office, 2013)

Hanoi

Hanoi province is bordering the downstream of Cau River in the South of the basin The city is also located on the mouth of Red River and Da River Being a metropolitan city, the province has high population volume, 4.7 million, and 14639 per persons km2 of population density over the total area of 319.56 km2 The average

temperature is 29.2 °C and the average precipitation is 1800 mm (Hanoi Geography, 2014)

Bac Ninh

Bac Ninh is located just 30 km North-East of the Capital Hanoi The province having 823 km2 can be recognized as the smallest province in Vietnam The total population is about 1.54 million in 2015 with 1676 persons per km2 population density The average recorded temperature is 23.6 °C and the average precipitation is 1926 mm Bac Ninh is famous for industrial hub especially foreign investment and trade (BAC NINH Province - A Hi-Tech Development Hotspot, 2017)

Lang Son

Lang Son is a mountainous province that borders Quang Ninh Province in the North and Bac Giang Province in the South, and Bac Kan Province in the West The province area is 8310 km2 with 831887 population (i.e., 102 population density) in 2009 Forest and agriculture are the dominant land covers The province is dominated by tropical monsoon climate The average recorded temperature varies 17 to 22 °C while the average precipitation is approximately 1200 – 1600 mm (Lang Son Statistics Office, 2021)

Bac Giang

Bac Giang Province has 3895.59 km2 that borders Lang Son Province in the North and Bac Ninh Province in the West As of 2019, the total population is about 1.8 million with a population density of 483 persons per km2 It composed of 72% of the mountainous districts The average annual temperature is 23.3°C and the average annual rainfall is 1915 mm The industry-construction sector contributes the largest share of the province GDP followed by agricultural sector (Wu A., 2021)

2.2 Hydrological Modelling

Allaby & Allaby (1999) defined hydrological modelling as the process of characterization of real hydrologic features and system with the integration of physical models, mathematical equations and computer simulations The model helps understand

Soil and Water Assessment Tool (SWAT) is one of the popular models to study watershed hydrology in relation to climate change SWAT has built a number of proven hydrological researches, up to 36, in Vietnam (Tan et al., 2019) Tran et al (2017) simulated SWAT model whereas rainfall had a significant relationship with daily total nitrogen load in the Cau River Basin Thai et al (2017) applied SWAT model in the Cau watershed to study erosion and stream flow under the impact of climate change whereas different pattern of soil loss and increased river discharge were predicted for the future Bui et al (2019) used integrated SWAT and QUAL2K for water quality modelling to overcome data scare challenges in the Cau watershed Chuong et al (2014) evaluated water quality by SWAT model in Ta Trach watershed in central Vietnam S Shrestha et

al (2018) coupled SWAT and climate model in the Songkhram River Basin, Thailand

to study climate and land use impact on hydrology and water quality It was found that the stream flow was associated with future climate and nitrate-nitrogen had a strong relationship with land cover S Shrestha et al (2016) investigated the water resource potential under different time frames in the Indrawati River Basin, Nepal This study discovered that river discharge pattern and intensity was varied consistently with future climate scenarios

– 2014 In the North-East region, the increased days of maximum temperature greater

than 35 °C is observed during 1960 – 2014 (Ngu et al., 2016) The annual changes of

temperature were provided in figure 2.1. 

Figure 2.1 Annual Changes of Temperature (Ngu et al., 2016)

The slightly increased amount of precipitation was received in the country scale

but a decline annual precipitation ranging from 5.8 to 12.5 % in 57 years (1958 – 2014)

was observed in Northern Vietnam But in South central and Southern regions, the

precipitation was increased to 19.8 from 6.9 % and 18.8% respectively Regarding

seasonal variation, in table 2.1, a decreased pattern of precipitation was observed in the

autumn and the significant rainfall occurred during the spring in the North while winter

and spring tend to receive increased precipitation in Southern Vietnam (Ngu et al., 2016)  

Table 2.1 Variation of Precipitation (%) during 1958 – 2014

Climatic Regions

Spring (MAR – MAY)

Summer (JUN – AUG)

Autumn (SEP – NOV)

Winter (JAN – FEB)

15

Climate change and future projection for Vietnam was developed using five regional climate models; Weather Research and Forecast for climate projection (clWRF), Providing Regional Climates for Impact Studies (PRECIS), Conformal Cubic Atmospheric Model (CCAM), Regional Climate Model (RegCM), and Atmospheric General Circulation Model/Meteorological Research Institute (AGCM/MRI) Long term baseline period (1986 – 2005) was considered to forecast future climate change of different centuries; the early term (2016 - 2035), middle term (2046 – 2065), and (2080 – 2099) as the ending period of 21st century (2080 – 2099) Overall, it was predicted that under the RCP 4.5 scenario, the surface temperature tends to increase 1.9 – 2.4 °C and precipitation would increase over 20% in the North (Ngu et al., 2016) In the mid-century under the RCP 4.5 scenario, the annual temperature would increase 1.6 – 1.7 °C

in which 1.2 – 1.6 °C in winter season, 1.3 – 1.6 °C in spring season, 1.6 – 2.0 °C in summer, 1.6 – 1.9 °C in autumn The average annual maximum and minimum temperature tend to increase 1.7 – 2.7 °C and 1.4 – 1.6 °C The winter rainfall tends to decrease 10% in maximum, spring precipitation likely to increase 10%, summer rainfall

is expected to increase 5 - 15% and the autumn rainfall is likely to increase 15 – 35%

In addition to these verifications, the climate change-based impact projections indicated in the IPCC AR5 are the latest challenge The current situation in the basin, where the area is being developed for not only as the agricultural area but also as a modernized industrial zone, makes it imperative to propose a sustainable water resources policy and strategy

2.4 Water Stress Assessment

There are a variety of methods to quantify the vulnerability of the basin in terms

of water stress These are Falkenmark indicator, the Green-Blue Water Scarcity Index, Water Stress Index, Smakhin Water Supply Stress Index, Stream flow-based index, Water Scarcity Index, Water Supply Sustainability Risk Index, Watershed Sustainability Index, Green Water Stress Index and Green Water Scarcity Index Detailed technical reviews were clearly described in the technical report of Xu and Hu (2017)

Water stress refers to those who do not have access to reliable and sufficient water supply to meet their daily routine Water scarcity can be defined as the region in which

a group of people is undergoing lack of sufficient water supply for a period of time (Rijsberman et al., 2006) The index based on ratio of withdrawal to available water resource is popular with the indicator range of low (<0.1), moderate (0.1 – 0.2), medium (0.2 – 0.4) and high-water stress (>0.4) (Vorosmarty et al., (2005) & Xu and Hu (2017))

According to IPCC AR4 projection, there is an increase in global average temperature but a reduction in precipitation thus would alter water supply to consumption ratio (Solomon et al., 2007) Bromand (2015) predicted water stress in Kabul River Basin, Afghanistan, through a coupled SWAT model and climate models Koiso et al (2019) evaluated quantitative water supply and demand balance using HSPF (Hydrological Simulation Program – Fortran) Model in Mekong River Basin under RCP 8.5 scenario whereas the changes in water resource by climate change would affect water stress intensity Yamamua et al (2018) developed water resource assessment tool

in global-scale dimensions This study was conducted in Mekong River Basin and Murray-Darling Basin and stated that downstream regions of Mekong River Basin was

at high risk of water stress stating that more than 80% of Murray-Darling Basin area already started high water stress since 2020s The resulted indicators showed the level

of water stress will be more significant in the future Sun et al (2008) predicted water stress through water supply stress index and water scarcity stress index ratio in the Southeastern United States It was discovered that increased population had risen the intensity of water stress while changes in future land use had little impact on water resource availability

In Vietnam, general water stress assessment was carried out at 16 major river basins by 2030 Water Resource Group using water exploitation index (WEI) in 2016 WEI is the ratio of annual withdrawal of freshwater to long-term water resource potential WEI indicates the pressure on water resource availability regarding the abstraction of municipal, agriculture and industrial sectors with the threshold values 0

to 100% (i.e., no stress is less than 10%, low stress is 10 to 20%, stressed is 20 to 40%

and severe stress is recognized if the value is greater than 40%) In figure 2.2, dry season WEI of Red-Thai Binh showed low stress (19%) in 2016 and tend to reach stressed level

17

Figure 2.2 Water Stress Levels in Vietnam (2030WRG, 2017)

2.5 Water Quality Evaluation

Generally, Vietnam can be divided into seven major regions; Northwest, Northeast, Red River delta, North Central Coast, South Central Coast, Central Highlands, North East of Mekong and Mekong River Delta Basically, the upstream water quality

is reportedly good, but downstream water quality is deteriorated after flowing through residential areas, industrial clusters, craft villages and agricultural areas due to the direct discharge of untreated wastewater and improper utilization of pesticides and fertilizers

As a result, downstream water quality in Northeast region, Red River Delta, South Central Coast and Northeast of Mekong region were deteriorated Distribution of seven regions, the status of water quality and the associated pollution sources were provided

in figure 2.3 (2030WRG, 2017)

The Water Quality Index (WQI) is an algebraic expression thereby interpreting multiple pollution variables into a single score The subjective weight corresponding to each parameter is assigned, the important level is evaluated and give out a dimensional number (Garcia, 2017) The WQI is being used in both developed and developing countries to support decision makers through raising public awareness and participation

Being effective as a tool in assisting sound water resource management, the application

of WQI was widely used in India, Chile, England, Wales, Taiwan, Australia, Malaysia etc (Phu, 2019) Vietnam adopted the first ever WQI application in 2011 and improved version was disclosed to the public and applied since 2019

In November 12, 2019, with the purpose to encourage the effort in the evaluation

of surface water quality, the Vietnamese Government issued a new Vietnamese National Water Quality Index (VN_WQI) calculation method (Decision No 1460 / QD-TCMT 2019) This new VN_WQI is an improved version of the former WQI stipulated on July

1, 2011 (Decision No.879/QD-TCMT)

In the updated VN_WQI, the involvement of subjective purpose (i.e., either

pesticide or heavy metal module) was allowed to integrate in the calculation method Therefore, the new VN_WQI involves diverse water quality parameters than the former WQI (2011) that was based on only 9 parameters; WT, pH, BOD5, COD, turbidity, TSS, NH4+ - N, PO4- - P and coliform (Pham et al., 2017 & Son et al., 2020) But physical parameters like turbidity and total suspended solid were no longer involved in the new VN_WQI It was noteworthy that the new VN_WQI brought more detailed classification than the previous method of calculation with six range of classification while the former had only five categories However, the index value remains the same, ranging from 0 to 100 Generally, it can be easily understood that the higher WQI indicates the better water quality

The applicability of WQI is widely popular in Vietnam Son et al (2020) examined water quality in Cau river using the combined method of water quality index, comprehensive pollution index, organic pollution index, and trace metal pollution index

It was found that seasonal and spatial distribution greatly affected river water quality whereas organic pollutions were observed in many places showing higher risk of eutrophication especially in downstream regions Phu (2019) used WQI to assess the status of water quality in Luy River in Binh Thuan Province This study revealed the effectiveness of WQI in investigating seasonal river water quality with regard to spatial distribution of watershed Hong (2018) performed the WQI based assessment on Tien

19

River quality in Southern Vietnam to find out long term annual and seasonal changes of water quality

Statistical tools and analysis are effective to treat water quality data Zeinalzadeh

& Rezaei (2017) applied principal component analysis (PCA) and WQI (National Sanitation Foundation) in Shahr Chai agricultural watershed in Iran in which declined stream flow, pollution load and agricultural discharge were associated with the health

of downstream water quality Yamazaki et al (2017) examined the river water quality

of Rekifune River and Satsunai River watersheds in Japan with multivariate statistical analysis of PCA and cluster analysis whereas the significant pollution load from the tributaries under different landscape poses a strong influence on the major river Ibrahim

& Ali (2016) used the combined method of PCA, Cluster Analysis and WQI to evaluate water quality of Tigris River in Iraq in which the point sources, spatial similarity of pollutant characteristics and necessary purpose of treatments were discovered Mishra (2010) used PCA to extract multiple river water quality parameters of the River Ganges

in India into the important parameters and the discharge sources Gajbhiye et al (2015) examined surface water quality in Jabalpur City in India and stated that PCA was supportive to identify the main variability in water quality variance Tripathi & Singal (2019) also applied PCA method to select the most important water quality parameters for WQI calculation in the River Gang of India Kamble & Vijay (2011) analyzed the water quality in coastal region of Mumbai in India whereas the seasonal water quality was identified with different pollution levels by cluster analysis Garcia (2017) used a variety of WQIs and PCA to evaluate nutrient and organic loads in Acude Macela Macela Reservoir in Brazil

It was found that all the applied WQI studies in the study basin were conducted with the previous version of calculation method The new version involved integration

of diverse water quality parameters which is more suitable to assess water quality status Moreover, multivariate statistical analysis such as PCA and CA are proved to be effective to provide the status of river water quality in terms of spatial and temporal variations among three sub-basins

Figure 2.3 River Water Quality in Vietnam (2030WRG, 2017)

2.6 Summary

There is no research related to water stress assessment assisted by hydrological model in the UPTBRB while the majority of the previous studies were focused on stream

flow, erosion and others The use of spatial and time series data (i.e., far from current

situation) still present the challenges to reflect recent changes of watershed hydrology and climate change This is necessary to evaluate the current status of Cau River water quality and also to establish water quality database for Thuong and Luc Nam River Basins The application of the updated WQI and multivariate statistical analysis would bring a wider insight into the evaluation of basin water quality

Firstly, the model divides the targeted basin into multiple sub-basins, and hydrological response units (HRUs) are further generated based on the similar geomorphological and hydrological features Therefore, the model requires a number of input data; topography, land cover, soil layer and climatic data of the basin And then SWAT performs simulation of physical process; runoff, nutrient and sediment transport under continuous time In addition, SWAT showed a profound tool to model even in the data scare basin and also to assess the impact on water resources under variables of interest; land use, climate change etc This study used Arc interface of SWAT 2012 that served as an extension in ArcGIS

In SWAT, the water balance of hydrological cycle is calculated using the following equation 3.1 with regard to the hydrological cycle in the basin

SW t = SW 0 + ∑ R day – Q uaff – E a – w sweep – Q gw ) (3.1) Where SW t is the final calculated soil water content (mm H2O) in a given time (days)

SW 0 is the initial amount (mm H2O) of soil water content, t is the time (days),

R day indicates the amount of precipitation on day i (mm H2O) Qsuf is the surface runoff amount on day i, Ea the amount of evapotranspiration on day i (mm H2O), wsweep is the amount of water percolating through the soil profile on day i (mm H2O) and Qgw is the amount of return flow on day i

The surface runoff of the basin is estimated using the following equation 3.2

CN 1 = CN 2 -

Where, CN1, CN2, and CN3 are different moisture condition in which the slope in CN2

is less than 5% The topographic factor like slope can affect the CN, thus the following equation 3.6 is used in case of higher slope

Where, the CN 2s is the adjusted curve number with slope, in which CN 2 and CN 3 are also adjusted with the default slope 5% The slp is the average slope of the basin

23

3.1.1 Preparation of In-put Data for Model Set-up

Unlike other hydrological models, SWAT model set up requires a number of input parameters; Digital Elevation Model (DEM), Land Cover (LC), soil classification

map, weather data (i.e., maximum and minimum temperature, precipitation, solar radiation, relative humidity, wind speed etc.,) This research collected the best available

data in the basin

i Digital Elevation Model (DEM)

DEM is the most fundamental data source for hydrologic modeling DEM is essential to delineate the basin boundary and also to classify elevation class and sub-basin division This study used the SRTM3 DEM (version 3) with the spatial resolution

of 30 meters grid cells developed by the United States Geological Survey (USGS) and the National Geospatial-Intelligence Agency (NGA) The database was downloaded from the webpage of http://srtm.csi.cgiar.org/srtmdata/ Figure 3.1 is the DEM of UPTBRB

Figure 3.1 DEM of the UPTBRB

ii High Resolution Land Use and Land Cover

High Resolution Land Use and Land Cover (HRLULC) was used to classify the basin land use patterns HRLULC is a product of Earth Observation Research Center/Japan Aerospace Exploration Agency (EORC/JAXA) that was developed for the entire mainland of Vietnam for the year of 2015 Having a high spatial resolution of 15 meters and detailed land cover classes, it is suitable for SWAT hydrologic modelling thereby enabling to assign the specific land use pattern (EORC/JAXA, 2016)

The HRLULC was downloaded from the webpage of the www.eroc.jaxa.jp The collected dataset was extracted into the basin scale, and then analyzed under six main categories to provide a quick understanding of general land cover status; barren, agricultural land, forest, urban/built-up, water body and others (See figure 3.2)

It was found that the upstream region of the basin was occupied by the forest type whereas agricultural land was identified as the prominent land use type found in almost every part of the basin The percentage of residential land was found to be low in the upstream region but some clusters were recorded while approaching to the downstream regions But the proportion is very small when compared to forest and cultivated areas

A few proportions of water body and barren land were also observed The share of others type like grass/shrub land was also found

According to the observed statistics, the dominant land cover was agriculture with 42.33%, forest-mixed cover stood the second place with 40.14%, and the range area occupied 7.61%, and residential land, water body and barren land were 4.12%, 2.69% and 3.11% respectively (See figure 3.3)

25

Figure 3.2 High Resolution Land Use and Land Cover Map of the UPTBRB (2015)

Figure 3.3 Major Land Cover Statistics in the UPTBRB

The resulted land cover classification layers were treated in corrdance with SWAT land cover codes Due to having high resoulation and accuracy, the dataset showed a very detailed classifcation classses, thus the corresponding SWAT land use and land cover codes were assigned carefully as shown in table 3.1

Table 3.1 SWAT Land Use Code and Statistics (2015)

SWAT Land

iii Soil Database

The required soil database for the UPTBRB was downloaded from the webpage

of http://www.fao.org/geonetwork/ The collected soil map was extracted based on the

basin boundary and spatial distribution of soil classes were analyzed (See figure 3.4)

The detailed results of analysis were illustrated in the figure 3.5

Figure 3.4 Soil Map of the UPTBRB

27

Figure 3.5 The Proportion of Soil Classes observed in the UPTBRB

iii Weather Data

Daily observed weather data were collected from the Institute of Meteorological and Hydrology (IMH) under MONRE Since this study was aimed to find out changes

in recent year, thus 15 years (2005 – 2019) of necessary data were collected A total of five climate variables in daily format were collected from 11 meteorological stations located in the basin; maximum temperature (°C), minimum temperature (°C), precipitation (mm), relative humidity (%), wind speed (m/s) and sunshine hours (hr) where solar radiation was computed based on sunshine hours for one station The names

of data collected meteorological stations are Bac Kan, Thai Nguyen, Vinh Yen, Tam Dao, Bac Giang, Bac Son, Hiep Hoa, Huu Lung, Dinh Hoa, Son Dong and Luc Ngan (See figure 3.6) However, the daily record of river discharge data was available only in the Gia Bay hydrological station at Thai Nguyen Province The following table 3.2 summarized all the data used and detail information such as sources, data type and time series etc., were provided

Table 3.2 Summary for Preparation of Model Input-Data

Sr Data Type Source Spatiotemporal Resolution Number of Stations (Period) Spatial data

IMH, MONRE Point/Daily 11 (2005 – 2019)

3 Precipitation IMH, MONRE Point/Daily 11 (2005 – 2019)

4 Relative

Humidity

IMH, MONRE Point/Daily 11 (2005 – 2019)

6 Solar Radiation Calculation Point/Daily 1 (2005 – 2019) Hydrological data

1 River Discharge IMH, MONRE Point/Daily 1 (2008 – 2019)

Figure 3.6 Location of Hydro-meteorological Stations

3.2.2 HRU Analysis

After dividing sub-basins, the analysis of Hydrological Response Units (HRUs) were further studied Spatial data of land cover, soil map and slope classification were further assigned and SWAT generated a total number of 347 HRUs (See figure 3.7)

Figure 3.7 Formation of HRUs

3.2.3 Integration with Weather Database

SWAT weather database was prepared using five parameters; maximum temperature, minimum temperature, precipitation, relative humidity and wind speed from 11 stations but solar radiation was computed for 1 station and integrated into SWAT model

SWAT model for the UPTBRB was successfully simulated for the period of (2005 – 2019) whereas the first three years (2005 – 2007) was used as a warming period

to maintain model performance SWAT Error Check Tool was used to examine the simulated hydrological parameter value

3.3 SWAT Model Calibration and Validation

The output of SWAT model was calibrated and validated through the Soil and Water Assessment Tool Calibration and Uncertainty Procedure (SWAT-CUP) with Uncertainty in Sequential Uncertainty Fitting (SUFI-2) algorithm Calibration is the effort of trying to get the best model simulation match with the observed data through the adjustment of input parameters The objective function was optimized by the process

of calibration Validation is a process in which the calibrated parameters were treated with an independent set of data without any further changes Sensitivity analysis helps

to define the most the sensitive parameters on the hydrological process of the basin thereby influencing the model output or the objective function More detailed explanations were available in SWAT User’s Manual (Abbaspour, 2012)

3.4 Model Performance Evaluation

The overall performance of SWAT model was evaluated by several statistical parameters The Nash Sutcliff Efficiency (NSE) was used as the major objective function due to its popularity in the hydrological field but other parameters like the Percentage Bias (PBIAS), and the ratio of root mean square error (RSR) were also considered

The NSE determines the degree of fitness of the model between the simulated and observed data with the value ranging from -∞ to 1 The PBIAS indicates the tendency of the simulated value to determine larger or smaller than the observed value The optimum value is zero and positive value indicates model underestimation while the negative sign indicates over estimation The RSR represents the performance of the model through the ration of root mean square error (RMSE) to the standard deviation in which the higher value is associated with the better performance occupying the optimum value of zero (Singh et al., 2004 & S Shrestha et al., 2016)