Geospatial Analyses of Groundwater Depletion and Contamination: Multiscale - Global, Regional and Local Analyses

Mississippi State University Scholars Junction

Theses and Dissertations Theses and Dissertations

8-9-2019

Geospatial analyses of groundwater depletion and contamination: Multiscale - global, regional and local analyses

Aynaz Lotfata

Follow this and additional works at: https://scholarsjunction.msstate.edu/td

Recommended Citation Lotfata, Aynaz, "Geospatial analyses of groundwater depletion and contamination: Multiscale - global, regional and local analyses" (2019). Theses and Dissertations. 2418. https://scholarsjunction.msstate.edu/td/2418

This Dissertation - Open Access is brought to you for free and open access by the Theses and Dissertations at Scholars Junction. It has been accepted for inclusion in Theses and Dissertations by an authorized administrator of Scholars Junction. For more information, please contact [email protected]. Template C with Schemes v4.0 (beta): Created by L. Threet 2/5/19

Geospatial analyses of groundwater depletion and contamination: Multiscale - global, regional

and local analyses

By TITLE PAGE Aynaz Lotfata

A Dissertation Submitted to the Faculty of Mississippi State University in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy in Earth and Atmospheric Sciences in the Department of Geosciences

Mississippi State, Mississippi

August 2019

2019

Geospatial analyses of groundwater depletion and contamination: Multiscale - global, regional

and local analyses

By APPROVAL PAGE Aynaz Lotfata

Approved:

______Shrinidhi S. Ambinakudige (Major Professor)

______John C. Rodgers, III (Committee Member)

______Padmanava Dash (Committee Member)

______Prem B. Parajuli (Committee Member)

______Renee M. Clary (Graduate Coordinator)

______Rick Travis Dean College of Arts & Sciences

Name: Aynaz Lotfata ABSTRACT Date of Degree: August 9, 2019

Institution: Mississippi State University

Major Field: Earth and Atmospheric Sciences

Major Professor: Shrinidhi S. Ambinakudige

Title of Study: Geospatial analyses of groundwater depletion and contamination: Multiscale - global, regional and local analyses

Pages in Study: 174

Candidate for Degree of Doctor of Philosophy

The overarching objective of this dissertation was to study groundwater resources on global, local, and regional scales. The first objective of this dissertation was to analyze the groundwater nitrate contamination in the Edwards-Trinity and the Southern High-Plains aquifers of Texas. The second was to study groundwater quality in terms of seawater intrusion in the

California Coastal Basin, Upper Floridian, and North Atlantic Coastal Plain aquifers. This dissertation also provided a comprehensive overview of the groundwater level in basins at the global scale and further analyzed agricultural activities on groundwater storage in small and large basins.

To achieve first objective, Ordinary Least Square (OLS) and Geographically Weighted

Regression (GWR) models were used to study the relationship between groundwater nitrate contamination and land use. This dissertation further identified dominant groundwater types using

USGS well data and to estimate the extent of seawater intrusion in terms of dominant ions and ocean salinity in the United States coastal aquifers. Finally, groundwater storage anomaly was quantified using Gravity Recovery and Climate Experiment (GRACE) derived variations in total

Terrestrial Water Storage (TWS) and the Global Land Data Assimilation System (GLDAS). Land

cover data representing a percentage of irrigated lands using groundwater resources was used to study agricultural activities on groundwater storage.

Groundwater nitrate contamination was positively associated with cotton production in

Southern High-Plains and Edwards-Trinity aquifers. The nitrate concentrations tended to increase as the well-depth decreased in both aquifers. Results showed that the dominant ions in the study area were Na+ and Cl- . The study concluded that Na-Cl and mixed Ca-Mg-Cl were dominant water types in the United States’ coastal aquifers. Results also indicated that seawater intrusion is occurring in the US coastal aquifers. Groundwater depletion has increased in southern Asia, western North America, and southwestern Europe due to groundwater withdrawal for agricultural use. However, farming practice is not the main reason for groundwater scarcity in South America,

Africa, and Australia.

DEDICATION

This dissertation is dedicated to the memory of my grandparents. Also, dedicated to my beloved father Dr. Nader Lotfata, mother Mrs Farah Amirmotallebi and my brothers Dr. Aydin

Lotfata and Yasin Lotfata, other family members and friends who supported me.

ACKNOWLEDGEMENTS

I would like to sincerely thank my supervisor, Dr. Shrinidhi Ambinakudige for his guidance and support throughout this study, and especially for his confidence in me. I learned from his insight a lot and I believed I learned from the best. I would also like to thank Dr. Rodgers, Dr.

Dash and Dr. Parajuli for serving as committee members on my dissertation and for their constructive criticism that significantly improved the quality of this dissertation. I would like to thank Dr. Parajuli his comments and questions were very beneficial in completion of my research work. I thank Dr. Rodgers who always assured me with his encouraging words. To Dr. Dash, I was grateful for his support during my education.

I express my thanks to Dr. Domenico “Mimmo” Parisi, Dr. Steven Grice, and Dr. Michael

Taquino for providing research assistantship for three years in National Strategic Planning and

Research Analysis Center (NSPARC). To Dr. John Brooks, I was grateful for the interpretation of some of the results presented in this dissertation. I thank the entire faculty and staff that helped me complete this journey. Also, I am thankful to all my colleagues and friends at the Department of

Geosciences, especially Devon Flicker and Katerina Sergi. I will always be indebted to National

Strategic Planning and Research Analysis Center (NSPARC) and the Department of Geosciences,

Mississippi State University for supporting my research through a financial assistantship without which this dissertation would have never been possible.

iii

TABLE OF CONTENTS

DEDICATION ...... ii

ACKNOWLEDGEMENTS ...... iii

LIST OF TABLES ...... vii

LIST OF FIGURES ...... vii

CHAPTER

I. INTRODUCTION ...... 1

1.1 Background ...... 1 1.2 Estimating Groundwater Quality and Groundwater Quantity ...... 3 1.3 Why Study Groundwater Quality in Coastal and Inland Areas of the U.S.? ...... 8 1.4 Why Study Groundwater Quantity Throughout the World? ...... 9 1.5 Study Objectives ...... 10 1.5.1 Factors Affecting the Spatial Pattern of Nitrate Contamination in Texas Aquifers (Chapter-II) ...... 11 1.5.2 Degradation of Groundwater Quality in the Coastal Aquifers of the United States (Chapter-III) ...... 12 1.5.3 Investigating Global Groundwater Variations in Basins of the World (Chapter-IV) ...... 13 1.6 References ...... 16

II. FACTORS AFFECTING THE SPATIAL PATTERN OF NITRATE CONTAMINATION IN TEXAS AQUIFERS ...... 17

2.1 Abstract ...... 17 2.2 Introduction ...... 18 2.3 Background ...... 21 2.4 Study Area ...... 23 2.5 Data and Methods ...... 26 2.5.2 Multivariate Statistical Analysis ...... 28 2.5.3 Geographically Weighted Regression ...... 29 2.6 Results and Discussion ...... 30 2.6.2 Multivariate Statistical Analysis ...... 35 2.6.2 Geographically Weighted Regression ...... 38 2.7 Conclusions ...... 43 iv

2.8 References ...... 46

III. DEGRADATION OF GROUNDWATER QUALITY IN THE COASTAL AQUIFERS OF THE UNITED STATES ...... 53

3.1 Abstract ...... 53 3.2 Introduction ...... 53 3.3 Objectives ...... 56 3.4 Background ...... 57 3.5 Study Area ...... 60 3.5.1 Upper Floridian Aquifer ...... 61 3.5.2 California Coastal Plain Aquifer ...... 62 3.5.3 North Atlantic Coastal Plain Aquifer ...... 63 3.6 Data and Method ...... 64 3.6.2 Piper Diagram Groundwater Quality Indices: (GQI (dom) & GQI (mix)) ...65 3.6.3 Groundwater Quality Indices (fsea, GQI (fsea) & GQI Seawater Intrusion (swi)) ...... 67 3.6.4 Geostatistical Analysis ...... 69 3.7 Results and Discussions ...... 70 3.7.1 Piper Diagram ...... 71 3.7.2 Groundwater Quality Indices ...... 73 3.7.3 Seawater Intrusion Indices ...... 77 3.7.4 Geostatistical Analysis ...... 83 3.8 Concluding Remarks ...... 88 3.9 References ...... 91

IV. INVESTIGATING GROUNDWATER VARIATIONS IN BASINS OF THE WORLD ...... 99

4.1 Abstract ...... 99 4.2 Introduction ...... 100 4.3 Objectives ...... 102 4.4 Background ...... 102 4.5 Data and Method ...... 107 4.5.1 GRACE‐Derived Total Water Storage Data ...... 107 4.5.2 Global Land Data Assimilation System Output ...... 108 4.5.3 Irrigation Data ...... 109 4.5.4 Analysis of Groundwater Depletion ...... 111 4.6 Results and Discussions ...... 112 4.6.1 Trends in Groundwater Storage Anomalies ...... 112 4.6.2 Groundwater Storage Anomalies ...... 127 4.6.3 Trends in Groundwater Storage Anomalies, Basin Size, and Irrigated Lands ...... 134 4.7 Conclusion ...... 141 4.8 References ...... 144

V. CONCLUSION ...... 153

5.1 Discussions On The Overall Outcome and Scope Of Future Research ...... 153 5.2 Significance of This Study ...... 157 5.3 Future Research ...... 158 5.4 References ...... 160

APPENDIX

A. TABLE ...... 163

B. FIGURE ...... 167

LIST OF TABLES

Table 1.1 Illustrating natural and human factors affecting groundwater quality (Yee and Souza 1987)...... 2

Table 2.1 Nitrate (N) concentration in the Southern High-Plains aquifer and the Edwards- Trinity aquifer...... 32

Table 2.2 Land use patterns in the Edwards-Trinity aquifer from 2008 to 2017...... 33

Table 2.3 Land use patterns in the Southern High-Plains aquifer from 2008 to 2016...... 34

Table 2.4 Average land use patterns in one-kilometer buffer around wells from 2008 to 2017...... 35

Table 2.5 Results of the OLS regression analysis in the Southern High-Plains aquifer...... 36

Table 2.6 Results of the OLS regression analysis in the Edwards-Trinity aquifer...... 37

Table 2.7 GWR Coefficients...... 39

Table 3.1 Illustrating hydro-chemical domains using GQI (mix) and GQI (dom) (Tomaszkiewicz et al. 2014): ...... 67

Table 3.2 Illustrating average groundwater quality indices in three principal aquifers of the U.S...... 74

Table 3.3 Average seawater fraction (fsea) and groundwater quality indices...... 78

Table 3.4 GQI (swi) ranges (Tomaszkiewicz et al. 2014)...... 83

Table 3.5 Moran I and descriptive statistics for seawater fraction (fsea)...... 84

Table 3.6 Semi-variogram models of fsea parameter for Universal Kriging...... 85

Table 4.1 Data sources...... 111

Table 4.2 Average groundwater trend and percentage irrigated areas using groundwater in North American basins...... 114

vii

Table 4.3 Average groundwater trend and percentage irrigated areas using groundwater in South American Basins...... 116

Table 4.4 Average groundwater trend and percentage irrigated areas using groundwater in African basins...... 119

Table 4.5 Average groundwater trend and percentage irrigated areas using groundwater in European basins...... 121

Table 4.6 Average groundwater trend and percentage irrigated areas using groundwater in Asian basins...... 124

Table 4.7 Average groundwater trend and percentage irrigated areas using groundwater in Australian basins...... 125

Table 4.8 Sample world basins covering shallow aquifers...... 134

Table 4.9 Errors in major basins of the world ...... 141

Table A.1 Illustrating ion concentrations of groundwater samples ...... 164

viii

LIST OF FIGURES

Figure 1.1 Illustrating factors affecting coastal aquifers (Ketabchi et al. 2016)...... 5

Figure 2.1 Study areas: the Edwards-Trinity and the High-Plains aquifers (WHYMAP 2011)..25

Figure 2.2 Texas land uses in 2017 (USDA 2017)...... 28

Figure 2.3 Illustrating nitrate and well-depth relation in the Edwards-Trinity aquifer (2.3a) and the Southern High-Plains aquifer (2.3b)...... 31

Figure 2.4 Average nitrate concentration from 2008 to 2017 (2.4a) in the Edwards-Trinity and (2.4b) the Southern High-Plains aquifers...... 33

Figure 2.5 Relationship between nitrate concentration and cotton cultivation in the Edwards- Trinity aquifer (2.5a) and the Southern High-Plains aquifer (2.5b)...... 34

Figure 2.6 Spatial patterns of GWR coefficients in the Southern High-Plains aquifer...... 40

Figure 2.7 Spatial patterns of GWR coefficients in the Edwards-Trinity aquifer...... 41

Figure 3.1 Map of the study area showing California Coastal Basin aquifer, Upper Floridian aquifer, and North Atlantic Plain aquifer...... 61

Figure 3.2 Piper Diagram illustrating general classifications of groundwaters types (Tomaszkiewicz et al. 2014)...... 66

Figure 3.3 Illustrating anions and cations composition in California Coastal Basin aquifer, Upper Floridian aquifer and North Atlantic Coastal Plain aquifer...... 71

Figure 3.4 Piper diagram illustrating general classification of groundwaters (mg/L) and seawater-freshwater mixing water types (mg/L) in the coastal aquifers of the U.S. ..72

Figure 3.5 Illustrating groundwater quality using Groundwater Quality Index (GQI(mix)) and Groundwater Quality Index (GQI(dom)) in the coastal aquifers of the United Sates...... 76

Figure 3.6 Illustrating fsea and GQI (swi) relationship...... 79

vii

Figure 3.7 Illustrating seawater intrusion using Seawater Fraction (fsea) for 240 coastal wells and Groundwater Quality Index (GQI(swi) for selected 107 sample wells ...... 82

Figure 3.8 Semi-variogram for the California Coastal Basin aquifer (8a), Upper Floridian aquifer (8b), and North Atlantic Coastal Plain aquifer (8c)...... 85

Figure 3.9 Illustrating Seawater Fraction (fsea) using Universal Kriging at Upper Floridian aquifer (3.9a), North Atlantic Coastal Plain aquifer (3.9b) and California Coastal Basin aquifer (3.9c& 3.9d)...... 87

Figure 4.1 Illustrating percentage irrigated areas using groundwater resources (FAO 2010). ..110

Figure 4.2 Illustrating global groundwater trend and p-value during 2002-2015 in North America...... 115

Figure 4.3 Illustrating global groundwater trend and p-value during 2002-2015 in South America...... 117

Figure 4.4 Illustrating global groundwater trend and p-value during 2002-2015 in Africa...... 119

Figure 4.5 Illustrating global groundwater trend and p-value during 2002-2015 in Europe. ....122

Figure 4.6 Illustrating global groundwater trend and p-value during 2002-2015 in Asia...... 124

Figure 4.7 Illustrating global groundwater trend and p-value during 2002-2015 in Australia. .126

Figure 4.8 Illustrating average monthly groundwater anomalies in the major Asian basins. ....128

Figure 4.9 Illustrating average monthly groundwater anomalies in the major European basins...... 129

Figure 4.10 Illustrating average monthly groundwater anomalies in the major African basins. .130

Figure 4.11 Illustrating average monthly groundwater anomalies in the major North American basins...... 131

Figure 4.12 Illustrating average monthly groundwater anomalies in the major South American basins...... 132

Figure 4.13 Illustrating groundwater trend variations by basin size and percentage irrigated lands using groundwater in Asia...... 135

Figure 4.14 Illustrating groundwater trend variations by basin size and percentage of irrigated land by using groundwater in Europe...... 136

Figure 4.15 Illustrating groundwater trend variations by basin size and percentage of irrigated land by using groundwater in Africa...... 137 viii

Figure 4.16 Illustrating groundwater trend variations by basin size and percentage of irrigated land by using groundwater in North America...... 138

Figure 4.17 Illustrating groundwater trend variations by basin size and percentage of irrigated land by using groundwater in South America...... 139

Figure B.1 Illustrating average monthly groundwater level changes in the major basins – basins in North America (a), basins in South America (b), basins in Africa (c), basins in Europe (d), and basins in Asia (e)...... 168

CHAPTER I

INTRODUCTION

1.1 Background

In most parts of the world, water is a scarce resource. Only 2.5 percent of the water on earth is fresh water and over two thirds of this is frozen in glaciers and polar ice caps (USGS 2016).

Water demand already exceeds supply in many parts of the world. It is estimated that about 70 percent of the global water supply is used for irrigation in agriculture. Furthermore, groundwater accounts for as much as 85 percent of the earth’s usable water for irrigation (Wada et al. 2012,

Gleeson et al. 2012). Groundwater provides approximately half of the total global domestic water demand and is a major provider of industrial and agricultural demands (Xiao et al. 2015).

Groundwater is best defined as water that exists underground in saturated zones beneath the soil surface and is a vital source of freshwater, accounting for about 30.l percent of the world’s available freshwater (USGS 2016). However, groundwater is rapidly consumed at unsustainable rates in many areas of the world and in some areas has resulted in the detrimental effects of seawater intrusion (Brovelli et al. 2007). The salinization of groundwater by seawater intrusion in coastal areas may cause decreases in crop production and threaten the health of coastal ecosystems

(Herbert et al. 2015). Groundwater contamination also occurs when anthropogenic products are introduced into it, thus causing it to become unsafe and unfit for human use (Eamus et al. 2016).

Groundwater plays an important role in the socioeconomic development and ecological health of

these dependent aquatic-based ecosystems. Table 1.1. explains natural and human factors affecting groundwater quality.

Table 1.1 Illustrating natural and human factors affecting groundwater quality (Yee and Souza 1987).

Natural Factors Natural Source Types of Contaminant Precipitation Dissolved gases, dust, and emission particles Infiltration through vegetation, Biochemical products, organic materials, color, and minerals swamps, or soil and rocks, seawater Aquifer rocks Minerals content Inter-aquifer mixing of cold water Minerals and gases and thermal water Human Factors Waste Source Type of Contaminant Agricultural activities Fertilizers, pesticides, and herbicides Mining operations Metallic trace elements and phosphates Nuclear facilities Radioactive chemicals, heat, and dissolved solids Urban activities (storm and Organic materials, dissolved solids, suspended solids, sanitary sewers, sewage-disposal detergents, bacteria, phosphate, nitrate, sodium, chloride, sulfate, metallic plants, cesspools and septic tanks, trace elements, and others and sanitary landfills) Industrial facilities (food Biochemical oxygen demand, suspended solids, sodium, chloride processors) Geothermal activities Heat, dissolved solids, fluoride, and metallic trace elements Hazardous waste- and toxic waste Toxic metals, hazardous chemicals, and organic disposal sites compounds

Salinization, contamination, and rapid groundwater depletion threatens groundwater resources (Custodio et al. 2016). Aquifer depletion, the leaching of fertilizers into aquifers, and seawater intrusion has all led to groundwater resource degradation. An acceptable level of groundwater quality is important for controlling excessive withdrawal potential and ensures the protection of healthy ecosystems (Lee et al. 2018). This dissertation assesses the groundwater

quality of the coastal and inland aquifers of the United States. Apart from examining the groundwater quality of these areas, the purpose of this dissertation was to quantify the groundwater levels throughout the world using Geographic Information System (GIS) and remote sensing data.

1.2 Estimating Groundwater Quality and Groundwater Quantity

Water quality analysis is one of the most important aspects of groundwater studies.

2+ 2+ + + - - - Groundwater consists of major ion elements comprised of Ca , Mg , Na , K , Cl , HCO3 , NO3 ,

2- and SO4 (Tomaszkiewicz et al. 2014, Appelo and Postma 2005). Natural and human factors affect groundwater quality (Yee and Souza 1987), and Nitrate-nitrogen contamination occurs naturally in groundwater (Lake et al. 2003).

The main sources of nitrates in groundwater are due to agricultural activities where the extensive uses of various fertilizers are known to cause groundwater nitrate contamination. The presence of a high amount of nitrates in the water can cause methemoglobinemia in humans (Ducci et al. 2010). Also, in an effort to more effectively identify this, Ducci (2018) presented a methodology to assess nitrate contamination in groundwater using mostly land use thematic maps and from statistical data in southern Italy. He highlighted the dependence of farmers who use fertilizers on their agricultural products, which consequently contain nitrate contaminants found in the groundwater of urbanized areas. Later on, Malki et al. (2017) carried out several samplings at different sites of the Plio-Quaternary aquifer of Chtouka located in the Southwestern part of

Morocco in order to investigate the spatial variation of groundwater quality. The temporal evolution of groundwater levels further indicated intensive, excessive amount of its withdrawal mainly in irrigated areas, where nitrate concentrations exceeded 50 mg/L.

Saidi et al. (2011) used the DRASTIC method and a hazard assessment to determine the degree of risk affecting the groundwater quality in the Souassi aquifer of Tunisian Sahel. Kerr-

Upal et al. (1999) used the secondary data to develop GIS layers for soil drainage and nitrogen application rates by land use systems in Ontario. Passarella et al. (2002) assessed the groundwater quality using Disjunctive Kriging (DK) in the Modena plain, northern Italy. The DK evaluates the conditional probability of pollutant concentration at a given time. Diodato et al. (2013) presented a non-parametric hydro-geostatistical approach for mapping nitrate hazard areas in the aquifers of the Campania Plain.

Multivariate statistical analysis was also used to identify the source of groundwater pollution in (Huang et al. 2010, Diodato et al. 2013, Ducci 2018). In the case of Huang et al.

(2010), they applied multivariate statistical methods, such as factor and cluster analysis, to search for the interrelation between water quality parameters, factors representing their characteristics, and possible pollution sources of groundwater quality in Pingtung Champaign, Taiwan. The results showed that the groundwater quality of hinterland was better than that of the coastal area, which may have been affected by seawater intrusion resulting in aquifer salinization.

The quality of groundwater should likewise be studied in the coastal areas. Coastal aquifers are principally affected by both natural and anthropogenic activities like sea level rise, erosion, encroachment, and over-withdrawal of groundwater (Fig.1.1).

Figure 1.1 Illustrating factors affecting coastal aquifers (Ketabchi et al. 2016).

Langevin et al. (2003), Brovelli et al. (2007), and Goswami and Clement (2007) used numerical modelling to simulate transient seawater intrusion. They estimated the density difference between freshwater and seawater and the inland‐sloping hydraulic gradient (Paniconi et al. 2001). Additionally, Strack (1976) studied the maximum withdrawal rate to avoid seawater movement into groundwater.

Many studies characterized water types of aquifers using a water quality index (Gopinath et al. 2019, Appelo and Postma 2005, Tomaszkiewicz et al. 2014). Appelo and Postma (2005) quantified seawater fraction in terms of groundwater chloride concentration and ocean salinity. 5

Tomaszkiewicz et al. (2014) developed the seawater intrusion index using a Piper diagram. Ayed et al. (2017) assessed the groundwater vulnerability to seawater intrusion using a parametric method of vulnerability assessment, known as, GALDIT in the Jeffara of Medenine coastal aquifer, of southeastern Tunisia. GALDIT considers five parameters of groundwater occurrence: aquifer hydraulic conductivity, height of groundwater level above sea level, distance from the shore, impact of seawater intrusion and thickness of the aquifer. Additionally, chemical, isotopic and geophysical electrical resistivity methods were used to detect seawater intrusions (Sathish et

- - al. 2011, Aris et al. 2007). Cl is a dominant ion of seawater and Bicarbonate (HCO3 ) and is the most abundant ion in groundwater. Furthermore, the ratio of Chloride-Bicarbonate ions in ground and seawater are very different, making the ratio between these two ions a useful index of the presence of seawater in groundwater (Fadili et al. 2015).

The quantity of groundwater storage should likewise be estimated. Advances in remote sensing have made this an ideal tool for monitoring water resources. The NASA Gravity Recovery and Climate Experiment (GRACE) satellite mission supplies approximate monthly variations of terrestrial water storage (TWS) information on the basis of measurements of the Earth’s gravity field (Wahr 2004). In GRACE data; atmospheric, oceanic circulation and solid Earth tides were removed to indicate TWS variations (Rodell et al. 2007).

Chinnasamy et al (2013) used seasonal and annual hydrologic signals obtained by GRACE and simulated soil moisture variations from land data assimilation systems to show groundwater depletion trends in northwestern Gujarat. Frappart and Ramillien (2018) discussed the application of GRACE to quantify the groundwater volume variations. Indeed, GRACE was the first remote sensing mission to provide temporal variations of TWS. TWS is the sum of the water masses that

contain snow, surface water, soil moisture, and groundwater at a spatial resolution of one hundred kilometers. GRACE-based groundwater changes presented significant aquifer depletion over large regions, such as the Middle East, the northwestern India aquifer, the northern China Plain aquifer, the Murray-Darling Basin in Australia, the High-Plains, and the California Central Valley aquifers in the United States of America (USA) (Frappart and Ramillien 2018). Famiglietti et al. (2011) studied GRACE data in the Central Valley of California and concluded that groundwater was depleted at an annual rate of 20.4 ± 3.9 mm Equivalent Water Height (EWH) in the valley.

Katpatal et al. (2018) and Khanbilvardi et al. (2016) used GRACE gravity models and the

Global Land Data Assimilation System (GLDAS) land-surface models as well as the actual field data to study groundwater level variations per GRACE pixel. Similarly, Gautam et al. (2017) used

10 years of data sets such as GRACE, soil moisture content, rainfall from 2003 to 2012, and bore- hole data of 42 wells to assess the groundwater level variations in regions of the Uttar Pradesh in

India. Finally, Chinnasamy and Ganapathy (2017) analyzed the long-term trends (14 years, from

2002 to 2014) in TWS and groundwater storage for Peninsular Malaysia, using GRACE.

As a consequence of these studies, it is essential to examine groundwater quality and quantity in order to understand the impacts of natural processes and anthropogenic activities on groundwater degradation and storage depletion. Changes in groundwater storage are influenced by groundwater withdrawal rates. Apart from groundwater quantity, it is essential to also monitor groundwater quality in order to limit the release of nitrogen fertilizers contained within the soil due to agricultural activities as well as any resulting sodium chloride contamination due to seawater intrusion.

The need to protect groundwater in a concerted effort at the local, regional, and global scales is highlighted in five chapters of this dissertation. Chapter-I covers the significance of study sites, their respective goals and a brief overview of the methodology used. Chapter-II investigates the groundwater nitrate contamination levels of various Texas aquifers. Chapter-III explains the groundwater degradation of multiple U.S. coastal aquifers due to seawater intrusion. Chapter-IV then reviews groundwater depletion throughout basins of the world, and finally, Chapter-V contains a brief discussion of the overall outcomes and a proposed scope for future research.

1.3 Why Study Groundwater Quality in Coastal and Inland Areas of the U.S.?

Seawater intrusion associated with groundwater over-withdrawal and the lowering of groundwater levels has occurred in the regions of the: California Coastal Basin aquifer, Upper

Floridian aquifer, and North Atlantic Coastal Plain aquifer of the United States. Principal coastal aquifers in the United States lack a comprehensive assessment of groundwater quality in terms of seawater intrusion. Many numerical and analytical models were used to model a seawater- groundwater interface zone due to sea level rise and groundwater over-withdrawal (Lu et al et al.

2013, Riva et al. 2018, Ataie-Ashtiani et al. 1999). A groundwater quality index was quantified based on different quality parameters to assess the suitability of groundwater for drinking and irrigation purposes. However, this did not include seawater salinity to evaluate the groundwater quality in terms of seawater intrusion in the coastal aquifers of the United States.

Groundwater quality samples obtained from USGS monitoring wells are recorded at a specific time and are not provided for all aquifer monitoring wells. The California Coastal Basin,

Upper Floridian, and North Atlantic Coastal Plain aquifers were chosen for the coastal areas and the Edwards-Trinity and the Southern High-Plains aquifers were selected inland to study

groundwater quality and to better understand the groundwater degradation of coastal and inland aquifers. Increased nitrate concentrations in drinking water have been associated with adverse health effects, such as methemoglobinemia (Rivett et al. 2006). Major aquifers that have been identified as having exceeded their nitrate contamination levels are: Seymour, Southern High-

Plains, and Edwards-Trinity aquifers. In the state of Texas, the Texas Water Development Board

(TWDB) has provided a water quality database for groundwater in Texas wells. The purpose of the program is to monitor changes in groundwater quality over time in major and minor aquifers in the state. TWDB did not provide a comprehensive monthly data set for nitrate variations of wells in major aquifers. The database contains water quality data for about 55,000 wells. Among all of the major aquifers, the Southern High-Plains and Edwards-Trinity had a comprehensive data set to further analyze (TWDB).

A decadal assessment of trends in the concentrations of nitrogen from 1991 to 2017 showed a significant increase of nitrate content in groundwater in the United States (Cao et al. 2017).

Moreover, nitrate concentrations in deep aquifers are likely to increase during the next decade as nitrates mobilize downward from shallow aquifers (Nolan et al. 2002).

1.4 Why Study Groundwater Quantity Throughout the World?

Groundwater is a scarce resource in most parts of the world. Groundwater volume variations are measured using GRACE and GLDAS, which study groundwater depletion rates at different sized basins throughout the world. Aquifer systems with diverse recharge and discharge rates provide examples of responses to local climate change, global warming, and human activities.

Groundwater accounts for 85 percent of the global supply that is used for irrigation (Wada et al.

2012, Gleeson et al. 2012). However, over-withdrawal of groundwater for agricultural activities is

not a worldwide reason for groundwater scarcity (MacDonald et al. 2012). Allocation of groundwater in trans-boundary basins can cause socio-environmental conflicts resulting in water scarcity (Nijsten et al. 2018). The sustainable management of the unevenly distributed groundwater resources holds the key to overcoming water scarcity. Balancing water availability and use within basins is equally important because of limited hydrological terrestrial ecosystems in small-scale basins (Mukherjee et al. 2018).

Additionally, groundwater declines would lead to the local and regional depletion of aquifers, which would have adverse consequences for agricultural productivity and human health

(Stromberg et al. 1996). The baseflow reduction of streams and wetlands due to declines in groundwater levels causes permanent changes in ecosystems (Stromberg et al. 1996). In fact, while water demands increase in a warmer environment, climate change exacerbates the situation in aquifers where precipitations rates are decreased (Taylor et al. 2012). Understanding groundwater storage and its natural variability is essential for better preserving and managing natural resources on a global scale (Famiglietti and Rodell 2013).

1.5 Study Objectives

The first objective of this dissertation is to analyze the groundwater nitrate contamination in the Edwards-Trinity and the Southern High-Plains aquifers of Texas. The second is to study groundwater quality in terms of seawater intrusion in the California Coastal Basin, Upper

Floridian, and North Atlantic Coastal Plain aquifers. Furthermore, it aims to estimate groundwater storage variations across world basins. The following is a brief overview of the aforementioned objectives and the methodology used to achieve them.

1.5.1 Factors Affecting the Spatial Pattern of Nitrate Contamination in Texas Aquifers (Chapter-II)

Chapter-II explores in-depth how human activities affect groundwater quality. Agricultural activities cause significant changes on land, and those activities have adverse effects on groundwater resources. The chemicals that accumulate in the soil zone slowly move toward groundwater and deteriorate its quality. The application of fertilizers to crops can result in groundwater contamination. Many studies denote significant groundwater nitrate contamination throughout the U.S., particularly in shallow or unconfined aquifers (Van der Schans et al. 2009,

Lockhart et al. 2013). In these cases, groundwater contaminants can exceed the maximum acceptable levels for public use.

Specific objectives of Chapter II are as follows:

❖ to analyze the extent of nitrate concentration in groundwater in the Southern High-Plains

and Edwards-Trinity aquifers of Texas in the United States;

❖ to determine what types of land use have the most significant impacts on nitrate

concentrations of groundwater in the Southern High-Plains and Edwards-Trinity aquifers

of Texas in the United States.

To achieve the goals, the relationship between nitrate concentrations in groundwater and land use was examined using a multivariate regression analysis and a Geographically Weighted

Regression model (GWR). Groundwater nitrate concentrations were obtained from the Texas

Water Development Board (TWDB) database on groundwater quality. In this study, 192 groundwater samples, which include 69 groundwater samples from the Southern High-Plains aquifer and 123 groundwater samples from the Edwards-trinity aquifer were chosen to study groundwater nitrate contamination levels.

1.5.2 Degradation of Groundwater Quality in the Coastal Aquifers of the United States (Chapter-III)

Chapter-III explains that Seawater Intrusion (SWI) is a significant threat to freshwater resources in coastal aquifers around the world. Seawater intrusion is associated with a lowering of groundwater levels due to local climate conditions and human activities. For instance, when too much groundwater is withdrawn from coastal aquifers, seawater intrusion occurs due to the disturbed balance between inland freshwater and the ocean. Saltwater has intruded into many of the coastal aquifers of the U.S. from California to the Florida Gulf Coast and Georgia coast to the

Long Island.

Specific objectives of Chapter III are as follows:

❖ to identify dominant groundwater types in coastal aquifers using USGS well data;

❖ to estimate the extent of seawater intrusion in terms of dominant ions and ocean salinity in

U.S. coastal aquifers using various groundwater quality indices;

❖ to interpolate the seawater intrusion estimates to analyze the spatial patterns of seawater

intrusion into coastal aquifers using geostatistical techniques.

To achieve these goals, Piper diagrams were used to display patterns in the water chemistry of groundwater and to represent any trends in major ions. A Groundwater Quality Index GQI (mix) was also used to characterize any mixing of fresh groundwater and seawater. Additionally, a

Groundwater Quality Index (Dominant) GQI (dom) was used to characterize chemical patterns in groundwater, where 0 represents Ca-Cl groundwater and 100 represents Na-HCO3 water type of aquifers. A simple tool to identify seawater intrusion is the seawater fraction (fsea) with values ranging from 0 to 1, where 0 represents freshwater and 1 represents the saline water type within the aquifers. Seawater fraction was quantified in terms of groundwater chloride and ocean salinity

concentration levels. A Groundwater Quality Index GQI (fsea) was then quantified in terms of chloride concentration and ocean salinity is quantified by using fsea. However, GQI (fsea) may not be an appropriate index independently. The combination of GQI (fsea) and GQI (mix) results can be a more illustrative index for seawater-groundwater mixing. GQI (swi) was also used to characterize seawater intrusion in groundwater, where 0 represents high seawater intrusion and

100 represents freshwater in the aquifer. Furthermore, to interpolate sea fsea point measurements, as they were quantified based on chloride concentrations and ocean salinity values, a geostatistical analysis was used.

1.5.3 Investigating Global Groundwater Variations in Basins of the World (Chapter-IV)

Chapter-IV presents evidence that groundwater depletion is not an uncommon reality in the basins of the world where groundwater supports the human activities and ecosystems.

Groundwater depletion occurs when groundwater exploitation exceeds the natural groundwater recharge for large areas and within long time periods. A realistic picture of global groundwater conditions must be based on data aggregation from a variety of widely distributed organizations.

However, inconsistencies among the collected local data make global groundwater studies difficult. Remote sensing technologies (such as GRACE satellite data) provide a novel means for accurately estimating groundwater depletion levels and can also serve as an indicator of groundwater depletion levels on a global scale.

Specific objectives of Chapter III are as follows:

❖ to provide a comprehensive overview of the groundwater levels in basins on a global scale;

❖ to analyze the agricultural activities on groundwater storage in small and large

transboundary basins.

To achieve these goals, GRACE derived variations of Terrestrial Water Storage (TWS) containers, which were quantified at the Center for Space Research at the University of Texas at

Austin. As GRACE data does not directly provide the results of groundwater measurements, the

GLDAS, which contains results for accumulated snow and soil moisture levels, was used to quantify groundwater measurements (GRACE 2018). Furthermore, irrigation data representing a percentage of irrigated lands using groundwater resources was used.

Each monthly GRACE grid refers to the surface mass deviation for that month relative to a baseline average from January 2004 to December 2009 (Bettadpur 2007). TWS variations include contributions of changes to the Groundwater (GW) and land surface water in terms of Soil

Moisture (SM) and Snow Water Equivalent (SWE), as expressed in the eq. (Bettadpur 2007,

Chambers 2006 )[1.1].

∆TWS=∆GW + ∆SM +∆SWE [1.1]

A total of 143 months of data from TWS GRACE derived anomalies was quantified between April 2002 and December 2015, at the Center for Space Research (CSR) at the University of Texas in Austin (Bettadpur 2007, Chambers 2006). Data was also used from the GLDAS version1 (GLDAS-1) 1.0-degree resolution covering a period from April 2002 to December 2015.

A GLDAS grid refers to the total accumulated snow and soil moisture variations for that month relative to a baseline average obtained from January 2003 to December 2007. In order to ensure a valid analysis of the data, anomalies from GRACE and GLDAS should be within the same time period. The GRACE data time baseline was converted from January 2004 to December 2009 to

January 2003 to December 2007. This is done by averaging the monthly gravity data reported by

CSR between January 2003 and December 2007 and removing the mean value from each month to compute the anomalies (GRACE 2018).

The focus of this dissertation was to study groundwater resources on global, local, and regional scales to understand the serious challenges associated with obtaining reliable groundwater quality and quantity data.

1.6 References

Appelo, C.A.J., and Postma D. 2005. Geochemistry, groundwater and pollution. London: CRC Press.

Aris, A. Z., Abdullah, M. H., Ahmed, A., and Woong, K. K. 2007.Controlling factors of groundwater hydrochemistry in a small island’s aquifer. International Journal of Environmental Science Technology 4 (4): 441-450.

Ataie-Ashtiani, B., Volker, R. E., and Lockington, D. A. 1999. Tidal effects on seawater intrusion in unconfined aquifers. Journal of Hydrology 216:17-31.

Ayed, B., Jmal, I., Sahal, S., and Bouri, S. 2017. Assessment of groundwater vulnerability using a specific vulnerability method: case of Maritime Djeffara shallow aquifer (Southeastern Tunisia). Arabian Journal of Geosciences 6(10): 30-35.

Bettadpur, S. 2007. Level-2 Gravity Field Product User Handbook (Rev. 2.3, February 20, 2007) Center for Space Research, University of Texas at Austin.

Brovelli, A., Mao, X., and Barry, D.A. 2007. Numerical modeling of tidal influence on density- dependent contaminant transport. Water Resource Research 43.

Cao, P., Lu, C., and Yu, Z. 2017. Historical nitrogen fertilizer use in agricultural ecosystem of the continental united states during 1850-2015: application rate, timing, and fertilizer types. Earth System Science 10: 969-984.

Chambers, D. P.2006. Evaluation of new GRACE time‐variable gravity data over the ocean. Geophysical Research Letter 33: LI7603.

Chinnasamy, P., Hubbart, J.A., and Agoramoorthy, G. 2013. Using remote sensing data to improve groundwater supply estimations in Gujarat, India. Earth Interactions 17: 1-17.

Chinnasamy, P., and Ganapathy, R. 2017. Long term variations in water storage in Peninsular Malaysia. Journal of Hydrology informatics 20(5).

Custodio, E., Cabrera, M. D. C., Poncela, R., Puga, L.O., Skupien, E., and del Villar, A. 2016. Groundwater intensive exploitation and mining in Gran Canaria and Tenerife, Canary Islands, Spain: hydrogeological, environmental, economic and social aspects. Science of the Total Environment 557-558: 425-437.

Diodato, N., Esposito, L., Bellocchi, G., Vernacchia, L., Fiorillo, F., and Guadagno, F.M. 2013. Assessment of the spatial uncertainty of nitrates in the aquifers of the Campania Plain (Italy). American Journal of Climate Change 02(02): 128-137.

Ducci, D. 2018. An easy-to-use method for assessing nitrate contamination susceptibility in groundwater. Geofluids.

Ducci, D., Corniello, A., and Sellerino, M. 2010. Hydro-stratigraphical setting and groundwater quality status in alluvial aquifers: the low Garigliano River Basin (Southern Italy), case study. Groundwater Quality Sustainability 01:(197-203).

Eamus, D., Fu, B., Springer, A.E., and Stevens L.E. 2016. Groundwater dependent ecosystems: classification, identification techniques and threats. In: Jakeman Barreteau O., Hunt R.J., Rinaudo J.D., Ross A. (eds) Integrated Groundwater Management. Springer, Cham.

Famiglietti, J.S., Lo, M., Ho, S., Bethune, J.,Anderson, K.J., Syed, T. H., Swenson, S. C., de Linage, C. R., and Rodell, M. 2011. Satellites measure recent rates of groundwater depletion in California's Central Valley. Geophysical Research Letters 38 (3).

Famiglietti, J.S., and Rodell, M., 2013. Water in the balance. Science 340:1300-1301. Accessed [12/03/2018]: http://dx.doi.org/10.1126/science.1236460.

Fadili, A., Mehdi, K., Riss, J., Najib, S., Makan, A., and Boutayab, K. 2015. Evaluation of groundwater mineralization processes and seawater intrusion extension in the coastal aquifer of Oualidia, Morocco: hydro-chemical and geophysical approach. Arabian Journal of Geosciences 8(10).

Frappart, F., and Ramillien, G. 2018. Monitoring groundwater storage changes using the gravity recovery and climate experiment (grace) satellite mission: a review. Remote Sensing 10(6): 829.

Gautam, P.K., Arora, S., Kannaujiya, S., and et al. 2017. Sustain. Water Resource Management 3: 441.

Gleeson, T., Wada, Y., Bierkens, M. F. P., and van Beek, L. P. H. 2012. Water balance of global aquifers revealed by groundwater footprint. Nature 488: 197-200.

Gopinath, S., Srinivasamoorthy, K., Saravanan, K., Prakash, R., and Karunanidhi, D. 2019. Characterizing groundwater quality and seawater intrusion in coastal aquifers of Nagapattinam and Karaikal, South India using hydrogeochemistry and modeling techniques. Human and Ecological Risk Assessment: An International Journal. DOI: 10.1080/10807039.2019.1578947.

Goswami, R. R., and Clement, T. P. 2007. Laboratory-scale investigation of saltwater intrusion dynamics. Water Resources Research 43.

Gravity Recovery and Climate Experiment (GRACE) land 2018. Accessed [12/12/2018]: http://grace.jpl.nasa.gov, supported by the NASA Measures Program.

Herbert, E.R., Boon, P., Burgin, A.J., Neubauer, S.C., Franklin, R.B., Ardón, M., Hopfensperger, K.N., Lamers, L.P.M., and Gell, P. 2015. A global perspective on wetland salinization: Ecological consequences of a growing threat to freshwater wetlands. Ecosphere 6(10):1- 43.

Huang, Y., Yang, C., Lee, Y., Tang, P., and Hsu, W. 2010. Variation of groundwater quality in seawater intrusion area using cluster and multivariate factor analysis. Sixth International Conference on Natural Computation (ICNC 2010) 3021-3025.

Katpatal, Y. B., Rishma, C., and Singh, C.K. 2018. Sensitivity of the Gravity Recovery and Climate Experiment (GRACE) to the complexity of aquifer systems for monitoring of groundwater. Hydrogeology Journal (26): 933-943.

Ketabchi, H., Mahmoodzadeh, D., Ataie‐Ashtiani, B., and Simmons, C. T. 2016. Sea‐level rise impacts on seawater intrusion in coastal aquifers: Review and integration. Journal of Hydrology 535: 235-255.

Kerr-Upal, M., Van Seters, T., Whitehead, G., Price, J., and Stone, M. 1999. Assessing the risk of groundwater nitrate contamination in the region of waterloo, Ontario. Canadian Water Resources Journal 24(3): 225-233.

Khanbilvardi, R., Ganju, A., Rajawat, A.S., Chen, J.M., Verma, A., Kumar, A., and Kumar, S. 2016. Analysis of groundwater anomalies using GRACE over various districts of Jharkhand. Proc. of SPIE Vol. 9877.

Kim, S. H., Hejazi, M., Liu, L., Calvin, K., Clarke, L., Edmonds, J., Kyle, P., Patel, P., Wise, M., and Davies, E. 2016. Balancing global water availability and use at basin scale in an integrated assessment model. Climate Change 136(2): 217-31.

Lake, L.R., Lovett,A., Hiscock, K.M., Beston, M., Foley, A., Sunnenberg, G., Evers, S., and Fletcher, S. 2003. Evaluating factors influencing groundwater vulnerability to nitrate pollution: developing the potential of GIS. Journal of Environmental Management 68(3):315-28.

Langevin, C. D., Shoemaker, W. B., and Guo, W. 2003. MODFLOW-2000, The US Geological Survey modular ground-water model-documentation of the SEAWAT-2000 version with the variable-density flow process (VDF) and the integrated MT3DMS transport process (IMT). U.S. Geological Survey Open File Report 03-426.

Lee, E., Jayakumara, R., Shrestha, S., and Han, Z. 2018. Assessment of transboundary aquifer resources in Asia: Status and progress towards sustainable groundwater management. Journal of Hydrology: Regional Studies 20:103-115.

Lockhart, K.M., King, A.M., and Harter, T. 2013. Identifying sources of groundwater nitrate contamination in a large alluvial groundwater basin with highly diversified intensive agricultural production. Journal of Contaminant Hydrology 151:140-154.

Lu, C. H., Chen, Y. M., Zhang, C., and Luo, J. 2013. Steady-state freshwater-seawater mixing zone in stratified coastal aquifers. Journal of Contaminant Hydrology 505: 24-34.

MacDonald, A.M., Bonsor, H.C., Ó Dochartaigh, B.E., and Taylor, R.G. 2012. Quantitative maps of groundwater resources in Africa. Environmental Research Letters 7 (2): 024009.

Malki, M., Bouchaou, L., Hirich, A., Brahim, Y.A., and Choukr-Allah, R. 2017. Impact of agricultural practices on groundwater quality in intensive irrigated area of Chtouka-Massa. Morocco Science Total Environment 574: 760-770.

Mukherjee, A., Nath Bhanja, S., and Wada, Y. 2018. Groundwater depletion causing reduction of baseflow triggering Ganges river summer drying. Scientific Reports 8:12049.

Nijsten, G.J., Christelis, G., Villholth, K.G., Brauned, E., and Gaye, C.B. 2018. Transboundary aquifers of Africa: review of the current state of knowledge and progress towards sustainable development and management. Journal of Hydrology. DOI: 10.1016/j.ejrh.2018.03.004.

Nolan, B. T., Hitt, K. J. and Ruddy, B. C. 2002. Probability of nitrate contamination of recently recharged groundwaters in the conterminous United States. Environment, Science and Technology 36:2138-2145.

Paniconi, C., Khlaifi, I., Lecca, G., and et al. 2001. Transport in Porous Media, 43: 3. Accessed [12/12/2018]: https://doi.org/10.1023/A:1010600921912.

Paniconi, C., Khlaifi, I., Lecca, G., Giacomelli, A., and Tarhouni, J. 2001. Modeling and analysis of seawater intrusion in the coastal aquifer of eastern Cap‐Bon, Tunisia, Transp. Porous Media 43:3-28.

Passarella, G., Vurro, M., D'Agostino, V., Giuliano, G., and Barcelona, M.J. 2002. A probabilistic methodology to assess the risk of groundwater quality degradation. Environmental Modeling & Assessment 79(1): 57-74.

Rodell, M., Chen, J.,Kato, H.,Famiglietti, J.S.,Nigro,J., and Wilson,C.R. 2007. Estimating groundwater storage changes in the Mississippi River basin (USA) using GRACE. Hydrogeology Journal 15(1): 159–166.

Riva, M., Guadagnini, A., and Dell’Oca, A. 2018. Probabilistic assessment of seawater intrusion under multiple sources of uncertainty. Advances in Water Resources (75)93-104.

Rivett, M., Drewes,J., Barrett, M., Chilton, J., Appleyard, S., Dieter, H.H., Wauchope, D., and Fastner, J. (2006). Chemicals: Health relevance, transport and attenuation, in Protecting Groundwater for Health: Managing the quality of Drinking-water sources, edited by O. Schmoll, O. Howard, J. Chilton and I. Chorus. London: IWA Publishing.

Saidi,S., Bouri,S., Ben Dhia, H., and Anselme, B. 2011. Assessment of groundwater risk using intrinsic vulnerability and hazard mapping: Application to Souassi aquifer, Tunisian Sahel. Agricultural Water Management 98(10): 1671-1682.

Sathish S., Elango L., Rajesh R., and Sarma V.S. 2011. Application of three-dimensional electrical resistivity tomography to identify seawater intrusion. Earth Science India 4(I): 21-28.

Strack, O. D. L. 1976. A single-potential solution for regional interface problems in coastal aquifers. Water Resource Research 12:1165-1174.

Stromberg, J.C., Tiller, R., and Richter, B. 1996. Effects of groundwater decline on riparian vegetation of semiarid regions: the San Pedro, Arizona. Ecological Applications 6 (1): 113- 131.

Taylor, R.G., Scanlon, B., Döll, P., Rodell, M., van Beek, R., Wada, Y., Longuevergne, L.,Leblanc, M., Famiglietti, J.S., Edmunds, M., Konikow, L., Green, T.R., Chen, J., Taniguchi, M., Bierkens, M.F.P., MacDonald, A., Fan, Y., Maxwell, R.M., Yechieli, Y., Gurdak, J.J., Allen, D., Shamsudduha, M., Hiscock, K., Yeh, P.J.-F., Holman, I., and Treidel, H. 2012. Groundwater and climate change. Nat. Climate Change. Accessed [01/02/2019]: http://dx.doi.org/10.1038/nclimate1744.

Texas Water Development Board (TWDB) Groundwater Data. Accessed [02/05/2019]: http://www.twdb.texas.gov/groundwater/data/index.asp.

Tomaszkiewicz, M., Abou Najm, M., and El-Fadel, M. (2014) Development of groundwater quality index for seawater intrusion in coastal aquifers. Environmental Modelling & Software 57:13-26.

US Geological Survey (USGS) 2016. Groundwater Availability Report. Accessed [12/12/2018]: https://www.sciencebase.gov/catalog/item/5661981ae4b06a3ea36c56fc.

Van der Schans, M.L., Harter, T., Leijinse, A., Mathews, M.C., and Meyer, R.D. 2009. Characterizing sources of nitrate leaching from an irrigated dairy farm in Merced County. California Journal of Contaminant Hydrology 110: 9-2.

Wada, Y., van Beek, L. P. H., and Bierkens, M. F. P. 2012. Non-sustainable groundwater sustaining irrigation: a global assessment. Water Resource Research 48.

Wahr, J.,Swenson, S., Zlotnicki, V., and Velicogna, I. 2004. Time-variable gravity from GRACE: first results. Geophysical Research Letters 31(11).

Xiao, R., He, X., Zhang, Y., Ferreira, V.G., and Chang, L. 2015. Monitoring groundwater variations from satellite gravimetry and hydrological models: A comparison with in-situ measurements in the mid-atlantic region of the United States. Remote Sensing 7(1): 686- 703.

Yee, J., and Souza, W. 1987. Quality of Ground Water in Idaho. USGS Water Supply Paper 2272. Denver, US: Government Printing Office.

CHAPTER II

FACTORS AFFECTING THE SPATIAL PATTERN OF NITRATE CONTAMINATION IN

TEXAS AQUIFERS

2.1 Abstract

The elevated level of nitrate in groundwater is a serious problem in Texas aquifers.

Identifying the factors affecting the groundwater nitrate concentration is imperative to control groundwater quality. The aim of this study was to determine factors that have significant impacts on the elevated groundwater nitrate concentrations in the Southern High-Plains and the Edwards-

Trinity aquifers. The multivariate statistical analysis (Ordinary Least Square (OLS)) and

Geographically Weighted Regression (GWR) models were used to study the relationship between groundwater nitrate contamination and land use of the study areas. Results showed that in these aquifers, extent of nitrate concentration was depended on well-depth, cotton cultivation, shrubland, and grassland areas. Groundwater nitrate concentration was positively associated with cotton cultivation. The nitrate concentration in the groundwater increased as the cotton production increased in the Edwards-Trinity and the Southern High-Plains aquifers. The nitrate concentrations exceeded 2 mg/L in all 69 wells in the Southern High-Plains aquifer and 123 wells in the Edwards-

Trinity aquifer. The nitrate concentrations tended to increase as the well-depth decreased in both the Southern High-Plains and the Edwards-Trinity aquifers.

2.2 Introduction

Fertilizers and pesticides are essential to increase agricultural production (Ikerd 1999), and nitrogen, phosphorus, and potassium are the three main nutrients used in agriculture. Nitrogen fertilizer is an essential input for agricultural sustainability; however, groundwater nitrate pollution is a worldwide problem (Lake et al. 2003, Schröder et al. 2004, Birkinshaw and Ewen 2000,

Kyllmar et al. 2004). Nitrogen fertilizers have been used in agricultural lands of the United States since the 1850s (Cao et al. 2017). According to the Food and Agriculture Organization (FAO), the consumption of fertilizers was 112.5 kilograms per ha in 2005 and increased to 137 kilograms per ha in 2015 in the United States (Cao et al. 2017). About 11.5 million tons of nitrogen was used annually as fertilizer in agricultural areas of the United States (Puckett 2016), and in 2015, 268 tons of cotton and 6,123 tons of corn were infused with nitrogen fertilizer in the United States

(U.S. NASS 2018). The amount of nitrogen fertilizers used in agriculture depends on the type of crops grown, soil quality, and the local climate (Liu et al. 2005).

Nitrate is soluble and has high mobility, as well as the potential to move through an unsaturated zone by leaching (Chowdary et al. 2005). Many studies have shown the impacts of agricultural crop production on nitrate concentrations in groundwater (Harter et al. 2002, Jordan and Smith 2005, Dunn et al. 2005, Liu et al. 2005). Harter et al. (2002) assessed nitrate and salt leaching to shallow groundwater in semiarid vulnerable hydrogeologic regions, and they found that shallow groundwater vulnerability to nitrate contamination varied spatially. Jordan and Smith

(2005) employed a Geographic Information System (GIS) to estimate losses of nitrate from agricultural lands to groundwater based upon livestock data and a leaching factor in Northern

Ireland. Dunn et al. (2005) predicted nitrate concentrations for ungauged catchments in order to fill gaps in monitoring data and offer guidelines in relation to policy development in Scotland. Liu 18

et al. (2005) used land use data, well depth, septic tank use, distance from the well, and livestock and cropping activities around wells to predict nitrate concentration in groundwater in Thailand.

They found that nitrate contamination of groundwater occurs in predictable patterns (U.S. EPA

2001). Additionally, cropland regions with well-drained soils are most at risk for groundwater nitrate contamination (Wick et al. 2012).

Natural groundwater generally has a nitrate concentration of less than 2 mg/L (Muller and

Helsel 1996) and the nitrate maximum contamination threshold is 10 mg/L (U.S. EPA 1995). The adverse effects of chemical fertilizers and pesticides on both humans and the environment cannot be overlooked, especially in relation to groundwater (Igbedioh 1991). Consumption of nitrate in drinking water causes low oxygen levels in the blood, which is a potentially fatal condition (Ward et al. 1996), and along with health risks, groundwater contaminated with toxic and chemical constituents takes several years to clean up (U.S. EPA 2001).

Nitrate contents in groundwater exceed 10 mg/L in the Southern High-Plains aquifer,

Edward-Trinity, Seymour, and Southern Gulf Coast aquifers in Texas (Scanlon et al. 2004). More than 2 million ha of land rely on groundwater to irrigate crops in Texas (USDA 2019). High levels of nitrate in groundwater in coastal versus inland areas of Texas pose a serious threat to groundwater quality and pose a potential hazard to sustainable crop production. The highest nitrate concentrations were found under irrigated agriculture, while moderate levels of nitrate originated under dryland agriculture and very low levels of nitrate were found under rangeland settings (Texas

State Soil and Conservation Board 2012).

Many studies have investigated nitrate groundwater contamination in Texas (Kreitler 1975,

Kreitler and Jones 1975, Kreitler and Browning 1983, Bartolino 1994, Fryar et al. 2000, Hudak

2018, Juntakut et al. 2019, Opsahl et al. 2017). Kreitler (1975) developed a nitrogen isotope technique for distinguishing different sources of nitrate contamination; studies of nitrate contamination in alluvial aquifers in central Texas indicated that fertilizers are major sources of groundwater contamination (Stewart et al.1994, Daniel 1997). Hudak (2018) assessed that high nitrate concentration in the Seymour aquifer is related to ongoing agricultural practices, including the cultivation of native grassland, the application of fertilizers, and irrigation with nitrate- contaminated groundwater. Concentrations of nitrate ranged from 30.5 mg/L in 1999 to 100.5 mg/L in 2015 in Texas aquifers (Hudak 2018). Juntakut et al. (2019) studied the long-term effect of the agricultural, vadose zone, and climatic factors on nitrate contamination. Spatial statistics and regressions were used to investigate the relationships between groundwater nitrate concentrations and several potential natural and anthropogenic factors, including soil drainage capacities, vadose zone characteristics, crop production areas, and irrigation systems (Juntakut et al. 2019). Areas with a thicker vadose zone and larger saturated thickness of the aquifer have relatively lower nitrate concentrations (Juntakut et al. 2019). Opsahl et al. (2018) studied water quality changes under a range of hydrologic conditions and in contrasting land cover settings (rural and urban) in the Edwards-Trinity aquifer in Texas. They found that the Edwards aquifer is susceptible to pollution, and land cover, aquifer hydrogeology, and changes in hydrologic settings that affect the vulnerability.

The objectives of this study were to analyze the extent of nitrate concentration in groundwater in the Southern High-Plains and Edwards-Trinity aquifers of Texas in the United

States and to identify land uses that have significant impacts on nitrate concentrations of

groundwater using multivariate statistical analysis and GWR models in the Southern High-Plains and Edwards-Trinity aquifers in Texas.

2.3 Background

Groundwater serves as the only drinking source in many parts of the world, and it is also the world’s most extracted raw material with withdrawal rates in the range of 982 km3/year (Shukla and Saxena 2018, Margat and Gun 2013). Nitrogen occurs naturally in groundwater; however, nitrogen fertilizers and manure applied to soil can increase the amount of nitrate in groundwater

(Schilling and Wolter 2001, Liu et al. 2005, Shamrukh et al. 2001, Harter et al. 2002, Almasri and

Kaluarachchi 2004, Chowdary et al. 2005). Increased nitrate levels (more than 2 mg/L) in groundwater used for drinking can cause adverse health effects (Mueller and Helsel 1996); drinking water with nitrate concentrations of 4 mg/L increases the risk of non-Hodgkin’s lymphoma (Ward et al. 1996).

Precipitation can further transport dissolved nitrogen into groundwater (Almasri and

Kaluarachchi 2005). Various factors can cause nitrogen fertilizer to leach into groundwater, such as land use, on-ground nitrogen loading, recharge of water used in irrigation water, soil nitrogen dynamics, soil chemical characteristics, and soil depth (Vinten and Dunn 2001, Dunn et al. 2005,

Almasri and Kaluarachchi 2003). It will be dispersed with the groundwater flow once nitrate leaches into groundwater (Shamrukh et al. 2001, Almasri and Kaluarachchi 2005). Identification of areas with heavy on-ground nitrogen loadings is significant for land use planners and decision- makers (Tesoriero and Voss 1997).

Statistical analysis was a commonly used technique (Wick et al. 2012, Nolan et al. 2002,

Nolan and Stoner 2000, Nolan 2001, Scanlon et al. 2004, Shrestha and Luo 2017) to study land

use activities’ impacts on groundwater contamination. Regression analysis was used to determine factors influencing groundwater nitrate contamination (Wick et al. 2012). Logistic regression was used to study the effects of nitrogen fertilizers and pesticides on groundwater contamination

(Nolan et al. 2002, Nolan and Stoner 2000, Nolan 2001). Scanlon et al. (2004) used multivariate logistic regression to study nitrate concentrations in shallow wells (≤ 30 m). Shrestha and Luo

(2017) used Principal Component Analysis (PCA) and Geographically Weighted Regression

(GWR) to find the most effective environmental variables that control groundwater nitrate concentration in the Central Valley (CV) aquifer.

Many studies have identified sources of groundwater nitrate contamination based on the local analysis in various parts of the United States. (Van der Schans et al. 2009, Lockhart et al.

2013). Van der Schans et al. (2009) provided a map of the spatial variation associated with the impact of dairy farming on groundwater nitrogen levels; they measured local groundwater heads and nitrate concentrations in a monitoring well network. Variability in crops, soil type, and depth to groundwater contributes to the large variability in nitrate occurrence across the underlying aquifer system. Lockhart et al. (2013) sampled two hundred wells and studied vadose zone thickness, soil type, well proximity to dairies, and dominant land use near the well to examine the spatial variation for groundwater nitrate contamination.

Nitrate is an essential part of the environmental nitrogen cycle, and biological nitrogen fixation accounts for 67.25 percent of nitrogen fixed per year. The remaining 32.75 percent is fixed through non-biological processes, such as atmospheric lightning and industrial discharges

(Miyamoto et al. 2008, Shukla and Saxena 2018). The nitrogen fixed in the soil undergoes the

- process of ammonification and gets converted to ammonium (NH3 ), which is further converted

- - first to nitrite (NO2 ) and then to nitrate (NO3 ). Rain and irrigation water can leach nitrate into groundwater as nitrate-containing compounds are generally soluble and readily percolate into groundwater (Self and Waskom 2014).

Ecosystems have the potential to retain nitrogen accumulation and transfer it to other ecosystems (Shukla and Saxena 2018). Agricultural systems have low nitrogen accumulation potential, but the high transfer potential shows a risk of pollution in other ecosystems, especially transferring into the groundwater. A forest’s natural tendency is to retain and accumulate nitrogen in the soil pool (Shukla and Saxena 2018); as such Miller et al. (2005) documented that the runoff immediately after the rain had 150-400 mg/L of nitrate in groundwater. The increasing nitrate concentration has adverse effects on various ecosystems and may cause algal bloom, eutrophication, and an increase in NOx emissions (Galloway et al. 2003, Zhou et al. 2011).

Additionally, several minerals include Tobelite, Niter, and Nitratine, have nitrogen in their lattice, and this nitrogen is converted to nitrate when it is released from the lattice through the weathering process (Holloway and Dahlgren 2002). The nitrate formed during the weathering process may leach into the aquifer and pollute groundwater (Holloway and Smith 2005, Shukla and Saxena

2018). Collectively, croplands and other agricultural activities are the major factors associated with elevated nitrate concentration in groundwater (Azizullah et al. 2011).

2.4 Study Area

Nitrogen fertilizer application is higher in Texas, Mississippi, and Georgia (Gronberg and

Spahr 2012). We studied the groundwater nitrate contamination in two major aquifers in Texas

(Fig.2.1), the Southern High-Plains and the Edwards-Trinity aquifers. Nitrate concentration in the

Southern High-Plains aquifer exceeded the EPA designated threshold of 10 mg/L in drinking water

resources since the 1980s (U.S. EPA 2001). The High-Plains aquifer is the principal source of water in the agricultural areas in the United States. The aquifer extends eight states with most of its area in Nebraska, Texas, and Kansas (Fig.2.1), and the aquifer is divided into the Northern

High-Plains, the Central High-Plains, and the Southern High-Plains areas (USGS 1984). The

Southern High-Plains aquifer has moderate precipitation, low humidity, and high evaporation

(Sanderson and Frey 2014).

About 3 percent of the High-Plains aquifer comprises water exceeding 1,000 mg/L dissolved solids, most of which are in Texas (USGS 1984). In most areas of the High-Plains aquifer, where the concentration of dissolved solids exceeds 1,000 mg/L, the chemical composition of the water is influenced by the underlying bedrock. Most of the High-Plains aquifer consists of sand, gravel, clay, and silt and has a maximum thickness of 800 feet. The unconsolidated sediment has enough permeability to allow contaminants to flow through the High-Plains aquifer (USGS

1984).

The long-term trend in groundwater contamination in the Southern High-Plains aquifer is due to excessive agricultural activities which affect the sustainable use of groundwater. Higher levels of nitrates were found in the shallow wells of the Southern High-Plains aquifer (Chaudhuri and Ale 2014). Approximately half of the Southern High-Plains groundwater used in the agriculture consisted of four dominant crops: cotton, corn, sorghum, and wheat (Scanlon et al.

2004).

The Edwards-Trinity aquifer is characterized by high hydraulic conductivity transmissivity

(Lindgren 2006). The aquifer is located in west-central Texas (Fig.2.1). The Edwards-Trinity aquifer has a land use with thin soils and is principally used as rangeland where the agricultural

lands have decreased in recent years (Musgrove et al. 2016). The Edwards-Trinity aquifer is composed of sandstone (upper zone) and carbonate rock (lower zone). Karst features make the groundwater system vulnerable to contamination because of the rapid mobility of groundwater through high-porosity voids in the Edwards-Trinity aquifer (White 1988). The topography contains a gently rolling plain to the east and moderately hilly country to the west. The amount of recharge and discharge differs significantly from county to county within the area because of such features as topography (USGS 1984).

Figure 2.1 Study areas: the Edwards-Trinity and the High-Plains aquifers (WHYMAP 2011).

2.5 Data and Methods

The relationship between nitrate concentrations in groundwater and land use was examined using multivariate regression analysis and geographically weighted regression model. A database of nitrate measurements from private wells located in Texas was compiled from the Texas Water

Development Board (TWDB 2019). TWDB has provided a water quality database for groundwater in Texas wells. The purpose of the program is to monitor changes in groundwater quality over time in major and minor aquifers in the state. TWDB did not provide a comprehensive monthly data set for nitrate variations of wells in all major aquifers. Among all of the major aquifers, the Southern

High-Plains and Edwards-Trinity had a comprehensive data set to further analyze (TWDB). We selected samples from private wells because they are typically shallower than monitoring wells

(Ransom et al. 2018). Shallow groundwater is more vulnerable to nitrate contamination. Other

- - forms of nitrogen, such as ammonia (NH3 ) and nitrite (NO2 ) were not used in this study because of low concentrations (Shamrukh et al. 2001). A total of 192 groundwater samples were used, which include 69 groundwater samples from the Southern High-Plains aquifer and 123 groundwater samples from the Edwards-trinity aquifer. The TWDB database prepared for water quality samples between 1980 and 2017; however, only the last ten years of recorded water quality data between 2008 and 2017 years were used, which were collected in the spring and summer seasons from 2008 to 2017. However, water quality data for 69 wells in the Southern High-Plains aquifer was available only for 2008, 2011, 2012, and 2016.

Information on land use was compiled from the National Land Cover Data (NLCD). Land use data updated over the last ten years between 2008 and 2017 were utilized (USDA NASS 2018).

This product was developed by the Multi-Resolution Land Characteristics (MRLC) consortium.

The development of the NLCD was directed by USGS and EPA (National Land Cover Database). 26

The NLCD was derived from images acquired by the Landsat Thematic Mapper sensor. It is a land cover dataset that contains a detailed legend of crop types (e.g., corn, wheat, soybeans, cotton, rice), and produced by the U.S. Department of Agriculture (USDA), and National Agricultural

Statistics Service (NASS) with a spatial resolution of 30 m. NLCD is a 56-class land cover classification scheme (Boryan et al. 2011). The minimum mapping unit of this data was 0.4 ha and this means that small patches of vegetation can be missed in the modeling process (USDA 2017).

We performed an analysis to relate groundwater nitrate to surrounding land uses. Generally, northern and northeastern Texas has cotton as the dominant land use class. While, central Texas has more shrublands, and the eastern Texas has more grasslands. Corn, sorghum, and winter wheat are some of the other agricultural crops grown in this region. The percent areas were quantified into three land use and land cover categories: cotton, grassland, and shrubland within a 1-kilometer radius of each well (Fig.2.2).

Figure 2.2 Texas land uses in 2017 (USDA 2017).

2.5.2 Multivariate Statistical Analysis

Ordinary least square regression (OLS) was used in this study to investigate the relationship between groundwater nitrate contamination and land use activities. Two models were utilized – one for the Southern High-Plains aquifer and one for the Edwards-Trinity aquifer. Well-depth, cotton, shrubland, and grassland were employed as independent variables; the nitrate concentration was used as the dependent variable expressed in Equation [2.1]:

Y= β0+ β1 X1 + β2 X3 + …. [2.1]

where Y is a dependent variable which denotes nitrate concentration and X1, X2, X3, and

X4 are independent variables which indicate well-depth, cotton, shrubland, and grassland, respectively.

2.5.3 Geographically Weighted Regression

The geographically weighted regression (GWR) model was further used to determine the effect of spatial heterogeneity on the explanatory variable. GWR assumes that the relationship varies across the study area (non-stationary). This method models the spatially varying relationships by generating individual regressions for each data point, where nearby observations are given more weight. This method minimizes the spatial autocorrelation in the residuals and generates local coefficient maps to observe the spatial heterogeneity over the area (Fotheringham et al. 2002). GWR models have been used to find spatial relationships. Kundzewicz and Doll

(2009) studied the relationship between climate change scenarios and groundwater decreases in areas with a high social vulnerability index (Kundzewicz and Doll 2009). The GWR was also used to test the impact of agricultural intensity, industrial and tourism concentration, and urban growth on the water quality index (Yu et al. 2012).

The GWR model was used to analyze the spatially varying relationship between well- depth, groundwater nitrate concentration, and land uses. The GWR model is an expansion on the traditional multiple linear regression model by estimating the local parameters as follows [2.2]:

Yi = β0 (u i, v i) + β1 (u i,v i) xi1 + ⋯ + βnx in + ε I [2.2]

where Yi is the dependent variable at location (ui, vi), (ui, vi) indicates the coordinates of the i th data point and β0 and β1 are continuous functions of (u, v) at the point i, and xin is the nth independent variable at location (ui, vi) (Fotheringham et al. 2002). The coefficients (β0 and β1) in

GWR are functions of spatial location compared with the traditional multilinear regression model.

Nearer observations for each regression point have a greater effect on estimating the local set of coefficients (βo and β1) than observations farther away in the GWR model. GWR allows coefficients to vary over the study area and a set of coefficients can be estimated at any location so that a coefficient surface can be visualized and questioned for relationship heterogeneity

(Fotheringham et al. 1998). The Gaussian kernel was used in GWR, and the weight for the

Gaussian kernel function is assessed as follows [2.3]:

Wij = exp[-(dij/b)2 ] [2.3]

where dij is the Euclidean distance between point i and the data point j, and b is a fixed bandwidth size defined by a distance metric measure. The Gaussian kernel weight continuously and regularly decreases from the center of the kernel but never reaches zero. The Gaussian kernel is appropriate for fixed kernels since it can mitigate the risk of there being no data within a kernel

(Fotheringham et al. 2002). The Akaike information criterion was used to assess the goodness of fit for models [2.4].

(AICc): AICc = Deviance + 2k [n/ (n-k-1)] [2.4]

where n is the number of data points and k is the number of parameters in the model. Lower values for AICc indicate better model fits.

2.6 Results and Discussion

The nitrate concentrations exceeded 2 mg/L in all 69 wells in the Southern High-Plains aquifer and 123 wells in the Edwards-Trinity aquifer. The nitrate concentrations tended to increase as the well-depth decreased in both the Southern High-Plains and the Edwards-Trinity aquifers

(Fig. 2.3).

180 90 160 80 140 70 120 60 100 50 80 40

60 30 Nitrate (mg/L) Nitrate 40 (mg/L) Nitrate 20 20 10 0 0 0 200 400 600 800 1000 0 50 100 150 200 Well-Depth(m) 2-3a Well-depth (m) 2-3b

Figure 2.3 Illustrating nitrate and well-depth relation in the Edwards-Trinity aquifer (2.3a) and the Southern High-Plains aquifer (2.3b).

In the Southern High-Plains aquifer, the maximum nitrate concentration was 81 mg/L, and the minimum nitrate concentration was 2 mg/L (see Table 2.1). Fifty wells contained nitrate concentrations greater than 5 mg/L, and 35 wells in the Southern High-Plains aquifer exceeded the maximum nitrite contamination threshold of 10 mg/L (see Table 2.1). In the Southern High-Plains, all 69 wells were located in an unconfined aquifer, where they are most susceptible to contamination.

In the Edwards-Trinity aquifer, the maximum nitrate concentration was 157.6 mg/L, and the minimum nitrate concentrations was 2 mg/L (see Table 2.1). Here, 111 wells contained a nitrate concentration greater than 5 mg/L; furthermore, 58 wells exceeded the maximum contamination threshold of 10 mg/L. Out of the 123 wells that we analyzed, 92 wells were found in an unconfined aquifer. Collectively, the nitrate concentrations exceeded the maximum 10 mg/L in about 50 percent of the wells of the Southern High-Plains and the Edwards-Trinity aquifers.

Croplands were the major factors associated with high nitrate concentration in groundwater. For example, nitrogen fertilizers were used at an average rate of 61.65 kg ha-1 of cotton in Texas in 2017 (U.S. NASS 2017). Nitrogen accumulation in the soil leaches into the aquifers through runoff, and especially transfers into shallow aquifers (Shukla and Saxena 2018).

Table 2.1 Nitrate (N) concentration in the Southern High-Plains aquifer and the Edwards- Trinity aquifer.

Aquifers Max N Min N Number of Number of Number of Number of Number of (mg/L) (mg/L) wells wells wells Confined Unconfined (N>2(mg/L)) (N>5(mg/L)) (N>10(mg/L)) Wells Wells

Southern 81 2 69 50 35 - 69 High-Plains

Edwards- 157.6 2 123 111 58 33 92 Trinity

The results of the trend analysis of groundwater nitrate contamination are presented in Fig.

2.4. The nitrate concentration exceeded the EPA threshold of 10 mg/L for many years in wells of both the Edwards-Trinity and the Southern High-Plains aquifers (Fig. 2.4). As shown in Fig. 2.4a, the nitrate concentration decreased at a moderate rate (-0.12) from 2008 to 2017 in the Edwards-

Trinity aquifer. The nitrate concentration decreased at a high rate (-3.10) from 2008 to 2016 in the

Southern High-Plains aquifer, as illustrated in Fig. 2.4b. However, in every year, the nitrite concentration exceeded the EPA threshold of 10 mg/L (EPA 2001).

As agriculture contributes significantly to the nitrate concentration in groundwater, we analyzed the land use pattern within 1 km of each sample well. In the Edwards-Trinity aquifer, cotton, shrubland, and grassland are the dominant land use patterns within 1 km of the sample wells. These major land use patterns varied from 26.6 percent (in 2012) to a maximum of 87

percent in 2015 (see Table 2.2). A change in the cropping pattern will affect the amount of fertilizer applied and thus affect the nitrate concentration in the groundwater. However, the overall nitrate concentration (Fig. 2.4a) stayed above the threshold in every year from 2008 to 2016. Similarly, in the Southern High-Plains, cotton, shrubland, and grassland are also the dominant land use patterns within 1 km of the sample wells. These land use patterns varied from 51.2 percent in 2008 to 70.2 percent in 2012 (see Table 2.3). In every year from 2008 to 2016, the nitrate concentrations were well above the threshold (Fig. 2.4b).

20 35 - 30 15 25

10 20 Plains - 15 5

High 10 Trinity Aquifer Aquifer Trinity 5

0 Nitrate (mg/L) Southern Southern Nitrate (mg/L) Nitrate (mg/L) Edwards Nitrate (mg/L) 0 2008 2011 2012 2016 2.4a 2.4b

Figure 2.4 Average nitrate concentration from 2008 to 2017 (2.4a) in the Edwards-Trinity and (2.4b) the Southern High-Plains aquifers.

Table 2.2 Land use patterns in the Edwards-Trinity aquifer from 2008 to 2017.

2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

Cotton 0.6% 0.3% 6.7% 17.3% 0.2% 0.5% 1.7% 11.7% 25.6% 5.4%

Shrubland 32% 25.8% 57.8% 38.4% 21% 23.8% 35.1% 62.6% 28.7% 39%

Grassland 7.2% 4.7% 4.7% 16.8% 5.4% 8% 4.5% 12.7% 1% 18.2%

Total 39.8% 30.8% 69.2% 72.5% 26.6% 32.3% 41.3% 87% 55.3% 62.6%

Table 2.3 Land use patterns in the Southern High-Plains aquifer from 2008 to 2016.

2008 2011 2012 2016 Cotton 16.6% 26.7% 35.3% 32.5% Shrubland 18.8% 36% 17.6% 16% Grassland 15.8% 24.4% 17.3% 20.6% Total 51.2% 87.1% 70.2% 69.1%

The nitrate concentrations in the wells in the Edwards-Trinity aquifer had less nitrate concentration as compared with the wells in the Southern High-Plains aquifer. This could be because the Southern High-Plains had a higher percentage of land used for growing cotton compared with that of the Edwards-Trinity aquifer. The unconsolidated sand and gravel have enough permeability to allow contaminants to flow through the High-Plains aquifer. In both aquifers, the nitrate concentration trend correlated well with the percentage area under the cotton crop (Figs. 2.5a and 2.5b). The amount of nitrate that percolates to groundwater can be highly variable between different crops or land use types. This is owing to the differences in crop nutrient demands, soil and climate conditions, and farm management techniques (Ransom et al. 2018).

180 y = 1.2x + 14.7 90 y = 0.15x + 13.7 R² = 0.49 R² = 0.05 160 80 140 70 120 60 100 50 80 40

60 30 Nitrate (mg/L) Nitrate

40 20 Nitrate (mg/L) Nitrate 20 10 0 0 0 20 40 60 80 100 0 50 100 Cotton Area (%) 2.5a Cotton Area (%) 2.5b

Figure 2.5 Relationship between nitrate concentration and cotton cultivation in the Edwards- Trinity aquifer (2.5a) and the Southern High-Plains aquifer (2.5b).

From the average of area under different land use for all the years within the 1 km buffer around each of the sampled wells, it is clear that farming is predominant in the Southern High-

Plains aquifer (Table 2.4). On an average between 2008 and 2016, 18ha (5 percent) corn, 104ha

(31 percent) cotton, 8.5ha (2 percent) sorghum, 26.4ha (6.8 percent) wheat, 56.4ha (19 percent) grassland, and 51.4ha (20 percent) shrubland were found in the Southern High-Plains aquifer.

Contrarily, the Edwards-Trinity aquifer region is predominantly a shrub land with shrubland (190ha (60 percent)), cotton (32.4 ha (12 percent)), corn (10.2ha (3 percent)), and grassland (34.1 ha (10 percent) (see Table 2.4). Nitrogen is the crucial nutrient for influencing grassland yields (Kidd et al. 2017). The application of nitrogen fertilizer to increase grassland yields and the addition of animal waste to soil were the sources of nitrate in the groundwater under shrubland and grassland in the Edwards-Trinity aquifer.

Table 2.4 Average land use patterns in one-kilometer buffer around wells from 2008 to 2017.

Southern High-Plains Edwards-Trinity Land Use Patterns Area (ha) Area (ha) Corn 18 (5%) 10.2 (3%) Cotton 104(31%) 32.4 (12%) Sorghum 8.5(2%) - Wheat 26.4 (6.8%) - Grassland 56.4 (19%) 34.1 (10%) Shrubland 51.1(20%) 190 (60%)

2.6.2 Multivariate Statistical Analysis

This study then evaluated the groundwater nitrate contamination using multivariate regression analysis. Sixty-nine wells of the Southern High-Plains aquifer and 123 wells of the

Edwards-Trinity aquifer were chosen to study the factors controlling groundwater nitrate concentrations.

Ordinary Least Square (OLS) results showed that the elevated nitrate concentrations were significantly negatively related to well-depth, positively related to the percent of the cotton area, and negatively related to the percent of the grassland area in the Southern High-Plains aquifer (see

Table 2.5). The inverse relationship between well-depth and nitrate concentration may be related to the fact that shallow, unconfined aquifers are most susceptible to nitrate contamination, as the average well-depth in the Southern High-Plains was 64 m.

Therefore, increased nitrate contamination was associated with the increased cotton production, which covers 31 percent of the area around the wells in the Southern High-Plains aquifer. Furthermore, the inverse relationship between decreased groundwater nitrate concentration and grassland may be associated with less cultivation of grassland in the Southern

High-Plains aquifer (see Table 2.5).

Table 2.5 Results of the OLS regression analysis in the Southern High-Plains aquifer.

Multivariate Statistical Analysis Descriptive Statistics Coefficient Std. Error Mean Std. Deviation (Constant) 33.78 7.25 Well-Depth -0.05* 0.02 64 m 96.12 Cotton 0.03** 0.08 31% 25.60 Shrubland 0.08 0.10 16.38% 21.56 Grassland -0.34*** 0.12 15.18% 16.99 Adjusted R Square 0.23 - - - AIC C 978 - - - Number of 69 69 Observation *** p<0.001; **p<0.01; * p<0.05

In the Edwards-Trinity aquifer (see Table 2.6), the OLS results showed that elevated nitrate concentrations were significantly negatively related to well-depth and positively related to the

percent of the cotton area, the percent of the grassland area, and the percent of the shrubland area.

The inverse relationship between well-depth and nitrate concentration may be associated with the fact that shallow, unconfined aquifers are most vulnerable to nitrate contamination. The average well-depth in the Edwards-Trinity aquifer is 136 m. The increased nitrate contamination is a problem related to increased cotton production, which covers 9.6 percent of the areas around wells.

Apart from the increased cotton production, the elevated nitrate contamination was associated with the increased shrubland and grassland. Grassland covers 7.4 percent, and shrubland covers 47.7 percent of the areas around the wells. An adjusted R square (.45) explained the variations of the nitrate concentration by well-depth, cotton production, shrubland, and grassland in the Edwards-

Trinity aquifer (see Table 2.6).

Table 2.6 Results of the OLS regression analysis in the Edwards-Trinity aquifer.

Multivariate Statistical Analysis Descriptive Statistics Coefficient Std Error Mean Std. Deviation (Constant) 11.1*** 3.76 - - Well-Depth -0.006** 0.003 136 m 481.3 Cotton 1.2*** 0.12 9.6% 19.6 Grassland 0.21* 0.13 7.4% 11.2 Shrubland 0.05** 0.04 47.7% 30.9 Adjusted R Square 0.45 - - - AICC 887 - - - Number of 123 123 Observations *** p<0.001; **p<0.01; * p<0.05

Therefore, increased levels of nitrate contamination found in the wells of both aquifers can be attributed to the use of nitrogen fertilizer to improve grassland and cotton productivity. Animal waste was a source of nitrate contamination in groundwater under shrubland and grassland, as well

(Galloway et al. 2008).

2.6.2 Geographically Weighted Regression

A simple linear regression model did not completely account for the spatial variability in the nitrate concentration and explanatory variables. The low value of an adjusted R square (see

Tables 2.5 and 2.6) indicates that a simple linear regression had a low goodness of fit.

Geographically weighted regression models were developed to consider the spatial heterogeneity in the nitrate concentration and explanatory variables as a result.

The geographically weighted regression addresses the problem of spatial heterogeneity

(Fotheringham et al. 2003) that occurs when relationships between variables vary among neighborhoods with different characteristics. The results of the GWR models are presented in

Table 2.7. The table shows the minimum and maximum values within the study area. The models indicate large differences between the minimum and maximum values within the study area.

Figures 2.6 and 2.7 show the spatial distribution of the GWR coefficients of cotton, shrubland, grassland, and well-depth in the Southern High-Plains and Edwards-Trinity aquifers. Figure 2.6 and 2.7 also denotes the spatial distribution of the R square and SE (standard error) for each model.

The GWR models in all cases indicate that each independent variable had a significant influence on the nitrate concentration, but that the effect varied spatially (Figs. 2.6 and 2.7). Cotton and shrubland had minimum coefficient values in negative integers and maximum values in positive integers in the Southern High-Plains aquifer. All of the independent variables had minimum coefficient values in negative integers and maximum values in positive integers in the Edwards-

Trinity aquifer (see Table 2.7).

Table 2.7 GWR Coefficients.

Southern High- Plains Edwards -Trinity (Model 2) (Model 1) Min Max Min Max Constant 16.67 48.43 7.10 16.08 Well-depth -0.11 -0.01 -0.02 0.004 Cotton -0.04 0.17 -14.64 2.47 Shrubland -0.11 0.05 -0.05 0.51 Grassland -0.38 -0.21 -14.86 0.80 Adjusted R Square 0.26 0.52 AIC 850 843

Fixed Bandwidth, Type=AICC

The values of the OLS and GWR models were compared to test the robustness of the models. Compared with the OLS models (see Table 2.5 and Table 2.6), the GWR models (see

Table 2.7) have lower AIC scores and higher adjusted R2 values, which shows the robustness of the GWR models over the OLS models.

Figure 2.6 Spatial patterns of GWR coefficients in the Southern High-Plains aquifer.

Figure 2.7 Spatial patterns of GWR coefficients in the Edwards-Trinity aquifer.

Therefore, as seen in Fig.2.6 and 2.7, in the Southern High-Plains and the Edwards-Trinity aquifers, the elevated groundwater nitrate concentrations are linked to well-depth, cotton production, grassland, and shrubland. Throughout the Southern High-Plains aquifer, as the well- depth decreased, the nitrate concentration in the groundwater increased. Fig. 2.6 shows that the nitrate concentration in the groundwater increased in the eastern and northeastern parts of the

Southern High-plains aquifer as cotton production and shrubland increased; however, the nitrate concentration and grassland have an inverse relationship. The nitrate concentration in the groundwater decreased as the grassland land increased throughout the aquifer (Fig.2.6).

Throughout the Edwards-Trinity aquifer, as the well-depth increased, the nitrate concentration in the groundwater increased. In central-northwestern Edwards-Trinity aquifer, the cotton production and nitrate concentration were positively associated. Likewise, throughout the

Edwards-Trinity aquifer, grassland and the nitrate concentration in the groundwater were positively associated (Fig.2.7).

Nitrate concentrations in wells vary with location and well-depth. The mapping of a groundwater nitrate concentration can be useful in groundwater resource management. An estimation of the magnitude of the nitrate leaching from agricultural activities into the water of aquifers is essential to design appropriate policies that monitor agricultural practices.

The risk of groundwater nitrate contamination varies across the Southern High-Plains and

Edwards-Trinity aquifers. Groundwater nitrate contamination is related to population density, well-drained soils, and highest cropland areas. Nitrogen losses from the agricultural activities end up as high groundwater nitrate concentrations, which have adverse consequences on human and environmental health, such as through the eutrophication of groundwater-dependent ecosystems

(Zhou et al. 2011); however, Esmaeili et al. (2014) demonstrated that using high nitrate concentration groundwater for irrigation can reduce the necessity for inorganic fertilizers and reduce the cost of cultivation and nitrate contamination.

The source of nitrate in Texas groundwaters is uncertain. A variety of natural and human factors affect groundwater nitrate contamination (Scanlon et al. 2004). A high nitrate content in the groundwater can be used in irrigation, and the application of nitrogenous fertilizers can be considerably minimized for crop production.

Accuracy and Limitation

The TWDB dataset used different data gathering methods over the years that could have introduced inaccuracy in the data. The TWDB also reports that recording of nitrate as N and NO3 is not consistent in the data. This may have introduced some inaccuracy in our final results.

Similarly, the well depth was recorded based on the "interpretation" (memory) of the person who originally reported the data to an agency which might have included some error in the well depth data (http://www.twdb.texas.gov/groundwater/faq/faqgwdb.asp).

The accuracy of NLCD data also varied from year to year. The accuracy of 2017 NLCD data was 87 percent, while 2016 NLCD accuracy was 88 percent. Accuracy of 2015, 2014, 2013,

2012, 2011, 2010, 2009, 2008, and 2007 NLCD data were 88 percent, 85 percent, 88 percent, 89 percent, 87 percent, 89 percent, 88 percent, 89 percent, and 86 percent respectively (Yang et al.

2018, Wickham et al. 2017, Wickham et al. 2013). Groundwater nitrate contamination includes complex hydrogeochemical processes that cannot be fully understood through use of Land Use and Land Cover data. It is suggested that geological, hydrological and atmospheric parameters need to be included to further assess the extent of nitrate concentration.

2.7 Conclusions

The study identified spatial variations in the groundwater nitrate concentration of the

Southern High-Plains and the Edwards-Trinity aquifers. The nitrate concentration showed significant spatial variation. The relation of elevated nitrate concentration with well-depth, cotton, shrubland, and grassland varied spatially. Nitrate concentration associated with well-depth, land use, and land cover patterns does not vary in the same way. For example, the nitrate contaminant level associated with grassland was increased in the Edwards-Trinity aquifer compared with the

Southern High-Plains aquifer because grassland is a potential factor to groundwater contamination

in the Edwards-Trinity aquifer. Nitrate concentration increased as cotton production increased in the Southern High-plains and Edwards-Trinity aquifers. Increased nitrate concentration in the

Southern High-Plains aquifer is related to cotton production because farming is a major cause of groundwater pollution in the Southern High-Plains aquifer. There are certain elements of each land use pattern that can contribute to the increasing groundwater nitrate contamination because the combined land uses around wells influence the nitrate contamination of groundwater.

Cotton production and grassland receive the highest levels of nitrogen fertilizer, and groundwater nitrate budgets are associated with each of the hydrological characteristics, land use, and land cover around the wells in the study; therefore, it is important to contain spatial variations of nitrate concentration when analyzing groundwater nitrate contamination.

Nitrogen fertilizer has long been recognized as an important indicator of soil productivity.

This study has shown that a nitrogen fertilizer application to the grasslands in the Edwards-Trinity aquifer is increased compared with the Southern High-Plains aquifer. It seems that nitrogen fertilizer is mostly used in the farming practices in the Southern High-Plains aquifer compared with the Edward-Trinity aquifer, where nitrogen fertilizer is mostly used to improve grassland yields. Our study indicates nitrate contaminants in the east part of the Southern High-Plains and northwest part of the Edwards-Trinity aquifers. Meanwhile, nitrate is one of the most pervasive contaminants in the groundwater in Texas.

The groundwater nitrate concentration must be understood within a geographic framework to ensure the sustainable use of groundwater, which is important for local management purposes.

The analysis should include local spatial variations of elements such as hydrologic characteristics and the land use activities if groundwater nitrate contamination causes adverse effects on human

and ecosystem health. Groundwater nitrate contamination should be assessed to adjust farm- management practices. It is vital to define nitrate high-risk aquifers in order to prioritize regulatory oversight and assistance efforts in these areas.

2.8 References

Almasri, M.N., and Kaluarachchi, J.J. 2003. Regional variability of on-ground nitrogen loading due to multiple land uses in agriculture dominated watersheds. Proceedings of the 7th International Conference on Diffuse Pollution and Basin Management. Dublin, Ireland.

Almasri, M.N., and Kaluarachchi, J.J. 2004. Implications of on-ground nitrogen loading and soil transformations on ground water quality management. Journal of the American Water Resources Association (JAWRA) 40 (1):165-186.

Almasri, M.N., and Kaluarachchi, J.J. 2005. Multi-criteria decision analysis for the optimal management of nitrate contamination of aquifers. Journal of Environmental Management 74: 365-381.

Azizullah, A., Khattak, M.N.K., Richter, P., and Hader, D.P. 2011. Water pollution in Pakistan and its impact on public health -a review. Environment International 37:479-497.

Bartolino, J. R. 1994. Source of nitrate nitrogen in the Seymour Aquifer, Knox County, Texas. AGU Abs. Spring Meeting, May 23-27, Baltimore, Maryland.

Birkinshaw, S.J., and Ewen J. 2000. Nitrogen transformation component for SHETRAN catchment nitrate transport modelling. Journal of Hydrology 230:1-17.

Boryan, C., Yang, Z., Mueller, R., and Craig, M. 2011. Monitoring US agriculture: The US department of agriculture, national agriculture statistics service, cropland data layer program. Geocarto International 26:341-358.

Cao, P., Lu, C., and Yu, Z. 2017. Historical Nitrogen Fertilizer Use in Agricultural Ecosystem of the Continental United States during 1850-2015: Application rate, Timing, and Fertilizer Types. Earth System Science 10: 969-984.

Chaudhuri, S., and Ale, S. 2014. Long-term (1930–2010) Trends in groundwater levels in Texas: Influence of soils, landcover and water use. Science of the Total Environment 490:379-90.

Chowdary, V.M., Rao, N.H., and Sarma, P.B.S. 2005. Decision support framework for assessment of non-point-source pollution of groundwater in large irrigation projects. Agriculture Water Management 75:194-225.

Daniel, J. A. 1997. Effectiveness of animal waste containment in a Texas playa. Environmental and Engineering Geoscience 3:563-572.

Dubrovsky, N.M., Burow, K.R., Clark, G.M., Gronberg, J.M., Hamilton, P.A., Hitt, K.J., Mueller, D.K., Munn, M.D., Nolan, B.T., Puckett, L.J., Rupert, M.G., Short, T.M., Spahr, N.E., Sprague, L.A., and Wilber, W.G.2010. The quality of our Nation’s waters-Nutrients in the Nation’s streams and groundwater, 1992–2004: U.S. Geological Survey Circular 1350, 174 p.

Dunn, S.M, Vinten, A.J.A., Lilly, A., DeGroote, J., and McGechan, M. 2005. Modelling nitrate losses from agricultural activities on a national scale. Water Science & Technology 51(3- 4):319-27.

Esmaeili, A, Moor, F., and Keshavarzi, B. 2014. Nitrate contamination in irrigation groundwater, Isfahan, Iran. Environmental Earth Sciences 72(7): 2511-2522.

Fotheringham, A.S., Brunsdon, C., and Charlton, M. 2002. Geographically weighted regression: the analysis of spatially varying relationships. Chichester: Wiley.

Fotheringham, A.S., Brunsdon, C., and Charlton, M.E. 2003. Geographically Weighted Regression: The Analysis of Spatially Varying Relationships. New York, USA: Wiley.

Fotheringham, A.S., Charlton, M.E., and Brunsdon, C. 1998. Geographically weighted regression: a natural evolution of the expansion method for spatial data analysis. Environment and Planning A 30 (11):1905-1927.

Fryar, A. E., Macko, S. A., Mullican, W. F., Romanak, K. D., and Bennett, P. C. 2000. Nitrate reduction during ground-water recharge, Southern High Plains, Texas. Journal of Contaminant Hydrology 40(4):335- 363.

Galloway, J.N., Aber, J.D., Erisman, J.W., Seitzinger, S.P., Howarth, R.W., Cowling, E.B., and Cosby, B.J. 2003. The nitrogen cascades. Bioscience 53(4):341-356.

Galloway, J. N., Townsend, A. R., Erisman, J. W., Bekunda, M., Cai, Z., Freney, J. R., Martinelli, L. A., Seitzinger, S. P., and Sutton, M. A. 2008. Transformation of the nitrogen cycle: Recent trends, questions and potential solutions. Science 320: 889-892.

Gronberg, J.A.M., and Spahr, N.E. 2012. County-level estimates of nitrogen and phosphorus from commercial fertilizer for the conterminous United States,1987-2006. Scientific Investigations Report 2012-5207, USGS.

Harter, T., Davis, H., Mathews, M., and Meyer, R. 2002. Shallow groundwater quality on dairy farms with irrigated forage crops. Journal of Contaminant Hydrology 55: 287-315.

Holloway, J.M., and Dahlgren, R.A. 2002. Nitrogen in rock: occurrences and biogeochemical implications. Global Biogeochemical Cycles 16(4):1-65.

Holloway, J.M., and Smith, R.L. 2005. Nitrogen and carbon flow from rock to water: regulation through soil biogeochemical processes, Mokelumne River watershed, California, and Grand Valley, Colorado. Journal of Geophysical Research 110.

Hudak, P.F. 2018. Spatial and temporal trends of nitrate, chloride, and bromide concentration in an alluvial aquifer, North‐Central Texas, USA. Environment Quality Management 27(4): 79-86.

Igbedioh, S.O. 1991. Effects of agricultural pesticides on humans, animals, and higher plants in developing countries. Archives of Environmental & Occupational Health 46(4):218-24.

Ikerd, J. 1999. Sustainable agriculture as a rural economic development strategy. Accessed [12/05/2018]: http://www.ssu. missouri.edu/faculty/JIkerd/papers/sa-cdst.htm.

Jordan, C., and Smith, R.V. 2005. Methods to predict the agricultural contribution to catchment nitrate loads: designation of nitrate vulnerable zones in Northern Ireland. Journal of Hydrology 304(1-4):316-29.

Juntakut, P., Daniel, A.B., SnowcErin, D., and Haackerd Chittaranjan Ray, M.K. 2019. The long- term effect of agricultural, vadose zone and climatic factors on nitrate contamination in Nebraska’s groundwater system. Journal of Contaminant Hydrology 220: 33-48.

Kidd, J., Manning, P., Simkin, J., Peacock, S., and Stockdale, E. 2017. Impacts of 120 years of fertilizer addition on a temperate grassland ecosystem. PloS one 12(3).

Kreitler, C. W. 1975. Determining the source of nitrate in groundwater by nitrogen isotope studies, Bureau of Economic Geology. University of Texas at Austin. Investigations Report 83:57.

Kreitler, C. W., and Jones, D. 1975. Natural soil nitrate: the cause of the nitrate contamination of groundwater in Runnels County, Texas. Ground Water 13: 53-61.

Kreitler, C. W., and Browning, L. 1983. Nitrogen isotope analysis of groundwater nitrate in carbonate aquifers: natural sources versus human pollution. Journal of Hydrology 61:285- 301.

Kundzewicz, Z. W., and Doll, P. 2009. Will groundwater ease freshwater stress under climate change? Hydrological Sciences Journal 54: 665-675.

Kyllmar, K., Martensson, K., and Johnsson, H. 2004. Model-based coefficient method for calculation of N leaching from agricultural fields applied to small catchments and the effects of leaching reducing measures. Journal of Hydrology 304(1-4):343-54.

Liu, A., Ming J., and Ankumah R. O. 2005. Nitrate contamination in private wells in rural Alabama, United States. Science of the Total Environment 346:112-20.

Lindgren, R.J. 2006. Diffuse-flow conceptualization and simulation of the Edwards aquifer, San Antonio region, Texas. U.S. Geological Survey Scientific Investigations Report 5319:48.

Margat, J., and Gun, J.V.D. 2013. Groundwater around the world, a geographic synopsis. Boca Raton: CRC Press.

Miller, W.W., Johnson, D.W., Denton, C., Verburg, P.S.J., Dana, G.L., and Walker, R.F. 2005. In conspicuous nutrient laden surface runoff from mature forest Sierra watersheds. Air Water Soil Pollution 163:3-17.

Miyamoto, C., Ketterings, Q., Cherney, J., and Kilcer, T. 2008. Nitrogen fixation, agronomy fact sheet series. Accessed [04/05/2019]:http://nmsp.cals.cornell.edu/publications/factsheets/factsheet39.pdf.

Mueller, D.K., and Helsel, D.R. 1996. Nutrients in the Nation’s waters - Too much of a good thing. US Geological Survey. Circular 1136:1-2.

Musgrove, M., Opsahl, S.P., Mahler, B.J., Herrington, C., Sample, T.L., and Banta, J.R. 2016. Source, variability, and transformation of nitrate in a regional karst aquifer: Edwards aquifer, central Texas. Science of the Total Environment 568:457-469.

Nolan, B. T. 2001. Relating nitrogen sources and aquifer susceptibility to nitrate in shallow ground waters of the United States. Ground Water 39:290-299.

Nolan, B. T., and Stoner, J. D. 2000. Nutrients in groundwaters of the conterminous United States. Environmental Science & Technology 34:1156-1165.

Nolan, B. T., Hitt, K. J., and Ruddy, B. C. 2002. Probability of nitrate contamination of recently recharged groundwaters in the conterminous United States. Environmental Science & Technology 36:2138-2145.

Opsahl, S.P., Musgrove, M., and Slattery, R.N. 2017. New insights into nitrate dynamics in a karst system gained from high-frequency optical sensor measurements. Journal of Hydrology 546: 179-188.

Puckett, L.J. 2016. Nonpoint and Point Sources of Nitrogen in Major Watersheds of the United States. U.S. Geological Survey Water-Resources Investigations Report 94-4001.

Ransom, K. M., Bell, A. M., Barber, Q. E., Kourakos, G., and Harter, T. 2018 A Bayesian approach to infer nitrogen loading rates from crop and land-use types surrounding private wells in the Central Valley, California. Hydrology and Earth System Sciences 22: 2739-2758.

Sanderson, M.R., and Frey, R.S. 2014. Structural impediments to sustainable groundwater management in the High-Plains Aquifer of western Kansas. Agriculture and Human Values 1-17.

Scanlon, B. R., Reedy, R. C., and Kier, K. B. 2004. Evaluation of nitrate contamination in major porous media aquifers in Texas. Final Contract Report to the Texas Commission on Environmental Quality 48 p.

Schröder, J.J., Scholefield, D., Cabral, F., and Hofman, G. 2004. The effects of nutrient losses from agriculture on ground and surface water quality: the position of science in developing indicators for regulation. Environmental Science & Policy 7:15-23.

Schilling, K.E., and Wolter, C.F. 2001. Contribution of base flow to nonpoint source pollution loads in an agricultural watershed. Ground Water 39 (1):49-58.

Self, R., and Waskom, R.M. 2014. Nitrates in drinking water, fact sheet no. 0.517 crop series soil. Accessed [04/04/2019]: http:// extension.colostate.edu/docs/pubs/crops/00517.pdf.

Shrestha, A., and Luo,W. 2017. Analysis of groundwater nitrate contamination in the central valley: Comparison of the geodetector method, principal component analysis and geographically weighted regression. International Journal of Geo-Information 6(10):297.

Shukla, S., and Saxena, A. 2018. Global Status of Nitrate Contamination in Groundwater: Its Occurrence, Health Impacts, and Mitigation Measures. In: Hussain C (eds), Handbook of Environmental Materials Management. Springer, Cham.

Shamrukh, M., Corapcioglu, M., and Hassona, F. 2001. Modeling the effect of chemical fertilizers on ground water quality in the Nile Valley Aquifer, Egypt. Ground Water 39 (1):59-67.

Stewart, B. A., Smith, S. J., Sharpley, A. N., Naney, J. W., McDonald, T., Hickey, M. G., and Seweeten, J. M. 1994. Nitrate and other nutrients associated with playa storage of feedlot wastes, paper presented at Proc. Playa Basin Symposium, Texas Tech University Lubbock.

Tesoriero, A.J., and Voss,F.D. 1997. Predicting the probability of elevated nitrate concentrations in the Puget Sound Basin: implications for aquifer susceptibility and vulnerability. Groundwater 35(6):1029-39.

Texas State Soil & Conservation board 2012.Accessed [12/12/2018]: https://www.tsswcb.texas.gov/groundwater-nitrogen-source-identification-and- remediation-texas-high-plains-and-rolling-plains.

Texas Water Development Board (TWDB) Groundwater Data. Accessed [01/02/2019]: http://www.twdb.texas.gov/groundwater/data/index.asp.

U.S. Environmental Protection Agency (U.S. EPA) 2001. Source water protection practices bulletin: Managing small-scale application of pesticides to prevent contamination of drinking water. Washington, DC: Office of Water.

U.S. Environmental Protection Agency (U.S. EPA) 1995. Drinking water regulations and health advisories. Washington, DC: Office of Water.

U.S. Geological Survey (USGS) 1984. Geohydrology of the High-Plains aquifer in parts of Colorado, Kansas, Nebraska, New Mexico, Oklahoma, South Dakota, Texas, and Wyoming. Accessed [03/04/2019]: https://pubs.usgs.gov/pp/1400b/report.pdf.

U.S. Department of Agriculture (USDA) 2019. Irrigation and water use. Accessed [04/04/2019]: https://www.ers.usda.gov/topics/farm-practices-management/irrigation-water-use/.

U.S. Department of Agriculture (USDA) 2019. Land Use and Land Cover Estimates for United States. Accessed [ 06/14/2019]: https://www.ers.usda.gov/about- ers/partnerships/strengthening-statistics-through-the-icars/land-use-and-land-cover- estimates-for-the-united-states/.

U.S. National Agricultural Statistics Service (U.S. NASS) 2018. Agricultural Chemical Use Program. Accessed [04/04/2019]: https://www.nass.usda.gov/Surveys/Guide_to_NASS_Surveys/Chemical_Use/.

Van der Schans, M.L., Harter, T., Leijinse, A., Mathews, M.C., and Meyer, R.D. 2009. Characterizing sources of nitrate leaching from an irrigated dairy farm in Merced County, California. Journal of Contaminant Hydrology 110:2-9.

Vinten, A.J.A., and Dunn, S.M. 2001. Assessing the effects of land use on temporal change in well water quality in a designated nitrate vulnerable zone. Science of the Total Environment 265:253-68.

Ward, M. H., Mark, S. D., Cantor, K. P., Weisenburger, D. D.,Correa-Villasenor, A., and Zahm, S. 1996. Drinking water nitrate and the risk of non-Hodgkins lymphoma. Epidemiology 7: 465- 471.

White, W.B. 1988. Geomorphology and hydrology of karst terrains. New York: Oxford University Press.

Wick, K., Heumesser, C., and Schmid, E. 2012. Groundwater nitrate contamination: factors and indicators. Journal of Environmental Management 111(3): 178-86.

Wickham, J.D., Stehman, S.V., Gass, L., Dewitz, J., Fry, J.A., and Wade, T.G. 2013. Accuracy assessment of NLCD 2006 land cover and impervious surface. Remote Sensing of Environment 130: 294–304.

Wickham, J., Stehman, S.V., Gass, L., Dewitz, J.A., Sorenson, D.G., Granneman, B.J., Poss, R.V., and Baer, L.A. 2017. Thematic accuracy assessment of the 2011 National Land Cover Database (NLCD). Remote Sensing of Environment 191:328–341.

World-wide Hydrogeological Mapping and Assessment Programme (WHYMAP) 2011. River and Groundwater Basins and Aquifers of the World. Accessed [04/04/2019]: https://www.whymap.org/whymap/EN/Maps_Data.

Yang, L. et al. 2018. A new generation of the United States National Land Cover Database: Requirements, research priorities, design, and implementation strategies. Journal of Photogrammetry and Remote Sensing 146:108–123.

Zhou, W., Sun, Y., Wu, B., Zhang, Y., Huang, M., Miyanaga, T., and Zhang, Z. 2011. Autotrophic denitrification for nitrate and nitrite removal using sulphur-limestone. Journal of Environmental Sciences 23(11):1761-1769.

DEGRADATION OF GROUNDWATER QUALITY IN THE COASTAL AQUIFERS OF THE

UNITED STATES

3.1 Abstract

The objective of this study was to analyze groundwater types and the extent of seawater intrusion in terms of groundwater samples’ ion concentrations using USGS monitoring wells. The extent of seawater intrusion varies among localities; therefore, Groundwater Quality Indices (fsea,

GQI (dom), GQI (mix), GQI (swi)) were used to estimate the spatial distributions of groundwater types in the U.S. coastal aquifers. Results showed that the dominated groundwater chemistry in the study area was Na+ and Cl- . Results indicated that seawater intrusion of groundwater occurred in the U.S. coastal aquifers. Based on the results of GQI(mix) and GQI (dom), most of the wells fell under domain II in the Piper diagram indicating that they were of the Na-Cl water type. Na-Cl and mixed Ca-Mg-Cl were determined to be the two dominant groundwater types in the coastal aquifers of the U.S.. Seawater intrusion was evident in the Palm Beach and Miami-Dade counties of Florida, the San Diego and Santa Barbara areas of California, and within Virginia and New

Jersey on the east coast. As the sea level continues to rise further inland, Na-Cl concentration in groundwater will increase particularly in aquifers along low-lying coasts.

3.2 Introduction

Assessing groundwater quality is vital for groundwater-dependent ecosystems and human communities (Eamus et al. 2016). Groundwater salinization is a process of soil degradation that 53

decreases soil fertility, which then affects the ecological health of lands (Herbert et al. 2015).

Groundwater accounts for 25 percent of freshwater withdrawals in the United States (Maupin et al. 2014). The need for high-quality drinking water supplies becomes vital as the U.S. population continues to grow, especially as natural processes and human activities alter the balance of ionic concentrations in the aquifers. The balance of water and salt in coastal aquifers, where the land- ocean interaction establishes a narrow interface zone is altered due to seawater intrusion and over- withdrawal of groundwater (Cooper et al. 1964). Seawater intrusion is a natural process in coastal areas; however, anthropogenic activities can exacerbate the incidence of seawater intrusion

(Cooper et al. 1964, Zhang et al. 2011, Fatoric and Chelleri 2012). Seawater intrusion and anthropogenic activities affect groundwater quality; these activities may include industrial discharges, urban activities, agriculture, and disposal of waste (Peters and Meybeck 2000, Huizer et al. 2018).

Seawater intrusion is defined as the movement of seawater into groundwater aquifers

(Freeze and Cherry 1979, Eissa et al. 2018). The increase of soluble salts’ accumulation above natural levels is known as salinization. The contact between seawater and groundwater in coastal aquifers deteriorates the groundwater quality by increasing salinity levels to above drinking and irrigation standards for water (Peters and Meybeck 2000). According to the Environmental

Protection Agency (U.S. EPA 1995), drinking water should not include more than 500 mg/L of total suspended solids (TSS) (U.S. EPA 1995). Because seawater contains almost 30,000 mg/L of

TSS, a small amount of seawater in groundwater may cause groundwater contamination. Chloride

- + 2- 2+ 2+ + (Cl ), sodium (Na ), sulfate (SO 4), magnesium (Mg ), calcium (Ca ), and potassium (K ) are the most abundant ions of seawater and constitute approximately 95 percent of all sea salts (USGS

+ + 2+ 2+ - - - - 2016a). The four cations (Na , K , Mg , Ca ) and four anions (Cl , HCO3 , SO4 , NO3 ) are

often used to analyze the groundwater quality (Moujabber et al. 2006, Appelo and Postma 2005).

Ocean salinity has been documented to range from 34 to 37 parts per thousand (Poisson and Papaud

1983). Change in groundwater chemistry because of seawater intrusion generally shows a shift from low-mineralized Ca-HCO3, Ca-Na-HCO3 (Appelo and Postma 2005), and Ca-Cl water types to high-mineralized Na-Cl groundwater type (Narany et al. 2014).The presence of mineralized water underground and its emergence to the surface threatens the world’s food supply produced by irrigated agriculture (Morris et al. 2003).

Populations in coastal U.S. counties have grown to 94 million people in 2016, which is about 29 percent of the total U.S. population. The largest populations settled on the Pacific coastline followed by the Atlantic coastline and the Gulf of Mexico in 2016 (U.S. Census Bureau

2015). About 37 percent of the world’s population lives within 100 km of the coast; similarly, two- thirds of the world’s cities are located on coastlines (Neumann et al. 2015). From 2010 to 2015, groundwater use in the United States increased by 8.3 percent, while surface water use decreased by 13.9 percent (USGS 2016b). More than 40 million people, including most rural populations, obtain drinking water from domestic wells in the United States (Alley et al. 1999).

Groundwater geology, hydrology, geochemistry, and chemical uses on land vary throughout the United States (Pande and Moharir 2018, Cooper et al. 1964). In the western United

States, people use groundwater for irrigation, but this supply is at risk of contamination by nitrates, pesticides, and chemicals. Artificial chemicals and geologic sources play roles in the groundwater contamination in the North Atlantic Coastal Plain aquifer while carbonate rock in southeastern

Florida is a low permeability aquifer protected from artificial and geologic contaminants (USGS

2014). Groundwater over-withdrawal for public supply in the eastern United States and groundwater withdrawal for industrial use (e.g. irrigation) in the western and southeastern United

States has led to groundwater depletion and seawater intrusion (Appelo and Postma 2005).

Dausman and Langevin (2005) found that high Cl- concentration in groundwater is the major problem in Broward County, Florida. The primary water source for Broward County is the

Biscayne aquifer, and a surplus of Cl- in the Biscayne aquifer is due to the decline of the groundwater levels to below the mean sea level (Werner et al. 2013). Similarly, in California, the overuse of groundwater combined with multi-year drought (e.g., no rainfall and high evapotranspiration) has resulted in the Cl- concentration in groundwater exceeding the maximum contaminant level (CA DWR 1958).

The previous studies (e.g. Abarca and Clement 2009, Boufadel 2000, Riva et al. 2018,

Mantoglou 2003) did not explain seawater intrusion in the U.S. coastal aquifers using groundwater quality indices. In this study, chemical compositions of groundwater were analyzed to identify water types in coastal aquifers and the extent of seawater intrusion; USGS monitoring wells were used to create the water quality indices.

3.3 Objectives

The specific objectives of this research are:

(a) to identify dominant groundwater types in the coastal aquifers using USGS well data;

(b) to estimate the extent of seawater intrusion in terms of dominant ions and ocean salinity in the

U.S. coastal aquifers using various groundwater quality indices, and;

(c) to interpolate the seawater intrusion estimates to analyze the spatial patterns of seawater intrusion in coastal aquifers using geostatistical techniques.

3.4 Background

Seawater naturally moves into groundwater through tides and waves due to the hydraulic connection between groundwater and seawater (Appelo and Postma 2005, Barlow and Wild 2002).

Seawater intrusion is associated with a lowering of the groundwater table, (i.e. decreasing the hydraulic gradient sloping toward the sea) (Barlow and Wild 2002). When the groundwater table head is below the sea-level head, seawater actively moves into aquifers (Cooper et al. 1964). As a result, seawater flows inland when the dynamic balance of freshwater and seawater in coastal aquifers is disturbed by groundwater withdrawals (Appelo and Postma 2005). Reduced groundwater levels lead to a decrease in groundwater flow to the sea, causing seawater to intrude into the coastal aquifers (Purwoarminta et al. 2018). The hydrogeologic settings and groundwater withdrawals can cause different progressions of seawater intrusion into coastal aquifers (Abarca and Clement 2009, Boufadel 2000, Riva et al. 2018, Mantoglou 2003). Saline water contamination occurs through several ways, including: lateral intrusion from the ocean, upward intrusion from deeper saline zones of a groundwater system, downward intrusion from coastal waters (such as estuaries) and tidal driven seawater flooding of low-lying coastal lands (Ataie-Ashtiani et al.

1999). Because of future sea-level rise, an increase in aquifer salinization is also expected in many coastal plains globally (Sherif and Singh 1999, Döll 2009, Carretero et al. 2013).

The extent of seawater intrusion into an aquifer depends on several factors, including the total rate of groundwater that is withdrawn from an aquifer compared to the total freshwater recharge to the aquifer (Barlow and Reichard 2010, Raicy et al. 2012, Capaccioni et al. 2005,

Giambastiani et al. 2007, Trabelsi et al. 2007, Somay and Gemici 2009). Furthermore, the distance between the locations of groundwater discharge and seawater (Taniguchi et al. 2002), the geologic structure of an aquifer (including structural features such as faults, folds, and bounding submarine

canyons), the hydraulic properties of an aquifer (including the interconnectivity of coarse-grained units within multi-layered aquifer systems), and the presence of confining units that may prevent seawater from moving vertically toward or within the aquifer are all contributing factors (Werner et al. 2013).

Several analytical and semi-analytical expressions have been developed to help explain seawater intrusion (Strack 1976, Dagan and Zeitoun 1998, Cheng et al. 2000, Bakker 2006,

Nordbotten and Celia 2006, Park et al. 2009). For example, Strack (1976) studied the maximum withdrawal rate to avoid seawater movement into groundwater. Dagan and Zeitoun (1998) highlighted the density-dependent flow of seawater and found that more dense seawater moves towards less dense freshwater. A line boundary between groundwater and seawater is protected with denser seawater underlying freshwater. The over-withdrawal of groundwater can cause the denser seawater to be drawn toward the freshwater zones of the aquifer (Dagan and Zeitoun 1998).

Cheng et al. (2000) used the analytical solution of the sharp-interface seawater intrusion model in their study to estimate groundwater withdrawal rates for an existing multiple well in a coastal aquifer with the structured messy genetic algorithm. Bakker (2006) studied the flow in the freshwater zone and flow in the seawater zones simultaneously to investigate the pressure and normal flux across the interface.

Numerical simulations also provide valuable insights into seawater intrusion. Lu et al.

(2013) investigated a set of laboratory experiments and numerical simulations to investigate the effect of geological stratification on seawater-freshwater mixing. Riva et al. (2018) considered four key parameters controlling the process: gravity number, permeability anisotropy ratio and transverse and longitudinal numbers. The relationship between the tidal fluctuations and groundwater table oscillations were analyzed by Ataie-Ashtiani et al. (1999), Lu et al. (2013), and

Pool et al. (2014). Studies showed that the water table has a significant relation to the mean sea level and the distance of the point of observation from the sea. Misut and Voss (2007) analyzed the impact of aquifer storage and recovery practices on the transition zone associated with the seawater wedge in the New York City aquifer. Kerrou et al. (2013) analyzed the effects of uncertainty in the permeability and distribution of groundwater withdrawal rates on seawater intrusion in the Korba aquifer in the Republic of Tunisia.

Some studies have used this kind of spatial analysis to describe and analyze the various spatial patterns in water quality (Kitsiou et al. 2002, Huang et al. 2001). Kitsiou et al. (2018) used the spatial analysis method to present the spatial distribution of N-NO3, N-NH4 and chlorophyll-a

(chl-a) concentrations measured in seawater samples collected in coastal areas. Kumari et al.

(2018) used an Inverse Distance Weighted (IDW) interpolation technique to study the distribution of various water pollutants in Ulagalla cascade, Sri Lanka. Tomaszkiewicz et al. (2014) used the indicator kriging method to spatially analyze a groundwater quality index at a karstic aquifer along the eastern coast of the Mediterranean Sea.

Later on, Werner and Gallagher (2006) studied seawater intrusion in the coastal aquifer of the Pioneer Valley in Australia, indicating the advantages of combining hydrogeological and hydro-chemical analyses to understand salinization processes. Pattern diagrams were used to characterize the chemical patterns of groundwater and the most commonly used of these is the

Piper diagram (Piper 1944). The Piper diagram reveals similarities and differences among groundwater samples because those with similar qualities will incline to plot together as groups

(Todd 2001). Results from multiple analyses (i.e., several groundwater wells in a region) can be plotted on the same diagram and then interpreted to document the chemical character or the hydro-

chemical patterns (Back 1983) as well as possible mingling of freshwater with seawater (Arslan et al. 2012).

Ions exchanged between groundwater and seawater during seawater intrusion defines the groundwater types and groundwater salinity levels. Seawater and freshwater have a different hydrochemistry. Seawater is characterized by an almost uniform chemistry whereas Cl- and Na+

2+ - make up approximately 84 percent of the total ionic composition. Ca and HCO3 are dominant ions for freshwater (Richter and Kreitler 1993). The levels of Cl- concentration, total dissolved solids (TDS), and electrical conductivity (EC) are generally used as common indicators to identify an increase in salinity in groundwater (Singhal and Gupta 2010, Rhoades et al. 1992). Magnesium, calcium, sodium, and sulfate are also major ions in seawater that can be traced to groundwater after the seawater intrudes.

3.5 Study Area

Groundwater quality samples obtained from USGS monitoring wells are recorded at a specific time and are not provided for all aquifer monitoring wells. The California Coastal Basin,

Upper Floridian, and North Atlantic Coastal Plain aquifers were chosen for the coastal areas to identify dominant groundwater types and estimate the extent of seawater intrusion (Fig.3.1).

Figure 3.1 Map of the study area showing California Coastal Basin aquifer, Upper Floridian aquifer, and North Atlantic Plain aquifer.

3.5.1 Upper Floridian Aquifer

The Upper Floridian aquifer is one of the most productive aquifers in the world (USGS

2016b). It contains a thick sequence of carbonate rocks (limestones and dolomites) that lies beneath all of Florida and southern Georgia, as well as parts of South Carolina and Alabama. Groundwater withdrawals from the Floridan aquifer have caused long-term water-level declines of more than

3m in several areas, including southeastern and northeastern Florida and coastal Georgia (Miller

1986).

An important component of southeastern Florida’s hydrologic system is the highly productive Biscayne aquifer, which is the source of water supply in Miami-Dade, Broward, and

Palm Beach counties, respectively. In most of the Florida Peninsula, Biscayne aquifer, and southeast Florida, groundwater withdrawals exceed discharge (Megdal and Petersen-Perlman

2018), while the average annual precipitation is 138 cm in this region. The aquifer is unconfined, and the water table responds quickly to recharge evapotranspiration, and withdrawal from wells

(Renken et al. 2008).

The southwest coast is characterized by high transmissivity rates at which groundwater flows horizontally through an aquifer. Recharge rates in the southwest coast are relatively higher than Biscayne aquifer, averaging 50.8 cm year-1 (Megdal and Petersen-Perlman 2018), and the depth of wells typically varies from 243 m to 365 m below the land surface. Withdrawals from the aquifer system added up to more than 158 m3 /s in 2000 and in many areas, it is the sole source of freshwater (Maupin and Barber 2005).

3.5.2 California Coastal Plain Aquifer

The California Coastal Basin aquifer extends from northern San Francisco Bay through the

Los Angeles and Orange County coastal plain. California’s coast is divided into four hydrologic regions: North Coast, San Francisco Bay, Central Coast, and South Coast (Crossett et al. 2004).

The North Coast has a coastline of about 550 km. The North Coast receives the highest amount of precipitation. Annual rainfall ranges from 25.4cm to over 250cm, and groundwater is used for agricultural and domestic needs (CA DWR 2013b). The San Francisco Bay is the smallest hydraulic region in California. The region imports about 70 percent of its water and gets the remaining 30 percent from local sources (CA DWR 2013c), and groundwater represents about 5 percent of water used in the region. The Central Coast is home to about 4 percent of the state’s population and uses groundwater to meet about 80 percent of its domestic, municipal, and agricultural water demands, making it the most groundwater-reliant region in the state (CA DWR 62

2013a). The South Coast is the most populated region in California and is home to some of the largest cities in the U.S., including Los Angeles and San Diego. Groundwater withdrawal in the

South Coast dates back over 100 years, and seawater intrusion has been documented in the region

(CA DWR 2013d).

The California coastal plain aquifer contains unconsolidated sand and gravel that underlies all southern and central California. The unconsolidated deposits thicknesses range from only a few meters to as much as 304.8m (USGS 2016b). The aquifer is identified by high transmissivity values up to 920 m2/d, and groundwater withdrawal rates are estimated at 0.23 m3 /s-cubic meters /year.

Meanwhile, well-depth is ranging from 30.47m to 213.36m beneath the land surface in the

California Coastal Basin aquifer (USGS 2016b).

3.5.3 North Atlantic Coastal Plain Aquifer

The North Atlantic Coastal aquifer extends from the southern border of North Carolina northward through Long Island. About 800 million gallons per day were withdrawn from unconfined and confined aquifers in the North Atlantic Coastal Plain to meet the need of public water supply for more than 21 million people in 2017 (EPA 2017).

The surficial aquifer system is a primary source of public drinking water supply in populated Long Island and southern New Jersey (USGS 2016a). The surficial North Atlantic

Coastal Plain aquifer consists of semi-consolidated and unconsolidated sand and gravel. The relatively coarse and permeable sediments that compose the surficial aquifer system are in general less than 30.48m thick (USGS 2016a). The aquifer is characterized by moderate transmissivity values of less than 92.90m2 /d as shown by the low decline in flow rate during the dry season, and well-depths exceed 228 m. Chemicals are easily transported down to the unconfined aquifers due to permeable sands and gravels of the surficial aquifer system (USGS 2016a). 63

3.6 Data and Method

The U.S. Geological Survey (USGS) National Water Information System (NWIS) database has collected groundwater quality data for national wells since 1988. This dataset includes time series data on levels of Cl- (mg/L), magnesium (mg/L), sodium(mg/L), calcium (mg/L), potassium

(mg/L), and bicarbonate (mg/L) ions. In this study, the last recorded date for ion data was used to estimate coastal aquifers water types. Each sample well presented an ion level for a different month and year because the last recorded date of sampling differed among wells across the study areas.

Data from 240 wells within 10 kilometers of US coasts were used to map the seawater intrusion in terms of Cl- concentration. Cl- is easily traceable due to the conservative nature of the anion

(Appelo and Postma 2005), i.e., it does not strongly participate in rock-water interactions and is recalcitrant to such reactions (Fadili et al. 2015). Furthermore, indices and diagrams showing the

+ + + - 2+ 2- - relationships between the major elements (Na , Mg , K , Cl , Ca , SO4 , and HCO3 ) of 107 monitoring wells were chosen to understand the groundwater mineralization process against seawater intrusion (Appendix Table A.1).

Monthly ocean salinity data from 1968 to 2009 were acquired from the National Oceanic and Atmospheric Administration (https://www.nodc.noaa.gov/cgi-bin/OC5/WOA09/woa09.pl).

The ocean surface salinity data is used to estimate seawater intrusion. Using the “Near” tool in

ArcGIS, we linked the USGS well to the nearest 1-degree grid points. We then matched the months of ocean salinity data to the same months of the well data.

The following equation was used to convert the groundwater Cl- concentration to salinity level: salinity (ppt) = 0.00180665 * Cl- (mg/L) (Brankovits et al. 2017). Freshwater Cl- threshold was acquired from the U.S. EPA (2017). The EPA has identified 250 mg/L as a Cl- threshold level in drinking water.

3.6.1 Chemical Compositions

The ionic composition of water can be used to classify it into ionic types based on the dominant dissolved cation and anion. The dominant dissolved ion must be greater than 50 percent of the total. For example, water classified as sodium-chloride type water contains more than 50 percent of the total cation milliequivalents as sodium and more than 50 percent of the total anion milliequivalents as Cl-. If no cation or anion is dominant (greater than 50 percent), the water is classified as mixed (USGS 2016a).

3.6.2 Piper Diagram Groundwater Quality Indices: (GQI (dom) & GQI (mix))

Piper diagrams were used to highlight and group ions in groundwater exceeding maximum thresholds for public use and the agricultural activities in addition to classifying aquifers into various classes based on ionic types (Tomaszkiewicz et al. 2014, Wu et al. 2017). The diamond field of the Piper diagram can be divided into six differing domains (Fig.3.2): I, II, III, IV, V, and

VI representing Ca-HCO3, Na-Cl, mixed Ca-Na-HCO3, mixed Ca-Mg-Cl, Ca-Cl, and Na-HCO3 water types, respectively (Fig.3.2) (Sarath Prasanth et al. 2012).

Figure 3.2 Piper Diagram illustrating general classifications of groundwaters types (Tomaszkiewicz et al. 2014).

Freshwater is represented in domain I whereas saline water, including seawater, is in domain II. Simple mixing of freshwater and seawater is designated by a horizontal line across the center of the diagram, represented numerically by the Groundwater Quality Index (GQI (mix)) as expressed in the equation [3.1] (Tomaszkiewicz et al. 2014):

2+ 2+ - GQI (mix)= [ (Ca + Mg ) / total cations + (HCO3 )/total anions] *50 (in mg/L) [3.1]

The resulting index GQI (mix) can range from 0, representing highly saline water, to 100, representing highly freshwater (Domain I). Other domains are defined using GQI(mix) concurrently with Groundwater Quality Index (Dominant) (GQI (dom)). GQI (dom) is expressed by the following equation [3.2] (Tomaszkiewicz et al. 2014):

+ + - GQI(dom)= [ (Na + K ) / total cations + (HCO3 )/total anions] *50 (in mg/L) [3.2]

where 0 represents Ca-Cl water (domain V) and 100 represents Na-HCO3 (domain VI) water. As it is shown in Table 3.1, GQI (mix) and GQI (dom) are used to determine chemical patterns at the water of monitoring wells and their corresponding domains in the Piper diagram.

Table 3.1 Illustrating hydro-chemical domains using GQI (mix) and GQI (dom) (Tomaszkiewicz et al. 2014):

Domain GQI piper (mix) GQI piper (dom)

I 50-100 25-75

II 0-50 25-75

III 25-75 50-75

IV 25-75 25-50

V 25-75 0-25

VI 25-75 75-100

3.6.3 Groundwater Quality Indices (fsea, GQI (fsea) & GQI Seawater Intrusion (swi))

The seawater fraction in groundwater was assessed using chloride concentration provided by Appelo and Postma (2005). This index was used in many studies (Mahlknecht et al. 2017,

Tiwari et al. 2019, Panteleit et al. 2011, Arslan 2013, Edet and Okereke 2001, Elewa et al. 2013

Bakari et al. 2012) to estimate the seawater and freshwater mixing in the coastal aquifers. Tiwari

et al (2019) found that the groundwater samples were in the moderate saline zone to highly saline zone in the island of Favignana in Italy. Accordingly, a simple tool to identify seawater intrusion is the seawater fraction (fsea) with values ranging from 0 to 1 wherein 0 represents high freshwater, and 1 represents high seawater intrusion in the aquifer. fsea is quantified in terms of groundwater

Cl- concentration and ocean salinity (Appelo and Postma 2005) as expressed in equation [3.3]

(Appelo and Postma 2005):

fsea = [ (salinity level -freshwater salinity threshold)/ (seawater salinity– freshwater

salinity threshold)] [3.3]

where salinity level is the groundwater salinity, and seawater salinity is ocean salinity.

Groundwater Quality Index in terms of Cl- concentration and ocean salinity is quantified by using equation [3.4] (Tomaszkiewicz et al. 2014). The resulting index GQI (fsea) can range from 0, representing high seawater intrusion, to 100, representing highly freshwater.

GQI(fsea) = (1- fsea) *100 [3.4]

However, GQI (fsea) may not be an appropriate index independently. It fails to identify most hydro-geochemical reactions associated with seawater intrusion, such as cation exchanges, which can affect the chemical compositions of groundwater subject to seawater intrusion. The combination of fsea and GQI (mix) can be a more illustrative index for seawater-groundwater mixing; therefore, as expressed in equation [3.5] (Tomaszkiewicz et al. 2014), GQI (swi) is used to characterize seawater intrusion in groundwater, where 0 represents high seawater intrusion, and

100 represents freshwater in aquifer.

GQI (swi) = [(GQI (mix) + GQI(fsea)] / 2 [3.5]

3.6.4 Geostatistical Analysis

The geostatistical analysis was used to interpolate fsea point measurements, as they are quantified based on Cl- concentration, and ocean salinity values. Kriging interpolation was used to obtain the spatial distribution of seawater fractions in coastal aquifers. It was used to estimate the value of a variable at a location based on the known values of the same variable at other locations, and kriging techniques predict values from interpolation without bias and with minimum variance.

It assigns weights to nearest neighbors and is expressed by the following equation [3.6] (Rajaee et al. 2016).

N X(s0) = ∑̭ ƛ i Ȥ(si) [3.6] i=1 where Z(si) refers to the measured value at the ith location, λ i indicates an unknown weight for the measured value at the i-th location, s0 denotes the prediction location, and N refers to the number of measured values.

Spatial structure and geostatistical analysis are based upon an experimental semi- variogram, which is a function of the lag distance between two observation points and fitted to a mathematical model. The commonly used spherical and exponential models were evaluated by determining several parameters including the nugget, semi-variance at h=0, range, and lag distance where the semi-variance becomes constant, known as the sill. The Root Mean Square Standardized

Error (RMSE) is used to evaluate the model performance (Li et al. 2015).

There are different types of Kriging techniques, such as Ordinary Kriging, Universal

Kriging, Indicator Kriging, and Co-Kriging. Trends in the data suggested we should use Universal

Kriging, and spatial trends or drifts display a tendency for the values (Isaaks and Srivastava 1989).

Before using kriging, the distribution patterns of fsea point measurements were identified in the

study area. Global Moran’s I Spatial Autocorrelation was used to evaluate whether the pattern is clustered, dispersed, or random (Dale and Fortin 2002).

3.7 Results and Discussions

A total of 107 USGS wells were selected from three aquifers located along the Pacific

Ocean, Gulf of Mexico, and Atlantic Ocean to study the major-ion composition of groundwater.

The results of the analysis of ion-types (Fig.3.3) indicated that sodium (Na+) accounted for 54 percent of the anions and Cl- accounted for 78 percent of the cations in the California Coastal Basin aquifer. In the Upper Floridian aquifer, sodium accounted for 77 percent of anions and Cl- accounted for 82 percent of the cations. Similarly, in the North Atlantic aquifer, 80 percent of the anions are comprised of sodium ions (Na+), and 87 percent of the cations are Cl- ions (Fig.3.3).

Therefore, all three aquifers can be classified as Na-Cl water types based on the dominant dissolved cations and anions in these aquifers.

Figure 3.3 Illustrating anions and cations composition in California Coastal Basin aquifer, Upper Floridian aquifer and North Atlantic Coastal Plain aquifer.

3.7.1 Piper Diagram

The major-ion compositions of 107 of the wells from the three aquifers are displayed in the Piper diagram (Fig. 3.4). The Piper diagram displays the chemical patterns of water from selected wells (Fig.3.4).

Figure 3.4 Piper diagram illustrating general classification of groundwaters (mg/L) and seawater-freshwater mixing water types (mg/L) in the coastal aquifers of the U.S.

After plotting GQI (mix) and GQI (dom) on the Piper diagram, most of the wells fell under domain II of the Piper diagram indicating that they were of the Na-Cl water type. Twenty-nine out of thirty-five monitoring wells of the California Coastal Basin, twenty-nine out of thirty-one monitoring wells of the Upper Floridian, and thirty-one out of forty-one monitoring wells of the

North Atlantic Coastal Plain aquifers fell under domain II (Fig.3.4). Three out of thirty-five monitoring wells of the California Coastal Basin, two out of thirty-one monitoring wells of the

Upper Floridian, and four out of forty-one monitoring wells of the North Atlantic Coastal Plain aquifers fell under domain V (Fig.3.4). Additionally, two out of thirty-five monitoring wells of the

California Coastal Basin and four out of forty-one monitoring wells of the North Atlantic Coastal

Plain aquifers fell under domain IV. Furthermore, one out of thirty-five wells of the California

Coastal Basin and one out of forty-one monitoring wells in the North Atlantic Coastal Plain aquifer fell under domain VI. Lastly, one out of forty-one monitoring well fell under domain I in the North

Atlantic Coastal Plain aquifer (Fig.3.4).

3.7.2 Groundwater Quality Indices

The average GQI (mix) and GQI (dom) index values for all wells in each of the three aquifers are presented in Table 3.2. For the California Coastal Basin aquifer, the average GQI

(mix) index value was 17.76 and the average GQI (dom) index value was 40.47. Based on GQI

(mix) index, the wells in the California Coastal Basin aquifer are in domain II, and based on GQI

(dom) index, the wells in this aquifer are also in domain I, II, and IV (see Table 3.1). Similarly, in the Upper Floridian aquifer, the average GQI (mix) index value was 49 and the GQI (dom) index value was 43.4. These values also suggest that on average, wells in the Upper Floridian aquifer fall within domain II, III, IV, V, and VI (see Table 3.1). Furthermore, in the North Atlantic Coastal

Plain aquifer, the average GQI (mix) value was 20 and the average GQI (dom) value was 47.8.

These values indicate that on average, wells in the North Atlantic Coastal Plain aquifer are in domain I, II, and IV (see Table 3.1).

Table 3.2 Illustrating average groundwater quality indices in three principal aquifers of the U.S.

Regions Number of GQI(mix) GQI(dom) Na+(mg/L ) Cl- Wells (mg/L)

California Coastal Basin 35 17.76 40.47 1813.17 3058 Aquifer

Upper Floridian Aquifer 31 49 43.45 1433 3244.93

North Atlantic Coastal Plain 41 20 47.84 1986.76 3840.94 Aquifer

Further, we categorized GQI (mix) into three classes, where: GQI (mix) below 25 indicates high Na-Cl, between 25 and 50 indicates moderate Na-Cl, domain III, IV, V, and VI water types, and above 50 indicates a Ca-HCO3 water type. Similarly, we categorized GQI (dom) into three classes where: GQI (dom) below 25 indicates Ca-Cl type, between 25 and 50 indicates Ca-Mg-Cl, domain I, II types, and above 50 indicates mixed Ca-Na-HCO3 , Na-HCO3 , and domain I, II types

(Fig.3.5).

In the California Coastal Basin aquifer, we used twenty monitoring wells in San Diego county and fifteen monitoring wells in Santa Barbara county to analyze groundwater types. The

GQI (mix) indicated that all monitoring wells in San Diego and Santa Barbara fell under the Na-

Cl water type. Further, all monitoring wells in San Diego were classified as high Na-Cl and all wells in Santa Barbara were classified as having a moderate amount of Na-Cl contamination

(Fig.3.5a). The GQI (dom) indicated that in the California Coastal Basin aquifer, all thirty-five monitoring wells fell under Ca-Mg-Cl, domain I, and II types (Fig.3.5b).

In the Upper Floridian aquifer, we used thirty-one monitoring wells to analyze groundwater types. Results indicated that all thirty-one monitoring wells fell under Na-Cl water type. Twenty-

nine out of thirty-one (93 percent) monitoring wells were classified as having a high concentration of Na-Cl and two out of thirty-one (7 percent) wells were classified as having moderate concentration of Na-Cl (Fig.3.5c). Further, based on GQI (dom) twenty-seven out of thirty-one

(85 percent) wells were classified as Ca-Mg-Cl types and three out of thirty-one (8 percent) monitoring wells were classified as Ca-Na-HCO3 and Na-HCO3 types, and one out of thirty-one (7 percent) monitoring well was classified as having Ca-Cl (Fig.3.5d).

In the North Atlantic Coastal Plain aquifer, we used forty-one monitoring wells to analyze groundwater types. Results indicated that all forty-one monitoring wells fell under Na-Cl water type. Thirty-one out of forty-one (75 percent) monitoring wells were classified as having high concentrations of Na-Cl, eight out of forty-one (20 percent) monitoring wells were classified as having a moderate concentration of Na-Cl, and two out of forty-one (5 percent) wells were classified as having Ca-HCO3 types (Fig.3.5e). Further, based on based on GQI (dom), twenty-six out of forty-one (63 percent) monitoring wells were classified as having Ca-Mg-Cl types. Eleven out of forty-one (27 percent) monitoring wells were classified as having Ca-Na-HCO3 and Na-

HCO3 types, and four out of forty-one (10 percent) wells were classified as having Ca-Cl types

(Fig.3.5f).

Figure 3.5 Illustrating groundwater quality using Groundwater Quality Index (GQI(mix)) and Groundwater Quality Index (GQI(dom)) in the coastal aquifers of the United Sates.

In summary, our analysis found Na-Cl as the common pollutant in all North Atlantic

Coastal Plain, California Coastal Basin, and Upper Florida aquifers.

3.7.3 Seawater Intrusion Indices

Fsea, GQI (fsea), and GQI (swi) are used to estimate seawater intrusion levels of wells within various coastal aquifers. Seawater intrusion was estimated using 107 selected sample wells.

Table 3.3 shows that in the California Coastal Basin aquifer, the average GQI (fsea) value was

89.3 indicating a high level of freshwater and less seawater intrusion. Because GQI (fsea) estimated seawater intrusion based only upon Cl- concentration levels, this is not a true estimation of the extent of seawater intrusion. Therefore, we calculated the average GQI (swi), which was

53.5 indicating that there is a significant level of seawater intrusion in monitored wells (see Table

3.3).

In the Upper Floridian aquifer, the average GQI (fsea) value was 90.7 indicating a high level of freshwater and less seawater intrusion. Because GQI (fsea) estimated seawater intrusion based only upon Cl- concentration levels, this is not a true estimation of the extent of seawater intrusion. Therefore, we calculated average GQI (swi) which was 54.4 indicating that there is a significant level of seawater intrusion in monitored wells (see Table 3.3).

In the North Atlantic Coastal Plain aquifer, the average GQI (fsea) value was 91.2 indicating a high level of freshwater and less seawater intrusion. Because GQI (fsea) estimated seawater intrusion based only upon Cl- concentration levels, this is not a true estimation of the extent of seawater intrusion. Therefore, we quantified average GQI (swi) which was 55.6 indicating that there is a significant level of seawater intrusion in monitored wells (see Table 3.3).

Table 3.3 Average seawater fraction (fsea) and groundwater quality indices.

Aquifers Number fsea* GQI(fsea)* GQI(swi)* of Wells

California Coastal Basin 35 .1 89.34 53.55 Aquifer

Upper Floridian Aquifer 31 .09 90.72 54.43

North Atlantic Coastal 41 .08 91.26 55.66 Plain Aquifer

*fsea=1(high seawater intrusion); GQI(fsea)=100 (high freshwater); GQI(swi)= 100 (high freshwater)

Although U.S. coastal aquifers have large Cl- concentrations as a result of groundwater salinization, an increase in ionic content due to seawater-groundwater mixing, cannot be quantified based only upon Cl- concentration levels. Because a trivial increase in Cl- concentrations can almost triple the salinity of groundwater (Jones et al. 1999), even though it has a higher fsea.

Consequently, the extent of seawater intrusion needs to be assessed based upon ion mixing processes. The GQI (swi) indicates that wells have been exposed to seawater intrusion in the

California Coastal, Upper Floridian, and North Atlantic Coastal aquifers.

Figure 3.6 shows an inverse relationship between fsea and GQI (swi). In the California

Coastal Basin aquifer, Upper Floridian, and North Atlantic Coastal Plain aquifers, as fsea levels decreased, GQI (swi) content significantly increased (Fig.3.6). Fsea indicates that a trivial amount of dissolved salts in seawater are found in groundwater. Many wells in all three aquifers had fsea values close to 0 and high GQI (swi) values. This indicates that sea water intrusion is not that severe, but not completely ruled out as a principal factor for groundwater saline contamination.

Therefore, seawater-groundwater mixing as it is quantified by GQI (swi) and fsea shows that

dissolved salts in seawater are found to some extent in the water of the three aforementioned coastal aquifers.

Cardona et al. (2004) found that fsea mixed groundwater ranges from 0.001 to 0.04 in coastal aquifers of the Pacific Ocean. Similarly, Price and Swart (2006) noted fsea of mixed groundwater is 0.17 in coastal aquifers of the Atlantic Ocean. Our results indicated that in the

California Coastal Basin aquifer fsea was 0.1 which exceeded the fsea values (0.04) estimated by

Cardona et al. (2004) in the coastal aquifers of the Pacific Ocean.

Figure 3.6 Illustrating fsea and GQI (swi) relationship.

Further, we analyzed the spatial patterns of fsea and GQI (swi) within all three aquifers.

The extent of seawater movement into aquifers has varied among localities and hydrogeological characteristics. As it is shown in Fig.3.7, we grouped fsea into two major classes where 0 to 0.5

indicates low seawater intrusion and 0.5 to 1 indicates high seawater intrusion. Furthermore, we classified GQI (swi) into four major classes where 1 to 20 indicates severe seawater intrusion, 20 to 40 equals high seawater intrusion, 40 to 60 equals moderate seawater intrusion, and 60-80 equals low seawater intrusion.

Within the California Coastal Basin, thirty-three out of thirty-five (94 percent) monitoring wells had a trivial level of fsea, which ranged from 0 to 0.5. Two out of thirty-five (6 percent) wells had severe sea water intrusion as they had fsea value over 0.5 (Fig.3.7a).

Based on GQI(swi) which considers all ions, (Fig. 3.7b) two out of thirty-five (6 percent) monitoring wells in the California Coastal Basin had a severe seawater intrusion, twenty out of thirty-five (57 percent) monitoring wells had moderate seawater intrusion, and thirteen out of thirty-five (37 percent) wells had slight levels of seawater intrusion.

In the Upper Floridian aquifer, twenty nine out of thirty-one (96 percent) monitoring wells had fsea values ranging from 0 to 0.5. Two out of thirty-one (3 percent) wells had a severe seawater intrusion as they had fsea value over 0.5 (Fig. 3.7c). However, GQI(swi) values (Fig. 3.7d) showed that one out of thirty-one (3 percent) monitoring wells in the Upper Floridian had severe seawater intrusion, twenty-seven out of thirty-one (85 percent) wells had moderate seawater intrusion, and four out of thirty-one (12 percent) wells had a slight seawater intrusion.

In the North Atlantic Coastal Plain aquifer, forty out of forty-one (97 percent) monitoring wells had fsea values ranging from 0 to 0.5. One out of forty-one (2 percent) well had severe seawater intrusion as it has an fsea value of over 0.5 (Fig.3.7e). Additionally, GOI (swi) (Fig. 3.7f) showed that one out of forty-one (2 percent) monitoring well in the North Atlantic Coastal plain had a severe seawater intrusion, twenty nine out of forty-one (71 percent) wells had moderate

levels of seawater intrusion, and eleven out of forty-one (27 percent) monitoring wells had slight seawater intrusion levels.

Figure 3.7 Illustrating seawater intrusion using Seawater Fraction (fsea) for 240 coastal wells and Groundwater Quality Index (GQI(swi) for selected 107 sample wells

Meanwhile, the unconsolidated and semi consolidated sediment have enough permeability to allow seawater to flow through the California Coastal Basin, Upper Floridian, and North

Atlantic Coastal Plain aquifers as Biscayne aquifer in the southeastern Upper Floridian aquifer

(USGS 2016a). The shallow karst Biscayne aquifer is a primary source of drinking water for the southeastern Upper Floridian aquifer. The Biscayne aquifer of coastal Upper Floridian aquifer is prone to seawater intrusion because this area is near sea level (USGS 2014).

Furthermore, in our study, GQI (swi) ranged from 53.5 to 55.6 (see Table 3.3) in monitoring wells which is very similar to the results of previous studies (Price and Swart 2006,

Cardona et al. 2004), where GQI (swi) ranged from 4.8 to 79.9 in aquifers that had different levels of seawater intrusion (see Table 3.4).

Table 3.4 GQI (swi) ranges (Tomaszkiewicz et al. 2014).

Water type GQI(swi) based on worldwide literature Min Max Mean Freshwater 73.5 90.1 82.7 Mixed groundwater 47.8 79.9 63.4 Saline Groundwater 4.8 58.8 27.5 Seawater 3.1 9.2 5.8

3.7.4 Geostatistical Analysis

fsea values from 240 groundwater monitoring wells were analyzed using the Universal

Kriging technique to interpolate fsea values for the locations where we do not have monitoring wells. First, Moran’s I was calculated to assess whether the data has a negative or positive correlation or a random distribution. The Moran’s I values in the California Coastal aquifer (0.51),

Upper Floridian aquifer (0.79), and North Atlantic Coastal aquifer (0.54) were positive and

significant. These values indicate a clustering of wells that have similar Cl- concentrations (see

Table 3.5).

Table 3.5 Moran I and descriptive statistics for seawater fraction (fsea).

Moran I Spatial Statistics Descriptive Statistics

Number Moran’s I Z-score p-value Mean Std. of wells (fsea) Deviation (fsea)

fsea (California 98 .51 5.11 <.001 .09 .25 Coastal Aquifer)

fsea (Upper Floridian 88 .79 12.73 <.001 .13 .16 Aquifer)

fsea (North Atlantic 56 .54 6.86 <.001 .13 .23 Coastal Plain Aquifer)

RMSE was used to examine the performance of the interpolation models, and a semi- variogram was used to describe the structure of spatial variability (see Table 3.6 and Fig. 3.8). The

RMSE of fsea interpolation was low, which indicates less residual variance and a better model fit.

A semi-variogram was used to characterize higher spatial autocorrelation in one direction. There are spatial dependencies in the Cl- concentrations.

Table 3.6 Semi-variogram models of fsea parameter for Universal Kriging.

Nugget Sill Range RMSE California .0005 .01 .02 .06 Coastal Basin Aquifer

Upper Floridan .00007 .005 .03 .07 Aquifer

North Atlantic .00004 .004 .0004 .03 Coastal Plain Aquifer

Figure 3.8 Semi-variogram for the California Coastal Basin aquifer (8a), Upper Floridian aquifer (8b), and North Atlantic Coastal Plain aquifer (8c).

The fsea maps represented in Fig. 3.9 were obtained by the Kriging interpolation of 240 monitoring wells. The maps showed that the highest seawater intrusions are observed at the coastal fringe, while the inland seawater fractions become weaker. This suggests that the existence of seawater intrusion is responsible for groundwater salinization near the ocean.

The results indicate significant presence of Cl- in the California Coastal Basin, Upper

Floridian, and North Atlantic Coastal Plain aquifers. Palm Beach, Miami-Dade, and Broward coasts of the Upper Floridian aquifer seems to have higher Cl- concentration in groundwater (Fig.

3.9a). The Gloucester, Virginia Beach coasts (Virginia); Monmouth (New Jersey); and Queens

(New York) in the North Atlantic Coastal plain aquifer (Fig. 3.9b) also seems to have significant

Cl- concentration in groundwater. The seawater fraction became weaker despite the proximity of the ocean toward the northern North Atlantic coastal aquifer. This may be a consequence of the low number of monitoring wells and the low aquifer exploitation in this area part (Fig. 3.9b). Cl- concentrations are further elevated in northern parts of the study area in Santa Barbara and eastern parts of the study area in San Diego (Figs. 3.9c and 3.9d).

Accuracy and Limitation

Different dataset used this study had different accuracies. USGS’s water-level measurements at wells is accurate to 0.03 m (USGS 2005). The total groundwater samples passed through the filter were accurately determined to ±1 mL (USGS 2005). Regarding ocean salinity data, the satellites provide large-scale sea surface salinity measurements at accuracies of ±0.5 and

±0.2 psu respectively, at temporal scales of weeks to months over spatial scales of 50-100 km

(https://www.sciencedirect.com/topics/earth-and-planetary-sciences/sea-surface-salinity).

Seawater intrusion includes complex hydrogeochemical processes that cannot be fully understood through use of seawater intrusion indices. It is suggested that more monitoring wells and hydrological parameters need to be included to further assess the extent of seawater intrusion.

3.8 Concluding Remarks

This study used groundwater quality indices (GQI(mix), GQI(dom)) and seawater intrusion indices (GQI(swi), GQI(fsea)) to assess the groundwater quality. Seawater intrusion entails complex hydrogeochemical processes that cannot be fully captured through GQI (fsea) and fsea.

The seawater intrusion index (GQI (swi), derived by combining the seawater fraction index (fsea) and the freshwater–seawater mixing index (GQI(mix)) of the Piper Diagram, were used to show seawater intrusion in the coastal aquifers of the U.S. Geostatistical analysis was additionally used to estimate seawater intrusion in the unmeasured locations. Seawater intrusion indices were used across the world to estimate seawater intrusion. Similar to previous studies (Mahlknecht et al.

2017, Tiwari et al. 2019, Laaksoharju et al. 2009, Price and Swart 2006, Cardona et al. 2004,

Pulido-Leboeuf 2004, de Montety et al. 2008, Anderson et al. 2005, Panteleit et al. 2011, Arslan

2013, Edet and Okereke 2001, Elewa et al. 2013, Bakari et al. 2012, Pujari and Soni 2009) this

study also found that seawater intrusion occurrence in our study area. Likewise, Price and Swart

(2006) found fsea value of 0.17 in coastal aquifers in Atlantic Ocean. fsea value was 0.05 in coastal aquifers in Mexico (Cardona et al. 2004). In Mediterranean Sea, fsea value was 0.01 as estimated by Pulido-Leboeuf (2004). Likewise, fsea value was 0.9 in Isefjord Bay in Denmark (Anderson et al. 2005). In North Sea coastal aquifers in Germany, fsea value was 0.06 (Panteleit et al. 2011). fsea value was 0.05 in coastal aquifers of Black Sea (Arsalan 2013). In the coastal aquifers of

Atlantic Ocean in Nigeria, fsea value was 0.12 ( Edet and Okereke 2001). In the groundwater in coastal aquifers in Egypt, fsea value was 0.25 (Elewa et al. 2013). fsea value was .08 in the groundwater in coastal aquifers in Tanzania (Bakari et al. 2012). All these studies concluded that there was a seawater intrusion of groundwater in their study areas. Compared with previous studies, in this study, fsea value is 0.1 in the groundwater in the coastal aquifers of the California

Coastal Basin aquifer, it is 0.09 in the Upper Floridian aquifer, and 0.08 in the North Atlantic

Coastal Plain aquifer. Furthermore, based on previous studies (Price and Swart 2006, Cardona et al. 2004), GQI (swi) ranged from 4.8 to 79.9 in aquifers that were exposed to seawater intrusion.

In this study, GQI (swi) ranged from 53.5 to 55.6 in the California Coastal Basin, Upper Floridian and North Atlantic Coastal Plain aquifers indicating seawater intrusion in these aquifers.

Accordingly, seawater intrusion is evident in the Palm Beach and Miami-Dade counties of Florida, the San Diego and Santa Barbara areas of California, and within Virginia and New Jersey on the east coast.

Monitoring wells located within 10 km perpendicular to the ocean have Na-Cl and mixed

Ca-Mg-Cl water types. Elevated levels of Na+ and Cl- occur naturally, due to the hydraulic proximity of the aquifers to seawater. Na-Cl concentration in groundwater will increase, particularly in shallow, sandy aquifers along low-lying coasts, as the sea level continues to rise 89

further inland. Increased Na-Cl concentrations in U.S. aquifers can affect local vegetation and soil quality. Soil salinity can be increased at a very rapid rate, due to the persistent use of Na-Cl contamination in groundwater. Meanwhile, the components of groundwater can also be modified due to the influence of several factors, including the interaction between water and the aquifer material and the influences of human activities, such as the return of flow irrigation and over- withdrawal of the aquifers.

3.9 References

Abarca, E., and Clement, T. P. 2009. A novel approach for characterizing the mixing zone of a saltwater wedge. Geophysics Research Letter 36.

Alley, W.M, Reilly, T.E., and Franke, O.L. 1999. Sustainability of groundwater resources, vol. 1186. US Department of the Interior. US Geological Survey.

Appelo, C.A.J., and Postma D. 2005. Geochemistry, groundwater and pollution. London: CRC Press.

Arslan, H., Cemek, B., and Demir, Y. 2012. Determination of seawater intrusion via hydro- chemicals and isotopes in Bafra Plain, Turkey. Water Resource Management 26 (13): 3907-3922.

Ataie-Ashtiani, B., Volker, R. E., and Lockington, D. A. 1999. Tidal effects on sea water intrusion in unconfined aquifers. Journal of Hydrology 216:17-31.

Back, W., and Freeze, R.A. 1983. Chemical hydrogeology. Benchmark Papers in Geology. Stroudsburg, PA: Hutchinson Ross.

Bakker, M. 2006. Analytic solutions for interface flow in combined confined and semi-confined, coastal aquifers. Advance Water Resource 29: 417-425.

Bakari, S.S., Aagaard, P., Vogt, R.D. Ruden, F., Johansen, I., and Vuai, S.A. 2012. Delineation of groundwater provenance in a coastal aquifer using statistical and isotopic methods, Southeast Tanzania. Environmental Earth Science 66(3):889-902.

Barlow, P.M., and Reichard, E.G. 2010. Saltwater intrusion in coastal regions of North America. Hydrogeological Journal 18(1):247-260.

Barlow, P.M., and Wild, E.C. 2002. Bibliography on the occurrence and intrusion of saltwater in aquifers along the Atlantic coast of the United States. US Geological Survey Open-File Rep 02-235.

Boufadel, M. C. 2000. A mechanistic study of nonlinear solute transporting a groundwater-surface water system under steady state and transient hydraulic conditions. Water Resource Research (36): 2549-2565.

Brankovits, D., Pohlman, J., Niemann, H., Leigh, M.B., Leewis, M.C., Becker, K.W., Iliffe, T.M., Alvarez. F., Lehmann, M.F., and Phillips, B. 2017. Methane- and dissolved organic carbon- fueled microbial loop supports a tropical subterranean estuary ecosystem. Nature Communications 8: 1-12.

California Department of Water Resources (CA DWR) 2013a. California water plan update central coast hydrologic region. California Department of Water Resources. Regional Reports Vol. 2.

California Department of Water Resources (CA DWR) 2013b. California water plan update north coast hydrologic region. California Department of Water Resources. Regional Reports Vol.2.

California Department of Water Resources (CA DWR) 2013c. California water plan update San Francisco Bay hydrologic region. California Department of Water Resources. Regional Reports Vol. 2.

California Department of Water Resources (CA DWR) 2013d. California water plan update south coast hydrologic region. California Department of Water Resources. Regional Reports Vol. 2.

California Department of Water Resources (CA DWR) 1958. Seawater intrusion in California. Bulletin 63. Sacramento, CA: California Department of Water Resources.

Capaccioni, B., Didero, M., Paletta, C., and Didero, L. 2005. Saline intrusion and refreshening in a multilayer coastal aquifer in the Catania Plain (Sicily,Southern Italy): dynamics of degradation processes according to the hydrochemical characteristics of groundwaters. Journal of Hydrology 307: 1-16.

Carretero, S., Rapaglia, J., Bokuniewicz, H., and Kruse, E. 2013. Impact of sea-level rise on saltwater intrusion length into the coastal aquifer, Partido de La Costa, Argentina. Continental Shelf Research 62: 62-70.

Cardona, A., Calrrillo-Rivera, J., Huizar-Alvarez, R., and Graniel-Castro, E. 2004. Salinization in coastal aquifers of arid zones: an example from Santo Domingo, Baja California Sur, Mexico. Environmental Geology 45(3): 50-366.

Cheng, A. H., Halhal D., Naji, D., and Ouazar, D. 2000. Pumping optimization in saltwater- intruded coastal aquifers. Water Resource Research 36.2155-2165.

Cooper, H., Kohout, F., Henry, H., and Glover, R. 1964. Sea Water in Coastal Aquifers. U.S. Geological Survey Water-Supply Paper 1613-C.

Crossett, K.M., Culliton, T.J., Wiley, P.C., and Goodspeed, T.R. 2004. Coastal trends report series: Population trends along the coastal United States: 1980-2008. National Oceanic and Atmospheric Administration. Coastal Trends Report Series.

Dagan, G., and Zeitoun, D. G. 1998. Seawater-freshwater interface in a stratified aquifer of random permeability distribution. Journal of Contaminant Hydrology 29:185-203.

Dale, M.R.T., and Fortin, M. 2002. Spatial autocorrelation and statistical tests in ecology. Écoscience 9(2):162-167.

Dausman, A., and Langevin, C.D. 2005. Movement of the saltwater interface in the surficial aquifer system in response to hydrologic stresses and water-management practices, Broward County, Florida. U.S. Geological Survey Scientific Investigations Report 2004- 5256. U.S. Geological Survey.

Döll, P. 2009. Vulnerability to the impact of climate change on renewable ground-water resources: a global-scale assessment. Environmental Research Letter 4.

Eissa, M.A., Dreuzy, J.R., and Parker, B. 2018. Integrative management of saltwater intrusion in poorly constrained semi-arid coastal aquifer at Ras El-Hekma, Northwestern Coast, Egypt. Groundwater for Sustainable Development 6:57-70.

Fatoric, S., and Chelleri, L. 2012. Vulnerability to the effects of climate change and adaptation: the case of the Spanish Ebro Delta Ocean. Coastal Management 60:1-10.

Freeze, R. A., and Cherry, J. A. 1979. Groundwater, Prentice Hall, Inc.

Giambastiani, B.M., Antonellini, M., Essink, G.H.O., and Stuurman, R.J. 2007. Saltwater intrusion in the unconfined coastal aquifer of Ravenna (Italy): a numerical model. Journal of Hydrology 340:91-104.

Huizer, S., Radermacher, M., de Vries, S., Oude Essink, G.H.P., and Bierkens, M.F.P. 2018. Impact of coastal forcing and groundwater recharge on the growth of a fresh groundwater lens in a mega-scale beach nourishment. Hydrology and Earth System Sciences 22:1065- 1080.

Huang, B., Jiang, B., and Li, H. 2001. An integration of GIS, virtual reality and the Internet for visualization, analysis and exploration of spatial data. International Journal of Geographical Information Science 15:439-456.

Isaaks, E., and Srivastava, R. 1989. An Introduction to Applied Geostatistics. New York: Oxford University Press.

Jones, B. Vengosh, A. Rosenthal, E., and Yechieli, Y. 1999. Geochemical investigations. Seawater Intrusion in Coastal Aquifers-Concepts, Methods and Practices, Springer.

Kerrou, J., Renard, P., Cornaton, F., and Perrochet, P. 2013. Stochastic forecasts of seawater intrusion towards sustainable groundwater management: application to the Korba aquifer (Tunisia). Hydrogeological Journal 21:425-440.

Kitsiou, D., Coccossis, H., and Karydis M. 2002. Multi-dimensional evaluation and ranking of coastal areas using GIS and multiple-criteria choice methods. The Science of the Total Environment 284:1-17.

Kitsiou, D., Patera, A., and Kostopoulou, M. 2018. A GIS methodology to support seawater quality assessment in coastal areas. Global Nest Journal 20(1):122-127.

Kumari, M. K. N, Sakai, K., Kimura, S., Nakamura, S., Yuge, K., Gunarathna, M.H.J.P, Ranagalge, M., and Duminda, D.M.S. 2018. Interpolation methods for groundwater quality assessment in tank cascade landscape: A study of Ulagalla cascade, Sri Lanka. Applied Ecology and Environmental Research 16(5): 5359-5380.

Laaksoharju, M., Skårman, E., Gómez, J. B., and Gurban, I. 2009. M3 version 3: User's manual SKB TR‐09‐09. Sweden: Svensk Kärnbränslehantering AB.

Li, H.Y., Webster, R. Z., and Shi, Z. 2015. Mapping soil salinity in the Yangtze delta: REML and universal kriging (E-BLUP). Geoderma (237-238):71-77.

Lu, C. H., Chen, Y. M., Zhang, C., and Luo, J. 2013. Steady-state freshwater-seawater mixing zone in stratified coastal aquifers. Journal of Hydrology 505:24-34.

Mantoglou, A. 2003. Pumping management of coastal aquifers using analytical models of saltwater intrusion. Water Resource Research 39(12):1335.

Mahlknecht, J., et al. 2017. Assessing seawater intrusion in an arid coastal aquifer under high anthropogenic influence using major constituents, Sr and B isot..., Sci Total Environ 587– 588:282-295.

Masterson, J.P., Pope, J.P., Monti, J., Nardi, M.R., Finkelstein J.S., and McCoy, K.J. 2015. Hydrogeology and Hydrologic Conditions of the Northern Atlantic Coastal Plain Aquifer System from Long Island, New York, to North Carolina. Scientific Investigations Report 2013-5133 Version 1., U.S. Geological Survey.

Maupin, M.A., Kenny, J.F., Hutson, S.S., Lovelace, J.K., Barber, N.L., and Linsey, K.S. 2014. Estimated Use of Water in the United States in 2010. U.S. Geological Survey Circular 1405: 64.

Maupin, M.A., and Barber, N.L. 2005. Estimated withdrawals from principal aquifers in the United States. U.S. Geological Survey.

Miller, J.A. 1986. Hydrogeologic framework of the Floridan aquifer system in Florida and in parts of Georgia, Alabama, and South Carolina. US Geological Survey Prof Pap 1403-B.

Misut, P. E., and Voss, C. I. 2007. Freshwater-saltwater transition zone movement during aquifer storage and recovery cycles in Brooklyn and Queens, New York City, USA. Journal of Hydrology 337:87-103.

Moujabber, M.E., Samra, B.B., Darwish, T., and Atallah, T. 2006. Comparison of different indicators for groundwater contamination by seawater intrusion on the Lebanese coast. Water Resource Management 20(2):161-180.

Morris, B.L., Lawrence, A.R.L., Chilton, P.J.C., Adams, B., Calow, R.C., and Klinck, B.A. 2003. Groundwater and its Susceptibility to Degradation: A Global Assessment of the Problem of Options for Management. Early Warning and Assessment Report series, RS. 03-3. United Nations Environment Programme, Nairobi, Kenya.

Narany, T.S., Ramli, M.F., Zaharin Aris, A., Azmin Sulaiman W.N., Juahir, H., and Fakharian, K. 2014. Identification of the Hydrogeochemical Processes in Groundwater Using Classic Integrated Geochemical Methods and Geostatistical Techniques, in Amol-Babol Plain, Iran. Scientific World Journal 15 p.

National Oceanic and Atmospheric Administration (NOAA). Accessed [05/06/2018]: https://www.nodc.noaa.gov/cgi-bin/OC5/WOA09/woa09.pl.

Neumann, B., Vafeidis, A. T., Zimmermann, J., and Nicholls, R. J. 2015. Future Coastal Population Growth and Exposure to Sea-Level Rise and Coastal Flooding - A Global Assessment. In: PLoS ONE 10(3).

Nordbotten, J. M., and Celia, M. A. 2006. An improved analytical solution for interface upconing around a well. Water Resource Research 42.

Pande, C.B., and Moharir, K. 2018. Spatial analysis of groundwater quality mapping in hard rock area in the Akola and Buldhana districts of Maharashtra, India. Applied Water Science 8(4):106. Park, N., Cui, L., and Shi, L. 2009. Analytical design curves to maximize pumping or minimize injection in coastal aquifers. Ground Water 47:797-80.

Peters, N.E., and Meybeck, M. 2000. Water quality degradation effects on freshwater availability: impacts of human activities. Water International 25(2):185-193.

Petersen-Perlman, J.D., and Megdal, S.B. 2018. Decentralized Groundwater Governance and Water Nexus Implications in the United States, Jurimetrics 59: 99-119.

Piper, A. M. 1944. A graphic procedure in the geochemical interpretation of water analyses. American Geophysical Union Transactions 25:914-928.

Poisson, A., and Papaud, A. 1983. Diffusion coefficients of major ions in seawater. Marine Chemistry 13(4):265-280.

Pool, M., Post, V. E. A., and Simmons, C. T. 2014. Effects of tidal fluctuations and spatial heterogeneity on mixing and spreading in spatially heterogeneous coastal aquifers. Water Resource Research 51:1570-1585.

Price, R.M., and Swart, P.K. 2006. Geochemical indicators of groundwater recharge in the surficial aquifer system: Everglades National Park, Florida, USA. Special Pap. Geological Society of America 404:251.

Purwoarminta, A., Moosdorf, N., and Delinom R.M. 2018. IOP Conf. Series: Earth and Environmental Science 118.

Raicy, M. C., Parimala Renganayaki, S., Brindha, K., and Elango, L. 2012. Mitigation of seawater intrusion by managed aquifer recharge, Managed Aquifer Recharge. Methods, Hydrogeological Requirements, Pre and Post treatment Systems 83-99.

Rajaee, T., Nourani, V., and Pouraslan, F. 2016. Groundwater Level Forecasting Using Wavelet and Kriging. Journal of Hydraulic Structures 2(2):1-21.

Renken, R.A., Cunningham, K.J., Shapiro, A.M., Harvey, R.W., Zygnerski, M.R., Metge, D.W., and Wacker, M.A. 2008. Pathogen and chemical transport in the karst limestone of the Biscayne aquifer: 1. Revised conceptualization of groundwater flow. Water Resource Research 44:399-416.

Rhoades, J.D., Kandiah, A., and Mashali, A. 1992. The Use of Saline Waters for Crop Production FAO.

Richter, B.C., and Kreitler, C.W. 1993. Geochemical Techniques for Identifying Sources of Groundwater Salinization. U.S.: CRC Press.

Riva, M., Guadagnini, A., and Dell’Oca, A. 2018. Probabilistic assessment of seawater intrusion under multiple sources of uncertainty. Advances in Water Resources 75: 93-104.

Sarath Prasanth, S., Magesh, N., Jitheshlal, K., Chandrasekar, N., and Gangadhar, K. 2012. Evaluation of groundwater quality and its suitability for drinking and agricultural use in the coastal stretch of Alappuzha district, Kerala, India. Applied Water Science 2 (3):165- 175.

Sherif, M.M., and Singh, V.P. 1999. Effect of climate change on seawater intrusion in coastal aquifers. Hydrological Process 13:1277-1287.

Singhal, B., and Gupta, R.P.2010. Applied Hydrogeology of Fractured Rocks. Netherlands: Springer.

Somay, M.A., and Gemici, Ü. 2009. Assessment of the salinization process at the coastal area with hydrogeochemical tools and geographical information systems (GIS): Selçuk plain, Izmir, Turkey. Water, Air, and Soil Pollution 201:55-74.

Strack, O. D. L. 1976. A single-potential solution for regional interface problems in coastal aquifers. Water Resource Research12:1165-1174.

Taniguchi, M., Burnett, W. C., Cable, J. E., and Turner, J. V. 2002. Investigation of submarine groundwater discharge. Hydrological Process 16:2115-2129.

Tiwari, A. K., Pisciotta, A., De Maio, M. 2019. Evaluation of groundwater salinization and pollution level on Favignana Island, Italy, Environmental Pollution 249:969-981

Todd, D.K. 2001. Groundwater Hydrology. Canada: John Wiley and Sons Publication.

Tomaszkiewicz, M., Abou Najm, M., and El-Fadel, M. 2014. Development of groundwater quality index for seawater intrusion in coastal aquifers. Environmental Modelling & Software 57:13-26.

Trabelsi, R., Zairi, M., and Dhia, H.B. 2007. Groundwater salinization of the Sfax superficial aquifer, Tunisia. Hydrogeological Journal 15:1341-1355.

U.S. Environmental Protection Agency (EPA) 1995. Drinking water regulations and health advisories. Washington, DC.: Office of Water.

U.S. Environmental Protection Agency (EPA) 2017. Water face sheet. Water Sense EPA.

U.S. Geological Survey (USGS) 2005. National Field Manual for the Collection of Water- Quality Data. Accessed [06/15/2019]: https://water.usgs.gov/owq/FieldManual/compiled/NFM_complete.pdf.

U.S. Geological Survey (USGS) 2014. Origins and Delineation of Saltwater Intrusion in the Biscayne Aquifer and Changes in the Distribution of Saltwater in Miami-Dade County, Florida.

U.S. Geological Survey (USGS) 2015. Estimated use of water in the United States. Accessed [09/07/2018]: https://water.usgs.gov/watuse/wuto.html.

U.S. Geological Survey (USGS) 2016a. Groundwater Quality Report. Accessed [06/07/2018]: https://pubs.usgs.gov/wri/wri024045/htms/report2.htm.

U.S. Geological Survey (USGS) 2016b. Groundwater Availability Report. Accessed [08/08/2018]: https://www.sciencebase.gov/catalog/item/5661981ae4b06a3ea36c56fc.

U.S. Census Bureau 2015. Coastal areas. Accessed [09/12/2018]: https://www.census.gov/topics/preparedness/about/coastal-areas.html.

Werner, A. D., Bakker, M., Post, V. E. A., Vandenbohede, A., Lu,C., Ataie-Ashtiani, B., Simmons, C. T., and Barry, D. A.2013. Seawater intrusion processes, investigation and management: recent advances and future challenges. Advances in Water Resources 51:3-26.

Werner, A.D., and Gallagher, M.R. 2006. Characterization of seawater intrusion in the Pioneer Valley, Australia using hydrochemistry and three-dimensional numerical modelling. Hydrogeological Journal 14:1452-1469.

Wu, J., Wang, L., Wang, S., Tian, R., Xue, C., Feng, W., and Li, Y. (2017). Spatiotemporal variation of groundwater quality in an arid area experiencing long-term paper wastewater irrigation, Northwest China. Environmental Earth Sciences 76(13): 460.

Zhang, E., Savenije, H., Wu, H., Kong, Y., and Zhu, J. 2011. Analytical solution for salt intrusion in the Yangtze Estuary, China, Estuarine. Coastal and Shelf Science 91: 492-501.

INVESTIGATING GROUNDWATER VARIATIONS IN BASINS OF THE WORLD

4.1 Abstract

Land use can accelerate the depletion of groundwater resources that support humans and ecosystem services on a global scale. The expansion of irrigated lands, which accounts for approximately 85 percent of worldwide water consumption, is a reason for groundwater depletion.

The objective of this study was to provide an overview of groundwater depletion in basins of the world where small scale basins are highly vulnerable to groundwater depletion. The agricultural activities on the groundwater level further analyzed. In addition to irrigation data, data from the

Gravity Recovery and Climate Experiment (GRACE) satellite and the Global Land Data

Assimilation System (GLDAS) from April 2002 to December 2015 were used. Results indicated that groundwater depletion mostly occurred in southwestern Europe, southwestern Australia, southern South America, western United States, and central Asia, e.g., groundwater decreased at the rate of -0.10 cm month-1 in the Indus basin. These findings collectively demonstrated small size basins are the most water-scarce basins of the world, and global groundwater security cannot be protected without considering groundwater seasonal and monthly variations and land use on groundwater resources.

4.2 Introduction

Groundwater and soil moisture collectively account for over 98 percent of global freshwater resources, and more than two billion people depend on groundwater (Hiscock 2005).

Groundwater use around the world includes irrigation purposes, food production, industrial use, and domestic use (Döll et al. 2012, Shah et al. 2000), and it accounts for 85 percent of global irrigation water (Wada et al. 2012, Gleeson et al. 2012). Total groundwater use in agricultural activities is much higher in Asia and North America (GWP 2012). The increased rate of groundwater use has caused groundwater depletion as these resources were used to increase crop production by 250 percent during 1970-2000 in Asian and North American countries (IAH 2015).

Groundwater withdrawal rates vary in different parts of the world. In Africa, 86 percent of withdrawal water is used in agriculture (FAO 2016), and groundwater also accounts for 52 percent of freshwater used in irrigation in the Middle East. Twenty-five percent of withdrawn groundwater is used in the agricultural activities in eastern and southern Asia, and groundwater also accounts for 23 percent of freshwater used in irrigation in eastern Europe (FAO 2016). Previous studies

(Wada et al. 2010, Gleeson et al. 2012) demonstrated that eight out of ten countries with the high groundwater extraction are in Asia, including India, China, Nepal, Bangladesh, and Pakistan; half of the world’s total groundwater use in these countries is associated with the largest population and intensive agricultural activities (Gleeson et al. 2012).

Ecosystems are entirely dependent on groundwater; it supports soil moisture in the root zone and provides water to rivers, lakes, and wetlands (Fan et al. 2013, Wösten et al. 2008, Lee et al. 2018, Naumburg et al. 2005, Clifton and Evans 2001). Groundwater contamination and depletion are increasing because of the overuse and misuse of groundwater in many parts of the world (Narasimhan 2009). Long-term groundwater extraction has led to groundwater drought,

100

aquifer loss, and seawater intrusion in the coastal zones (Fienen and Arshad 2016) especially, shallow groundwater is exposed to contaminants in populated urban areas and irrigated lands. An example of this can be seen in coastal zones of Australia, where seawater intrusion caused the salinization of groundwater resources (Appelo and Postma 2005, Greene et al. 2016).

The Groundwater Markup Language (Boisvert and Brodaric 2011), Aquastat (FAO 2013), and the First Groundwater Interoperability Experiment (Open Geospatial Consortium Inc. 2013) serve as organizations that bring nations together to protect groundwater. These organizations try to improve data consistency when estimating groundwater depletion; however, data interpretation and data collection methods remain inconsistent across different countries (Giordano 2009).

Three-dimensional models are used to collect groundwater data throughout the world, such as

Numerical Simulation Models, including HydroGeoSphere (Therrien et al. 2012), GSFLOW

(Markstrom et al. 2008), and MIKE SHE (DHI Software 2012). Gravity Recovery and Climate

Experiment (GRACE) satellite data has allowed scientists to measure the groundwater storage on a large scale in the last decade; GRACE data values are supplied by measuring the Earth’s gravity

(Tapley et al. 2004), which is detected from a distance between two coordinated satellites that are generally separated by about 220 km (Tapley et al. 2004). GRACE observes gravity anomalies which are converted to monthly variations in terrestrial water storage (Rodell et al. 2009); therefore, variations in gravity are generally a function of changes in water storage (Ramillien et al. 2006). The hot spots for groundwater depletion are recorded by GRACE satellite data and include the High-Plains of the United States (Scanlon et al. 2012), India (Rodell et al. 2009), the

Tigris and Euphrates river basins, western Iran, and North Africa (Voss et al. 2013).

101

4.3 Objectives

In previous studies (Wada et al. 2010, Gleeson et al. 2012), groundwater storages were estimated throughout the world using remote-sensing technologies and hydrologic models. This study used the GRACE satellite data to provide a comprehensive overview of global groundwater depletion and rise, and groundwater seasonal and monthly variations in the basins of the world.

This study also focused on the impacts on groundwater consumption, where groundwater resources provide essential support for land irrigation and where groundwater resources are at risk of contamination and depletion.

4.4 Background

Groundwater depletion and scarcity became a significant problem globally due to groundwater over-withdrawal after the 1980s (Changming et al. 2001). Groundwater withdrawal is high in arid and semiarid areas, in addition to regions high in population density with low rates of annual recharge (Gleeson et al. 2011). Studies have shown an approximate doubling of global groundwater withdrawal between 1960 and 2000 (Konikow 2011, Morris et al. 2003, Wada et al.

2010).

The basins and aquifers in the Middle East and North Africa are identified as regions with severe water scarcity. Population distribution is based upon access to groundwater for use in agricultural activities in these regions, such as in Iran, Iraq, and Egypt (Abu Zeid 2006). An increase in the number of irrigation wells and the rate of groundwater withdrawal has led to a declining level of groundwater in many aquifers in Iran (Nabavi 2018). Rates of groundwater depletion were reported as 283 km3 a-1 in parts of Iran, central Yemen, and southern Spain (Wada et al. 2010, Konikow 2011, Morris et al. 2003). Iraq’s groundwater aquifers are large scale alluvial deposits, and these aquifers have poor water quality because higher temperature and reduced 102

rainfall has led to increased groundwater depletion (Obeed Al-Azaw and Ward 2016). About 143 km3 groundwater resources were depleted from 2003 to 2009 in eastern Tigris and Euphrates River basin (Voss et al. 2013). The annual groundwater extraction in the Nile aquifer is about 4.6 billion m3; groundwater withdrawal occurs in various parts of Egypt and groundwater uneven distribution makes it impossible to have a comprehensive groundwater management policy for the country

(Elnasha 2014). Most Arab countries, such as Oman and Saudi Arabia, are also extremely dependent on groundwater resources (Fienen and Arshad 2016); as such, Saudi Arabia is the third principal per capita water user in the world and over-withdrawal of non-renewable groundwater resources has resulted in water scarcity (Kajenthira et al. 2011). The aquifer in northern and eastern

India is likewise experiencing severe depletion for crop irrigation; over 45 million ha of land in

India is irrigated by groundwater (Fishman et al. 2011). The depletion of groundwater also occurred in northern China due to the installation of private wells for crop irrigation (Wang et al.

2007). In the North China Plain, groundwater levels have been declined at rates of 1 m per year due to the irrigation of winter wheat and summer maize (Yang et al. 2017).

In southern Europe, surface water cannot provide water demand from agriculture, industry, and households. In the Segura River basin in southeastern Spain, the crop irrigation over the last four decades has resulted in groundwater depletion (Martínez-Granados 2017); similarly, in the

Murray-Darling aquifer, below average rainfall reduces the availability of groundwater supplies, which affected much of the 2000s in southeast Australia (ABS 2010). In western Australia, over- withdrawal of groundwater, low rainfall, and dry climate (ABS 2010).

Groundwater resources have depleted about 800 km3 in aquifers of the United States over the last century (Konikow and Kendy 2005). The High-Plains aquifer serves as a groundwater resource in the United States (Konikow 2011, Morris et al. 2003, Wada et al. 2010, Werner et al. 103

2013, Dennehy et al. 2002). Scanlon et al. (2012) reported that 36 percent of the groundwater in the United States came from High-Plains depleted during 1900-2000 for crop irrigation.

Groundwater depletion occurred in the Lower Mississippi River basin where groundwater resources were used in rice production (Reba et al. 2017). In North America, the Yukon River basin is a permafrost watershed where groundwater depletion occurs due to the irregular rainfalls

(Muskett and Romanovsky 2011). Furthermore, although ice cover tends to be more sensitive to air temperature variations at lower than at higher latitudes (Livingstone et al. 2010), remote sensing observations indicate that ice cover loss seems to be more rapid in very high latitude lakes

(Latifovic and Pouliot 2007).

Although South America contains four of the world largest rivers—the Amazon, Paraná,

Orinoco, and Magdalena—drought has been affecting Eastern Brazil since 2012, and this condition causes depletion of both surface water and groundwater storages (Getirana 2016). In previous studies (Rodell et al. 2009, Scanlon et al. 2016), large groundwater reductions are found in the

Ganges, Brahmaputra, Salado, Don, Sao Francisco, Ob, Yukon, Haunghe, Dnieper, and Indus basins, and also in the Tigris and Volga basins.

Water consumption, irrigation, and climate change pressures affect groundwater levels and their temporal patterns, thereby threatening vital ecosystems such as arable irrigation lands

(Kundzewicz et al. 2007, IPCC 2007). The Intergovernmental Panel on Climate Change (IPCC

2007) has estimated the increase in global average surface temperature by 0.740 C from 1906 to

2005, and it will affect the renewable groundwater resources and its level. The inherent time lag between stress and a detectable system response exacerbates the vulnerability of groundwater resources to human activities, such as agricultural activities (Kundzewicz et al. 2007).

104

The impacts of expected climate changes and the world demographic increase on groundwater resources should be estimated. Previous studies presented different methods to estimate groundwater recharge. Chen et al. (2002) and Serrat-Capdevila et al. (2007) used a simple linear function that contains precipitation and temperature. The rate and magnitude of change in temperature and precipitation will affect groundwater recharge (Poff et al. 2002, Treidel et al.

2012). Soil properties or vegetation type and water use can also affect groundwater recharge rates

(Van Roosmalen et al. 2007). Döll and Florke (2005) also estimated climate change impacts on groundwater recharge at the global scale; their results indicated a decrease in groundwater recharge

(by more than 70 percent) for northeastern Brazil, the western part of southern Africa, and areas along the south of the Mediterranean Sea by 2050s. They also predicted an increase in groundwater recharge (by greater than 30 percent) across large areas, including the Sahel, Northern China, and the Western United States, and Siberia by 2050s.

Previous studies developed different methods to manage and monitor global water resources. The Asian Development Bank (ADB) provided a national framework of water security measure, the national water security index (NWSI), which has been used to compare nations’ water security performance in Asia and the Pacific region (ADB 2016). The NWSI includes five key dimensions: household, economic, urban, and environmental security and resilience to water- related disaster. Aggarwal et al. (2014) proposed a water insecurity index (WII) applied on a regional scale in India (Aggarwal et al. 2014). The framework has six key dimensions: resource, access, consumption, capacity, environment, and climatic stress. Wutich and Ragsdale (2008) presented the concept of household water insecurity in three dimensions: inadequate water supply, insufficient access to water distribution systems, and dependence on seasonal water sources. Wada et al. (2014) introduced the water scarcity index (WSI) defined as the ratio of total water 105

withdrawal to water availability. Kaufmann et al. (2010) presented the world governance index; the values with higher index value lead to increase water security. Döll et al. (2012) assessed the effect of source-differentiated withdrawals on the different water storage basins by using the

WaterGAP global hydrological model (WGHM, resolution: 0.5◦ x 0.5◦ grid). Hoekstra et al. (2012) assessed global blue water scarcity on the basin level. Blue water scarcity is defined as the ratio of total blue water footprint by means of total consumptive use and blue water availability (available surface water and groundwater). Gleeson et al. (2012) assessed the groundwater footprint as an index presenting the intensity of groundwater resources usage of big aquifer systems worldwide.

The groundwater footprint was defined as the area-averaged annual withdrawal of groundwater divided by the recharge rate minus the groundwater contribution to streamflow.

Beyond groundwater quantity, groundwater quality concerns can arise when contaminants enter an aquifer. Groundwater is never free from impurities, and it typically is composed of dissolved constituents (e.g., salts, organic and inorganic chemicals) (Morris et al. 2003). Human activities add chemicals, salts, and microorganisms that influence groundwater quality (Foster et al. 2002, Morris et al. 2003), and deeper aquifers react with delay to terrestrial contaminants compared to the shallow aquifers (Fishman et al. 2011).

Human activity impacts the quantity and quality of groundwater resources and groundwater ecosystems. It is predicted that human activities’ impact on groundwater will continue, at least until 2025, unless new groundwater management policies change this state of concern (Danielopol et al. 2003).

106

4.5 Data and Method

4.5.1 GRACE‐Derived Total Water Storage Data

The Gravity Recovery and Climate Experiment (GRACE) RL05 satellite mission enabled remote sensing of Terrestrial Water Storage anomalies (TWS) at regional to continental scales.

Several versions of mapped TWS were provided via the GRACE Tellus website at the Center for

Space Research (CSR), GeoForschungsZentrum (GFZ), and Jet Propulsion Laboratory (JPL). The spatial sampling of all grids was 1 degree in both latitude and longitude, i.e., about 111 km at the equator (see Table 4.2).

The GRACE data provided global coverage of monthly anomalies in total water storage in terms of equivalent water height (cm). GRACE provided data grids of monthly anomalies with noise (Swenson and Wahr 2002), atmospheric pressure, a glacial isostatic adjustment (GIA) (Wahr and Zhong 2013) and mass change corrections (Velicogna and Wahr 2006, Swenson and Wahr

2006). A de-striping filter was applied to the data to minimize the effect of correlated errors, as well. In GRACE data, time series for large river basins have RMS uncertainties that are generally

<2 cm, while smaller basins have RMS uncertainties about 3-4 cm. Uncertainties are highest for the smaller basins. The GRACE data accuracy is 10-30 mm depending basin size and removal other water storage signals (Scanlon et al. 2016).

Moisture (SM) and Snow Water Equivalent (SWE), as expressed in the eq. (Bettadpur 2007) [4.1].

∆TWS=∆GW + ∆SM +∆SWE [4.1]

107

A total of 143 months of data from TWS GRACE-derived anomalies was quantified between April 2002 and December 2015, at the Center for Space Research (CSR) at the University of Texas in Austin (Bettadpur 2007, Chambers 2006).

4.5.2 Global Land Data Assimilation System Output

GRACE data does not directly provide information about groundwater measurements. The

Global Land Data Assimilation System (GLDAS) data was used along with GRACE data to quantify groundwater measurements. The GLDAS is a land surface assimilation system which aims to utilize satellite- and ground-based observational data products (see Table 4.2). It uses Noah

2.7.1 land hydrology model and data assimilation techniques in order to generate optimal fields of land surface state (e.g., soil moisture, snow, and surface temperature) and flux (e.g., evapotranspiration, sensible heat flux) products (Rodell et al. 2004).

Data was used from the GLDAS version1 (GLDAS-1) 1.0-degree resolution covering the period from April 2002 to December 2015. The GLDAS contains accumulated snow and soil moisture monthly anomalies that were used to separate groundwater storage from TWS provided by GRACE data. A GLDAS grid refers to the total accumulated snow and soil moisture variations for that month relative to a baseline average obtained from January 2003 to December 2007. In order to ensure a valid analysis of the data, anomalies from GRACE and GLDAS should be within the same time period. The GRACE data time baseline was converted from January 2004 to

December 2009 to January 2003 to December 2007. This is done by averaging the monthly gravity data reported by CSR between January 2003 and December 2007 and removing the mean value from each month to compute the anomalies (GRACE 2018).

108

4.5.3 Irrigation Data

Irrigation data provided by the State of the World’s Land and Water Resources for Food and Agriculture was used (FAO 2010). The spatial sampling of all grids was 0.07 degrees in both latitude and longitude (see Table 4.2). Figure 1 presents the percentage of irrigated lands using groundwater resources. Groundwater resources were intensively used in the United States, southern Europe, India, and Asia (Fig.4.1). The basins of the world were provided by the World- wide Hydrogeological Mapping and Assessment Programme (WHYMAP 2011) (see Table 4.2), and data from a total of 37 basins in Europe, 72 basins in Asia, 40 basins in Africa, 36 basins in

South America, 54 basins in North America, and 9 basins in Australia were used in this study.

109

Figure 4.1 Illustrating percentage irrigated areas using groundwater resources (FAO 2010).

110

Table 4.1 Data sources.

Data Data Year Source Description Type GRACE Raster April2002- GRACE Total water storage thickness unit is cm. Spatial Satellite Dec2015 resolution is 1-degree. Data provided from: https://grace.jpl.nasa.gov/data/get-data/land-water- content/ GLDAS Raster April2002- GRACE Total water (accumulated snow and soil moisture) is Dec2015 cm. Spatial resolution is 1-degree Data provide Land from: https://grace.jpl.nasa.gov/data/get-data/land- Surface water-content/ Model Basins Shape 2010 WHYMAP River and Groundwater Basins of the World (2011) file

Percentage Raster & 2007 Aquastat (FAO) Developed by Food and Agriculture Organization of irrigated Feature (FAO), Spatial resolution is 0.07 degree. Data area using File provided from: groundwater http://www.fao.org/nr/water/aquamaps/#map

4.5.4 Analysis of Groundwater Depletion

A simple linear regression statistical technique was used in R software to calculate the per- pixel linear-trend model in groundwater water storage (cm month-1) from April 2002 to December

2015. The p-value that represents a probability of the error when the trend differs from zero was then quantified. Linear trend estimation expresses data as a linear function of time, and the independent variable of each pixel is a time series of monthly driven groundwater anomalies. Gaps between monthly data vary from one month to two months. The simple linear regression was an appropriate statistical method to quantify a trend because there was not a large gap between time series (Barnes and Barnes 2015, Li et al. 2017). The simple linear regression model as expressed in Eq. [4.2].

111

y= β0+β1 t +ɛ [4.2]

where y was a dependent variable, which indicates groundwater storage, and t was the independent variable, which indicates times series, and β1 is a trend coefficient which varies from pixel to pixel.

After calculating the per-pixel trends in groundwater anomalies, zonal statics was conducted in a GIS to estimate the average trend in groundwater anomalies in each basin. Further, we also used zonal statistics to estimate the percent irrigated land that is using groundwater in each basin. Finally, we plotted the basin size, percent irrigated area, and groundwater trend to analyze if there is any relation among them.

4.6 Results and Discussions

4.6.1 Trends in Groundwater Storage Anomalies

The relationship among groundwater withdrawal, climate, and groundwater level is complicated and spatially variable; therefore, the similarity of trends between pixels is illustrated in two groups. Both groundwater rise and depletion indicated that the groundwater levels are changing throughout the world.

Groundwater storage trends were analyzed for each GRACE pixel over the study time period, 2002 to 2015, using a simple linear regression slope. Positive slopes indicate an increase in groundwater storage, while negative slopes indicate a groundwater decline. The trend map was then overlaid with the basin boundaries. Almost all the basins in the world have shown fluctuations in groundwater levels, indicating a mix of negative and positive statistical trends in the groundwater level of each basin.

North American Basins

112

Long-term groundwater level trends varied in magnitude and direction across the North

American basins during the study period. The average groundwater trend calculated from zonal statistics showed an overall decreasing trend (-0.008 cm month-1) in the Colorado River basin.

However, the per-pixel groundwater trend explained the significant (p<0.05) long-term declining trends in the southwestern part of Colorado River basin while south and southeastern parts had a significantly increasing trend. There was no significant change in other regions within Colorado basin. In this basin, about 3.2 percent land under irrigation uses groundwater for irrigation (see

Table 4.3).

Average groundwater trend increased at a rate of 0.02 cm month -1 in the Mississippi River basin. The per-pixel groundwater trend analysis explained that the north western part of the basin showed a significantly increasing trend while the area along the Mississippi river in Louisiana and

Mississippi showed a declining trend (Fig.4.2). In this basin, about 25.2 percent land under irrigation uses groundwater for irrigation (see Table 4.3).

Average groundwater trend in the Nelson River increased at a rate of 0.003 cm month-1 .

The per-pixel groundwater trend analysis showed long-term declining trends in the eastern Nelson

River basin while the central area in this basin showed increasing trends. In this basin, about 2.1 percent land under irrigation uses groundwater for irrigation (see Table 4.3).

Average groundwater trend in the Columbia River increased at a rate of 0.04 cm month-1 .

The per-pixel groundwater trend analysis showed long-term increasing trends in eastern and western parts of this basin while there was no significant change in other regions within this basin

(Fig.4.2). In this basin, about 10.6 percent under irrigation uses groundwater for irrigation (see

Table 4.3).

113

Average groundwater trend in the Mackenzie River basin increased at the a rate of 0.09 cm month-1 . The per-pixel groundwater trend analysis showed long-term increasing trends in central areas while the eastern and northwestern areas in this basin showed decreasing trends. In this basin, about 0.1 percent under irrigation uses groundwater for irrigation (see Table 4.3).

Table 4.2 Average groundwater trend and percentage irrigated areas using groundwater in North American basins.

Basins Average Groundwater Trend Irrigated lands using groundwater (%) Colorado River -0.008 3.2 Mississippi River 0.02 25.2 Nelson River 0.003 2.1 Columbia River 0.04 10.6 Mackenzie River 0.09 0.1

114

Figure 4.2 Illustrating global groundwater trend and p-value during 2002-2015 in North America.

South American Basins

Large regions with significant groundwater declines included the eastern and southern regions of South America. Average groundwater level trend decreased from 2002 to 2015 with a rate of -0.06 cm month-1 in the Sao Francisco basin. The per-pixel groundwater trend analysis showed long-term groundwater declining trends throughout this basin (Fig.4.3.). In this basin, about 0.8 percent under irrigation uses groundwater for irrigation (see Table 4.4).

Similarly, average groundwater storage trend decreased from 2002 to 2015 in the Salado basin in southern South America with a rate of -0.05 cm month-1. The per-pixel groundwater trend

115

analysis showed long-term groundwater declining trends throughout this basin (Fig.4.3). In this basin, about 0.2 percent under irrigation uses groundwater for irrigation (see Table 4.4).

Average groundwater level trend increased in the Amazon basin with a rate of 0.07 cm month-1. The per-pixel groundwater trend analysis showed significant long-term groundwater increasing trends in the central regions of the Amazon basin (Fig. 4.3). In this basin, about 0.2 percent under irrigation uses groundwater for irrigation (see Table 4.4).

Table 4.3 Average groundwater trend and percentage irrigated areas using groundwater in South American Basins.

Basins Average Groundwater Irrigated lands Trend using groundwater (%) Sao Francisco -0.06 0.8 Salado -0.05 0.2 Amazon 0.07 0.2

116

Figure 4.3 Illustrating global groundwater trend and p-value during 2002-2015 in South America.

African Basins

Average groundwater trends declined in the largest basins of Africa, including the Lake

Chad, with a rate of -0.02 cm month-1. However, the per-pixel trend analysis showed groundwater rises in northern Lake Chad while the southern area in this basin showed declining trends

(Fig.4.4.). In this basin, about 0.17 percent under irrigation uses groundwater for irrigation (see

Table 4.5). Average groundwater trends declined with a rate of -0.04 cm month -1 in the North

117

Interior basin in Africa. The per-pixel analysis showed long-term groundwater declining trends in northwestern and western regions of this basin while the eastern regions showed long-term increasing trends in this basin (Fig. 4.4.). In this basin, about 0.1 percent under irrigation uses groundwater for irrigation (see Table 4.5).

Groundwater significantly increased in the largest basins of Africa, including the Niger,

Nile, Zambezi, and Orange basins. The average groundwater trend increased with a rate of 0.04 cm month-1 in the Niger basin. The per-pixel groundwater trend explained the significant (p<0.05) long-term declining trends in the northern Niger basin while the southern area in this basin showed increasing trends (Fig.4.4.). In this basin, about 0.07 percent under irrigation uses groundwater for irrigation (see Table 4.5).

Similarly, average groundwater trend in the Nile basin increased at a rate of .01 cm month-

1 . The per-pixel groundwater trend explained the significant (p<0.05) long-term declining trends in the eastern and southwestern Nile basin while there was no significant change in other regions within this basin (Fig.4.4). In this basin, about 0.17 percent under irrigation uses groundwater for irrigation (see Table 4.5).

Average groundwater trend increased in the Zambezi basin with a rate of .12 cm month-1.

The per-pixel groundwater trend explained the significant (p<0.05) long-term increasing trends throughout this basin (Fig.4.4). In this basin, about .06 percent under irrigation uses groundwater for irrigation (see Table 4.5). Similarly, average groundwater trend increased in the Orange basin with a rate of 0.12 cm month-1 . The per-pixel groundwater trend explained the significant (p<0.05) long-term increasing trends throughout this basin while northwestern regions showed the significant long-term declining trends (Fig.4.4.). In this basin, about 0.06 percent under irrigation uses groundwater for irrigation (see Table 4.5). 118

Table 4.4 Average groundwater trend and percentage irrigated areas using groundwater in African basins.

Basins Average Groundwater Irrigated lands Trend using groundwater (%) Lake Chad -0.02 0.17 North Interior -0.04 0.10 Basin Niger 0.04 0.07 Nile 0.01 0.17 Zambezi 0.12 0.06 Orange 0.12 0.06

Figure 4.4 Illustrating global groundwater trend and p-value during 2002-2015 in Africa.

119

European Basins

Groundwater depletion furthermore occurred in central, eastern, and southwestern Europe, where populations are highly dependent on groundwater resources (Konikow 2011). Average groundwater level trend decreased in the Dnieper basin at a rate of -0.03 cm month-1. The per-pixel groundwater trend explained the significant (p<0.05) long-term declining trends throughout the basin (Fig.4.5.). In this basin, about 0.9 percent under irrigation uses groundwater for irrigation

(see Table 4.6).

The average groundwater level trend decreased at a rate of -0.02 cm month-1 in the Danube basin. The per-pixel groundwater trend explained the significant (p<0.05) long-term declining trends in the eastern and north western areas in the Danube basin while there was no significant change in other regions within this basin (Fig.4.5). In this basin, about 5.2 percent under irrigation uses groundwater for irrigation (see Table 4.6).

Average groundwater level trend in the Tejo basin decreased at a rate of -.05 cm month-1

.The per-pixel groundwater trend explained the significant (p<0.05) long-term declining trends throughout the Tejo basin (Fig.4.5.). In this basin, about 32.9 percent under irrigation uses groundwater for irrigation (see Table 4.6).

Average groundwater trend in the Seine basin decreased at a rate of -.06 cm month-1 . The per-pixel groundwater trend explained the significant (p<0.05) long-term declining trends throughout the basin (Fig.4.5). In this basin, about 56 percent under irrigation uses groundwater for irrigation (see Table 4.6). Also, the average groundwater level trend in the Loire basin decreased at a rate of -.04 cm month-1 (Fig. 4.2). The per-pixel groundwater trend explained the significant (p<0.05) long-term declining trends throughout the basin (Fig.4.5). In this basin, about

50.3 percent under irrigation uses groundwater for irrigation (see Table 4.6). 120

The average groundwater level trend increased at a rate of 0.02 cm month-1 in the Volga basin. The per-pixel groundwater trend explained the significant (p<0.05) long-term increasing trends throughout northern areas while southern parts had decreasing trends (Fig.4.5). In this basin, about 4.4 percent under irrigation uses groundwater for irrigation (see Table 4.6).

Table 4.5 Average groundwater trend and percentage irrigated areas using groundwater in European basins.

Basins Average Groundwater Irrigated lands Trend using groundwater (%) Dnieper -0.03 0.9 Danube -0.02 5.2 Tejo -0.05 32.9 Seine -0.06 56 Loire -0.04 50.3 Volga 0.02 4.4

121

Figure 4.5 Illustrating global groundwater trend and p-value during 2002-2015 in Europe.

Asian Basins

The average long-term groundwater level trends indicated declining groundwater storage in the Indus, Ganges, Brahmaputra, Tigris and Euphrates, and Huang He (Yellow River) basins.

The average groundwater trend in the Indus basin decreased at a rate of -0.10 cm month-1 . The per-pixel groundwater trend explained the significant (p<0.05) long-term declining trends almost throughout the basin (Fig.4.6). In this basin, about 21.1 percent under irrigation uses groundwater for irrigation (see Table 4.7).

The average groundwater trend in the Ganges basin decreased at a rate of -0.13 cm month-

1 . The per-pixel groundwater trend explained the significant (p<0.05) long-term declining trends 122

almost throughout the basin (Fig.4.6). In this basin, about 55.6 percent under irrigation uses groundwater for irrigation (see Table 4.7). The average groundwater trend in the Brahmaputra basin decreased at a rate of -0.11 cm month-1. The per-pixel groundwater trend explained the significant (p<0.05) long-term declining trends almost throughout the basin (Fig.4.6). In this basin, about 9 percent under irrigation uses groundwater for irrigation (see Table 4.7).

The average groundwater trend in the Tigris and Euphrates basin decreased at a rate of -

0.10 cm month-1. The per-pixel groundwater trend explained the significant (p<0.05) long-term declining trends almost throughout the basin (Fig.4.6). In this basin, about 7.5 percent under irrigation uses groundwater for irrigation (see Table 4.7). Similarly, the average groundwater trend in the Huang He (Yellow River) basin decreased at a rate of -0.03 cm month-1. The per-pixel groundwater trend explained the significant (p<0.05) long-term declining trends almost throughout the basin (Fig.4.6). In this basin, about 15.1 percent under irrigation uses groundwater for irrigation

(see Table 4.7).

Groundwater significantly increased in the largest basins of northern Asia, includes the OB and Lena basins. Average groundwater level trend increased at a rate of 0.07 cm month-1 in the

OB basin and at a rate of 0.07 cm month-1 in the Lena basin (see Table 4.7). The per-pixel groundwater trend explained the significant (p<0.05) long-term increasing trends throughout these basins (Fig.4.6). In OB basin, about 0.13 percent under irrigation uses groundwater for irrigation, and in the Lena basin, about 0.03 percent under irrigation uses groundwater for irrigation (see

Table 4.7).

123

Table 4.6 Average groundwater trend and percentage irrigated areas using groundwater in Asian basins.

Basins Average Groundwater Irrigated lands Trend using groundwater (%) Indus -0.10 9 Ganges -0.13 55.6 Brahmaputra -0.11 9 Tigris and -0.10 7.5 Euphrates Huang He -0.03 15.1 OB 0.07 0.13 Lena 0.07 0.03

Figure 4.6 Illustrating global groundwater trend and p-value during 2002-2015 in Asia.

Australian Basin

Average groundwater trend in the Murray basin increased with a rate of 0.01 cm month-1.

124

The per-pixel groundwater level trend explained the groundwater depletion in the southern Murray basin while the northern areas had significantly increasing trends (Fig 4.7). In these basins, about

0.2 percent under irrigation uses groundwater for irrigation (see Table 4.8).

Average groundwater trend in the Eyre Lake basin increased with a rate of 0.03 cm month-

1. The per-pixel groundwater level trend explained the groundwater depletion in the southeastern

Eyre lake basin while the northern areas had significantly increasing trends (Fig 4.7). In these basins, about .01 percent under irrigation uses groundwater for irrigation (see Table 4.8). Average groundwater trend in the West Coast basin decreased with a rate of -0.05 cm month-1. The per- pixel groundwater level trend explained the groundwater depletion throughout this basin (Fig 4.7).

In these basins, about 0.01 percent under irrigation uses groundwater for irrigation (see Table 4.8).

Table 4.7 Average groundwater trend and percentage irrigated areas using groundwater in Australian basins.

Basins Average Groundwater Irrigated lands Trend using groundwater (%) Murray 0.01 0.2 Eyre Lake 0.03 0.01 West Coast -0.05 0.1

125

Figure 4.7 Illustrating global groundwater trend and p-value during 2002-2015 in Australia.

Thus, groundwater storage typically increased in the central basins in North America and

South America, and the groundwater level increased in the southern basins in Africa. Groundwater storage increased in northern Asia, whereas the central and southern Asian basins experienced groundwater depletion. Additionally, groundwater storage depleted in the coastal basins in

Australia, North America, South America, and western Europe. It is important to identify the regions with a long-term declining groundwater trend on a global scale; it helps predict aquifer drought and seawater intrusion in coastal basins due to declining groundwater storage and hydraulic gradient, and the effects of sea-level rise.

126

4.6.2 Groundwater Storage Anomalies

Despite extensive negative and positive trends across the world, the average monthly groundwater storage anomalies varied by basins during the study period, 2002-2015 (Figs.4.8, 4.9,

4.10., 4.11, 4.12 and Appendix Fig. B.1).

Asian Basins

The Ganges basin was experiencing a groundwater decline at a rate of -0.13 cm month-1 from 2002 to 2015 (Fig. 4.8). The groundwater level reached a maximum in the Ganges basin in

December 2003, after that there was a declining trend overall. Groundwater reached the lowest point in May 2013 (Fig. 4.8). May is the hottest month in India, and therefore, these areas experience high evapotranspiration (IPCC 2014) during May.

Brahmaputra basin experienced a groundwater decline at a rate of -0.11 cm month-1 from

2002 to 2015. Groundwater storage reached a maximum level in November 2004 (Fig. 4.8).

Groundwater storage trend fluctuated until September 2008 but showed a continuously negative trend afterwards.

The Indus basin experienced a groundwater decline at a rate of -0.10 cm month-1 from 2002 to 2015. While the groundwater storage fluctuated frequently before 2011, the average groundwater storage showed negative anomalies from February 2011 to February 2015 (Fig. 4.8).

The Tigris and Euphrates basin similarly experienced a groundwater decline at a rate of -

0.10 cm month-1, from 2002 to 2015. The groundwater had negative average anomalies throughout the study period except for few months (Fig.4.8). Approximately 90 percent of the annual rainfall occurs between November and April in central Asia (IPCC 2014).

Asian basins depend much more heavily on groundwater during the dry seasons and drought (Lee et al. 2018). Groundwater anomalies showed that the water levels have gone down

127

in some areas and, at the same time, have risen in other areas. Collectively, a period of groundwater recovery was very short in several regions of Asia, due to climate conditions and excessive groundwater use.

Ganges Basin Groundwater Level Brahmaputra Basin Groundwater Level Anomalies Anomalies 20 40 10 20 Dec2003 0 0 -10 -20 Sep2008 Nov2004 Sep2008 May2013 -20

-40 (cm) Anomalies Montly

Monthly Anomalies Anomalies (cm) Monthly

200204 200303 200401 200411 200508 200605 200703 200712 200809 200907 201004 201102 201201 201212 201401 201501

200809 200303 200401 200411 200508 200605 200703 200712 200907 201004 201102 201201 201212 201401 201501 200204 Indus Basin Groundwater Level Anomalies Tigris and Euphrates River Basin 15 Groundwater Level Anomalies 10 5 20 10 0 0 -5 -10 -10 -20 June2006 Feb2011

-15 Feb2015 -30

Monthly Anomalies Anomalies (cm) Monthly

200204 200303 200401 200411 200508 200605 200703 200712 200809 200907 201004 201102 201201 201212 201401 201501

200303 200401 200411 200508 200605 200703 200712 200809 200907 201004 201102 201201 201212 201401 201501 200204 Figure 4.8 Illustrating average monthly groundwater anomalies in the major Asian basins.

European Basins

The Dnieper and Don basins in Europe experienced a groundwater decline during the study period, 2002-2015 (Fig. 4.9). The Dnieper basin experienced a groundwater decline at a rate of -

0.03 cm month-1; furthermore, the groundwater level reached a minimum level in the Dnieper basin in January 2012 and January 2014 (Fig. 4.9).

128

The Don basin experienced a similar groundwater decline at a rate of -0.03 cm month-1, and the groundwater level reached a minimum level in the Don basin in December 2012 (Fig.4.9), because the groundwater recharge decreased at the basins of the major rivers of European Russia as the winter runoff increased (Telegina 2014).

Dnieper Basin Groundwater Level 15 Don Basin Groundwater Level Anomalies Anomalies 10 10 5 5 0 0 -5 -5 -10 -10 -15 Nov2011 -15 Apr2011 Jan2014 Dec2012

-20 Jan2012 Anomalies (cm) Monthy -20

Monthly Anomalies Anomalies (cm) Monthly

201401 200204 200303 200401 200411 200508 200605 200703 200712 200809 200907 201004 201102 201201 201212 201501

200303 200401 200411 200508 200605 200703 200712 200809 200907 201004 201102 201201 201212 201401 201501 200204 Figure 4.9 Illustrating average monthly groundwater anomalies in the major European basins.

African Basins

The Lake Chad experienced a groundwater decline at a rate of -0.02 cm month-1 . The groundwater level reached a maximum in the Lake Chad basin in December 2012 in the Lake Chad basin (Fig.4.10).

The Nile basin experienced a groundwater increase at a rate of 0.01 cm month-1 . In the

Nile basin, the groundwater level reached a maximum level in January 2015 (Fig.4.10). However, the rainiest period is from July to September in central north Africa and from May to October in southern Africa (Future Climate for Africa 2016).

129

Lake Chad Basin Groundwater Level Nile Basin Groundwater Level Anomalies Anomalies 15 15 10 10 5 5 0 0 -5 -5 -10 Jan2015

Dec2012 Monthly Anomalies Anomalies (cm) Monthly

-10 Anomalies (cm) Monthly -15

200204 200303 200401 200411 200508 200605 200703 200712 200809 200907 201004 201102 201201 201212 201401 201501

200508 200303 200401 200411 200605 200703 200712 200809 200907 201004 201102 201201 201212 201401 201501 200204 Figure 4.10 Illustrating average monthly groundwater anomalies in the major African basins.

North American Basins

Groundwater depletion has been a significant concern of the Colorado River basin for over a decade (Christensen et al. 2004). The Colorado River basin experienced a groundwater decrease at a rate of -0.008 cm month-1 . Figure 4.11 shows that the Colorado River basin experienced its highest amount of groundwater storage decline in the September 2003; furthermore, the groundwater level had long-term negative anomalies in the Colorado River from January 2012 through December 2015 as compared to the previous years and months, indicating a risk of groundwater depletion and aquifer drought. From 2012 to 2015, the short-term groundwater recharge did not contribute to recovering the groundwater storage.

In the Mississippi River basin, the average groundwater anomalies indicated a groundwater storage increase at a rate of 0.02 cm month-1 (Fig. 4.11); however, the Mississippi River basin experienced its highest amount of groundwater decline in December 2005 and March 2013. During the February 2008 and September 2010, the positive groundwater anomalies indicated a groundwater storage recovery.

130

Colorado River Basin Groundwater Level Mississipi River Basin Groundwater Level Anomalies 10 15 5 10 0 5 -5 0 -5 -10 Feb2008 -10 Sep2010 Mar2013 Jan 2012 Anomalies (cm) Monthly

-15 Sep2003 -15 Dec2005

Monthly Anomlaies Anomlaies (cm) Monthly

201304 200204 200302 200311 200408 200504 200512 200608 200705 200801 200809 200906 201002 201010 201108 201206 201404 201502

200303 200401 200411 200508 200605 200703 200712 200809 200907 201004 201102 201201 201212 201401 201501 200204 Figure 4.11 Illustrating average monthly groundwater anomalies in the major North American basins.

South American Basins

The Orinoco basin experienced a groundwater increase at a rate of 0.03 cm month-1 . The

Orinoco basin experienced its highest rate of groundwater decline in March 2003 and February

2010 (Fig.4.12). The Tocantins basin experienced a groundwater decrease at a rate of -0.0005 cm month-1 . The Tocantins basin similarly experienced its highest rate of average groundwater decline in December 2003. The continuous short-term recharges contributed to groundwater recovery in the Tocantins and Orinoco basins of South America (Fig. 4.12). Groundwater anomalies in the Tocantins and Orinoco basins in South America do not illustrate long-term significant positive or negative groundwater trends (Fig. 4.12).

A depletion of groundwater resources has been identified in numerous global basins, suggesting that withdrawals have exceeded the natural recharge rates in critically important global freshwater supplies. Temporal changes in groundwater storage influence the complex interactions in global aquifer systems (Thomas et al. 2017). Long-term groundwater recovery is needed to sustain groundwater discharge through the spring and summer seasons, the base flow to surface 131

water, and agriculture discharge through withdrawal wells (Mukherjee et al. 2018). Our use of

GRACE to identify variations in groundwater storage explains the influence of groundwater use, the aquifer response to withdrawal, and climatic variabilities.

Tocantins Basin Groundwater Level Orinoco Basin Groundwater Level Anomalies Anomalies 40 30 20 20 0 10 0 -20 -10 -40 -20

-60 Dec2003 -30 Mar2003 Mar2010

Monthly Anomalies Anomalies (cm) Monthly

200712 200204 200303 200401 200411 200508 200605 200703 200809 200907 201004 201102 201201 201212 201401 201501 200303 200401 200411 200508 200605 200703 200712 200809 200907 201004 201102 201201 201212 201401 201501 200204 Figure 4.12 Illustrating average monthly groundwater anomalies in the major South American basins.

Shallow Aquifers

Apart from groundwater quantity, a shallow aquifer system that extends across the entire basin will be at a high risk of groundwater contamination (Foster and Hirata 1988). Agricultural activities would result in a deterioration of the groundwater quality in shallow aquifers. The overuse of fertilizers, which are used to improve agricultural yields, threatens the groundwater quality in the shallow aquifers. In Table 4.8, a list of major basins falling under shallow aquifers is provided along with the area under irrigation from groundwater resources. In the Indus basin,

21 percent of the irrigated land uses groundwater resources for irrigation. In the Brahmaputra basin, 9 percent of the irrigated land uses groundwater resources for irrigation. In the Amur basin,

13.3 percent of the irrigated land uses groundwater resources for irrigation. Similarly, in the

132

Mississippi River basin, 25.2 percent, and in the shallow aquifer of the Loire basin in Europe 50 percent of the irrigated land uses groundwater resources for irrigation. Large portions of land are used for agriculture in both the European and Asian basins; however, in the basins of Africa, South

America, and Australia, a very small percent of the irrigated lands use groundwater resources for irrigation (see Table 4.8).

Additionally, the development of crop production in small-scale basins and shallow aquifers results in poor water quality and water shortage, particularly in southwestern Europe. The

Loire, Guadiana, Tejo, Douro, and Guadalquivir basins in Europe are extensively used for agricultural activities (Page and Kaika 2003) (see Table 4.9), and these basins experienced on average they a groundwater decline from 2002 to 2015. Deeper aquifers react with delay to terrestrial contaminants compared with the shallow aquifers (Fishman et al. 2011).

133

Table 4.8 Sample world basins covering shallow aquifers.

Basins Region Basin Area Irrigated GRACE Aquifer Type (Km2) lands using Groundwater groundwater Trend Brahmaputra South Asia 478,671 9 Negative Shallow Basin Amur Basin South East 1,836,384 13.3 Negative Shallow Asia (China) Tigris & Middle East 630,208 7.5 Negative Shallow Euphrates Basin

Indus Basin South Asia 1,035,494 21.1 Negative Shallow Loire Basin Europe 103,675 50.3 Negative South Shallow and deep Guadiana Basin Europe 58,387 34.1 Negative Shallow Tejo Basin Europe 62,158 32.9 Negative Shallow Douro Basin Europe 84,948 32.6 Negative Shallow Trent Basin Europe 8,398 18.8 Negative Shallow Nile Basin Africa 3,515,996 0.2 Positive Shallow Mississippi North 2,839,927 25.2 Positive Shallow, River Basin America Complex Major Tocantins Basin South 1,473,811 0.1 Negative Shallow & Major America Parana Basin South 7637,889 0.6 Positive Shallow America Orinoco Basin South 1,182,946 0.1 Positive Shallow America Murray Basin Australia 4,468,930 0.4 Positive Shallow aquifer in south of basin

4.6.3 Trends in Groundwater Storage Anomalies, Basin Size, and Irrigated Lands

The average groundwater storage trends and an average percentage of irrigated areas using groundwater resources were calculated for each basin from 2002 to 2015. The average

134

groundwater storage varied according to the amount of irrigated land using groundwater resources and the size of the basins (Fig. 4.13). The basins in Asia saw an increase in groundwater trend as the basin size increased; however, groundwater depletion increased as the amount of irrigated land using groundwater resources increased (Fig. 4.13). Groundwater, therefore, is a valuable resource for agriculture, especially small-scale basins where groundwater depletion has the potential to disrupt societal and ecological functions.

3000000 70

2500000 60 ) 2 50 2000000 40 1500000 30

1000000 Groundwater

20 Basins' Basins' Areas(km

500000 10 Percentage Irrigated Areas by 0 0 -0.25 -0.2 -0.15 -0.1 -0.05 0 0.05 0.1 0.15 Average Groundwater Trend for Basins

Basin areas- Groundwater Trend Percentage Irrigated Areas using Groundwater-Groundwater Trend

Figure 4.13 Illustrating groundwater trend variations by basin size and percentage irrigated lands using groundwater in Asia.

The average groundwater trends vary according to the average amount of irrigated land using groundwater resources and the size of the basins in Europe (Fig.4.14). The rate of groundwater depletion decreased as the basin size increased. Also, the rate of groundwater depletion increased with the increase in percent irrigated land in the basins (Fig.4.14). Over- withdrawal of water in small-scale basins will deplete more water when compared with large-scale

135

basins where water resources are supported by hydrologic features, such as lakes and rivers

(Mukherjee et al. 2018, Döll et al. 2012). Therefore, small-scale Asian and European basins are at greater risk of water scarcity because water is not available in other portions of the aquifer and basin (Mukherjee et al. 2018).

450000 60 400000 50

350000

) 2 300000 40 250000 30 200000 150000 20 Basins' Basins' Area (km 100000 10 50000 0 0 -0.15 -0.1 -0.05 0 0.05 0.1 0.15 0.2

Average Groundwater Trend for Basins Percentage Irrigated Area using Groundwater Basin areas- Groundwater Trend Percentage Irrigated Areas using Groundwater-Groundwater Trend

Figure 4.14 Illustrating groundwater trend variations by basin size and percentage of irrigated land by using groundwater in Europe.

The basins in Africa are commonly shared by two or more countries, and the groundwater depletion occurred in the large basins (Fig.4.15). However, increases in the basin size coincided with an increase in groundwater depletion. Additionally, groundwater depletion did not increase, due to the increased average irrigated land using groundwater resources (Fig.4.15). This may be due to the fact that in Africa, percentage of land under irrigation is negligible. The reason for water scarcity in the African basins is a lack of technological and strategic planning to use groundwater

136

resources (Nijsten et al. 2018). Overloaded agricultural activities and groundwater over- withdrawal are not reasons for water scarcity in the basins of Africa (MacDonald et al. 2012).

6000000 0.7

5000000 0.6 )

2 2 0.5 4000000 0.4 3000000 0.3

2000000 Groundwater

0.2 Basins Basins 'Areas(km

1000000 0.1 Percenathe Irrigated Area using 0 0 -0.04 -0.02 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 Average Groundwater Trend for Bains

Basin areas-Groundwater Trend Percentage Irrigated Areas using Groundwater-Groundwater Trend

Figure 4.15 Illustrating groundwater trend variations by basin size and percentage of irrigated land by using groundwater in Africa.

Groundwater storage is decreasing in various portions of the North American basins in response to groundwater withdrawal. The results have indicated that groundwater trends increased with the increase in basin size, whereas groundwater depletion increased with increase in land area under irrigation from groundwater resources (Fig. 4.16).

137

3000000 45

40 2500000

35 )

2 2 2000000 30

25 1500000 20

1000000 15 Basins'Areas(km 10 500000 5

0 0 -0.06 -0.04 -0.02 0 0.02 0.04 0.06 0.08 0.1 0.12 Average Groundwater Trend for Basins Percentage Irrigated Areas by using Groundwater Basin areas-Groundwater Trend Percentage Irrigated Areas using Groundwater-Groundwater Trend

Figure 4.16 Illustrating groundwater trend variations by basin size and percentage of irrigated land by using groundwater in North America.

Agricultural activities are a dominant land use in the eastern and southeastern portions of

South America. Overall land area under irrigation using groundwater resources is typically low in

South America. Contrary to Asia, Europe and North America, groundwater did not decline as the land under irrigation by groundwater resources increased. Average groundwater storage increased as the basin sizes increased (Fig.4.17). It would, therefore, appear that the groundwater resources supported by hydrologic features such as lakes and rivers increased as basin sizes increased.

138

12000000 8

7 10000000

6 ) )

2 2 8000000 5

6000000 4 Groundwater 3 4000000

2 Basins' Basins' Areas(km

2000000 Percentage Irrigated Lands using 1

0 0 -0.2 -0.15 -0.1 -0.05 0 0.05 0.1 0.15 Average Groundwater Trend for Basins Basin areas-Groundwater Trend Percentage Irrigated Areas using Groundwater-Groundwater Trend

Figure 4.17 Illustrating groundwater trend variations by basin size and percentage of irrigated land by using groundwater in South America.

The size of the basins and the rate of groundwater use should, therefore, be treated as important units for planning and managing water resources. Ratios of groundwater use in the irrigated land of the basins should be rearranged based upon basin size. The severity of the water crisis threatens small-sized basins due to the intensive agricultural activities in Asian, South

American, and southwestern European basins.

Agricultural activities should be planned to sustain a groundwater balance over the entirety of a basin. Because groundwater moves slowly through the small spaces within rocks, between rocks, and between loose materials, such as sand and gravel, it may take years, or even centuries, to move from one location to another location (USGS 2016). Therefore, intensive agricultural activities may lead to local groundwater depletion, and it may disrupt the natural flow of the underground water in the long-term (USGS 2016). A comprehensive model of the land use-basin

139

size- groundwater storage in the basin scale is imperative. Basins’ commonalities and differences should be surveyed across local, national, and continental borders to understand water balance at the global level.

Accuracy and Limitation

In GRACE data, the errors decrease as the radius increases, falling from 38 mm at 500 km to 15 mm at 1000 km. The user can be 68.3 percent confident the true mass value variations in

GRACE data (Wahr et al. 2006). Total error estimates obtained by summing measurement and leakage errors are provided via GRACE Tellus website for major basins of the World (see Table

4.10). Scaling factor error for GLDAS-1 Noah LSM is less than 1 (0.92) (Long et al. 2016).

In the irrigation land database used in this study, in countries with a high resolution of irrigation statistics at sub-nation levels uncertainty is less than in countries where data were available at the national scale only (Siebert et al. 2007).

140

Table 4.9 Errors in major basins of the world

Basins Combined Measurement and Combined Measurement Leakage Errors in GLDAS and Leakage Errors in GRACE

Amazon 11.5 11.2

Mississippi 10.7 8.2

Ob 9.5 7.1

Lena 9.6 9.5

Niger 11.6 12

Amur 11 12.1

Mackenzie 11 10.5

Ganges 14 17.2

Vola 10.7 9.2

Yokun 16.6 20.1

Murray 17.7 20.3

Indus 26.5 21.3

Loire 17.4 25.5

Columbia 18.2 21.9

Colorado 15.5 12.8

Danube 16.5 14.8

Orinoco 30.1 27.6

(Landerer and Swensen 2012).

4.7 Conclusion

The issues outlined in this paper highlighted global groundwater depletion. The average groundwater storage varies according to the amount of irrigated land using groundwater resources and the size of the basins. Groundwater depletion has increased in southern Asia, western North

141

America, and southwestern Europe due to groundwater withdrawal for agricultural use. However, farming practice is not the only reason for groundwater scarcity in South America, Africa, and

Australia, because other factors, including local climate conditions, access to technologies, and efficient management, may cause water scarcity.

Increasing water needs caused by rapidly growing populations and global climate change affect groundwater storage. Land use and climate change can accelerate the depletion of groundwater resources that support humans and ecosystem services on a global scale. Humans have employed large-scale changes on the terrestrial biosphere, primarily through agriculture, i.e., groundwater in the semi-arid and arid basin has been decreasing by approximately 150 km3 a -1 due to increasing withdrawal (Field and Michalak 2015). Changes in the frequency and intensity of regional precipitation patterns will also affect groundwater resources. The impacts of humans and climate change on the hydrologic cycle must be studied. Furthermore, the global water security index should be developed by using GRACE satellite technologies to monitor water resources.

GRACE satellite technologies contribute to the identification of monthly and seasonal groundwater depletion and recharge.

In this study, the depletion of groundwater resources has been identified in several global basins, suggesting that withdrawals have exceeded the natural recharge rates in critically important global freshwater supplies. However, the influences of the direct and indirect consequences of climate variability on groundwater resources should not be ignored. This study was a preliminary study to analyzing the agricultural activities on groundwater storage in small and large basins.

According to the 2030 Water Resources Group, one third of the world’s population will settle near basins whose water scarcity will be larger than 50 percent by 2030 (U.S. GWS 2012).

142

Tracking the evolution of the global groundwater variations can not only contribute to defining a global disorder impairing water quantity and quality, but it can also inform accurate management and ecosystem restoration. It can help (1) estimate the severity of losses, (2) predict the projection of groundwater quantity and quality over time, (3) identify and detect the stability, resistance, and resilience of groundwater resources over time, and (4) select appropriate management strategies and ecosystem restoration interventions over the course of time (Kaushal et al. 2017). Therefore, tracking the evolution of the groundwater variations in basins is imperative to estimate the seasonal and monthly variations of groundwater and project groundwater quantity over time in a basin, because water stress will increase in many regions throughout the world, including the western

United States, northern Africa, southern Africa, the Middle East, Australia, and parts of south Asia and China (U.S. GWS 2012).

143

4.8 References

Abu Zeid, M. 2006. The Middle East water report. The 4th World Water Forum, Mexico City, Mexico.

ABS — see Australian Bureau of Statistics. Aggarwal, S., Punhani, G., and Kher, J. 2014.Hotspots of household water insecurity in India’s current and future climates: Association with gender inequalities. The Journal of Gender Water. 3:34-41.

Asian Develop Bank (ADB) 2016. Strengthening water security in Asia and the Pacific; Asian Water Development.

Appelo, C.A.J., and Postma, D. 2005. Geochemistry, groundwater and pollution. UK, London: CRC Press.

Barnes, E. A., and Barnes, R. J. 2015. Estimating linear trends: Simple linear regression versus epoch differences. Journal of Climate 28.

Bettadpur, S. 2007. Level-2 Gravity Field Product User Handbook Center for Space Research. University of Texas at Austin.

Boisvert, E., and Brodaric, B. 2011. Groundwater Markup Language (GWML)-Enabling Groundwater Data Interoperability in Spatial Data Infrastructures. Journal of Hydroinformatics 14(1):93.

Changming, L., Jingjie, Y., and Kendy, E. 2001. Groundwater exploitation and its impact on the environment in the North China plain. Water International 26(2):265-272.

Chambers, D. P. 2006.Evaluation of new GRACE time‐variable gravity data over the ocean. Geophysical Research Letters 33.

Chen, Z., Grasby, S., and Osadetz, K. 2002. Predicting average annual groundwater levels from climate variables: an empirical model. Journal of Hydrology 260:102-117.

Christensen, N., Wood, A. W., Voisin, N., Lettenmaier, D. P., and Palmer, R. N. 2004.The effects of climate change on the hydrology and water resources of the Colorado River Basin. Climate Change 62: 337-363.

Danielopol, D.L., Griebler, C., Gunatilaka, A., and Notenboom, J. 2003. Present state and future prospects for groundwater ecosystems. Environmental Conservation 30(02):104-130.

144

Dennehy, K.F., Litke, D.W., and McMahon, P.B. 2002.The high plains aquifer, USA: groundwater development and sustainability. Geological Society, London, Special Publications 193(1):99-119.

DHI Software 2012. MIKE SHE. Accessed [01/01/2018]: http://www.dhisoftware.com/Products/WaterResources/MIKESHE.aspx.

Döll, P., Hoffmann-Dobrev, H., Portmann, F. T., Siebert, S., Eicker, A., Rodell, M., Strassbert, G., and Scanlon, B. R. 2012. Impact of water withdrawals from groundwater and surface water on continental water storage variations. Journal of Geodynamics 59-60:143-56.

Döll, P., Hoffmann-Dobrev, H., Portmann, F. T., Siebert, S., Eicker, A., Rodell, M., Strassberg, G., and Scanlon, B. R., 2012. Impact of water withdrawals from groundwater and surface water on continental water storage variations. Journal of Geodynamics 59-60: 143-156.

Döll, P., and Florke, M. 2005. Global-scale estimation of diffuse groundwater recharge: model tuning to local data for semi-arid and arid regions and assessment of climate change impact. Frankfurt Hydrology Paper.

Elnashar, W.Y. 2014. Groundwater management in Egypt. Journal of Mechanical and Civil Engineering 11(4).

Fan, Y., Li, H., and Miguez-Macho, G. 2013. Global Patterns of Groundwater Table Depth. Science 339:940.

Fienen, M.N., and Arshad, M. 2016. The International scale of the groundwater issue. In: Jakeman A.J., Barreteau O., Hunt R.J., Rinaudo JD., Ross A. (eds) Integrated Groundwater Management. Springer, Cham.

Field, C.B., and Michalak, A.M. 2015. Water, climate, energy, food: inseparable & indispensable. Daedalus 1447-17.

Fishman, R.M., Siegfried, T., Raj, P., Modi, V., and Lall, U. 2011. Over-extraction from shallow bedrock versus deep alluvial aquifers: reliability versus sustainability considerations for India’s groundwater irrigation. Water Resource Research 47(12).

Food and Agriculture Organization of the United Nations (FAO) 2017. Water for Sustainable Food and Agriculture. Accessed [12/12/2018]: http://www.fao.org/3/a-i7959e.pdf.

Food and Agriculture Organization of the United Nations (FAO) 2010. AQUASTAT –FAO’s global information system on water and agriculture. Accessed [12/01/2018]: http://www.fao.org/nr/Aquastat.

Foster, S.S.D., and Hirata, R. 1988. Groundwater pollution risk assessment. Pan American Centre for Sanitary Engineering and Environmental Sciences. Lima 73. 145

Foster, S., Hirata, R., Gomes, D., D’Elia, M., and Paris, M. 2002. Groundwater quality protection: a guide for water utilities, municipal authorities, and environment agencies. Washington, DC: World Bank.

Future Climate for Africa 2016. Africa’s Climate Helping Decision‑Makers Make Sense of Climate Information. Accessed [02/12/2019]: http://www.futureclimateafrica.org/wp- content/uploads/2016/11/africas-climate-final-report-4nov16.pdf.

Getirana, A. 2016. Extreme Water Deficit in Brazil Detected from Space. Journal of Hydrometeorology 17:591-599.

Gleeson, T., Smith, L., Moosdorf, N., Hartmann, J., Dürr, H.H., Manning, A.H., van Beek, L.P.H., and Jellinek, A.M. .2011. Mapping permeability over the surface of the earth. Geophysical Research Letters.

Gleeson, T., Wada, Y., Bierkens, M. F. P., and van Beek, L. P. H. 2012. Water balance of global aquifers revealed by groundwater footprint. Nature 488:197-200.

Global Water Partnership (GWP) 2012. Groundwater resources and irrigated agriculture: making a beneficial relation more sustainable. Perspectives Paper. Elanders: Sweden.

Gravity Recovery and Climate Experiment (GRACE) land 2018. Accessed [04/05/2019]: http://grace.jpl.nasa.gov, supported by the NASA Measures Program.

Greene, R., Timms, W., Rengasamy, P., Arshad, M., and Cresswell, R. 2016. Soil and aquifer salinization: toward an integrated approach for salinity management of groundwater. In: Jakeman A.J., Barreteau O., Hunt R.J., Rinaudo JD., Ross A. (eds) Integrated Groundwater Management. Springer, Cham.

Haacker, E. M. K., Kendall, A.D., and Hyndman, D.W. 2016. Water level declines in the high- plains aquifer: predevelopment to resource senescence. Groundwater 54:2.

Hiscock, K.M. 2005.Hydrogeology: Principles and Practice. Blackwell: Oxford.

Hoekstra, A. Y., Mekonnen, M. M., Chapagain, A. K., Mathews, R. E., and Richter, B. D. 2012. Global monthly water scarcity: blue water footprints versus blue water availability: PLoS ONE 7(2): e32688.

International Association of Hydrologist (IAH) 2015. Strategic overview series, food security & groundwater. S Foster & G Tyso (eds). International Association of Hydrogeologists. Accessed [04/05/2018]: https://iah.org/wp-content/uploads/2015/11/IAH-Food-Security- Groundwater-Nov-2015.

146

Intergovernmental Panel on Climate Change (IPCC) 2014.Asia. Accessed [04/05/2019]: https://www.ipcc.ch/site/assets/uploads/2018/02/WGIIAR5-Chap24_FINAL.pdf.

Intergovernmental Panel on Climate Change (IPCC) 2007.Climate Change - The physical science basis (vol. 1), impacts, adaptation and vulnerability (vol. 2), mitigation of climate change (vol. 3) and synthesis report, the fourth assessment report of the Cambridge: Cambridge University Press.

Kajenthira, A., Siddiqi, A., and Anadon, L.D. 2012. A new case for promoting wastewater reuse in Saudi Arabia: bringing energy into the water equation. Journal of Environmental Economics and Management 102:184-192.

Konikow, L.F., and Kendy, E. 2005. Groundwater depletion: a global problem. Hydrogeology Journal 13(1): 317-320.

Konikow, L.F. 2011. Contribution of global groundwater depletion since 1900 to sea-level rise. Geophysical Research Letters 38(17).

Kundzewicz, Z.W., Mata, L.J, Arnell, N.W., Döll, P., Kabat, P., Jimenz, B, Miller, K.A., Oki, T., Sen, Z., and Shiklomanov, I.A. 2007. Freshwater resources and their management. Climate Change 2007: Impacts, Adaptation and Vulnerability, Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, ML. Parry, O.F. Canziani, J.P. Palutikof, P.J. van der Linden and C.E. Hanson, Eds., Cambridge University Press, Cambridge, UK, 173-210.

Kaushal, S.S., Gold, A.J., and Mayer, P., M. 2017. Land use, climate, and water resources-global stages of interaction. Water 9: 815. doi:10.3390/w9100815.

Kaufmann, D., Kraay, A., and Mastruzzi, M. 2010. The worldwide governance indicators: methodology and analytical issues. World Bank Policy Research Working Paper No. 5430 World Bank.

Landerer, F., and Swenson S. 2012. Accuracy of scaled GRACE terrestrial water storage estimates. Water Resource Research 48: W04531, doi: 10.1029/2011WR011453.

Latifovic, R., and D. Pouliot D.2007.Analysis of climate change impacts on lake ice phenology in Canada using the historical satellite data record, Remote Sensing Environment 106:492– 507.

147

Leng, G., Tang, M. Q., Gao, H., and Leung, L.R. 2014. Modeling the Effects of Groundwater-Fed Irrigation on Terrestrial Hydrology over the Conterminous United States. Journal of Hydrometeorology 15:957-972.

Li, J.L., Thompson, D.W.J., and Barnes, E.A. 2017. Quantifying the Lead Time Required for a Linear Trend to Emerge from Natural Climate Variability. Journal of Climate 30.

Livingstone, D. M., Adrian, R., Blenckner, T.,George, G., and Weyhenmeyer, G.A. .2010. Lake ice phenology, in The Impact of Climate Change on European Lakes, edited by G. D. George, pp. 51–61, Springer, Dordrect, Netherlands.

Long, D., Chen, X., Scanlon, B. R., Wada, Y., Hong, Y., Singh, V. P., Chen, Y., Wang, C., Han, Z., and Yang, W. 2016. Have GRACE satellites overestimated groundwater depletion in the Northwest India Aquifer? Scientific Reports 6: 24398.

Martínez-Granados, D., and Calatrava, J. 2017. Combining economic policy instruments with desalinisation to reduce overdraft in the Spanish Alto Guadalentín aquifer. Water Policy 19(2): 341-357.

MacDonald, A.M., Bonsor, H.C., Ó Dochartaigh, B.E., and Taylor, R.G. 2012. Quantitative maps of groundwater resources in Africa. Environmental Research Letters 7 (2): 024009.

Markstrom, S.L., Niswonger, R.G., Regan, R.S., Prudic, D.E., and Barlow, P.M. 2008. GSFLOW- coupled groundwater and surface-water FLOW model based on the integration of the Precipitation-Runoff Modeling System (PRMS) and the Modular Ground-Water Flow Model (MODFLOW-2005). Techniques and Methods 6-D1. Reston.

Morris, B.L., Lawrence, A.R., Chilton, P., Adams, B., Calow, R.C., and Klinck, B.A. 2003. Groundwater and its susceptibility to degradation: a global assessment of the problem and options for management. UNEP early warning and assessment vol report series RS. Nairobi.

Mukherjee, A., Nath Bhanja, S., and Wada, Y. 2018. Groundwater depletion causing reduction of baseflow triggering Ganges river summer drying. Scientific Reports 8:12049.

Muskett R.R., and Romanovsky V. E. Groundwater storage changes in arctic permafrost watersheds from GRACE and in situ measurements. Environ Res Lett. 2009;4:045009.

Narasimhan, T. N. 2009. Groundwater: from mystery to management. Environmental Research Letters 4.

Nabavi, E. 2018. Failed policies, falling aquifers: Unpacking groundwater over abstraction in Iran. Water Alternatives 11(3): 699-724.

148

Naumburg, E., Mata-Gonzalez, R., Hunter, R., Mclendon, T., and Martin, D. 2005. Phreatophytic vegetation and groundwater fluctuations: A review of current research and application of ecosystem response modeling with an emphasis on Great Basin vegetation. Environmental Management 35:726–740.

Nijsten, G.J., Christelis, G., Villholth, K.G., Brauned, E., and Gaye, C.B. 2018.Transboundary aquifers of Africa: review of the current state of knowledge and progress towards sustainable development and management. Journal of Hydrology. DOI: 10.1016/j.ejrh.2018.03.004.

Open Geospatial Consortium Inc. 2013.Groundwater interoperability experiment 2. Accessed [12/24/2018]:http://external.opengeospatial.org/twiki_public/HydrologyDWG/Groundwa terInteroperabilityExperiment2.

Obeed Al-Azawia, A.A., and Ward, F.A. 2016. Groundwater use and policy options for sustainable management in Southern Iraq. International Journal of Water Resources Development 33(4). Page, B., and Kaika, M. 2003.The EU water framework directive: part 2 policy innovation and the shifting choreography of governance. European Environment 13:328-343.

Poff, N.L., Brinson, M.M., and Day, J.W. 2002. Aquatic Ecosystems and Global Climate Change. Potential Impacts on Inland Freshwater and Coastal Wetland Ecosystems in United States. Pew Center on Global Climate Change, Arlington.

Ramillien, G., Frappart, F., Guntner, A., Ngo-Duc, T., Cazenave, A., and Laval, K. 2006. Time variations of the regional evapotranspiration rate from Gravity Recovery and Climate Experiment (GRACE) satellite gravimetry. Water Resources Research 42: W10403.

Reba, M.L., Massey, J.H., Adviento-Borbe, M.A., Leslie, D., Yaeger, M.A.,Anders, M., and Farris, J. 2017. Aquifer depletion in the lower Mississippi River Basin: Challenges and solutions. J. Contemp. Water Res. Educ. 162(1):128–139. doi:10.1111/j.

Rodell, M.,Houser, P. R., Jambor, U., Gottschalck, J., Mitchell, K., Meng, C.-J., Arsenault, K., Cosgrove, B., Radakovich, J., Bosilovich, M., Entin, J. K., Walker,J. P., Lohmann, D., and Toll, D. 2004. The Global Land Data Assimilation System. Bulletin of the American Meteorological Society 85 (3): 381-394.

Rodell, M., Velicogna, I., and Famiglietti, J.S. 2009. Satellite-based estimates of groundwater depletion in India. Nature 460:999-1002.

Scanlon, B.R., Faunt, C.C., Longuevergne, L., Reedy, R.C., Alley, W.M., McGuire, V.L., and McMahon, P.B. 2012. Groundwater depletion and sustainability of irrigation in the US High Plains and Central Valley. Proceedings of the National Academy of Sciences of the United States of America 109(24):9320-9325.

149

Scanlon, B. R., Z. Zhang, H. Save, D. N. Wiese, F. W. Landerer, D. Long, L. Longuevergne, and J. Chen 2016. Global evaluation of new GRACE mascon products for hydrologic applications, Water Resource Research 52, 9412–9429, doi:10.1002/ 2016WR019494.

Shah, T., Molden, D., Sakthivadivel, R., and Seckler, D. 2000. The global groundwater situation: overview of opportunities and challenges. Colombo: International Water Management Institute.

Siebert, S., Döll, P., Feick, S., Hoogeveen, J., and Frenken, K. 2007. Global Map of Irrigation Areas version 4.0.1. Johann Wolfgang Goethe University, Frankfurt am Main, Germany / Food and Agriculture Organization of the United Nations, Rome, Italy.

Serrat-Capdevila, A., Valdes, J. B., Perez, J. G., Baird,K., Mata, L. J., and Maddock, J. 2007. Modeling climate change impacts and uncertainty on the hydrology of a riparian system: The San Pedro Basin (Arizona/Sonora). Journal of Hydrology 347: 48-66.

Swenson, S. C., and Wahr, J. 2006. Post‐processing removal of correlated errors in GRACE data. Geophysical Research Letters 33.

Swenson, S., and Wahr, J. 2002. Methods for inferring regional surface mass anomalies from GRACE measurements of time‐variable gravity. Journal of Geophysical Research 107(B9): 2193.

Tapley, B.D., Bettadpur, S., Watkins, M., and Reigber, C. 2004. The gravity recovery and climate experiment: mission overview and early results. Geophysical Research Letters 31(9): L09607.

Telegina, E.A. 2014. Spatial and temporal variations of winter discharge under climate change: Case study of rivers in European Russia, Remote Sensing and GIS for Hydrology and Water Resources. IAHS Publication.

Therrien, R., McLaren, R.G., Sudicky, E.A., and Park, Y-J 2012. HydroGeoSphere: a three- dimensional numerical model describing fully integrated subsurface and surface flow and solute transport. Groundwater Simulations Group. Waterloo, Ontario: Canada.

Thomas, B. F., Caineta, J., and Nanteza, J. 2017. Global assessment of groundwater sustainability based on storage anomalies. Geophysical Research Letters 44(11): 445-11.

Treidel, H., Martin-Bordes, J.J., and Gurdak, J.J. (Eds.) 2012. Climate Change Effects on Groundwater Resources: A Global Synthesis of Findings and Recommendations. International Association of Hydrogeologists (IAH)-International Contributions to Hydrogeology. Taylor & Francis Group, CRS Press: New York.

U.S. Geological Survey (USGS) 2016. Groundwater Quality Report. Accessed [11/06/2018]: https://pubs.usgs.gov/wri/wri024045/htms/report2.htm. 150

U.S. Global Water Security 2012. Accessed [05/01/2019]: https://www.dni.gov/files/documents/Newsroom/Press%20Releases/ICA_Global%20Wat er%20Security.pdf.

Van Roosmalen, L., Christensen, B.S.B. and Sonnenborg, T.O. 2007. Regional differences in climate change impacts on groundwater and stream discharge in Denmark. Vadose Zone Journal 6 (3): 554-571.

Velicogna, I., and Wahr, J. 2006. Acceleration of Greenland ice mass loss in spring 2004. Nature 443:329-331.

Voss, K.A., Famiglietti, J.S., Lo, M., Linage, C., Rodell, M., and Swenson, S.C. 2013. Groundwater depletion in the Middle East from GRACE with implications for transboundary water management in the Tigris-Euphrates-Western Iran region. Water Resources Research 49(2):904-914.

Wada, Y., van Beek, L. P. H., and Bierkens, M. F. P. 2012. Non-sustainable groundwater sustaining irrigation: a global assessment. Water Resources Research 48: W00L06.

Wada, Y., van Beek, L.P.H., van Kempen, C.M., Reckman, J.W.T.M., Vasak, S., and Bierkens, M.F.P. 2010. Global depletion of groundwater resources. Geophysical Research Letters 37: L20402.

Wada, Y., Wisser, D., and Bierkens, M. F. P. 2014.Global modeling of withdrawal, allocation and consumptive use of surface water and groundwater resources. Earth System Dynamics 5:15-40.

Wahr, J.,Swenson, S., and Velicogna, I. 2006. Accuracy of GRACE mass estimates, Geophysical Research Letters 33:L06401, doi:10.1029/2005GL025305.

Wahr, A, G., J., and Zhong, S. 2013. Computations of the viscoelastic response of a 3-D compressible Earth to surface loading: an application to Glacial Isostatic Adjustment in Antarctica and Canada. Geophysical Journal International (192):557-572.

Wang, J., Huang, J., Rozelle, S., Huang, Q., and Blanke, A. 2007. Agriculture and groundwater development in northern China: trends, institutional responses, and policy options. Water Policy 9(S1):61-74.

Werner, A.D., Zhang, Q., Xue, L., Smerdon, B.D., Li, X., Zhu, X., Yu, L., and Li, L. 2013. An initial inventory and indexation of groundwater mega-depletion cases. Water Resources Management 27(2):507-533.

Wutich, A., and Ragsdale, K. 2008. Water insecurity and emotional distress: Coping with supply, access, and seasonal variability of water in a Bolivian squatter settlement. Social Science Medicine 67:2116-2125. 151

Wösten, H.M., Clymans, E., Page, S.E., Rieley, J.O., and Limin, S.H. 2008. Peat-water interrelationships in a tropical peatland ecosystem in Southeast Asia. CATENA 73: 212- 224.

World-wide Hydrogeological Mapping and Assessment Programme (WHYMAP). 2011. River and Groundwater Basins of the World. Accessed [01/24/2019]: https://www.whymap.org/whymap/EN/Maps_Data/Rgwbw/rgwbw_node_en.html.

Yang, X. L., Chen, Y. Q., Steenhuis, T. S., Pacenka, S., Gao, W. S., Ma, L., and Sui, P. 2017. Mitigating Groundwater Depletion in North China Plain with Cropping System that Alternate Deep and Shallow Rooted Crops. Frontiers in plant science 8:980.

152

CONCLUSION

5.1 Discussions On The Overall Outcome and Scope Of Future Research

The first objective of this dissertation was to analyze the groundwater nitrate contamination in the Edwards-Trinity and the Southern High-Plains aquifers of Texas. The second was to study groundwater quality in terms of seawater intrusion in the California Coastal Basin, Upper

Floridian, and North Atlantic Coastal Plain aquifers. Furthermore, it aimed to estimate groundwater storage variations across world basins. Understanding both groundwater quality and quantity levels is critical for taking the appropriate measures to protect human and environmental health. Population growth, as well as agricultural and industrial activities all, threaten the quality and quantity of usable groundwater, which in turn threatens sustainable crop production (Cao et al. 2017).

Various aquifers are affected with elevated nitrate levels across the United States (Puckett

2016). Nitrogen based fertilizers have been used in U.S. agricultural lands since the 1850s (Cao et al. 2017). The consumption of fertilizers in the United States was 112.5 kilograms per ha in 2005 and increased to 137 kilograms per ha in 2015 according to the Food and Agriculture Organization

(Cao et al. 2017). In chapter-II, the first focus of this study was to define the relationship between increased nitrate concentration levels and land use activities. This inquiry involved a spatial analysis of groundwater contamination by nitrates since nitrogen sources vary spatially. Results indicated that nitrate contamination increases with decreasing well-depth. The accumulated

153

nitrogen fertilizers in the soil infiltrate the Southern High-Plains aquifer due to nitrogen fertilizers used for cotton production. Also, results revealed that a contamination of the groundwater beneath the grassland was due to the overapplication of nitrogen fertilizers in order to improve grassland yields.

High nitrate concentration levels are directly associated with substantial land use activities.

Nitrate concentration levels have been observed to rise in groundwater at concentrations above 2 mg/L after monitoring the wells of the Edwards-Trinity and Southern High-Plains aquifers.

Concentrations above a 2 mg/L threshold reflect anthropogenic contributions (Nolan et al.1997).

Nitrate concentration levels also collectively exceeded the maximum contaminant level in about

50 percent of the wells that were monitored within the Edwards-Trinity and Southern High-Plains aquifers.

Groundwater contamination occurs when anthropogenic chemicals infiltrate groundwater and make it unsafe for human use. Apart from anthropogenic chemical groundwater contamination, the encroachment of seawater into various aquifers causes groundwater salinization and makes it unfit for human use (Morris et al. 2003, Peters and Meybeck 2000). The extent of seawater intrusion into an aquifer depends on several factors, including: the total rate of extracted groundwater from an aquifer (Barlow and Reichard 2010, Raicy et al. 2012, Capaccioni et al.

2005, Giambastiani et al. 2007, Trabelsi et al. 2007, Somay and Gemici 2009), the distance between the locations of groundwater discharge and seawater (Taniguchi et al. 2002), the geologic structure of an aquifer (including structural features such as faults, folds, and bounding submarine canyons), the hydraulic properties of an aquifer (including the interconnectivity of coarse-grained units within multi-layered aquifer systems), and the presence of confining units that may prevent seawater from moving vertically toward or within the aquifer (Werner et al. 2013).

154

In chapter-III, this study identified dominant groundwater types in coastal aquifers using

USGS well data. Na-Cl and mixed Ca-Mg-Cl were determined to be the two most dominant groundwater types, for three major aquifers within the U.S., including: the California Coastal Basin aquifer, Upper Floridian aquifer, and North Atlantic Coastal Plain aquifer. Furthermore, this study estimated the extent of seawater intrusion in U.S. coastal aquifers using various groundwater quality indices. Results indicated that water within U.S. coastal aquifers has the largest chloride concentration levels. However, ionic strength in the water is seen to increase as a result of seawater-groundwater mixing, which cannot be quantified through observing only chloride concentration levels. The extent of seawater intrusion needs to be assessed by observing ions’ mixing processes as well. Results of the measured ionic exchange levels indicated that many wells have components of seawater found in the North Atlantic, Upper Floridian and California Coastal

Basin aquifers. GQI (swi) ranging from 53.5 to 55.6 in the water of monitoring-wells in the

California Coastal Basin, Upper Floridian and North Atlantic Coastal Plain aquifers indicated groundwater seawater contamination. Based on previous studies (Price and Swart 2006, Cardona et al. 2004), GQI (swi) ranges from 4.8 to 79.9 where aquifers were exposed to seawater intrusion.

Results further indicated that Cl- is a major pollutant of the California Coastal, Upper

Floridian, and North Atlantic coastal plain aquifers and the largest chloride contamination of groundwater was found in the Palm Beach, Miami-Dade, and Broward coasts of the Upper

Floridian aquifer, Gloucester, Virginia Beach coasts (Virginia); Monmouth (New Jersey); and

Queens (New York) in the North Atlantic Coastal plain aquifers. The seawater fraction became weaker despite the proximity of the ocean towards the northern North Atlantic coastal aquifer.

This may be a consequence of a reduced number of monitoring-wells and a decreased level of aquifer exploitation in the northern North Atlantic coastal aquifer. Additionally, chloride

155

concentrations were found to be further elevated in most groundwater resources in the San Diego and Santa Barbara coastal regions.

Groundwater quantity problems create major challenges facing the world today, which result in groundwater drought and water scarcity. Groundwater plays an important role in providing water for use in agriculture, drinking water, and industrial purposes (Fienen and Arshad

2016). Global groundwater resources have been exploited excessively and for a long-time; as such, groundwater withdrawal has resulted in significant groundwater depletion levels and drought of the world basins (Wada et al. 2012, Gleeson et al. 2012). In chapter-IV, the third focus of this dissertation was to provide a global overview of groundwater depletion using GRACE remote sensing technology. The study suggested that to make groundwater use more efficient, there is a need to understand the amount of groundwater storage that is available for irrigated lands, other resources and basin sizes. The study underlined that small-size basins are more vulnerable to groundwater depletion compared to large-scale ones. Given that large-scale basins include several hydrologic features, such as lake and rivers, they can support the water needs of human communities during times of groundwater shortage and scarcity. Groundwater storage trends and the percentage of irrigated areas using groundwater were calculated for each basin from 2002 to

2015. Groundwater balances varied according to the amount of irrigated lands that utilized groundwater resources in basins and also their size was also seen as a factor. Basins in Asia, Europe and South America saw a decrease in groundwater depletion levels as basin size increased.

However, it was found that the amount of irrigated land using groundwater resources increased in

Asian basins considerably as groundwater depletion levels increased. It was also discovered that groundwater depletion occurs everywhere independent of basin size. Lastly, groundwater

156

depletion was observed to more harshly threaten biological and human communities in the smaller- size basins of Asia, Europe and South America.

Moreover, the study highlighted sustainable groundwater use in an entire basin. Excessive groundwater use in a portion of a basin might cause disruption to groundwater balance within an aquifer. A decrease of groundwater levels in shallow aquifers has also had a severe impact on the health of nearby rivers. Ecosystems including, terrestrial vegetation, river base flow systems, and wetlands depend on groundwater resources. A series of long-term groundwater depletion levels in the Brahmaputra, Amur Basin, Tigris and Euphrates, and Indus basins threaten terrestrial ecosystems.

5.2 Significance of This Study

The focus of this dissertation was to investigate the sources and spatial variations of groundwater contaminants in multiple inland and coastal aquifers primarily using groundwater quality indices and multivariate statistical techniques. An understanding of the dominant factors responsible for groundwater contamination helped advance our understanding about the influence of contaminants on the overall water quality of coastal and inland groundwater dependent ecosystems. The extent of seawater intrusion into aquifers will change as future global warming predictions forecast a greater variation in sea level (Rice et al. 2012, Hussain and Javadi 2016).

Groundwater nitrate concentrations will similarly vary through surface runoff and rainfall as climate change scenarios predict a variation in precipitation patterns (Miller et al. 2005). The increasing nitrate concentration levels have adverse effects on various ecosystems and may cause algal bloom, subsequent eutrophication, and also increases in NOx emissions (Galloway et al.

2003, Zhou et al. 2011). A variation in temperature and precipitation patterns globally will also likely affect groundwater recharge. An understanding of groundwater balance helped to further 157

our understanding of groundwater variations within small and large basins associated with agricultural activities. The influence of population growth and variations in temperature and precipitation patterns on groundwater storage and land use change will have a large impact on a global scale (Pulido-Velazquez et al. 2015).

5.3 Future Research

More research is required to understand the sources of nitrate contamination levels of systems. Sources of nitrate pollution can be divided into two main groups; nonpoint (diffuse), and point-source pollution. The application of large-scale agricultural fertilizer is the largest nonpoint source of pollution affecting groundwater quality. On the contrary, point sources are single and identifiable sources of contamination, mainly affecting localized areas. It is also important to include precipitation level as a factor within these systems and the ability of soils to transmit contaminants from the surface of the land as a key parameter for understanding overall groundwater nitrate contamination levels in any given area (Scanlon et al. 2004). Furthermore, shallow wells can be susceptible to nitrate contamination after flooding.

Further research is additionally required to understand the sources of Na-Cl contamination of coastal aquifers. Anthropogenic factors and climatic conditions can likewise have impacts on elevated Na-Cl contaminant levels. The input of Na+ and Cl- from road salt, seawater intrusion, animal, and human waste in rural areas and leaking landfills has degraded groundwater quality in some municipal and private wells in the United States (Panno et al. 2005). Therefore, more research is required to correlate groundwater quality to Land Use/Land Cover (LU/LC) in order to better understand the response of aquifers to human stress levels. For instance, green spaces can prevent the flooding and pollution of groundwater. These infrastructures are important components that can contribute to the success of green cities. 158

Groundwater over-consumption is a major challenge faced by the world’s population.

Impacts of groundwater depletion are exacerbated by local climate variabilities. It is thus important to study the combined effects of climate change patterns and measured groundwater withdrawal levels on groundwater resources. However, it is important to note that groundwater consumption is not a major reason for water scarcity in some regions of the world if the right technology is utilized to use groundwater more efficiently and equitably. Because groundwater scarcity occurs in transboundary basins and aquifers, the effective management of transboundary waters can provide equal access to groundwater resources.

The issue outlined in this dissertation assesses groundwater studies associated with groundwater quality and quantity levels on local, regional, and global scales. Use of groundwater as a resource has increased dramatically over the last century, making the balance between human needs and the ecosystem difficult to maintain, especially in arid and semi-arid areas wherein people are highly dependent on groundwater resources (Fienen and Arshad 2016). Three key international and local groundwater issues are: groundwater depletion, degradation of groundwater quality, and transboundary groundwater conflicts. Groundwater salinization and groundwater nitrate contamination levels due to seawater intrusion and the overapplication of nitrogen fertilizers render reductions in the usable volume of groundwater resources, as evidenced by groundwater quality. Furthermore, advances in remote sensing have helped advance our knowledge of groundwater storage variations across the world. Local direct measurements of groundwater quality, and estimating groundwater storage using remote sensing technologies thereby provide an opportunity to inter-scale and integrate the management of groundwater resources where groundwater, human, and ecosystems are all interconnected.

159

5.4 References

Barlow, P.M., and Reichard, E.G. 2010. Saltwater intrusion in coastal regions of North America. Hydrogeological Journal 18(1):247-260.

Capaccioni, B., Didero, M., Paletta, C., and Didero, L. 2005. Saline intrusion and refreshing in a multilayer coastal aquifer in the Catania Plain (Sicily, Southern Italy): dynamics of degradation processes according to the hydrochemical characteristics of groundwaters. Journal of Hydrology 307:1-16.

Fienen, M.N. and Arshad. M. 2016. The International scale of the groundwater issue. In: Jakeman A.J., Barreteau O., Hunt R.J., Rinaudo JD., Ross A. (eds). Integrated Groundwater Management. Springer, Cham.

Galloway, J.N., Aber, J.D., Erisman, J.W., Seitzinger, S.P., Howarth, R.W., Cowling, E.B., and Cosby, B.J. 2003. The nitrogen cascades. Bioscience 53(4):341-356.

Gleeson, T., Wada, Y., Bierkens, M. F. P., and van Beek, L. P. H. 2012. Water balance of global aquifers revealed by groundwater footprint. Nature 488:197-200.

Hussain, M. S., and Javadi, A. A. 2016. Assessing impacts of sea level rise on seawater intrusion in a coastal aquifer with sloped shoreline boundary. Journal of Hydro-Environment Research 11: 29-41.

Morris, B.L., Lawrence, A.R.L., Chilton, P.J.C., Adams, B., Calow, R.C., and Klinck, B.A. 2003. Groundwater and its susceptibility to degradation: A global assessment of the problem of options for management. Early Warning and Assessment Report series, RS. 03-3. United Nations Environment Programme, Nairobi, Kenya.

160

Nolan, B.T., Ruddy, B.C., Hitt, K.J., and Helsel, D.R. 1997. Risk of Nitrate in Groundwaters of the United States A National Perspective. Environmental Science Technology 31:2229- 2236.

Panno, S.V., Hackley,K.C., Hwang, H.H., Greenberg, S., Krapac, I.G., Landsberger, S., and O’Kelly, D.J. 2005. Data table for characterization and identification of the sources of NaCl in natural waters. Illinois State Geological Survey Open File Series 2005-1. Champaign, Illinois.

Peters, N.E., and Meybeck, M. 2000. Water quality degradation effects on freshwater availability: impacts of human activities. Water International 25(2):185-193.

Pulido-Velazquez, M., Peña-Haro, S., García-Prats, A., Mocholi-Almudever, A. F., Henriquez- Dole, L., Macian-Sorribes, H., and LopezNicolas, A. 2015. Integrated assessment of the impact of climate and land use changes on groundwater quantity and quality in the Mancha 5 Oriental system (Spain). Hydrology and Earth System Sciences 19:1677-1693.

Puckett, L.J. 2016. Nonpoint and Point Sources of Nitrogen in Major Watersheds of the United States, U.S. Geological Survey Water-Resources Investigations Report 94-4001.

Raicy, M. C., Parimala Renganayaki, S., Brindha, K., and Elango, L. 2012. Mitigation of seawater intrusion by managed aquifer recharge, managed aquifer recharge: methods, hydrogeological requirements. Pre and Post treatment Systems 83-99.

Rice, K. C., Hong, B., and Shen, J. 2012. Assessment of salinity intrusion in the James and Chickahominy Rivers as a result of simulated sea-level rise in Chesapeake Bay, East Coast, USA. Journal of Environmental Management 111 (0): 61-69.

Taniguchi, M., Burnett, W. C., Cable, J. E., and Turner, J. V. 2002. Investigation of submarine groundwater discharge. Hydrological Processes 16: 2115-2129.

Trabelsi, R., Zairi, M., and Dhia, H.B. 2007. Groundwater salinization of the Sfax superficial aquifer, Tunisia. Hydrogeological Journal 15:1341-1355.

161

Wada, Y., van Beek, L. P. H., and Bierkens, M. F. P. 2012. Non-sustainable groundwater sustaining irrigation: a global assessment. Water Resource Research 48: W00L06.

Werner, A.D., Bakker, M., Post, V.E., Vandenbohede, A., Lu, C., Ataie-Ashtiani, B., and et al. 2013. Seawater intrusion processes, investigation and management: recent advances and future challenges. Advance Water Resource 51:3-26.

162

TABLE

163

Table A.1 Illustrating ion concentrations of groundwater samples

- 2+ 2+ + + 2- - Date HCO3 Ca (mg/L) Mg (mg/L) Na K (mg/L) SO4 Cl (mg/L) (mg/L) (mg/L) (mg/L) 6/25/2011 136 435 285 2000 35 1260 2160 6/12/2005 100 158 193 1890 86.3 1210 2990

8/9/2010 462 19.9 10.2 10.4 12.9 940 8/12/2010 10 5.3 2.81 135 1.41 18.9 214 6/16/2011 20 35.1 7.21 95.9 2.81 21.3 217

12/16/2012 4 1000 230 23 730 4300 6/27/2017 4 43.3 8.01 332 2.88 34.3 616 6/27/2017 5 16.4 5.84 390 3.89 18.4 647

2/6/2017 8 24.3 5.98 0.83 5.58 218 10/9/2003 237 34.8 33.2 58.1 30.8 0.1 144 8/21/2018 660 424 228 349 9.21 1290 663

7/29/2009 142 50 32 44 2.3 78 11/21/2006 140 287 200 1250 24 1020 2250

4/22/2014 534 878 8120 4940 1770 14800

10/21/2003 448 36.5 54.2 38.3 0.5 1550

1/26/2018 24 11.5 363 15.3 19.8 464

5/16/2012 534 16.2 10.7 1110 26.9 1580 5/15/2012 852 24.8 31 1830 52.1 163 2670 5/10/2012 121 1040 458 9370 106 578 18500 5/17/2012 735 19.3 21.2 1540 45.4 164 2200 5/16/2012 851 24.6 31.4 1830 50.5 159 2660 5/30/2002 13 906 243 6860 77.5 465 12800

10/20/2005 26.5 483 4.86 50.3 336 10/19/2005 17 6.39 1.21 474 2.56 206 516

10/17/2005 10.8 5.86 779 4.1 396 790 1/9/2005 400 438 515 1400 23.3 855 3570 8/9/2016 491 394 298 376 4.58 1460 548 8/20/2018 458 205 105 242 9.39 447 478

8/21/2009 70 59.7 1400 41.1 447 1910 8/22/2018 467 180 160 667 28.9 384 1370

11/4/2007 66 54 550 26 5.1 950 7/27/2009 17 241 985 6150 190 1520 11800

164

Table A.1 (continued)

12/12/2002 1810 19400

1/6/2006 210 520 4500 150 660 8100

6/24/2002 75 5780

8/15/2008 1300 160 1900 5700

7/16/2009 45 1100 55 790 9.5 3200

9/8/2007 54.5 31.4 9.68 0.65 559 5/4/2006 78 32.5 6.37 508 8.58 3.29 815

3/2/2017 118 77.6 440 6.92 121 1310

10/7/2010 5.89 6.14 958 24.4 121 1210 3/17/2005 518 3.53 1.72 409 11.6 36.4 352

8/8/2002 1040 745 11700 263 11 22500 5/25/2011 490 2.57 2.42 608 18.8 65.2 729 7/21/2010 758 9.18 11.8 908 30.1 135 1060 7/10/2009 553 4.01 2 635 13.2 87.6 736 9/26/2001 1260 26.8 11.5 1250 29.8 175 1230 9/11/2003 484 36.1 65.1 1390 53.2 2.3 2100 5/12/2004 1040 8.89 8.76 1310 29.5 1.8 1540 5/13/2004 1020 49.5 67.2 2590 65.5 22.3 3770 9/8/2003 470 88 81.9 1400 52.1 134 2370

7/9/2010 100 3.79 737 4.31 170 1180 7/12/2010 100 94.5 2.52 624 4.11 149 1090

2/28/2002 156 86.9 494 6.95 120 1130 8/18/2006 390 340 110 1200 18 210 2400 8/18/2006 542 280 98 480 6.3 750 680 6/12/2018 160 1390 658 4010 17.9 1370 9470

6/14/2018 187 467 147 161 4.07 1220 6/17/2016 400 253 61.5 456 15.5 9.21 1050 6/11/2018 318 119 33.3 951 8.5 63.7 1550 6/14/2011 31 693 101 3970 31.6 9 8190 6/20/2012 28.9 423 144 3530 51.1 2.25 7250 5/17/2017 273 532 477 4210 82.5 811 8490 5/18/2017 275 530 668 5690 141 1170 11000 5/17/2017 206 452 134 1680 19.9 175 3330 3/29/2010 204 95 88 600 40 320 910 7/11/2018 225 527 210 2270 32.5 335 4580 9/21/2010 254 192 62.6 636 8.32 83.1 1380

12/19/2006 170 130 190 1300 400 2300

6/30/2000 100 79 360 22 58 860

165

Table A.1 (continued)

9/21/2010 236 160 34.8 485 10.3 99 1040 10/12/2015 246 284 107 1090 7.81 86.6 2070

9/12/2006 220 75 730 13 120 1500 10/26/2010 203 273 121 1200 19.2 206 2590 12/7/2015 259 315 282 2710 70.3 599 5070 7/21/2010 254 441 273 2850 52.9 510 6190

9/7/2004 29.8 47.9 286 17.6 1.74 501

5/17/2012 294 128 62.4 51.2 270 5090 5/11/2012 516 18.3 11 1160 26.8 120 1640 4/10/2013 370 13.1 5.84 869 15.1 72.8 1200 4/12/2013 282 43.9 14.9 1730 24 147 2680

5/7/2011 64.6 1.41 540 2.88 201 867

31/08/2010 147 29.6 628 9.82 237 1080 29/05/2009 30 7.83 4.92 732 6.73 135 919 12/15/2011 120 17.3 6.24 712 12.8 181 977

1/9/2015 16.1 15.8 954 19.5 1110 14/2/2017 45 876 2480 20900 691 4800 38200 1/31/2017 20 2050 2140 11600 122 3000 25200

4/12/2010 72 74 290 20 190 550 4/29/1971 140 154 97 600 14 424 1100

24/03/2000 56 58 140 13 57 350 30/08/2006 150 220 100 96 5.6 510 230 9/22/2010 411 214 228 1990 57.4 411 3990

25/01/2017 109 13.4 315 1180

26/03/2003 210 95.3 622 7.05 152 1440 20/06/2013 296 10.2 2.46 417 4.78 1.35 580 4/6/2009 366 260 97 240 5 190 740 6/12/2018 235 127 53.7 214 2.72 168 452 9/12/2003 266 36.7 38.9 381 33.9 16.8 625 9/8/2003 508 21.7 26.9 609 21.7 75.5 741 4/9/2013 465 4.46 2.45 516 13.5 58.9 494 7/26/2006 238 20 18 240 11 9.4 350

7/12/2011 67.7 3.06 508 4.54 193 816 9/28/2010 312 165 61.4 676 16.3 121 1350 5/10/2006 240 79 100 520 26 210 980 6/9/2006 240 42 64 360 22 300 420 5/26/2000 200 66 61 230 19 180 450

166

FIGURE

167

❖ North America

❖ South America

Figure B.1 (continued)

169

Figure B.1 (continued)

170

❖ Africa

Zambezi Basin Groundwater Level Anomalies 30

200602 200204 200212 200307 200401 200408 200502 200508 200608 200703 200709 200803 200809 200904 200910 201004 201010 201106 201201 201208 201304 201401 201409 201504 Monthly Anomalies (cm) Anomalies Monthly -10

-20

-30 c

Niger Basin Groundwater Level Anomalies

0 Monthly Anomalies MonthlyAnomalies (cm)

-5

200803 201201 200204 200212 200307 200401 200408 200502 200508 200602 200608 200703 200709 200809 200904 200910 201004 201010 201106 201208 201304 201401 201409 201504

-10 c

Figure B.1 (continued)

171

Orange Basin Groundwater Level Anomalies 10

-2

-6

-8 c

❖ Europe Tejo Basin Groundwater Level Anomalies

20 15 10 5 0

-5

200901 200911 200211 200304 200310 200403 200409 200502 200507 200512 200605 200610 200704 200709 200802 200807 200906 201004 201009 201103 201109 201203 201209 201304 201312 201406 201501 201508 -10 200204 -15

Monthly Anomalies (cm) Anomalies Monthly -20 -25 -30 d

Figure B.1 (continued)

172

Danube Basin Groundwater Level Anomalies 15

-5

200211 200709 201209 200204 200304 200310 200403 200409 200502 200507 200512 200605 200610 200704 200802 200807 200901 200906 200911 201004 201009 201103 201109 201203 201304 201312 201406 201501 201508

-10 Monthly Anomalies (cm) Anomalies Monthly -15

-20 d

Loire Basin Groundwater Level Anomalies 20 15 10 5 0

-5

200212 200307 200401 200408 200502 200508 200602 200608 200703 200709 200803 200809 200904 200910 201004 201010 201106 201201 201208 201304 201401 201409 201504 -10 200204 -15

-20 Monthly Anomalies (cm) Anomalies Monthly -25 -30 -35 d Figure B.1 (continued)

173

❖ Asia

Huang He (Yellow River) Basin Groundwater Level Anomalies

8 6 4 2 0

-2

200709 200212 200307 200401 200408 200502 200508 200602 200608 200703 200803 200809 200904 200910 201004 201010 201106 201201 201208 201304 201401 201409 201504

-4 200204 Monthly Anomalies (cm) Anomalies Monthly -6 -8 -10 e

Yangtze River (Chang Jiang) Basin Groundwater Level Anomalies 15

Monthly Anomalies (cm) Anomalies Monthly -5

200408 200502 200204 200212 200307 200401 200508 200602 200608 200703 200709 200803 200809 200904 200910 201004 201010 201106 201201 201208 201304 201401 201409 201504

-10

-15 e

Figure B.1 (continued)

174