An-Najah National University Faculty of Graduate Studies MANAGEMENT OF NITRATE CONTAMINATION OF GROUNDWATER USING LUMPED PARAMETER MODELS Prepared by Lubna "Mohammad Nayef" Abdullah Hajhamad Supervisor Dr. Mohammad N. Almasri Submitted in Partial Fulfillment of the Requirements for the Degree of Master in Water and Environmental Engineering, Faculty of Graduate studies, at An-Najah National University, Nablus, Palestine 2007 III Dedicated to My Parents My husband Sameer Al-Sheeb My Kids: Yazan, Zeina, and Rama My Brothers and Sisters. IV Acknowledgments First of all, praise be to Allah for helping me in making this thesis possible. I would like to express my sincere gratitude to Dr. Mohammad N. Almasri for his supervision, guidance and constructive advice. Special thanks also go to my defense committee. Thanks go also to those who helped in providing the data used in this research, mainly Water and Environmental Studies Institute (WESI). Special thanks go to Dr. Said Ghabayen from Gaza municipality for producing of many data used in this study. Thanks to my dear husband Sameer and my children for their love, patience and support. My parents, brothers and sisters, thank you for being a great source of support and encouragement. All my fellow graduate students, thank you. V Table of contents 1. INTRODUCTION 1 1.1 General ............................................................................................... 2 1.2 Justifications for the Selection of the Study Area .............................. 4 1.3 Research Question .............................................................................. 5 1.4 Research Objectives ........................................................................... 5 1.5 Research Outcome .............................................................................. 5 1.6 Research Output ................................................................................. 6 1.7 Thesis Organization ............................................................................ 6 2. DESCRIPTION OF THE STUDY AREA 8 2.1 Introduction ........................................................................................ 9 2.2 Topography ...................................................................................... 10 2.3 Climate ............................................................................................. 12 2.4 Rainfall Distribution ......................................................................... 13 2.5 Land Use ........................................................................................... 13 2.6 Soil Types ......................................................................................... 14 2.7 Water Resources ............................................................................... 15 2.8 Nitrate Pollution in the Groundwater of GCJC ................................ 15 2.9 Geology ............................................................................................ 18 2.10 Water Table Elevation ...................................................................... 20 3. LITERATURE REVIEW AND GENERAL BACKGROUND 22 3.1 Nitrate Contamination of Groundwater ........................................... 23 3.2 Health Problems Associated with Nitrate Contamination ............... 27 3.3 General Sources of Nitrate Contamination in Groundwater ............ 28 3.4 Management of Nitrate Contamination of Groundwater ................. 29 3.5 Groundwater Modeling .................................................................... 30 3.5.1 Introduction ................................................................................... 30 3.5.2 Model Definition ........................................................................... 31 3.5.3 Why Do We Need a Model in this study? .................................... 32 3.5.4 Main Output of Groundwater Modeling ....................................... 32 3.5.5 Mathematical Models .................................................................... 33 3.5.6 Initial and boundary conditions .................................................... 35 3.5.7 Model Calibration ......................................................................... 37 3.5.8 Sensitivity Analysis ....................................................................... 37 3.5.9 Lumped Parameter versus Distributed Groundwater Models ...... 39 4. METHODOLOGY 41 4.1 Introduction ...................................................................................... 42 4.2 Methodology description.................................................................. 42 5. MODEL DEVELOPMENT 45 VI 5.1 Introduction.......................46 5.2Developmentoftheconceptualquantity model.....................................47 5.2.1 Lateral Inflow (Gin).....................................................................48 5.2.2 Artificial recharge (QAr) ................................................................ 51 5.2.3 Recharge (R) ................................................................................. 52 5.2.4 Lateral outflow (Go) ...................................................................... 59 5.2.5 Water pumped for irrigation (QIrr) ................................................ 59 5.2.6 Water pumped for domestic purposes (QDO)…………...………59 5.3 Development of the conceptual quality model ................................ 59 5.3.1 Nitrate from lateral inflow (NO3Gin) ............................................. 60 5.3.2 Nitrate from artificial recharge (NO3QA) ..................................... 61 5.3.3 Nitrate from fertilizer surplus (NO3SURP) ................................... 62 5.3.4 Nitrate from recharge (NO3R) ....................................................... 63 5.3.5 Nitrate lost through lateral outflow (NO3Go) ............................... 67 5.3.6 Nitrate lost through irrigation (NO3Irr) ........................................ 67 5.3.7 Nitrate lost through domestic use of groundwater (NO3DO) ...... 67 5.3.8 Nitrate lost through denitrification (NO3DEN)............................ 68 5.4 Development of the mathematical models ....................................... 68 5.5 The numerical solution of the mathematical models ....................... 70 5.6 Model Calibration ............................................................................ 70 5.7 Sensitivity Analysis .......................................................................... 72 6. ANALYSIS AND DISCUSSION OF MODEL OUTPUT 77 6.1 Introduction ...................................................................................... 78 6.2 Quantity Model Output .................................................................... 78 6.3 Quality Model Output ...................................................................... 82 7. MANAGEMENT OF NITRATE CONTAMINATION OF THE GROUNDWATER OF THE STUDY AREA 87 7.1 Introduction ...................................................................................... 88 7.2 Proposed Management Options ....................................................... 88 7.2.1 Reduction of nitrate concentration in lateral inflow ..................... 90 7.2.2 Rehabilitation of the wastewater network..................................... 90 7.2.3 Full coverage of sewerage system ................................................ 90 7.2.4 Restriction on the use of fertilizers ............................................... 91 7.2.5 Combination of management options ........................................... 93 7.3 Results and discussion ...................................................................... 94 8. CONCLUSIONS AND RECOMMENDATIONS 97 8.1 Conclusions ...................................................................................... 98 8.2 Recommendations ............................................................................ 99 9. REFERENCES 101 VII List of figures Figure (1): Regional setting of Gaza Strip and the neighboring countries. .............. 9 Figure (2): The location of GCJC area within Gaza Strip. ...................................... 10 Figure (3): Different information categories for the GCJC area. ............................ 11 Figure (4): Average nitrate concentration in the study area for the years from 2000 to 2004. .................................................................................................. 17 Figure (5): Nitrate concentrations for different wells in the study area for the years from 2000 to 2004. ............................................................................... 18 Figure (6): A general cross section of GCA. ........................................................... 19 Figure (7): Time series of depth to water table for selected wells in the GCJC area. ................................................................................................................ 21 Figure (8): A schematic describing the proposed conceptual model of nitrogen loading and transformations. .......................................................................... 25 Figure (9): A Logic diagram for developing a mathematical model ....................... 35 Figure (10): Simulated change in hydraulic head resulting from change in parameter value. ............................................................................................. 38 Figure (11): A flowchart of the methodology. ........................................................ 43 Figure (12): Conceptual representation of the single-cell model. ........................... 46 Figure (13): Schematic of the overall model development. .................................... 47 Figure (14): Schematic of the overall flowchart of the development of the conceptual quantity model. ............................................................................ 48 Figure (15): Segmentation of model boundaries for the computation of lateral inflow and outflow. ........................................................................................ 50 Figure (16): Thiessen polygons of the rainfall stations for the GCJC area. ........... 53 Figure (17): Schematic of the overall flowchart of the conceptual quality model development. .................................................................................................. 60 Figure (18): The average nitrate concentration for observed and simulated values for years 2000 to 2003. ................................................................................... 72 Figure (19): Relative sensitivity coefficients of water table elevation for selected model parameters. Parameter IDs are as summarized in Table 5. ................. 75 Figure (20): Relative sensitivity coefficients of nitrate concentration for selected model parameters. Parameter IDs are as summarized in Table 5. ................. 76 Figure (21): The variability of the average water table elevation with time. ......... 79 Figure (22): The total input and output of water volume for the years from 2000 to 2003. ........................................................................................................... 79 Figure (23): Time series of total input and output of water volume for the study area. ................................................................................................................ 80 Figure (24): Pie chart of the components of groundwater inflow to the study area for the year 2003. ........................................................................................... 81 VIII Figure (25): Pie chart of the components of groundwater outflow from the study area for the year 2003..................................................................................82 Figure (26): The variability of the average nitrate concentration with time........83 Figure (27): The variability of the monthly variation in the change in nitrate mass in the groundwater of the study area. ............................................................. 84 Figure (28): Pie chart of the components of nitrate input to the groundwater of the study area for the year 2004. .................................................................... 86 Figure (29): Pie chart of the components of nitrate outflow from the groundwater of the study area for the year 2004. ................................................................ 86 Figure (30): The maximum nitrate concentrations in each year with different reduction percentages corresponding to (i) lateral inflow; (ii) percentages of leakage from wastewater network; (iii) percentages of cesspits; and (iv) percentages of fertilizer reduction. ................................................................. 92 IX List of tables Table (1): Annual rainfall data (in mm) for the relevant rainfall stations for the years from 2000 to 2004. ............................................................... 13 Table (2): Detailed categories of land use for the study area. This table is based on the land use map of the entire Gaza Strip as obtained from the PWA and later processed using GIS capabilities for the GCJC area. ...................................................................................................... 14 Table (3): Lumped categories of land use practices for the study area. ....... 14 Table (4): Classification of the soil types for the study area and the corresponding area. This table is based on the soil map of the entire Gaza Strip as obtained from the PWA and later processed using GIS capabilities for the GCJC area. ............................................................ 15 Table (5): Selected parameters for model sensitivity analysis. ..................... 74 Table (6): The water budget for the groundwater of the GCJC area in 2004……………………………………………………………...…80 Table (7): The nitrate budget for the groundwater of the GCJC area in 2003………………………………………………………....….......85 Table (8): The individual management options and their corresponding IDs93 Table (9): The different combinations between the individual management options with the corresponding IDs. .................................................... 94 Table (10): Summary of the results of the combined management options summarized in Table 7. ........................................................................ 95 X MANAGEMENT OF NITRATE CONTAMINATION OF GROUNDWATER USING LUMPED PARAMETER MODELS Prepared by Lubna "Mohammad Nayef" Abdullah Hajhamad Supervisor Dr. Mohammad N. Almasri Abstract Many regions all over the world depend entirely on groundwater resources for various uses. Nitrate contamination of ground water can cause methemoglobinemia. Evidence indicates that nitrate levels routinely exceeded the maximum contamination level (MCL) of 10 mg/L NO3-N in 90 percent of the water supply wells in the Gaza costal aquifer (GCA). In addition, elevated nitrate concentrations are encountered in Gaza city and Jabalia camp (GCJC). In order to simulate the occurrences of nitrate contamination in GCJC area, a single-cell model was developed. This model was employed to study different management options and to determine their efficiency in decreasing the nitrate contamination in the study area for a specified time horizon. Main findings of the research showed that there is an emerging need to manage the nitrate contamination problem in the groundwater of the study area and single management options are not effective when considered individually. As such, the combination of management options ought to be considered if nitrate concentration to drop below the MCL. 1 CHAPTER 1 INTRODUCTION 2 1.1 General Many regions all over the world depend entirely on groundwater resources for various uses (Babiker et al., 2003). Population growth and the increase in demand for water and food supplies place an increasing stress on the groundwater quality and quantity (Joosten et al., 1998). Over-abstraction of freshwater depletes the available quantity of groundwater. In addition, the increase in demand for food supplies may lead to groundwater contamination by nitrate since the major contributor to nitrate contamination in groundwater is the use of fertilizers associated with cropping activities (Konkow and Person 1985; Shamrukh et al., 2001). Gaza Coastal Aquifer (GCA) witnesses both quantity and quality problems due respectively to the overexploitation, excessive fertilization and raw wastewater leaching (Rosen et al., 1998; Refsgaard et al., 1999). GCA is an important source of water to over 1.4 million residents in Gaza Strip and is utilized extensively to satisfy agricultural, domestic, and industrial water demands (UNEP, 2003). Pollution of the groundwater in GCA is a major problem. Evidence indicates that nitrate levels routinely exceeded the MCL of 10 mg/L NO3-N in 90 percent of the water supply wells in the GCA (Almasri et al., 2005). Of the sources responsible for the elevated nitrate concentrations in GCA are the agricultural activities including the use of fertilizers, waste dumping, discharge of raw sewage, and irrigation with water contaminated by nitrate. GCA and the overlying soil are composed mainly of sands which indeed promote the vulnerability of GCA to contamination through the high 3 potential of nitrate leaching to groundwater. Since GCA is the main source of water for the residents of Gaza Strip, nitrate contamination of the aquifer is a public-health concern. A recent survey curried by the Ministry of Health shows that 124 of 640 infants (children under the age of 6 months) have methemglobin levels above 20 percent. The average concentration of nitrate in GCA is three times higher than the MCL. The degradation of groundwater quality in the GCA has stepped up public concern in recent years and has motivated the restoration and preservation of the aquifer especially when considering that most municipalities in Gaza Strip use groundwater without any treatment except for disinfection. To address the water-quality related issues and problems, the Palestinian Water Authority (PWA) in collaboration with the Environmental Quality Authority (EQA) has developed the first National Water Plan and the National Environmental Action Plan in part to better manage and preserve the water resources including groundwater and to set up policies and strategies that aim at protecting the Palestinian water resources. Such policies demand that the agricultural and industrial development to be in full compliance with the available water resources based on sustainable development and that pollution control measures should be introduced and ensured through enforcement if needed. As such, restoration efforts have intensified the need for developing protection alternative measures and management options such that the high contamination occurrences in the aquifer are reduced. That is, nitrate concentrations at the critical receptors are below the MCL. Such measures include the restriction on the use of fertilizers and the proper treatment and 4 disposal of wastewater. A major step in proposing and developing efficient protection alternatives is through the development of a mathematical model for the simulation of nitrate occurrences in the aquifer. In other words, we need to adopt a nitrate contamination management scheme that aims at minimizing nitrate concentration in groundwater such that the outcome of these management options is quantified using mathematical models. This research focuses on the analysis and modeling of nitrate contamination in the groundwater of Gaza City and Jabalia Camp (GCJC). The GCJC area is part of the Gaza Coastal Aquifer. The different components of the model will be elucidated and discussed. Thereafter, the model output for different proposed management options will be analyzed and presented. As part of the research work, the extent of nitrate contamination in the GCJC area will be analyzed spatially using ArcView geographic information systems (GIS) (Lasserce et al., 1999). 1.2 Justifications for the Selection of the Study Area Many reasons compelled the motivation to select GCJC as a study area. Among these reasons are the following: 1. GCJC has a large intensity of population where over half a million people live in it; 2. Groundwater contamination by nitrate is an on-going problem in GCJC; 3. A total of 39 municipal wells operate in GCJC for water supply; 5 4. The problem of nitrate pollution is attributed to internal and external sources. This indeed offers a good and realistic case for the management of groundwater contamination from nitrate; 5. GCJC suites the development of a lumped parameter model which was developed in this research to study the overall nitrate concentration due to current and future practices; and 6. Data availability for the selected study area. Jabalia Camp was taken in this research work with Gaza City since the water supply system is the same for both areas and the two areas are undergoing elevated nitrate concentration problem. 1.3 Research Question What would be the future overall nitrate concentration in GCJC due to current practices and potential management options? 1.4 Research Objectives The objectives of this research are the following: 1. To characterize and analyze nitrate occurrences in the groundwater of GCJC area; 2. To identify and quantify the probable sources of nitrate contamination in the groundwater of GCJC area; 3. To assess nitrate concentration in the groundwater of GCJC due to the adoption of protection measures. This specific objective entails the development of a mathematical model; and 4. To set up recommendations for efficient management options that can lead to aquifer recovery from nitrate pollution. 6 1.5 Research Outcome The following summarizes the research outcome: 1. Improve public awareness. Residents of GCJC area will gain an appreciation to the extent of the problem of nitrate contamination of groundwater; 2. Aid the decision makers. The developed mathematical model will definitely facilitate the decision making process in relation to the minimization of nitrate concentration in the groundwater of GCJC area; 3. Generalization. It is quite straightforward to generalize the application described herein to the aquifers of the West Bank; and 4. New insights. This research furnishes new insights and solutions to groundwater resources problems that involve nitrate contamination. In other words, this research is a contribution toward an efficient management of the Palestinian water resources. 1.6 Research Output The following summarizes the research output: 1. Analysis of the temporal distributions of nitrate concentration in the groundwater of GCJC area; 2. A groundwater mathematical model of nitrate concentration in GCJC area; 3. A set of recommended and verified management options to minimize groundwater nitrate contamination of GCJC area. 7 1.7 Thesis Organization The general structure of the thesis is as follows. Chapter II describes the study area. Chapter III provides related literature review and general background. Chapter IV demonstrates the general methodology. In chapter V, model development is elucidated. Chapter VI furnishes analyses and discussions regarding model output. In Chapter VII, preliminary management options are demonstrated and their efficiencies are assessed. Conclusions and recommendations are provided in Chapter VIII. 8 CHAPTER 2 DESCRIPTION OF THE STUDY AREA 9 2.1 Introduction The GCJC area is located in the north side of Gaza Strip, which is a narrow, low-lying stretch of sand dunes along eastern Mediterranean Sea. It forms the foreshore that slopes gently up to elevation of 105 m above main sea level (masl). Figure (1 depicts the regional setting of Gaza Strip and the surrounding countries. Figure (2 shows the location of the study area. Over 1.4 million Palestinians live in Gaza Strip; about one third of that lives in GCJC. The total area of GCJC is (58) km2. Figure 3 depicts different features of the GCJC area. Egypt Syria Jordan Lebanon Gaza Strip West Bank Neighboring countries Egypt Gaza Strip Historic Palestine Jordan Lebanon Syria West Bank Figure (1): Regional setting of Gaza Strip and the neighboring countries. 10 Figure (2): The location of GCJC area within Gaza Strip. 2.2 Topography The topography of the study area is characterized by elongated ridges and depressions, dry streambeds and shifting sand dunes. The ridges and depressions generally extend parallel to the coastline. The height of the land surface increases from west to east. The lowest height of the study area is zero which increases eastward gradually to 70-75 masl. Gaza Jabalia 500 0 500 100 0 Meters N EW S Gaza Strip Study area Jabalia Gaza 11 500 0 500 1000 Meters N EW S Water distribution network Study area Jabalia Gaza 0.5 1.0 0.0 1.5 -0.5 -1.0 -1 . 5 -2. 0 2.0 -2.5 -3.0 -4. 0 -3.5 -4.5 2.5 -5 .0 3.0 -1 .0 -1 .0 1 . 5 1.5 1.5 -4. 5 500 0 500 1000 Meters N EW S Study area Jabalia Gaza Contours of Water table elevation (msl) & & & & & & & &&& & &&&&&& # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $$ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $$ $ $ $ $ $ $ $$ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $$ $ $ $ $ $ $ $ $ $ $ $$ $$ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $$ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $$ $ $ $ $ $ $ $ $ $ $ $ $ $ $$ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $$ $ $ $ $ $ $$ $ $ $ $ $ $ $ $ $ $ $ $ $ $$ $ $ $ $ $ $ $ $ $ $ $$ $ $ $$ $ $$ $ $$ $ $ $$ $ $ $ $ $ $$ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $$$$$ $$ $ $ $ $ $ $$$$ $$ $ $ $ $ $ $$$ $ $ $ $ $$ $ $ $ $ $ $ $ $ $ $ $ $ $ $$$ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ % % % % % 500 0 500 1000 Meters N EW S Wells in Gaza and Jabalia # Municipal % Abandoned $ Agricultural & Piezometers Study area Jabalia Gaza #Y #Y #Y #Y #Y Shati Tuffah Jabalia Gaza City Gaza-Sout 500 0 500 1000 Meters N EW S #Y Rain stations Study area Jabalia Gaza 500 0 500 1000 Meters N EW S Land use Built-up areas Citrus Dates Field crops Fruits Grapes Greenhouses Horticulture Olives Open area Settlements Study area Jabalia Gaza 500 0 500 1000 Meters N EW S Soil distribution Dark brown / reddish brown Loess soils Sandy regosols Study area Jabalia Gaza Figure (3): Different information categories for the GCJC area. 12 500 0 500 1000 Meters N EW S Land use Agriculture Built-up areas Open area Settlements Study area Jabalia Gaza 55 40 50 45 60 35 30 25 20 15 10 5 6570 75 80 65 70 40 45 5 45 55 75 40 55 75 5 55 40 60 5 75 30 55 70 40 45 5 40 70 50 50 45 5 50 50 45 60 70 500 0 500 1000 Meters N EW S Study area Jabalia Gaza Contours of DEM (msl) & & & & & & & &&& & &&&&&& # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $$ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $$ $ $ $ $ $ $ $$ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $$ $ $ $ $ $ $ $ $ $ $ $$ $$ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $$ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $$ $ $ $ $ $ $ $ $ $ $ $ $ $ $$ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $$ $ $ $ $ $ $$ $ $ $ $ $ $ $ $ $ $ $ $ $ $$ $ $ $ $ $ $ $ $ $ $ $$ $ $ $$ $ $$ $ $$ $ $ $$ $ $ $ $ $ $$ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $$$$$ $$ $ $ $ $ $ $$$$ $$ $ $ $ $ $ $$$ $ $ $ $ $$ $ $ $ $ $ $ $ $ $ $ $ $ $ $$$ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ % % % % % -7 -6 -9 -7 -6 -5 -9 -8 -9 -6 -8 -7 -7 -5 -8 -6 -6 -5-8 -5 -7 -7 -7 -9 -7 -7 -6 -7 -8 -9 -8 -6 -9 -9 -15 -28 -46 -45 -63 -87-15 -19 -27 -10 -15 -15 -36 -59 -58 -30 -24 -37 -51 -53 -58 -59 -63 -64 -65 -15 -15 -15 -15 -34 -15 -10 -15 -15 -15 -32 -15 -15 -15 -15 -44 -13 -10 -15 -15 -15 -15 -10 -47-15 -41 -16 -15 -24 -16 -10 -11 -20 -50 -11 -13 -13 -14 -15 -15 -15 -15 -15 -15 -15 -15 -15 -15 -15 -15 -15 -15 -15 -15 -15 -15 -15 -15 -15 -15 -15 -15 -15 -15 -15 -15 -15 -15 -15 -15 -15 -15 -15 -15 -16 -26 -24 -15 -15 -15 -12 -15 -27 500 0 500 1000 Meters N EW S Elevation of bottom of screen (msl) # Municipal % Abandoned $ Agricultural & Piezometers Study area Jabalia Gaza Figure (3): Continue 2.3 Climate The Gaza Strip has a characteristically semi-arid climate. There are two well-defined seasons: the wet season starting in October and extending through March, and the dry season from April to September. Peak months for rainfall are December and January. The average mean daily temperature in Gaza Strip ranges from 25oC in summer to 13oC in winter. The annual average relative humidity is about 72 percent. Evaporation is high in 13 summer when there is always a water deficit. Winds prevail from the northwest but come from the southwest in winter. 2.4 Rainfall Distribution There are five rainfall stations in the study area and these are: Gaza City, Southern Gaza, Tuffah, Shati, and Jabalia. For efficient creation of Thiessen polygons (to be used later in model development), additional rainfall stations were considered that are located outside the study area. Rainfall data for these rainfall stations for the period from 2000 to 2004 are listed in Table (1). Table (1): Annual rainfall data (in mm) for the relevant rainfall stations for the years from 2000 to 2004. Name 2000 2001 2002 2003 2004 Beit Hanon 406 498 548 802 357 Beit Lahia 391 490 542 724 397 As-shati 425 479 522 627 343 Gaza City 335 512 544 599 385 Southern Gaza 368 564 661 791 503 Jabalia 389 540 566 693 374 At-tuffah 357 533 604 654 432 2.5 Land Use A land use map of the GCJC area is shown in Figure 3. The breakdown of land use by category is summarized in Table (2) and Table (3). Agricultural land occupies about 34% of the land surface and is the dominant economic sector in the study area. Built-up areas occupy 45% while almost 21% of the land is characterized as open area. 14 Table (2): Detailed categories of land use for the study area. This table is based on the land use map of the entire Gaza Strip as obtained from the PWA and later processed using GIS capabilities for the GCJC area. Land Use Category Area (km2) % Built-up areas 26.27 45 Citrus 5.24 9 Dates 2.60 4 Field crops 6.42 11 Fruits 2.86 5 Grapes 1.26 2 Greenhouses 0.51 1 Horticulture 1.08 2 Olives 0.09 0 Open area 12.21 21 Settlements 0.001 0 Table (3): Lumped categories of land use practices for the study area. General Area (km2) Agriculture 20.05 Built-up 26.27 Open area 12.21 Settlements 0.001 2.6 Soil Types In the study area, there are three types of soil and these are: dark brown/ reddish brown, loess soils, and sandy regosols. Table (4) summarizes the different soil types that exist in the GCJC area along with the total area for each type. As can be inferred from Table (4), sandy regosols type covers about 47% of the surface area followed by loess soils of approximately 21% of the surface area. 15 Table (4): Classification of the soil types for the study area and the corresponding area. This table is based on the soil map of the entire Gaza Strip as obtained from the PWA and later processed using GIS capabilities for the GCJC area. Soil Type Area (km2) Dark brown/ reddish brown 12.45 Loess soils 18.48 Sandy regosols 27.82 2.7 Water Resources There are an estimated 534 wells within the GCJC area (see Figure 3). The majority of these wells are privately owned and used for agricultural purposes. A total of 39 wells are owned and operated by municipalities and are used for domestic supply. The distribution of these wells is depicted in Figure (3). Agricultural wells are mostly drilled and installed as large diameter boreholes. Most agricultural wells in GCJC are shallow and extend only a few meters (5-10) below the water table. Municipal wells are deeper depending on location and distance from the coast. 2.8 Nitrate Pollution in the Groundwater of GCJC Pollution of the groundwater of GCJC area is a major problem. There are many sources of pollution and the aquifer is highly vulnerable to pollution. Many years of over-pumping have resulted in seawater intrusion and upcoming of saline groundwater. Furthermore, human activities including agriculture and inadequate waste management have increased groundwater contamination levels. Intensive cultivation and efforts to boost production 16 have led to excessive use of fertilizers, pesticides, herbicides and soil fumigants, while collection, treatment and disposal of wastewater and solid waste (including hazardous materials) are wholly inadequate in many areas (see Figure 4 and Figure 5). 17 500 0 500 1000 Meters N EW S NO3 2000 (mg/L - N) 2 - 9 9 - 17 17 - 24 24 - 31 31 - 39 39 - 46 46 - 53 53 - 61 61 - 68 No Data Study area Jabalia Gaza 500 0 500 1000 Meters N EW S NO3 2001 (mg/L - N) 7 - 32 32 - 57 57 - 83 83 - 108 108 - 134 134 - 159 159 - 184 184 - 210 210 - 235 No Data Study area Jabalia Gaza 500 0 500 1000 Meters N EW S NO3 2002 (mg/L - N) 3 - 12 12 - 21 21 - 29 29 - 38 38 - 47 47 - 55 55 - 64 64 - 73 73 - 81 No Data Study area Jabalia Gaza 500 0 500 1000 Meters N EW S [NO3-N] 2004 (mg/L - N) 7 - 15 15 - 24 24 - 33 33 - 41 41 - 50 50 - 59 59 - 67 67 - 76 76 - 85 No Data Study area Jabalia Gaza 500 0 500 1000 Meters N EW S NO3 - 2003 (mg/L - N) 7 - 13 13 - 18 18 - 24 24 - 30 30 - 35 35 - 41 41 - 46 46 - 52 52 - 58 No Data Study area Jabalia Gaza Figure (4): Average nitrate concentration in the study area for the years from 2000 to 2004. ( Source: Database of PWA) 18 0 10 20 30 40 50 60 70 80 90 100 110 2002 m g/ L (N O 3- N ) 0 10 20 30 40 50 60 70 80 90 100 110 2000 m g/ L (N O 3- N ) 0 10 20 30 40 50 60 70 80 90 100 110 2001 m g/ L (N O 3- N ) 0 10 20 30 40 50 60 70 80 90 100 110 2003 m g/ L (N O 3- N ) 0 10 20 30 40 50 60 70 80 90 100 110 2004 m g/ L (N O 3- N ) Figure (5): Nitrate concentrations for different wells in the study area for the years from 2000 to 2004. 2.9 Geology The aquifer of the GCJC consists of the Pleistocene age Kurkar Group (Gvirtzman, 1969) and recent (Holocene age) sand dunes. The Kurkar Group consists of marine and aeolian calcareous sandstone, reddish silty sandstone (‘hamra’), silts, clays, unconsolidated sands, and conglomerates. 19 Regionally, the Kurkar Group is distributed in a belt parallel to the coastline, from north of Haifa to the Sina in the south. Near the Gaza Strip, the belt extends about 15-20 km inland, where it unconformably overlies Eocene age chalks and limestones (the “Eocene”), or the Miocene- Pliocene age Saqiye Group, a 400-1000 meter thick sequence of marls, marine shales, and claystones. The transition from the Kurkar Group to the Saqiye Group is sometimes obscured by the presence of a thin, basal conglomerate. Figure (6) presents a generalized geological cross section of GCA. Figure (6): A general cross section of GCA. (Metcalf and Eddy, 2000) Clay formations or units within Gaza, and the coastal aquifer in general, are of two types: marine and fluvial. Marine clays are present along the coast, at various depths within the formation. They pinch out about 5 km from present coastline, and based on existing data, appear to become more important towards the base of the Kurkar Group. Three major clay layers extend inland about 2 to 5 km, depending on location and depth. GCA is composed of sands, calcareous sandstone and pebbles. Semi-per-meable and impermeable layers are sandwiched in between, dividing the system into sub-aquifers. This subdivision is especially developed in the western 20 23.3 23.5 23.8 Ja n- 00 M ar -0 0 M ay -0 0 Ju l-0 0 Se p- 00 N ov -0 0 Ja n- 01 M ar -0 1 M ay -0 1 Ju l-0 1 Se p- 01 N ov -0 1 Time D ep th to w at er (m ) 2D 24.0 25.0 26.0 27.0 28.0 29.0 30.0 31.0 Se p- 68 Se p- 71 Se p- 74 Se p- 77 Se p- 80 Se p- 83 Se p- 86 Se p- 89 Se p- 92 Se p- 95 Se p- 98 Se p- 01 Time D ep th to w at er (m ) E/32 0.0 5.0 10.0 15.0 20.0 25.0 30.0 N ov -7 1 N ov -7 4 N ov -7 7 N ov -8 0 N ov -8 3 N ov -8 6 N ov -8 9 N ov -9 2 N ov -9 5 N ov -9 8 N ov -0 1 Time D ep th to w at er (m ) D/6 34.5 35.0 35.5 36.0 36.5 37.0 Ja n- 01 Fe b- 01 M ar -0 1 A pr -0 1 M ay -0 1 Ju n- 01 Ju l-0 1 A ug -0 1 Se p- 01 O ct -0 1 N ov -0 1 Time D ep th to w at er (m ) R/24 part of the coastal plain, water level and quality. Further inland, the sub- aquifers effectively merge to form on system. All along the coast, there are areas of seawater intrusion due to over-pumping of the freshwater aquifer. 2.10 Water Table Elevation In order to spell out the variability of water table elevation with time in GCJC, selected time series were prepared from the available data and plotted accordingly as can bee seen from Figure (7). However and in order to arrive at meaningful impressions, depth to water table instead was computed and presented. 21 35.0 35.3 35.5 35.8 36.0 36.3 36.5 36.8 37.0 Ja n- 01 Fe b- 01 M ar -0 1 A pr -0 1 M ay -0 1 Ju n- 01 Ju l-0 1 A ug -0 1 Se p- 01 O ct -0 1 N ov -0 1 Time D ep th to w at er (m ) E/116 72.0 72.5 73.0 73.5 74.0 74.5 75.0 75.5 N ov -7 1 N ov -7 3 N ov -7 5 N ov -7 7 N ov -7 9 N ov -8 1 N ov -8 3 N ov -8 5 N ov -8 7 N ov -8 9 N ov -9 1 N ov -9 3 N ov -9 5 Time D ep th to w at er (m ) F/97 31.0 32.0 33.0 34.0 35.0 36.0 37.0 38.0 39.0 40.0 N ov -7 1 N ov -7 3 N ov -7 5 N ov -7 7 N ov -7 9 N ov -8 1 N ov -8 3 N ov -8 5 N ov -8 7 N ov -8 9 N ov -9 1 N ov -9 3 N ov -9 5 Time D ep th to w at er (m ) R/25C 0.0 10.0 20.0 30.0 40.0 50.0 60.0 Se p- 68 Se p- 71 Se p- 74 Se p- 77 Se p- 80 Se p- 83 Se p- 86 Se p- 89 Se p- 92 Se p- 95 Se p- 98 Se p- 01 Time D ep th to w at er (m ) R/60 69.5 70.0 70.5 71.0 71.5 72.0 72.5 73.0 73.5 74.0 N ov -7 1 N ov -7 4 N ov -7 7 N ov -8 0 N ov -8 3 N ov -8 6 N ov -8 9 N ov -9 2 N ov -9 5 N ov -9 8 N ov -0 1 Time D ep th to w at er (m ) R/84 51.0 52.0 53.0 54.0 55.0 56.0 57.0 58.0 59.0 Se p- 68 Se p- 71 Se p- 74 Se p- 77 Se p- 80 Se p- 83 Se p- 86 Se p- 89 Se p- 92 Se p- 95 Se p- 98 Se p- 01 Time D ep th to w at er (m ) R/38 Figure (7): Time series of depth to water table for selected wells in the GCJC area. (Source: Database of PWA) 22 CHAPTER 3 LITERATURE REVIEW AND GENERAL BACKGROUND 23 This chapter provides information about nitrate contamination of groundwater, health problems associated with nitrate contamination, management of nitrate contamination of groundwater, and about the groundwater model. 3.1 Nitrate Contamination of Groundwater Nitrogen (N) exists in the soil as nitrite (NO2), nitrate (NO3 -), ammonium (NH4 +), ammonia (NH3), and organic-nitrogen (organic-N). Ammonium is easily adsorbed on to the soil particles. Nitrate is the primary nitrogen species lost from soils by leaching due to its high mobility (Almasri and Kaluarachchi, 2004). Nitrate in water is present as a highly soluble salt. Standard water treatment practices do not affect nitrate concentrations in water (Bhumble, 1999; Mourabit et al., 2002). Nitrates from water can be removed by specialized water treatment technologies, which increase the cost of water treatment. Nitrate contamination of groundwater depends upon climate, fertilizer or manure management, soil, crop, and farming systems. A climate with rainfall exceeding evapotranspiration often leads to the infiltration of rainwater to groundwater. A portion of the water received through precipitation becomes surface runoff and is lost from the land to rivers or streams. When water moves on the surface of a soil, it dissolves some nitrates that are present in the surface layers of soils which may cause contamination to surface water. Another portion of the precipitation seeps into the soil and recharges the groundwater. This seeping water dissolves soil nitrates. Any excess nitrates that are present in this groundwater- 24 recharge zone will leach down to the groundwater and contaminate the aquifer (Bhumble, 1999). Nitrate is a world-wide problem which contaminates both soil and groundwater (Mitchell et al., 2003; Kraft and Stites, 2003; Liu et al., 2004). The US Environmental Protection Agency (US EPA) has established a maximum contaminant level (MCL) of 10 mg/L NO3-N (US EPA, 1995). Contaminated water by nitrate may cause methemoglobinemia in infants and stomach cancer in adults (Wolfe and Patz, 2002). Nitrate may indicate the presence of bacteria, viruses, and protozoa in groundwater if the source of nitrate is animal waste or effluent from septic tanks (Almasri and Kaluarachchi, 2004). Nitrogen applied through organic fertilizers or manure is converted to plant-available-nitrate by bacteria living in the soil. The growing plants uptake part of this nitrate. The growing bacteria also utilize nitrates. When sufficient decomposable organic matter is present, soil bacteria can remove a significant amount of nitrate through a process called immobilization. Another group of bacteria uses nitrates as a substitute for oxygen when oxygen is limited. These bacteria convert nitrate to gases such as nitrogen, nitrous oxide, and nitrogen dioxide. This is known as denitrification. Nitrate not taken up by crops or immobilized by bacteria into soil organic matter or converted to atmospheric gases by denitrification can leach from the root zone and possibly end up in groundwater (Bhumble, 1999). Figure (8) depicts a representation of the surface and subsurface activities related to nitrate application and leaching. 25 Figure (8): A schematic describing the proposed conceptual model of nitrogen loading and transformations. Among the many influencing factors, nitrate leaching from fertilizer use depends upon the fertilizer types, method of application, and climatic conditions. Nitrate leaching may be greater when a fertilizer contains the nitrate compared to the situations where ammoniacal nitrogen is the major component of a nitrogen-based fertilizer. Nitrate losses are likely to be more when all the nitrogen is applied in one application compared to split applications. Nitrogen fertilizers or manure used on a sandy soil are more vulnerable to leaching to groundwater than nitrogen used on a clay soil. Water moves 26 rapidly through sandy or other coarse-textured soils (Kraft and Stites, 2003; Babiker et al., 2003). The negative charge on the clay particles retains ammonium ions. This retention prevents them from leaching. Nitrate ions are negatively charged and are not retained by clay particles. More nitrates are lost by denitrification in clay soils than in sandy soils due to the low presence of oxygen in the clay soils as compared to sandy soils. (Stournaras, 1998; Bhumble, 1999; Rodvang et al., 2002). Soil thickness and distance between the root zone and groundwater also determine the vulnerability of an aquifer to pollution (see Figure 8). Nitrate leaching from shallow soils on fractured rocks such as limestone can cause extensive contamination of groundwater. Storage of manure in open fields with no protection from rain, direct discharge of manure overflow water to a stream, or leaking manure lagoons can all contribute to nitrate pollution of surface and groundwater (Bhumble, 1999; Liu et al., 2004). To estimate nitrate leaching, many approaches have been used. Some studies assumed a specific fraction of the on-ground nitrogen loading to leach as nitrate. Others have used soil nitrogen models to simulate the nitrogen dynamics in the soil. A few studies conducted simple yet efficient nitrogen mass balance calculations to estimate the nitrate leaching to groundwater (Meisinger and Randall, 1988). In general, accurate estimates of nitrate leaching are obtained when soil transformation models and nitrogen mass balance calculations are utilized. Once it is established that contamination of groundwater has occurred, management actions must be considered to restore the aquifer. Such actions 27 may imply the identification of areas that witness the contamination, characterization of sources responsible for contamination, development of alternative measures and options to restore the quality of the affected groundwater. Al-Agha (2004) showed in his study different environmental problems in Gaza Strip and he discussed approaches, measures and steps for an environmental management and legislation plan. Shomar (2006) showed that NO3-, Cl- and F- exceeded 2 to 9 times the World Health Organization (WHO) standards in 90% of the wells tested with maximum concentrations of 450; 3,000; and 1.6 mg/L, respectively. Abu Maila, El-Nahal, and Al- Agha (2004) investigated the seasonal variation in nitrate concentration to understand the mechanisms and parameters controlling this pollutants. 3.2 Health Problems Associated with Nitrate Contamination Nitrate contamination of fresh water can cause methemoglobinemia; a blood disorder to which infants are particularly susceptible. The risk comes from the reduction of NO3 to NO2, a process which occurs naturally in human saliva and in gastric fluid of infants. When NO2 is present, the blood compound hemoglobin is converted to methemoglobin, which cannot carry oxygen. In the blood of normal adults, enzymes convert the methemoglobin back to hemoglobin, but newborn infants and adults taking certain medications or with certain diseases do not have enough enzymes to make this reconversion. Symptoms of methemoglobinemia are bluish mucous membranes and digestive and respiratory problems. If the condition is severe, brain damage or death can result. In less severe cases or if diagnosed early, the condition 28 can be reversed. (University of Wisconsin Extension, 1983; USEPA, 1985b; Wolfe and Patz, 2002). Nitrites and nitrates have been linked to cancer, but the evidence thus far is inconclusive. The US federal drinking water standard MCL for NO3 is 10 mg/L of NO3- N. It may also be expressed as 45 mg/L NO3. The drinking water standard is based on the risk of methemoglobinemia to infants, the group at highest risk to this condition (USEPA, 1989). Nitrate is not just a problem for human health; domestic animals may also be adversely affected by high NO3 in water. Many plants and feeds are naturally high in NO3. If groundwater well is contaminated with NO3 and used to feed animals, NO3 poisoning is possible, particularly in ruminants such as cows or sheep. The University of Wisconsin suggests that NO3 levels above 40 mg/L in NO3 are risky for livestock, and water with more than 100 mg/L NO3 should not be used for livestock watering (University of Wisconsin Extension, 1983). 3.3 General Sources of Nitrate Contamination in Groundwater Sources of groundwater contamination by nitrate can be classified into point and non-point sources. Non-point sources such as fertilizer, dairy farms, manure application, leguminous crops, dissolved nitrogen in precipitation, irrigation return-flows, and dry deposition are considered the common non-point sources of nitrate. Point sources of nitrogen such as septic systems and cesspits can be major sources of nitrate pollution (Joosten et al., 1998; Stournaras, 1998; Rodvang et al., 2002; Mitchell et 29 al., 2003; Babiker et al., 2003). Septic tanks produce significant amount of nitrogen that leaches to the groundwater when there are no sewer systems. 3.4 Management of Nitrate Contamination of Groundwater In general, management alternatives for groundwater quality protection are practices designed to prevent further pollution or reduce the existing occurrences of pollution to acceptable levels (Almasri and Kaluarachchi, 2004). The agricultural best management practices to minimize NO3 inputs to groundwater encompass a broad and diverse area of crop and soil management options as well as socio-economic and possibly regulatory activities (Moore, 1979; Bear et al., 1992; Broeke and Putten, 1997). The guiding principle is to minimize the amount of NO3 in the rooting zone, especially during periods when leaching is likely to occur. This could involve multiple fertilizer applications; use of cover crops or deep-rooted crops; genetic selection to improve crop N-use efficiency; chemical additives that inhibit the rate of nitrification; slow-release of inorganic or organic fertilizers; the careful management of irrigation to minimize leaching; and inclusion of available N from NO3 in the rooting zone and N mineralized from organic matter, manure, and crop residues in N fertilizer recommendations. Considerable refinement of N-fertilizer recommendations, taking into account such factors as weather, N cycle, and level of management, is needed (Hasler, 1998; McLay et al., 2001; Oenema et al., 2004). Other solutions may be required in extreme cases. These could include land-use zoning to lower the density of cropland in, a tax, or legal res- 30 trictions on fertilizer use (Keeney and Follett, 1991; Meisinger and Delgado, 2002). Improvements in irrigation water-use efficiency with new irrigation tech- nologies and water needs forecasting can be expected as water costs rise. These approaches will also greatly improve N-use efficiency by lessening the amount NO3 leached (Zhangand and Jorgasen, 2004). Animal operations require special management of wastes to minimize NO3 pollution. These include proper management of feedlots to minimize nitrification, leaching, and application of the wastes to cropland at rates based on agronomic principles, including N needs of the crop (Anderson, 1978). It is critical to use system analysis to develop nitrate BMPs. Nitrogen- fertilizer recommendations using soil-plant mass balances (Meisinger, 1984) should aid greatly in lowering fertilizer use and environmental consequences of over fertilization without lowering greatly the economic returns. Developing these models requires multidisciplinary teams (Keeney and Follett, 1991). 3.5 Groundwater Modeling 3.5.1 Introduction Knowing and expecting the behavior of groundwater systems is not easy. Solutions of complex groundwater problems must involve formulating a correct conceptual model, selecting parameter values to describe spatial variability within the groundwater flow system, as well as spatial and temporal trends and past and future trends in water levels. Although some 31 decisions can be made using best engineering judgment, in many instances human reasoning alone is inadequate to synthesize the conglomeration of factors involved in analyzing complex groundwater problems. The best tool available to help groundwater hydrologists meet the challenge of prediction is usually a groundwater model. 3.5.2 Model Definition A Model can be defined as a simplified version of a real world system that approximately simulates the relevant excitation – response relations of the real–world system (Bear et al., 1992). Others define model as any tool that represents an approximation of a field situation. A mathematical model simulates groundwater flow indirectly by means of a governing equation thought to represent the physical processes that occur in the system, together with equations that describe heads or flows along the boundaries of the model (boundary conditions). For time-dependent problems, an equation describing the initial distribution of heads in the system also is needed (initial conditions). Mathematical models can be solved analytically or numerically. The set of commands used to solve a mathematical model on a computer forms the computer program or code. Models provide a framework for synthesizing field information and for testing ideas about how the system works. They can alert the modeler to phenomena not previously considered. They may identify areas where more field information is required. 32 3.5.3 Why Do We Need a Model in this study? Most groundwater modeling efforts are aimed at predicting the consequences of a proposed action. Models can be used in an interpretive sense to gain insight into the controlling parameters in a site-specific setting or as a framework for assembling and organizing field data and formulating ideas about system dynamics. Models can also be used to study processes in generic geologic settings. Generic models have been used to study lake-groundwater interaction. Generic modeling studies also may be helpful in formulating regional regulatory guidelines and as screening tools to identify regions suitable or unsuitable for some proposed action. 3.5.4 Main Output of Groundwater Modeling The basic processes that may be considered part of many groundwater problems include groundwater flow and solute transport. Groundwater flow is a process that can be modeled without consideration of solute transport. In this process we can find hydraulic head. Solute transport requires either simultaneous solution with or results from a groundwater flow model to find concentrations. This is because the movement (transport) of solutes is controlled partially by the groundwater movement. Solute transport models are used for a wide variety of groundwater quality problems, such as point source pollution (e.g. waste disposal wells), spread source pollution (e.g. landfills) or sea-water intrusion. 33 3.5.5 Mathematical Models Groundwater modeling begins with a conceptual understanding of the physical problem. The conceptual model usually consists of a set of assumptions that describe the system's composition, the transport processes, the mechanisms that govern them, and the relevant medium properties (Faust and Mercer; 1980; Bear et al., 1992). The next step in the modeling process is to express the conceptual model in the form of a mathematical model. The mathematical model contains the information as the conceptual one, but expressed as a set of equations which are amenable to analytical and numerical solutions. The solution of the mathematical equations yields the required predictions of the real-world system's behavior in response to various sources and/or sinks. The permeability of a porous medium, aquifer transmissivity, aquifer storativity, and porous medium dispersivity are examples of model coefficients. The numerical values of all the coefficients appearing in the model must be known unless that no model can be employed in any specified domain. To obtain the values of the coefficients, start by investigating the real-world aquifer system and find a period in the past for which information is available on: (i) initial conditions; (ii) excitations of the system; and (iii) observations of the response of the system. If such a period can be found, one can: (i) impose the known initial conditions on the model; (ii) excite the model by the known excitations of the real systems; and (iii) derive the response of the model to these excitations. In order to derive the model's response, one has to assume some trial values for the 34 coefficients and compare the response observed in the real system with that predicted by the model (Bear et al., 1992). The sought values of the coefficients are those that will make the two sets of values of state variables identical. Sensitivity analysis enables the modeler to investigate whether a certain percentage change in a parameter has any real significance, that is whether it is a dominant parameter or not. There are two methods of solutions, analytical and numerical. The preferable method is the analytical one, as once such a solution derived; it can be used for a variety of cases. A number of simplifying assumptions regarding the groundwater system are necessary to obtain an analytical solution. Once the conceptual model is translated into a mathematical model in the form of governing equations with associated boundary and initial conditions, a solution can be obtained by transforming it into a numerical model and writing a computer program (code) for solving it. Mathematical models consist of partial differential equations for groundwater flow and solute transport. The groundwater flow equations with appropriate boundary and initial conditions are used to analyze many groundwater problems, such as water supply. The solute transport equation is used with the groundwater flow equation to address pollution problems. These problems are not as well understood, especially the characterization of source terms and dispersion. These equations and their boundary conditions can be simplified and solved analytically. 35 More complex forms of the equations and boundary conditions may be solved numerically. Mathematical model of any physical system can be generalized as shown in Figure (9). CONCEPTUAL MODEL MATHEMATICAL MODEL NUMERICAL MODEL APPROXIMATE EQUATIONS NUMERICALLY RESULTING IN A MATRIX EQUATION THAT MAY BE SOLVED USING A COMPUTER ANALYTICAL MODEL SIMPLIFY EQUATION SO THAT SOLUTIONS MAY BE OBTAINED BY ANALYTICAL METHODS Figure (9): A Logic diagram for developing a mathematical model 3.5.6 Initial and boundary conditions The initial condition in general describes the state of the system at the beginning of simulation. For instance, the initial condition may describe the distribution of the contaminant in the groundwater within the domain at the beginning of simulation. A general form of initial condition can be written as ( ) ( )xf0t,xC == where ( )xf is a function defining the variation in concentration in the x direction at t=0. A common initial condition is C(x, t)=0 or C(x, t)=CI to provide a constant concentration within the system. Boundary conditions are mathematical statements specifying the dependent variable (head or concentration) or the derivative of the dependent variable (flux) at the boundaries of the problem domain. Correct selection of the boundary conditions is a critical step in model design. In steady-state 36 simulations, the boundaries largely determine the flow pattern. Boundary conditions influence transient solutions when the effects of the transient stress reach the boundary. In this case, the boundaries must be selected so that the simulated effect is realistic. Physical boundaries of groundwater flow systems are formed by the physical presence of an impermeable body of rock or a large body of surface water. Other boundaries form as a result of hydrologic conditions. These invisible boundaries are hydraulic boundaries that include groundwater divides. Hydrogeologic boundaries are represented by the following three types of mathematical conditions: Type 1. Specified head boundaries (Dirichlet conditions) for which head is given; Type 2. Specified flow boundaries (Neumann conditions) for which the derivative of head (flux) across the boundary is given. A no-flow boundary condition is set by specifying flux to be zero; Type 3. Head-dependent flow boundaries (Cauchy or mixed boundary conditions) for which flux across the boundary is calculated given a boundary head value. This type of boundary condition is sometimes called a mixed boundary condition because it relates boundary heads to boundary flows. There are several types of head- dependent flow boundaries. There are three types of boundary conditions in transport models: Concentrations are specified along a boundary; Concentration gradients are specified across a boundary, and Both concentrations along a boundary and concentration gradients across that boundary are specified. 37 3.5.7 Model Calibration The act of calibration standardizes a model. Many models are developed for specific situations and are, by definition, calibrated to that situation. Such models usually are not useful outside of their particular environment. The act of calibration is needed to increase the accuracy of the models. Calibration is the process of determining if there is any deviation from a standard observed (monitored) in order to compute a correction factor. The calibration procedure is theoretically very simple. It is simply running the model with normal inputs against items for which the actual values are known. These estimates are then compared with the actual values and the average deviation becomes the correction factor for the model. The actual data used for the calibration runs determines what type of calibration is done. In essence, the calibration factor obtained is really good only for the type of inputs that were used in the calibration runs. For a general total model calibration, a wide range of components with actual values need to be used. Better yet, numerous calibrations should be performed with different types of components in order to obtain a set of calibration factors for the various possible expected estimating situations. 3.5.8 Sensitivity Analysis A sensitivity analysis is the process of varying model input parameters over a reasonable range (range of uncertainty in values of model parameters) and observing the relative change in model response. Typically, the 38 observed changes in hydraulic head, flow rate or contaminant transport are noted. The purpose of the sensitivity analysis is to demonstrate the sensitivity of the model simulations to uncertainty in values of model input data. The sensitivity of one model parameter relative to other parameters is also demonstrated. Sensitivity analyses are also beneficial in determining the direction of future data collection activities. Data for which the model is relatively sensitive would require future characterization, as opposed to data for which the model is relatively insensitive. Model-insensitive data would not require further field characterization. If data are determined to be insensitive to variations in model input parameters, the modeler should assess the possible reasons for this insensitivity. Figure (10 depicts an example of results of sensitivity analysis. Change in Head with Model Parameter 0 10 20 30 40 50 60 0.5 0.6 0.7 0.8 0.9 1 1.2 1.4 1.6 1.8 2 Parameter Multiplier C ha ng e in H ea d More Sensitive Parameter Less Sensitive Parameter Figure (10): Simulated change in hydraulic head resulting from change in parameter value. 39 3.5.9 Lumped Parameter versus Distributed Groundwater Models Selecting the appropriate model for estimating impacts of nonpoint sources of pollution is a major task. An appropriate conceptual model should be sufficiently simple so as to be amenable to mathematical treatment, but it should not be too simple so as to exclude those features which are of interest to the investigation at hand. The information should be available for calibrating the model and the model should be the most economic one for solving the problem at hand (Bear, 1979). To completely model a system requires a very detailed knowledge of the physical properties and the processes governing groundwater movement. The virtue of a model rests in its ability to predict a general system from incomplete or partial data. The parsimonious model simplifies the representation of the physical structure and of the processes involved. Numerous types of models have been developed and used to predict water levels. The simplest are black box models that contain no spatial information, but can predict aquifer properties (Mercer and Faust, 1980). Lumped parameter models lack the spatial dimension in the equations describing flow and transport; consequently, only simple equations must be solved. These models offer the opportunity to simulate a given system with fewer data requirements for parameterization and calibration than their distributed counterparts. Lumped parameter models in groundwater applications generally were single cell models such as those developed by Gelhar and Wilson (1974) and Mercado (1976). 40 Distributed parameter models are normally chosen to increase the accuracy of predictions and to achieve a higher degree of spatial resolution. The more spatially and detailed the models, the more they have been difficult to calibrate and verify. In addition, input data must be developed for each cell; consequently, these models are not used to any great extent by regulatory agencies or other groups (Cary and Lloyd, 1984; Refsgaard et al., 1999). 41 CHAPTER 4 METHODOLOGY 42 4.1 Introduction Contamination of groundwater in the study area and elsewhere in Gaza Strip is a major problem due to the numerous sources of pollution and the high vulnerability of the aquifer to pollution. Human activities including excessive use of fertilizers, inadequate waste management, and disposal of raw wastewater have led to nitrate pollution of Gaza coastal aquifer. In order to understand the extent of the problem and to design efficient management options, a mathematical model was developed and utilized. In this chapter, a brief illustration of the methodology followed in carrying out the research work is provided. 4.2 Methodology description Figure (11 depicts the overall conceptual methodology followed in carrying out the research work. The methodology starts with the identification of the research objectives. This step is important since the objectives dictate to a great deal of extent the entire pathway of the work. So, I collected only the data which I need in my research work. In the process of data collection, I relied on different sources including internet, reports, journal articles, textbooks, and personal interviews. The study area which includes Gaza City and Jabalia Camp was selected based on different motivations, justifications and reasons as mentioned earlier in the first chapter. To gain insight regarding contamination extent from nitrate in the study area, GIS was employed. Using the outcome of the characterization of nitrate contamination extent and the data collected and with the aid of GIS 43 and MS Excel, the nitrate model was developed. Model development did pass through a variety of processes and steps including mainly the development of the conceptual model, mathematical model, calibration, and sensitivity analysis. Mathematical Model using MS Excel Set Up Research Objectives Selection of the Study Area Textbooks Internet Journal Articles Reports Data Collection Personal Communication Use of GIS Assessment of Nitrate Contamination Conceptual model - Data Proccesing Using GIS - Calibration - Sensitivity Analysis Development of Nitrate Model Model Results and Analysis Development of Management Options Conclusions and Recommendations Figure (11): A flowchart of the methodology. 44 MS Excel was used for the numerical modeling. Thereafter, model output was analyzed, assessed, and evaluated. The model was later used in the development and assessment of the management options to mitigate the impact of nitrate contamination of the groundwater of the study area. Based on the research carried out herein, conclusions and recommendations for future relevant research work were set up. 45 CHAPTER 5 MODEL DEVELOPMENT 46 5.1 Introduction The main objective from the development of the model is to find out the overall nitrate concentration of the aquifer as a function of time; that is C(t). The developed model is a single-cell lumped parameter model. The mass balance approach was used for both water and nitrate. This concept implies that the difference between what gets in and leaves out equals the change in the storage for the study area (model domain). Figure (12) depicts a schematic of the conceptual representation of the single cell lumped parameter model along with all the related parameters. Irrigation return flow W at er p um pe d fo r I rr ig at io n W at er p um pe d fo r D om es tic Denitrification Lateral outflow U ns at ur at ed Z on e Land Surface A qu ife r Lateral inflow Fertili zers R ainfall Surplus A rtifisial R echarge w ater Leakage W astew ater Leakage C esspits Figure (12): Conceptual representation of the single-cell model. 47 The developed model is comprised of two key components and these are the quantity (water) and quality (nitrate). The development of the nitrate model is the key target. However, the nitrate model development compels the prior development of the quantity model. This is because the nitrate mass in the aquifer depends on the available water quantity which can only be computed through the use of a quantity model. Figure (13 depicts the overall schematic for model development and the linkage between the water and the nitrate models. Water stresses Quantity model h(t) Quality model Nitrogen loadingC(t) Figure (13): Schematic of the overall model development. 5.2 Development of the conceptual quantity model To better comprehend the conceptual model development and model functionality, a flowchart was developed for the quantity model as shown in Figure (14. As depicted in Figure (12, Lateral inflow, artificial recharge, and natural recharge were classified as quantity model input. Lateral outflow and water pumped for irrigation and domestic purposes were classified as quantity model output. In the following, the details of the input parameters of the quantity model are illustrated. This section focuses on the 48 conceptual methodology used in the computation of the different input parameters pertaining to the quantity model. Water budget h(t) Quantity model Lateral inflow Artificial recharge Recharge Wastewater leakage Cesspits Rainfall Water leakage Irrigation return flowQuality model Lateral outflow Water pumped for domestic Water pumped for irrigation Figure (14): Schematic of the overall flowchart of the development of the conceptual quantity model. 5.2.1 Lateral Inflow (Gin) Lateral inflow is the subsurface flow that enters the model domain from its lateral boundaries and can be computed from the following equation (Darcy’s law): Gin = K × i × b × w × cos θ [1] where K: hydraulic conductivity (L/T) i: hydraulic gradient (-) b: aquifer saturated thickness (L) w: width of the aquifer (L) θ: angel between the flow direction and the imaginary line perpendicular to the boundary (º) 49 The hydraulic gradient can be computed from the groundwater elevation contour lines and is given by the following equation: x hi Δ Δ = [2] where Δh: the change in the hydraulic head (L) Δx: the distance between contour lines upon which Δh is measured (L) The aquifer average thickness is computed using the following equation: b = |Dp| + wt [3] where |Dp|: the absolute value of the distance from the sea level to the average bottom of the aquifer for the study area (L) wt: the average water table elevation from sea level (L). This value can be positive or negative Dp represents the average depths to the bottom of pumping wells in the study area. Since the model that is being developed is a lumped parameter model, Dp was determined by computing the weighted average of the depth of each pumping well within the study area after considering both well depth and pumping rate as shown in the following equation: ∑ ∑ = == n 1k k n 1k kk p Q DQ D [4] where Qk: the pumping rate for well k (L3/T) Dk: the depth of well k (L) 50 n: number of wells (-) In order to persuasively compute the lateral inflow that enters the model domain, the boundaries were discretized into segments as shown in Figure 15. For each segment, equation [1] was computed. Total lateral inflow equals the summation of all lateral inflows through all segments. %a %a %a %a %a %a%a %a %a %a %a %a %a %a %a %a%a %a%a %a %a %a %a %a %a %a %a %a %a %a %a %a %a%a %a %a %a %a %a %a %a %a %a %a %a %a %a 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 1617 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 3334 35 36 3738 39 40 42 43 44 45 46 47 500 0 500 1000 Meters N EW S %a The points of boundary Study area Jabalia Gaza Figure (15): Segmentation of model boundaries for the computation of lateral inflow and outflow. Equation [1] is based on Darcy's law. The hydraulic conductivity was assumed to be constant for the entire area and lies within a range of 20 to 80 m/d. To find the hydraulic gradient and the cross sectional area, a water table contour map was created using GIS. This map was obtained from a groundwater flow model for the entire Gaza Strip (see Figure 3) after 51 clipping it to the study area. In this map, the difference between any two contour lines is always constant which results in a constant change in the hydraulic head (Δh) between any two contour lines. Thus, the changes in the distances between the contour lines dictate the values of the gradient and also dictate the segmentation of the boundaries of the model domain. The angel between the flow direction and the imaginary line perpendicular to the boundary was measured at each segment. Flow directions were drawn perpendicular to the counter lines and the angels between these two lines were measured. The saturated thickness was assumed to decrease on monthly basis based on the average decline in the water table that was encountered in the past 30 years and amounts 10 cm per year. To find the aquifer saturated thickness, the distance from sea level to the water table (wt) and the distance from sea level to the bottom of the aquifer Dp were considered using equation [3]. Maps of Dp and wt (see equation [3]) were created using GIS and were used in the analysis and computation. 5.2.2 Artificial recharge (QAr) Artificial recharge is the amount of water injected intentionally into the aquifer in order to increase the water table elevation to serve a management objective for the mitigation of seawater intrusion into coastal aquifers. In our case and to the best of my knowledge, there are no artificial injection wells in the study area. As such, QAr was set to zero. However, model formulation is flexible in the essence that QAr appears in all computations with a zero value. 52 5.2.3 Recharge (R) Total recharge to the groundwater of the study area equals the summation of recharge from rainfall, irrigation return flow, wastewater leakage, leakage from water networks, and cesspits as depicted in Figure (14. Equation [5] was used to compute the over all recharge to the model domain as follows: R = Rra + RIr + RWWL + RWL + CSPT [5] where R: total recharge (L3) Rra: recharge from rainfall (L3) RIr: recharge from irrigation return flow (L3) RWWL: recharge from wastewater leakage (L3) RWL: recharge from water leakage (L3) CSPT: recharge from cesspits (L3) In the following subsections, all recharge components depicted in equation [5] are illustrated and explained. Recharge from rainfall In order to compute recharge from rainfall, the locations of rainfall stations were mapped using GIS. As such a GIS point shapefile of rainfall stations was created and used. For each station, the total monthly rainfall depth was computed based on the available daily values. Thiessen polygons were created for each station using GIS such that each transpired polygon was represented by a single station as shown in Figure (16. 53 #Y #Y #Y #Y #Y Shati Tuffah Jabalia Gaza City Gaza-Sout 500 0 500 1000 Meters N EW S #Y Rain stations Study area Jabalia Gaza Figure (16): Thiessen polygons of the rainfall stations for the GCJC area. In order to account for the recharge variability with soil type, each Thiessen polygon was intersected by the soil type shapefile using GIS to further divide each rainfall polygon to areas of different soil types that carry different fractions of recharge from rainfall. Total recharge from rainfall was then computed using the following equation [6]: ( )∑ = ××= z 1i iiira fraAraraR [6] where rai: monthly rainfall depth for each subdivided polygon i (L) Arai: area for each subdivided polygon (L2) frai: fraction of recharge for a specific soil type (-) z: total number of subdivided polygons (-) i: a specific subdivided polygon (-) 54 The estimation of the areas of the subdivided polygon (Arai ) was determined using GIS. Recharge from irrigation return flow Generally, not all the water used in irrigation is consumed by plants. In fact, a proportion of this may percolate beyond the soil zone and later recharge the aquifer. This recharge equals the multiplication of the total volume of water used for irrigation by the fraction of return flow as in the following equation: RIr = VIrr × δIrr [7] where VIrr: total volume of water used for irrigation in the study area (L3) δIrr: fraction of irrigation return flow that becomes recharge (-) In turn, VIrr can be computed using the following equation: [ ]∑ = ××= z 1i IrrIrrIrr )k(BINAdV [8] where dIrr: monthly irrigation rate for each crop type i (L); AIrr : area for each crop type (L2); BIN(k): a binary integer multiplication factor to account for the months that may receive irrigation water. This factor may have either the value of 1 or 0; The fraction of return flow from irrigation is within the range of 15 to 30% (Mercado, 1976). The area of each land use type (crop type) was obtained 55 using GIS. Based on the personal communications and the interviews, the monthly irrigation rates for each crop type were obtained (Eng. Mohammad Al-Hanbali and Dr. Hassan AbuQaoud, Personal Communication, 2006). Needless to mention that the land use map (see Figure (3) that was utilized herein reflects the different kinds of plantations. Since there are months without irrigation, irrigation in these months were nullified. To do so, the monthly irrigation rate was multiplied by a binary encoding scheme (see BIN(k) in equation [8]) where a value of 1 was used for the months when there was irrigation and 0 when otherwise. The multiplication of area by the monthly irrigation rate gives the monthly irrigation volume for each land use type. The summation of all these monthly volumes produces the total volume of the recharge from irrigation return flow. Recharge from wastewater leakage In this subsection, the quantification procedure of recharge from wastewater leakage from the sewage system is illustrated. This recharge equals the multiplication of the total volume of wastewater leakage from sewerage system by the fraction of wastewater recharge as illustrated in the following equation: RWWL = WWL ×δWWL [9] where WWL: total monthly wastewater leakage from the sewerage network (L3) δWWL: recharge fraction of wastewater leakage (-) 56 The monthly wastewater leakage is given by the following equation: WWL = POP × Wconsm × Φ × Ωww × PERSERV [10] where POP: total monthly population living within the study area (capita) Wconsm: per capita monthly water consumption (L3) Φ: percentage of water that becomes wastewater (%) Ωww : leakage percentage of wastewater from sewerage system (%) PERSERV: percentage of population serviced by the sewerage system (%) The monthly population living in the study area is computed using the following formulas: POPE = POPI × (1+GR)a POPM = 12 1 ( POPE – POPI) POP = POPI + POPM × STEP [11] where POPE: population at the end of the year (capita) POPI: initial population at the beginning of simulation (capita) POPM: monthly increase in population (capita) STEP: the month number for a specific year (-) GR: population growth rate (%) a: number of years for population projection (-) To find out the total volume of wastewater leakage from the sewerage system, the following issues were considered. First, the population in the 57 study area serviced by the wastewater collection network was estimated on monthly basis. Initial population was taken from the Palestinian Central Bureau of Statistics (www.pcbs.gov.ps). Personal communication was made to find out the percentage of population serviced by the sewage system (Dr. Said Ghabayen, formerly at PWA, 2005). Using a growth rate of 3.5%, monthly population were estimated. The second issue that was considered is the per capita water consumption. This was computed by considering the total monthly water consumption for the study area. The third issue was the determination of the fraction of wastewater leakage which was left to be determined through the calibration process though estimates from local reports provide a value of 10% (Metcalf and Eddy, 2000). The percentage of water that becomes wastewater was taken as 85%. Recharge from water leakage The water distribution network of the study area encounters leakage due probably to the lack of proper maintenance. It is expected that the leaking water will eventually recharge the aquifer. This recharge equals the multiplication of the total volume of water leakage and the fraction of the leakage that becomes recharge as can be seen from equation [12] and equation [13]. WL = PUMPDOM × Ωw [12] where WL: volume of leakage from the water network (L3). 58 PUMPDOM: volume of water pumped for domestic purposes on monthly basis (L3) Ωw: water leakage fraction from the network (-) The recharge to the aquifer from the leakage of water from the network is given by the following formula: RWL = WL × δWL [13] where δWL: fraction of water leakage that becomes recharge (-) Recharge from cesspits The recharge from cesspits equals the total wastewater leaching from cesspits multiplied by the fraction of wastewater that becomes recharge. Since the percentage of study area that is serviced by the sewage system is 90% then the percentage of study area with cesspits is 10%. Using a similar concept to that used in computing wastewater recharge, the total wastewater generated from cesspits was computed as shown in equation [14] and equation [15]. CSPT= WWcesspits × δCSPT [14] and WWcesspits = POP × Wconsm × Φ × (100 – PERSERV) [15] where WWcesspits: total wastewater generated from cesspits (L3) δCSPT: recharge fraction of wastewater from cesspits (-) 59 5.2.4 Lateral outflow (Go) Lateral outflow (Go) was computed using the same concept for determining lateral inflow (see equation [1]). 5.2.5 Water pumped for irrigation (QIrr) Water pumped for irrigation (QIrr) was only considered for irrigation purposes. It was estimated using equation [8]. 5.2.6 Water pumped for domestic purposes (QDO) Water consumed for domestic purposes (QDO) equals the population size multiplied by the per capita water consumption. To account for the actual amount being pumped from the aquifer, the following equation was used: )Ω × = w consm DO -(1 )W (POPQ [16] 5.3 Development of the conceptual quality model As mentioned earlier in this chapter and depicted in Figure (13, the quality model relies on the outcome of the quantity model; h(t). Mass balance of nitrate for the study area was employed in order to simulate the overall nitrate concentration. The sources of nitrate which were considered for the study area include lateral inflow, artificial recharge, fertilizer loading, and recharge. The denitrification, lateral outflow, groundwater pumped for domestic and 60 irrigation purposes were considered as the main sinks of nitrate for the study area. Figure (17 shows a flowchart that depicts the development of the conceptual quality model and the following subsections illustrate the parameters pertaining to the development of this model. Fertilizers C(t) Denitrification Lateral outflow Domestic groundwater Irrigation groundwater Lateral inflow Artificial recharge Recharge Wastewater leakage Water leakage Irrigation return flow Cesspits Rainfall Quality model Figure (17): Schematic of the overall flowchart of the conceptual quality model development. 5.3.1 Nitrate from lateral inflow (NO3Gin) The amount of nitrate (as mass) that enters the aquifer with lateral inflow from the surrounding areas (NO3Gin) can be calculated using the following two equations ([17] and [18]): NO3Gin = ∑ = z 1i jM [17] 61 Mj = Ginj × Cinj [18] where Mj: mass of nitrate entering the study area by lateral inflow through segment j (M) z: total number of segments comprising the boundaries of the model domain (-) Cinj: the average concentration of nitrate for segment j (M/L3) Using GIS, maps of average nitrate concentrations were created for years 2000 to 2004. The locations of nitrate wells were obtained using a GIS shapefile. For each year, the average nitrate concentration for each well was computed. After that, Thiessen polygons were created for each well such that each transpired polygon was represented by a single well and thus a single nitrate concentration value. This enabled the designation of nitrate concentration for lateral inflow that enters the study area through each segment (see Figure (15). These concentrations were multiplied by their corresponding lateral inflow values to obtain nitrate mass flux. Summing up these mass fluxes provides the amount of nitrate that enters the study area by lateral inflow. 5.3.2 Nitrate from artificial recharge (NO3QA) The amount of nitrate (as mass) that enters the aquifer through artificial recharge (NO3QA) can be calculated using the following equation: 62 NO3QA = QAr × CAr [19] where CAr: nitrate concentration in artificial recharge (M/L3) 5.3.3 Nitrate from fertilizer surplus (NO3SURP) The amount of nitrate (as mass) that enters the aquifer due to the fertilizer use in agricultural areas (NO3SURP) can be calculated from the following set of equations: NO3SURP = SURP × αFERT SURP = ( )∑ = ××− z 1i type )k(BINACONSFERT CONS = FERT × PERCONS [20] where SURP: total monthly mass of fertilizer surplus from all the agricultural land use types and corresponding crops (M) i: an indicator for land use type (crop type) FERT: amount of fertilizer applied for each type of land use per unit area (M/L2) αFERT: linear fraction for fertilizer that describes the transformations in the soil zone (-) Atype: area planted for each type of land use (L2) CONS: consumption of fertilizer for each crop (M/L2) PERCONS: percentage of fertilizers applied that would be taken up by plants (%) 63 BIN(k): a binary integer multiplication factor to account for the months of fertilization. This factor may have the value of either 1 or 0 To implement the set of equations in [20], many parameters must be determined. First of all, the amount of fertilizers being applied to different crop types ought to be determined. This peace of information was obtained through personal communication (Dr. Hassan Abuqauod, An-Najah National University, 2006). The area of each crop type was determined using GIS. An assumption was made that not all the fertilizers are taken up by plants and that a percentage of fertilizers are consumed. This leaves an amount that is ready to leach to groundwater. The set of equations in [20] are utilized for each crop type and the summation would give the total surplus of NO3 from fertilizers. 5.3.4 Nitrate from recharge (NO3R) Total nitrate that reaches the aquifer via recharge equals the summation of nitrate from rainfall, irrigation return flow, wastewater leakage, leakage from water networks, and cesspits. This can be expressed by the following equation: NO3R = NO3Rra + NO3Rir + NO3WWL + NO3WL + NO3CSPT [21] where NO3R: total mass of nitrate entering aquifer via recharge (M) NO3Rra: mass of nitrate entering aquifer via recharge from rainfall (M) 64 NO3Rir: mass of nitrate entering aquifer via irrigation return flow (M) NO3WWL: mass of nitrate entering aquifer via leakage of wastewater (M) NO3WL: mass of nitrate entering aquifer via leakage of water (M) NO3CSPT: mass of nitrate entering aquifer from cesspits (M) In the following subsections, a detailed description of all the components that appear in equation [21] is provided. Nitrate from rainfall recharge Nitrate that enters the aquifer from rainfall recharge can be estimated using the following equation: NO3Rra = Rra × Cp × αp [22] where Cp: nitrate concentration in rainfall (M/L3) αp: linear fraction for rainfall that describes the transformations in the soil zone (-) Equation [22] simply states that the value of recharge from rainfall (can be obtained using equation [6]) is multiplied by nitrate concentration in rainfall and the linear fraction for rainfall. Nitrate from irrigation return- flow recharge Nitrate that enters the aquifer from irrigation return flow can be estimated using the following equation: 65 NO3Rir = RQir × CIR × αIr [23] where CIR: nitrate concentration in irrigation return flow (M/L3) αIR: linear fraction for irrigation that describes the transformations in the soil zone (-) The value of CIR equals the initial nitrate concentration, C0 for the entire aquifer for the first time step (at the beginning of simulation). Thereafter and for each time step, nitrate concentration at the preceding time step is used. The value of recharge from irrigation return flow which can be obtained using equation [7] is multiplied by the nitrate concentration in irrigation return flow. After that, it has to be multiplied by the linear fraction of irrigation return flow. Nitrate from leakage of wastewater Nitrate that enters the aquifer from leakage of wastewater can be estimated using the following equation: NO3WWL = RWWL × CWWL × βWWL [24] where CWWL: total nitrogen concentration in the leakage of wastewater (M/L3) βWWL: linear fraction for wastewater leakage that describes the transformations in the soil zone (-) 66 The value of recharge from leakage of wastewater which can be obtained using equation [9] is multiplied by the concentration of total nitrogen in leakage of wastewater. After that it is multiplied by the linear fraction for rainfall. Nitrate from leakage of water Nitrate that enters the aquifer from leakage of water can be estimated using the following equation: NO3WL = RWL × CWL × αWL [25] where CWL: nitrate concentration in the leaking water (M/L3) αWL: linear fraction for water leakage that describes the transformations in the soil zone (-) The value of recharge from leakage of water which can be obtained using equation [13] is multiplied by the nitrate concentration in water that leaks from the water distribution network. This amount is then multiplied by the linear fraction of rainfall. Nitrate from cesspits Nitrate that enters the aquifer from cesspits can be estimated using equations [26]: NO3CSPT = NO3GENCSPT × βCSPT NO3GENCSPT = NGENCSPT × FraNO3N NGENCSPT = POP × PERUNSER ×NCAPITA [26] 67 where NO3GENCSPT: nitrate mass that originates in cesspits (M) βCSPT: linear fraction for cesspits that describes the transformations in the soil zone (-) FraNO3N: fraction of nitrogen from cesspits that becomes nitrate (-) NGENCSPT: total mass of nitrogen generated from the cesspits of the study area (M) NCAPITA: the generated mass of nitrogen per capita (M) 5.3.5 Nitrate lost through lateral outflow (NO3Go) The amount of nitrate lost from the aquifer through lateral outflow is computed using the following equation: NO3Go = Go × C × ε [27] where C: nitrate concentration in the aquifer (M/L3) 5.3.6 Nitrate lost through irrigation (NO3Irr) The amount of nitrate lost from the aquifer through water pumped from the aquifer for irrigation is given by the following equation: NO3Irr = QIrr× C [28] 5.3.7 Nitrate lost through domestic use of groundwater (NO3DO) The amount of nitrate lost from the aquifer through water pumped for domestic purposes is given by the following equation: 68 NO3DO = QDo× C [29] 5.3.8 Nitrate lost through denitrification (NO3DEN) The amount of nitrate lost by denitrification is given by equations [30], [31], and [32]: NO3DEN = Vw0 × λ × C [30] Vw= (h0+|Dp|) × A × n [31] t 693.0 =λ [32] where λ: denitrification rate (T-1) t: half-life time of nitrate (T) Vw: the monthly water volume in the aquifer at the beginning of each time step (L3) h0: water table elevation with reference to the sea level at the beginning of each time step (L) A: total area of the model domain (L2) n: aquifer porosity 5.4 Development of the mathematical models As mentioned earlier in this chapter, the development of the model of nitrate concentration in groundwater compels the development of a model for simulating the water table elevation (see Figure (13, Figure (14, and Figure (17). Two major equations were used to implement the mass 69 balance approach for the study area where the aquifer system was simulated as a single-cell lumped parameter model. The general mass balance equation for groundwater quantity and quality can be expressed by the following equation: QIN – QOUT = ΔSW (Water) QINCIN – QOUTCOUT = ΔSN (Nitrate) [33] For the groundwater quantity (the water table elevation), equation [33] would become as follows: Gin + QAr + Rra + RIr + Rwwl + Rwl + CSPT – G0 – Qirr – QDo = ∆SW ∆SW = Vw1 – Vwo Vwo = (ho + │Dfrom msl│)× Α × n Vw1 = ∆SW + Vwo h1 = Vw1 / A × n + Dfrom msl [34] For the groundwater quality (the nitrate concentration in the aquifer), equation [33] would become as follows: [ ] NSNO3Den NO3DoNO3Irr NO3Go NO3CSPT NO3WL NO3WWL NO3Rir NO3Rra NO3SURP NO3QA NO3Gin Δ=+++ −⎥ ⎦ ⎤ ⎢ ⎣ ⎡ ++ +++++