© Springer International Publishing AG 2018
Abdelazim M. Negm (ed.)Groundwater in the Nile Delta The Handbook of Environmental Chemistry73https://doi.org/10.1007/698_2018_253

Integrated Subsurface Thermal Regime and Hydrogeochemical Data to Delineate the Groundwater Flow System and Seawater Intrusion in the Middle Nile Delta, Egypt

Zenhom E. Salem1  , Osama M. Gaame2 and Taher M. Hassan2
(1)
Geology Department, Faculty of Science, Tanta University, Tanta, Egypt
(2)
Research Institute of Groundwater, National Water Research Centre, Cairo, Egypt
 
 
Zenhom E. Salem
1 Introduction
2 Hydrogeology of the Nile Delta
3 Methodology
4 Result and Discussion
4.1 Subsurface Thermal Regime
4.2 Hydrogeochemistry
5 Conclusions
6 Recommendations
References

Abstract

Several aquifers around the world are situated in the coastal zones and influenced by seawater intrusion. The development of populace in coastal territories and the conjugate increment in human, farming, and industrial activities have forced an increase in the needs for freshwater. The Quaternary aquifer in the Nile Delta is among the biggest groundwater aquifers in the world. Along its northern side, the aquifer is highly affected by the Mediterranean Sea. Because of the inordinate pumping in the course of the most recent couple of decades, the groundwater quality in the northern parts of the Nile Delta has been decreased extensively. Therefore, this chapter aims to trace the groundwater flow system and seawater intrusion in the study area using the multi-tracing technique. The integration between borehole temperatures and groundwater chemistry was good to conduct the aim of this study. Borehole temperature was measured in eight boreholes, and the groundwater was sampled from the same wells but sometimes from the shallow and deep zones. Tala well located to the south of the study area indicated the recharged fresh groundwater with downward flux of 0.8 m/year. The fresh groundwater started to discharge from south Tanta City till south Kafr Elsheikh City where the calculated upward fluxes were −0.1 to −0.5, − 0.35, and −0.23 m/year for Kafelarab, Nawag, and Elkarada wells. Hydrochemically, the groundwater in the area northern Kafr Elsheikh City is highly affected by seawater intrusion, and the measured temperature profiles are of discharge type, and their calculated upward fluxes were −0.6, −1.2, and −2.8 m/year for Kafr Mesaaed, Elhadady, and Motobes wells, respectively.

In comparison, temperature profile (Motobes well) affected by seawater intrusion has higher upward flux, while the freshwater recharge-type profile (Tala well) has lower downward flux. Hydrochemically, the seawater intrusion highly affected the wells from ElKarada wells to Motobes wells (northern Kafr Elsheikh City) and close to the Mediterranean Sea. Two types of saline water were recognized. The shallow groundwater is highly affected by seawater intrusion (TDS around 20 g/l), and the deeper groundwater is of hypersaline characters (80 g/l). These two types of saline water could deteriorate the groundwater quality in the Nile Delta in case of unresponsible severe pumping rates.

Keywords

Groundwater flow systemHydrogeochemistryNile DeltaSeawater intrusionSubsurface temperature

1 Introduction

There are three inquiries in hydrogeological research: (1) Where is the recharge area? (2) Where is the discharge area? (3) What is the pattern of the groundwater circulation in deep and shallow zones? Accordingly, the groundwater flow in vertical and horizontal directions could be classified “local” and “regional” groundwater framework (e.g., [13]). Local flow is restricted inside a single basin surrounded by topographic highs and vertically circulates in the shallow subsurface. Regional flow system is not mostly limited to a single basin and groundwater circulated to deeper zones. There are three techniques used to indicate the groundwater stream in a wide territory: measuring liquid potential, numerical modeling, and utilization of tracers. Six types of tracers were used for the groundwater investigation: colorimetry, photometry, mass spectrograph, electric conductivity, hydrochemistry, and temperature [4].

Seawater intrusion and submarine groundwater discharge (SGD) are commonly expected in coastal areas. Seawater intrusion (Fig. 1) is of fundamental interest for researcher and groundwater users in every coastal aquifer around the world. The seawater intrusion is a vital issue in arid and semiarid areas where the groundwater constitutes the major freshwater resources. About 3% of seawater mix with the freshwater in a coastal aquifer would render the freshwater unacceptable for human utilization. That is why the coastal aquifer groundwater should be intensively studied to keep up the dynamic freshwater-seawater interaction. The issue of saltwater/new water interface exists in Nile Delta because of overpumping for water system and seaside nature of its aquifer. Sherif et al. [5, 6], Amer and Sherif [7], Sherif and Singh [8], Sherif [9], and Sherif and Singh [10] used different techniques to study the seawater intrusion issue in the Nile Delta aquifer; however, none of them utilized subsurface thermal regime.
../images/437178_1_En_253_Chapter/437178_1_En_253_Fig1_HTML.png
Fig. 1

Illustrates the seawater intrusion in the continental fresh groundwater

The SGD considers the water output from a basin-scale hydrological cycle that represents an input into the ocean. SGD nearshore scale (Fig. 2) is the embayment or inner continental shelf scale operationally defined as extending 10 m to 10 km offshore and to depths of about 5–50 m below the seafloor, including the first confined or semi-confined portion of a submarine flow system. The shallowest submarine confining unit [11] permits primarily topographically driven the regional flow of fresh or brackish water offshore beneath the nearshore zone and sometimes entirely beneath lagoons, shallow estuaries, embayments, and barrier islands [12].
../images/437178_1_En_253_Chapter/437178_1_En_253_Fig2_HTML.png
Fig. 2

Schematic diagram of the embayment or inner shelf scale of submarine groundwater flow and discharge showing submarine flow of low-salinity water in the first confined aquifer and the zone of offshore discharge beyond the edge of the submarine confining unit [11]

The next larger scale is that of the entire continental shelf (Fig. 3), which may include multiple confined aquifer systems extending below the first confined aquifer to depths of 500 m or more below the seafloor and to the outer continental shelf edges, submarine canyon incisions, and even the continental slope, especially during sea-level low stands. The primary process driving flow at this scale is usually geothermal convection, which produces seawater recirculation through the shelf [1316]. Sediment compaction and associated dewatering, as well as brine-related processes, are also important in some settings [17]. It could be argued that there is actually more known about the occurrence of submarine groundwater systems at the shelf scale, including relict-reduced salinity groundwater than is known about intermediate embayment-scale submarine aquifer systems in some settings.
../images/437178_1_En_253_Chapter/437178_1_En_253_Fig3_HTML.png
Fig. 3

Schematic diagram of the continental shelf scale of submarine groundwater phenomena showing the variable position of the fresh-saline interface in multiple confined aquifers on the shelf, the variable widths of the mixed zone at the interface, the flow of saline water inward from the exposed edges of confined aquifers, and the upward movement of saline groundwater induced by geothermal heating at depth [11]

Cartwright [18] and Sakura [19] utilized subsurface temperature information to investigate the groundwater stream in a basin. Mostly, the typical geothermal gradient of a subsurface system is influenced by groundwater stream and changes in surface temperature. A hypothetical analysis of a distorted temperature field by geographically forced groundwater stream was introduced by Domenico and Palciauskas [20]. Their outcome demonstrates that groundwater temperature increments with depth affected by both thermal conduction and surface warming conditions. A temperature-depth profile without groundwater stream, for the most part, has a consistent slope with depth (Fig. 4a) and stratified thermal system (Fig. 4b) in the 2D cross section. At a similar height, the subsurface temperature in the recharge zone is brought down under the state of groundwater stream than under no stream condition. In the discharge zone, the temperature is higher with groundwater stream than under no stream conditions (Fig. 4c, d). The impact of groundwater flow and surface warming on subsurface temperature is delineated in Fig. 1e, f.
../images/437178_1_En_253_Chapter/437178_1_En_253_Fig4_HTML.png
Fig. 4

Schematic diagrams of the groundwater flow system and subsurface thermal regime under the condition of (a) and (b) no groundwater flow, (c) and (d) regional groundwater flow, and (e) and (f) regional groundwater flow with surface warming. Note: T, T o, and t are subsurface temperature, constant surface temperature, and time, respectively [21]

The chemical characteristics of the groundwater are mostly settled in the unsaturated zone. In contrast, less intense geochemical evolution occurs in the saturated zone, and changes take after dynamic changes in water quality toward regions of discharge. These procedures are time dependent and the chemical changes and, additionally, isotopic varieties might be utilized to give data on water stream pathways [22]. Utilizing multi-following strategy, Salem et al. [23, 24] prevailing to decide the connection between the groundwater and Shinano River in Nagaoka territory, Japan. In the present work, temperature and water chemistry were utilized for following the groundwater stream framework and seawater intrusion in the Nile Delta. Flow analysis is essentially in light of the idea of various leveled groundwater stream frameworks [2, 3].

2 Hydrogeology of the Nile Delta

In ancient times, the Nile River at Cairo formed a wide estuary that has been occupied by river deposits to frame the current fruitful delta of 250-km-wide base at the Mediterranean coast and around 160 km from Cairo (south) to sea coast (north). Seven branches of the Nile ran through the delta, but humans and nature have closed five of them [25]. The only two remaining channels are Rosetta, whose mouth is located just east of Alexandria, and Damietta, whose mouth is located at the northeast tip of the Delta. The other five mouths, which have been proven both historically and geologically to have existed [25].

The level of the lands of the delta runs between 17 m amsl at the southern limit and less than 1 m amsl at the northern parts [26]. The rice is cultivated every year in around 420,000 ha in the Nile Delta [27]. It is incorporated into summer trimming design, which additionally incorporates cotton and maize as principle summer crops in 2 or 3 years’ harvest turn. As rice cultivations use a huge quantity of water, it widely recharges the groundwater contrasted with alternate crops. In the northern territories, the rice is intensively cultivated compared to the other regions (Fig. 5). These regions are required to have hydrogeological characters not the same as alternate less rice developed regions as will be mentioned later.
../images/437178_1_En_253_Chapter/437178_1_En_253_Fig5_HTML.png
Fig. 5

Areal distribution of the rice cultivated lands in the Nile Delta (Modified after [27])

“The Nile Delta is one of the oldest intensely cultivated areas on earth. It is very heavily populated, with population densities up to 1,600 inhabitants per square kilometer” [25]. Current increments in the populace, together with upgraded ways of life and additionally the restricted freshwater sources in the Nile Delta, have made a more prominent request on water resources, requiring enhanced groundwater administration. Any new groundwater improvement in the Nile Delta should consider the potential outcomes of seawater intrusion and guarantee satisfactory control, with the counteractive action of saline intrusion being seen as perfect. Up till now, various groundwater investigations chiefly centered on seawater intrusion on the upper 100 m of the groundwater framework and expected salinities not surpassing that of Mediterranean water. There was no information on groundwater in the deeper parts of the Quaternary Nile Delta aquifer (depth up to 1,000 m). Recent studies gathered salinity estimations and found an across-the-board event of “hypersaline” groundwater: groundwater with salinities to a great extent surpassing that of seawater at depth higher than 400 m [28, 29]. This hypersaline groundwater incredibly influences the groundwater flow system and the new water capability of the aquifer. Engelen et al. [30] investigated the causes of the hypersaline groundwater and its transport system. They considered all the applicable salinization forms in the Nile Delta aquifer, over a time space of up to 2.5 million years, which is the time traverse in which the aquifer got deposited.

Several research works dealing with the hydrogeology and hydrogeochemistry of the Nile Delta aquifer were done covering the period from 1959 to 2017, among them: Zaghloul [31], El-Fayoumy [32], Shata [33], Anon [34], El-Dairy [35], El-Hefny et al. [36], Sallouma [37], Hamza et al. [38], Kotb [39], Mousa [40], Mousa [41], El-Shamy and Greish [42], Ezz and Deen [43], Sollouma and Gomaa [44], Salem et al. [29], Salem [45], Atwia [46], Shabaan [47], Frihy and Lawrence [48], El Banna [49, 50], El Banna and Frihy [51], El Banna and Frihy [52], El-Asmar and Hereher [53], Elewa et al. [54], Mabrouk et al. [55] and Shehata and El-Sabrouty [56], Salem et al. [57], Salem and El Bayumy [58, 59], Salem et al. [60], Salem and Osman [6163], and Salem et al. [6466].

The Nile Delta aquifer system is considered to be a leaky aquifer in the southern and middle parts and a free aquifer in the western and eastern borders, where the thickness of the top Holocene deposits reaches its minimum value. In the northern parts, the top Holocene deposits strongly retard upward discharge from the Plio-Pleistocene aquifer; therefore, the aquifer becomes of a confined type. The existing aquifer, in the study area, belongs to the main Nile Delta aquifer system. It is mainly formed of the Pleistocene graded sand and gravel, changing to fine sand intercalated with clay lenses. This aquifer is a semi-confined one. The thickness of the semi-confining layer is generally between 0 and 20 m and increases reaching a thickness of 70 m at the northern part of the Nile Delta. There are not too many studies with estimations of the hydraulic parameters for the overlying clay layer in literature. Farid [67] reported that its vertical hydraulic conductivity is at 0.0025 m/day, while Leaven [68] reported it at 0.0484 m/day. With slightly lower values, Wolf [69] reported it as 0.0011 m/day, and Arlt [70] at 0.0046 m/day. On the other hand, Sherif et al. [71] reported the vertical hydraulic conductivity to be about 0.67 m/day. Recently, Salem et al. [64] studied the petrophysical and the hydrogeological properties of this clay layer (Bilqas Formation) in the central part of the middle Nile Delta. They stated that “Bilqas Formation ranges in thickness from 3 m in the southwest direction to 31 m in the northeast direction. The shale content and porosity ranges are 54% to 97% and 21% to 55%, respectively. This layer has low values of permeability (16 × 10−9 to 78 × 10−9 mD) and hydraulic conductivity (<2 × 10−9 cm/s). The water salinity of this layer ranges from 200 to 1,600 mg/l.”

The Nile Delta Quaternary aquifer (Mit Ghamr Formation) is the main source of groundwater in the Nile Delta area [72]. It covers the whole Nile Delta. Its thickness varies from 200 m in the southern parts to 1,000 m in the northern parts [73] (Fig. 6). It constitutes variable proportions of sands, clays, and gravels with lateral variation and variable thickness. From those sections, the hydrogeological setting of the Nile Delta classified into one water-bearing zone (Quaternary) which involves confining thin beds (clay beds) that could prevent intrusion of seawater (Fig. 6). Dahab [75] estimated the hydraulic parameters of the aquifers. Transmissivity reaches its minimum values in the southwestern area, where it ranges between 2,000 and 3,000 m2/day. It reaches its maximum values in the middle and southeastern parts, where it ranges between 15,000 and 9,000 m2/day. Hydraulic conductivity values range between 60 and 70 m/day in the southwestern and northern parts. In the middle and southeastern parts, it reaches its maximum values, between 100 and 180 m/day. Effective porosity over the whole area ranges between 12 and 19%, which indicates that the aquifer is mainly composed of coarse sand and gravel. Recently, Salem et al. [64] studied the petrophysical and the hydrogeological properties of this Quaternary aquifer (Mit Ghamr Formation) in the central part of the middle Nile Delta. They stated that “in Mit Ghamr Formation, average shale content ranges from 4.5 to 22%. Numbers of scattered clay lenses are detected in different places with high intensity in the northeastern direction. Porosity ranges from 19 to 39%. High permeability values are recorded in this formation and ranged from 0.1 × 10−2 to 8.7 × 10−2 mD. The water salinity average values in this aquifer range from 220 mg to 2,100 mg/l. The calculated hydraulic conductivity values for this formation are of range 5.082 × 10−10 to 2.134 × 10−8 cm/s. In this layer, the increase in the shale content, the increase in porosity, decrease in the permeability and hydraulic conductivity, as well as the increase in salinity, are to the northern and northeastern directions.”
../images/437178_1_En_253_Chapter/437178_1_En_253_Fig6_HTML.png
Fig. 6

Locations and details of north-south hydrogeological cross sections in the Nile Delta area [60, 74]

The depth to the groundwater table in this aquifer ranges between 1–2 m in the north, 3–4 m in the middle, and 5 m in the south. Different estimated depths to groundwater table that have been reported by RIGW [76] and Morsy [77]. Direction of groundwater movement is northward to the sea. The irrigation and drainage networks are the main sources of groundwater recharge. The Rosetta branch, the Mediterranean Sea, coastal lakes, and pumped water used for irrigation, industrial, and domestic purposes are the main discharge forms.

3 Methodology

Flowchart illustrates the methodology of this work is shown in Fig. 7. Field work of this study includes groundwater sampling and borehole temperature measurement. The measured temperature-depth profiles were classified, and the subsurface temperature was modeled using 1D modeling techniques. The modeled and measured profiles were compared to estimate the vertical groundwater flux. The collected water samples were subjected to hydrochemical analysis and then graphically presented on Piper and Stiff diagrams. The results of subsurface temperature and hydrogeochemistry were integrated to get a complete image about the seawater-freshwater interaction. After that, conclusion and recommendation were written. Subsurface thermal measurements generally are made in the observation wells that are assumed to be in equilibrated thermal state between the water and the surrounding solid material. The Borehole temperature measurement technique was discussed in detail in Salem and El Bayumy [59]. Temperature measurements in this study were carried out in eight observation wells (Tala 2, Kafr Elarab 2, Nawag, Elkarada 2, Abu Mesaaed, Elhadady, Motobes 1, and Motobes 2) (Fig. 8). The equipment used for the measurements was a digital thermistor thermometer (resolution of 0.1°C) attached to a cable of 500 m length. Data were recorded from the water table to the bottom of the borehole.
../images/437178_1_En_253_Chapter/437178_1_En_253_Fig7_HTML.png
Fig. 7

Flowchart illustrates the methodology of the current study

../images/437178_1_En_253_Chapter/437178_1_En_253_Fig8_HTML.png
Fig. 8

Wells location map [29]

Water samples for chemical analysis were collected from all the observation wells (Tala 1 and 2, Kafr Elarab 1 and 2, Nawag, Elkarada 1 and 2, Abu Mesaaed, Elhadady, and Motobes 1 and 2) and two pumping wells in Qutor as well as seawater. The water table, electrical conductivity, TDS, as well as pH were measured in situ. All the chemical analysis of anions and cations were made in central laboratories of National Water Research Centre, Cairo. The hydrochemical analysis procedures were discussed in detail in Salem et al. [60]. Well, location map and the listed hydrochemical analysis are shown in Fig. 8 and Table 1. This chapter is a modified and updated version of Salem et al. [29], where 1D temperature modeling was added and data presentation and interpretation were updated.
Table 1

Chemical composition of the collected water samples [29]

Physicochemical parameters

Tala (1)

Tala (2)

Kafrelarab(1)

Kafrelarab(2)

Nawag

Qutor (1)

Qutor (2)

Elkarada(1)

Elkarada(2)

Kafr Mesaaed

Elhadady

Motobes (1)

Motobes (2)

Seawater

Total depth

m

100

320

100

280

45

45

100

45

75

85

90

140

430

 

Water table

m

3.9

3.9

3.5

3.5

3.0

ND

ND

1.6

0.2

−0.7

−0.5

0.0

−14.5

 

pH

7.85

7.9

9.51

9.32

8.16

8.65

7.1

10.71

8.03

7.47

7.08

5.87

6.69

7.5

TDS

mg/l

514

916

198

280

291

531

741

1,604

1,751

6,220

8,288

19,840

80,192

38,157

EC

mmhos/cm

0.802

1.429

0.31

0.447

0.454

0.845

1.48

2.51

2.74

9.72

12.95

31.00

125.3

59.523

CO3

mg/l

0.0

0.0

17.1

29.6

0.0

ND

ND

57.6

0.0

0.0

0.0

0.0

0.0

ND

HCO3

mg/l

365.9

551.8

53.2

40.7

44.8

122

188

234.0

200.0

23.2

29.0

11.6

174

157.0

Total alkalinity

mg/l

365.9

551.8

70.4

70.2

44.8

ND

ND

291.6

200.0

23.2

29.0

11.6

174

157

Ca

mg/l

71.4

92.3

14.86

17.1

18.6

60.4

54.4

31.85

40.5

293

472

539

2,150

513

Mg

mg/l

33

55.9

11.32

14.87

13.21

29

71.3

26.7

23.4

153

220

672

2,260

1,417

Na

mg/l

38.6

66

30.4

47.6

46.9

60

96

500

523

1,650

1,870

4,650

22,700

11,850

K

mg/l

11.6

19.4

7.42

8.4

7.5

5

5.75

39.7

16.4

26.9

37.2

151

737

220

Cl

mg/l

45.2

78.5

33.5

52.4

64.1

160

292

630

780

3,290

3,820

6,844

43,500

20,800

NO3

mg/l

<0.2

<0.2

<0.2

<0.2

0.91

9.4

ND

46.90

18.70

105.0

111

66

4,283

6.6

SO4

mg/l

57.20

92.40

44.10

57

73.50

67.00

42

68.20

55.40

100.2

520

5,237

908.4

3,260

Note: ND not determined

4 Result and Discussion

The two most illustrative parameters were utilized to be specific, temperature and hydrochemistry. The first of these tools gives data about vertical, rising, and downward flows. The second gives data about the arrangement of the water influenced, which thus gives data about the “palaeoprocesses,” the existence of salts in the medium, and the water origin.

4.1 Subsurface Thermal Regime

4.1.1 One-Dimensional Temperature Profiles

The vertical well logs demonstrate the thermal properties of the aquifers and enable separation of segments with generally contrasting advancements, as far as the presences or non-presence of warm water or streams of an alternate origin [78]. Temperature logs of groundwater additionally enabled the evaluation of the aquifer to be separated point by point, as for the presence of hot or cold water streams.

Regarding the temperature profiles of the present study (Fig. 9), they can be classified into two groups. The first group characterizes the area southern Kafr Elsheikh City, and the second group includes the wells located northern Kafr Elsheikh City. Based on the shape of the profiles, the first group includes recharge- and discharge-type profiles (Tala and Kafrelarab wells, respectively). This means in the study area, the fresh groundwater in the Nile Delta starts to flow southern Tanta City upward. The reason could be related to the existence of denser seawater in deeper zones which forced the light, fresh groundwater to flow upward. Kafr Elarab, Nawag, and Elkarada temperature logs have higher temperatures compared to Tala and Kafrelarab logs which indicate that the groundwater continues to discharge northward. The latter three profiles also have higher temperatures in the shallower zones; as these wells located in a cultivated land out of the urban effect so, such warm zones might be related to upward movement of the warm freshwater and its movement laterally northward in the shallow zones.
../images/437178_1_En_253_Chapter/437178_1_En_253_Fig9_HTML.png
Fig. 9

The measured temperature-depth profiles [29]

The second group of wells characterizes the northern part of Kafr Elsheikh City where Kafr Mesaaed, Elhadady, and Motobes wells (shallow and deep) are located. It has no significant profile type, but it seems different compared to the first group. Elhadady well and the shallow Rasheed well have warm water in their topmost parts. It also means that a discharge of the groundwater took place in the subsurface into the sea northward. For the same depth in the four wells, Kafr Mesaaed well has the lowest temperatures compared to the other three wells. This means that north Kafr Elsheikh City could be another recharge area different from the main recharge area of the Nile Delta aquifer. The later phenomenon happened due to excessive irrigation of highly cultivated rice fields in the northern part of the Nile Delta (Fig. 5). Motobes wells (shallow and deep) show different behavior in the shallow zone (Fig. 9). As the two wells lie close to the shoreline, this temperature difference could be related to the presence of two active groundwater flow systems one deeper and the other shallow.

The measured temperature profiles show three important observations as criteria for seawater intrusion in the Nile Delta. These observations are (1) although the horizontal distance between Nawag and Elkarada is about 30 km, their profiles met each other at a depth of 45 m below the surface; (2) in the same way, the distance between Kafrelarab well and Motobes wells is about 85 km, but the temperature difference at deep zones is less than 1°C; and (3) the profiles of Motobes and Elhadady wells are expected to meet each other at about 90 m below surface. These obtained three observations reflect an inland encroachment of seawater which is probably reached Nawag area in the deeper zones (northern Tanta City).

4.1.2 Temperature Cross Section

The vertical two-dimensional temperature distributions can be studied from combining the vertical logs of temperature made in different piezometers over the study area to show the spatial behavior of subsurface temperature along the fresh and saline groundwater flow direction from south to north (Fig. 10). As shown from the isothermal lines, Nile Delta basin in this study could be classified as mentioned before into two main areas: the area to the south of Kafr Elsheikh City and that northern Kafr Elsheikh City. The southern area shows the regional flow in the delta basin, where the freshwater is recharged south Tala, and it discharges starting from Kafrelarab well (south Tanta City) and flowing laterally in the shallow zones till Kafr Elsheikh (Elkarada wells). The northern area shows a local flow system where the water has recharged in the area lying south of Kafr Mesaaed well, and it discharges in the northern part, where Elhadady and Motobes wells are located. The recharge source of the groundwater in the latter area is expected to be related to agriculture and human activities (Fig. 5). Seawater intrusion could also be seen from Fig. 10 where the isothermal lines are nearly horizontal in the deep zones between Motobes and Nawag.
../images/437178_1_En_253_Chapter/437178_1_En_253_Fig10_HTML.png
Fig. 10

South-north cross section shows the subsurface thermal regime of the study area. Groundwater flow system is also indicated by the white arrows [29]

4.1.3 Model Calculations

To investigate the one-dimensional temperature profiles affected by groundwater flow, the estimation of subsurface temperature in the investigation region has been made utilizing distinctive vertical groundwater fluxes. Carslaw and Jaeger [79] got the mathematical solution for the subsurface temperature utilizing one-dimensional heat conduction-advection equation under the state of linear increment in surface temperature as presented in Eq. 1:

$$ {\displaystyle \begin{array}{ll}T\left(z,t\right)=&amp; {T}_{\mathrm{o}}+{T}_{\mathrm{G}}\left(z- Ut\right)+\left\{\left(b+{T}_{\mathrm{G}}U\right)/2U\right\}\\ {}&amp; \times \left[\left(z+ Ut\right){e}^{Uz/\alpha } \operatorname {erfc}\left\{\left(z+ Ut\right)/2{\left(\alpha t\right)}^{1/2}\right\}+\left( Ut-z\right) \operatorname {erfc}\left\{\left(z- Ut\right)/2{\left(\alpha t\right)}^{1/2}\right\}\right]\end{array}} $$
(1)

where b is the rate of the surface temperature increment; t is the time after semi-equilibrium condition [21]; T o is the surface temperature; T G is the general geothermal gradient; U = υc o ρ o/ in which υ is the vertical groundwater flux, c o ρ o is the heat capacity of the water, and is the heat capacity of the aquifer; α is the heat diffusivity of the aquifer; and erfc is the correlative error function. The modeling is restricted for semi-infinite layers with just vertical conduction and convection, and vertical groundwater flux is thought to be consistent with depth.

Setting T G is 0.19, 0.29, and 0.15°C/m for Tala and Kafrelarab; Nawag, Elkarada, and Kafr Mesaaed; and Motobes (shallow and deep) wells, respectively. T o values is 20.4°C for all wells. The utilized warm diffusivity (α) is 6 × 10−7 m2/s. For t meets 50 years, the utilized estimation of b is 0.05°C. In this manner, Salem and Osman [61] studied the change in the surface air temperature in the Nile Delta and detected warming trend in Damanhour weather station. Temperature depth profiles are modeled utilizing the above condition for various estimations of vertical groundwater motion (U) following Taniguchi et al. [21]. The comparisons between the modeled and the observed profiles are shown in Fig. 11a–c.
../images/437178_1_En_253_Chapter/437178_1_En_253_Fig11_HTML.png
Fig. 11

Binary diagram shows the comparison between the modeled and the measured temperature profile. Groundwater flux (U) values are either positive or negative indicating the downward or the upward movement, respectively

Recharge-type (downward groundwater flux) profile is only detected in Tala well (Fig. 11a) with U value equals 0.8 m/year. The other wells showed a discharge type (upward groundwater flux) with varying rates: −0.1 to −0.5, −0.35, −0.23, −0.6, −1.2, and −2.8 m/year for Kafelarab, Nawag, Elkarada, Kafr Mesaaed, Elhadady, and Motobes wells, respectively, as shown in Fig. 11a–c. These estimates indicate that fresh groundwater starts to move vertical upward southern Tanta City. The most dangerous recorded phenomenon is that seawater intrusion upward flux recorded in Motobes wells is higher than the downward recharge flux calculated in Tala well. Two observations were detected from this model. These observations include (1) the geothermal gradient decreases in the areas influenced by seawater intrusion like Motobes wells, and (2) recognized temperature profiles curvature in the shallow zones in most of the wells cannot be modeled by the current model because the groundwater flow is complicated in the shallow zones to be solved by this simple model.

4.2 Hydrogeochemistry

The hydrochemical compositions of the collected samples are variable (Table 1). In all groundwater samples, EC and TDS are increased with depth and from south to north. Deeper freshwater in Tala 2 and Kafrelarab 2 wells have higher EC and TDS compared to the shallow wells of the same areas. This phenomenon could be related to long time of water-rock interaction in deeper zones compared to the shallow ones. TDS increases from Nawag to Motobes wells in lateral and vertical directions. The most recognized fact is that groundwater of Motobes 1 (shallow) has TDS around 20,000 mg/l, while Motobes 2 (deep) has TDS value around 80,000 mg/l. That means the Nile Delta groundwater close to the Mediterranean Sea is affected by seawater intrusion in the shallow zones but is of brine hypersaline origin in the deeper zones. Shallow well values of pH are also variables ranging from 5.87 in Motobes 1 to 10.71 in Elkarada 1. According to Inskeep and Bloom [80] and Herman and Lorah [81], high pH values are related to calcite precipitation which occurs at pH > 8 and PCO2 < 0.1 atom. The mechanism of precipitation reaction is shown in Eq. 2:

$$ {\mathrm{Ca}}^{2+}+{{\mathrm{H}\mathrm{CO}}_3}^{-}+{\mathrm{OH}}^{-}\to {\mathrm{Ca}\mathrm{CO}}_3+{\mathrm{H}}_2\mathrm{O} $$
(2)

This reaction could happen in Kafrelarab and Nawag which contain the very low Ca2+ and HCO3 concentrations in contrast to Tala and Qutor. Presences of nitrates in groundwater usually indicate an infiltration of water as a result of agricultural and human activities. Groundwater from the area south Nawag has very low nitrate concentrations which means that clay layer of Bilqas Formation and the irrigation system do not lead to an excess of nitrate to reach groundwater. The wells located northern Kafr Elsheikh City have high nitrate concentration marking a great contribution from the irrigation system where those areas are highly cultivated with rice as shown in Fig. 5.

Two main chemical types are recognized in Piper diagram (Fig. 12). The first type is the freshwater samples from Tala wells (1 and 2), and the second type is represented as seawater type including seawater sample, Kafr Mesaaed, Elhadady, and Motobes (1 and 2) wells. The later samples are shifted up from the seawater sample in the direction of seawater intrusion proving primary salinity characters. Elkarada samples (1 and 2) show a little downward shift from the seawater sample as a result of a little replacement of Ca by Na forming NaHCO3-type water due to receiving calcium bicarbonate-rich water by surface infiltration. Although, Kafrelarab (1 and 2) and Nawag wells are freshwater but shifted toward seawater type. This could be related to loosing Ca and carbonates as calcite precipitation due to higher pH values. Qutor samples (1 and 2), on the other hand, are located toward seawater intrusion type as a result of ion exchange between Na of seawater with Ca and Mg from the aquifer material.
../images/437178_1_En_253_Chapter/437178_1_En_253_Fig12_HTML.png
Fig. 12

Piper diagram illustrates the hydrochemical processes [29]

To be able to understand the interaction between seawater intrusion and freshwater in Nile Delta aquifer, ion exchange processes should be explained. Appelo and Postma [82] explained the ion exchange processes. The infiltrated freshwater is dominant by Ca2+ and HCO3 ions. Cation exchangers in the aquifer, therefore, have mostly Ca2+ ions absorbed on the surfaces. In the intruded seawater, Na+ and Cl are the dominant ions, and the sediment in contact with seawater will have absorbed Na+ for the large part. When seawater intrudes in a coastal freshwater aquifer, an exchange of cations takes place according to Eq. 3:

$$ {\mathrm{Na}}^{+}+1/2\ \mathrm{Ca}\hbox{-} {\mathrm{X}}_2\to \mathrm{Na}\hbox{-} \mathrm{X}+1/2\ {\mathrm{Ca}}^{+2} $$
(3)

where X indicates the soil exchanger. Sodium is taken by the exchanger and Ca2+ released. The water type then changes from NaCl into a CaCl2 water type. The reverse process takes place with freshening, i.e., when freshwater flushes a salt water aquifer. Where Ca2+ is taken up from water in return for Na+ with a NaHCO3 water type as a result. Water chemistry can thus indicate upcoming of seawater or, conversely, that saltwater is flushed by freshwater. These two processes could happen completely or partially.

Figure 13 showed the vertical two-dimensional distribution of water chemistry along the south-north direction using Stiff diagrams. Regardless of the ionic concentrations, the present groundwater could be classified into three main types. The first one is recorded as Ca-Mg (HCO3)2 freshwater type in Tala wells. The second type has NaCl composition and including all the rest of the wells except Qutor wells. The third hydrochemical type encountered in Qutor wells is of CaCl2 in well 1 (at shallow depths) and MgCl2 type in well 2 (at deeper depths). The minor details of Stiff diagrams reflect the hydrochemical processes that happened in the Quaternary aquifer along the flow path as follows:
  1. 1.

    Kafrelarab and Nawag waters are freshwater of NaCl type; this may happen as a result of calcite precipitation during upward movement of calcium carbonate type in Tala due to an increase of pH.

     
  2. 2.

    Compared to the Motobes 1 which shows a typical seawater type, Kafr Mesaaed, Elhadady, and Motobes 2 glasses of water have a relative increase in Ca2+ and lower concentration of Na+ than Cl indicating seawater intrusion as mentioned by Appelo and Postma [82]. In the other hand, Elkarada samples show a relative increase in HCO3 , and Na+ is slightly greater than Cl. This could be related to freshwater infiltration from cultivated land.

     
  3. 3.

    In the same way, composition of Qutor groundwater (MgCl2 type in deeper zones and CaCl2 in shallow zones) could be explained as an indicator for seawater intrusion, and it might be located in the intermediate zone between the regional fresh groundwater and the intruded seawater in the Quaternary aquifer of the Nile Delta.

     
../images/437178_1_En_253_Chapter/437178_1_En_253_Fig13_HTML.png
Fig. 13

Stiff diagram shows the difference in the chemical types of the collected samples. (1) and (2) indicate the shallow and the deep groundwater at each site. (a) is south–north cross section and illustrates the average depth of the screens where the groundwater was sampled; color of the screen dots is the same of the related stiff diagrams

5 Conclusions

The integrated tracer technique between subsurface temperature and water chemistry was a good technique for tracing the groundwater flow system and seawater in the Nile Delta Quaternary aquifer. Two flow systems were detected: one is regional, and the other is local, and both of them are affected by seawater intrusion. The regional one is recharged south Tala and discharges northward in the area from Kafrelarab (south Tanta City) till ElKarada (south Kafr Elsheik City. The shape of the temperature profiles reflects a recharge type in Tala and discharge type in Kafrelarab, and the shape of the isothermal lines indicates a deep warm water flows upward to the shallow zones. Water chemistry is supporting this idea where the recharge area is characterized by CaHCO3 water type in Tala. Such water type is indicating an inland recharged freshwater. The discharge area of the system is characterized by NaCl type which is the most common water type due to calcite precipitation during upward flowing due to the change in pH. Recharge-type profile detected in Tala well has a groundwater flux (U) value equals 0.8 m/year. The other wells in this regional groundwater flow system showed a discharge type (upward groundwater flux) with varied U values: −0.1 to −0.5, −0.35, and −0.23 in Kafelarab, Nawag, and Elkarada wells, respectively. These estimates indicate that fresh groundwater starts to move vertically upward southern Tanta City.

Very low concentrations of nitrates in the groundwater of Tala and Kafrelarab and Nawag wells provide another evidence for being the water in the area south Nawag is contributed by deep circulating water. The shallow zones between Kafrelarab and Elkarada show a warming behavior as indicated from the temperature profiles. As those wells are not located in the urban areas, therefore, this thermal regime could be related to an upward and northward groundwater discharge.

The local flow system is identified from the cooling of the isothermal lines that are recognized in the area from north Kafr Elsheik till Motobes. This system is isolated from the regional one. The maximum depth of this local system is about 140 m in Kafr Mesaaed and 90 m in Elhadady. The recorded excess in nitrate concentrations indicates the agricultural and human surface activities. The temperature profiles measured in the area of this local system are of discharge value with upward flux (U) equals −0.6, −1.2, and −2.8 m/year for Kafr Mesaaed, Elhadady, and Motobes wells, respectively.

The subsurface thermal regime and the hydrogeochemical data in the study area gave an image about the spatial extend of seawater intrusion in the Nile Delta. The estimated criteria for seawater intrusion from the subsurface temperature are as follows:
  1. 1.

    The low-temperature difference between deep wells of Kafrelarab and Motobes wells in the deeper zones despite the long distance between them indicates nearly static seawater in the mentioned area.

     
  2. 2.

    The intersections between the temperature profiles. For example, Nawag and Elkarada profiles intersect at 45 m deep; Kafr Mesaaed and Elhadady profiles met Motobes profiles at 140 m and 90 m deep, respectively.

     
  3. 3.

    The nearly horizontal isothermal lines under most of the studied area in the deeper zones (till Nawag).

     
  4. 4.

    The low thermal gradient noticed in Motobes wells which located close to the Mediterranean Sea compared to other wells.

     

The abovementioned thermal criteria indicate the existence of a nearly static water body deeper in south Kafr Elsheikh City and shallower northern Kafr Elsheikh City which might be seawater intrusion.

Recognized hydrogeochemical criteria indicating seawater intrusion in the study area are as follows:
  1. 1.

    Groundwater in Qutor wells has both CaCl2 and MgCl2 water types, in the shallow and deeper zones, respectively. This indicates the existence of an intermediate zone between seawater and freshwater under Qutor area. This estimation is limited to 100 m depth which is the maximum depth of Qutor wells.

     
  2. 2.

    Seawater intrusion was also indicated from the partial ion exchange observed as a relative increase in Ca2+ and fewer Na+ than Cl as in Kafr Mesaaed and Elhadady wells.

     
  3. 3.

    In contrast to the regional groundwater flow system of the southern part and the local flow system in the northern part, Elkarada samples show a relative increase of HCO3 and higher Na+ compared to Cl. Such feature indicates partial ion exchange during freshwater propagation into the seawater. In another meaning, the groundwater under Elkarada is originally seawater mixed with a little of infiltrated freshwater.

     

Finally, the most dangerous information given from this research is that the groundwater in the Nile Delta is not affected only by seawater intrusion but also affected by hypersaline brine water inland propagation. The seawater affects the upper 200–250 m, but the hypersaline water was indicated in Motobes well which is of 420 m depth. The upward seawater intrusion flux rates in Motobes wells was 2.8 m/year which is much higher than the groundwater recharge flux at Tala well which was 0.8 m/day.

6 Recommendations

Governmental regulation of the pumping process from the Quaternary Nile Delta aquifer is urgently needed as the groundwater is not just facing the pollution from surface human activities and seawater intrusion but also suffers from the deeper hypersaline brine groundwater inland propagation. Integration between subsurface thermal and hydrogeochemical data is a good tool in recognition of the groundwater flow system and seawater intrusion.

Acknowledgment

This chapter is an update and revised version of the first draft which was presented at the 5th International Symposium on Geophysics, 2008, Tanta, Egypt. Also, the authors thank the editor Prof. Dr. Abdelazim Negm and the reviewers for their constructive remarks and comments.