2D and 3D visualization of aquifer sediments, surface water seepage and groundwater flow using DC-resistivity, DC-IP, and SP methods, West El-Minia, Egypt

In this study, the SP, DC-Resistivity and DC-TDIP methods were used and included 21 sites for determining the groundwater conditions and visualizing the aquifer sediments, surface water seepage, and groundwater flow in 2D and 3D. The results indicated the occurrence of Quaternary and Tertiary deposits, as well as shallow aquifer of Eocene fractured limestone and Quaternary deposits. The resistivity, chargeability and SP values were compared with each other and 2D and 3D visualization models were designed for these values to understand the sediment distribution and its geological and hydrogeological implications. High to medium, low and very low chargeability values indicated the compact sediments with high shale content, water-bearing layers and freshwater zones and dry sediments, respectively, and varied with the water amount, salinity and effective porosity. Low to very low values of SP indicated the occurrence and high accumulation of groundwater and an increase in shale, permeability and groundwater flow and source, and possibly salinity. The results showed that the resistivity, chargeability and SP values had an important role in separating the dry and saturated sediments and defining the fault zone, determining the source of surface water seepage, classifying the depths of high, medium and low groundwater accumulation and flow and permeability. The IP depended highly on the water amount and shale content and it was interesting in groundwater exploration with resistivity and in differentiating between groundwater and clays. The SP assisted in delineating the surface water seepage and groundwater source and flow direction, and predicting the sediments permeability.


Introduction
The DC-resistivity method and DC-TDIP method are widely used in hydro-geophysical investigations for obtaining highresolution information about the geological structure and the geological sediments of the subsurface. The structural findings are interesting in defining the units of hydro-stratigraphic, and the locations of major and minor fractures. They are also interested in estimating the electrical properties of the geological layers to distinguish rock types and their hydrogeological characteristics. Vanhala et al. (1992) and Weller et al. (1999) stated that IP method can differentiate the lithology of the host material from the resistivity method.
The geoelectrical methods used in this study include the SP method, DC-resistivity method, and DC-TDIP method. These methods aim to explore and determine the groundwater conditions in the study area, especially at shallow depths.

Topography and geomorphology
In general, the topography in the study area is heterogeneous due to the complex geological and structural conditions in the region because it is located between the Nile Valley and the western limestone plateau. Topographic map covering the area and surrounding area using digital elevation mode. The terrain ranges from 25 m (+ msl) in the east to ~ 225 m in the west, where the limestone plateau is located (Fig. 1,  right panel). The main geomorphological units in the study area are the limestone plateau to the west, old alluvial plains, young alluvial plains (Fig. 1, right panel), sand dunes, fanglomerates (limited extension in the western side), as well as surface drainage channels representing valleys (Said 1981). The limestone plateau is mainly composed of limestone and is dissected by faults, as shown in the surface geological map (Fig. 2, left panel) and surface drainage systems, as shown in the GIS imaging map (Fig. 1, left panel), which originated from rainfall and flooding, which are mostly parallel to subparallel and have directions in the WNW-ESE and NNW-SSE direction (Hassan et al. 1978;Ghareib 1990).
The old alluvial plains lie between the slope of the Eocene limestone plateau to the west and the cultivated lands, as shown in Fig. 1, left panel and right panel. These plains appear as wadi deposits and sediments and their general slope is from west to east at about 5 m/km. The young plains consist of layers of clay and silt, and the stratified sandy units are covered with gravel and form the cultivated lands. While sand dunes are found in belts of different shapes and are of the migratory type, as evidenced by the surface geological map (Fig. 2, left panel) (Abd El Mageed KSZ 1991).

Geology
Sediments composed of sand, graded gravel and silt along with rock fragments such as wadi deposits, limestone, limestone and sand dunes are the main components of the surface geology covering the surface of the study area. The age of these sediments ranges from the middle Eocene to the Quaternary as shown in the surface geological map (Fig. 2, left panel) (Said 1990). Quaternary deposits were represented in silts and clays from the Holocene sediments, about 3-12 m thick to the east of the study area. Pleistocene sediments consist of sand and gravel and are more pronounced in the west, north, and southwest of the study area with sanddivided limestone deposits. Its thickness ranges from 49 to 150 m and consists mostly of sand, gravel, silt and clay, then the western limestone plateau of the Rayan formation (Eocene limestone).
As for the subterranean deposits of the Tertiary age, they are divided into deposits of the Eocene and Pliocene periods. The Eocene sediments are divided into the upper and middle Eocene sediments with a thickness of ~ 285 m (Tamer et al. 1974). The middle Eocene sediments consist of the Rayan formation with a thickness of ~ 30 m. It is a limestone intercalated with shale, sandy shale and marl, then the Beni Suef formation consisting of limestone, marl and shale. The upper Eocene deposits include limestone and shale Maadi formations with a thickness of about 85 m. Yousif et al. (2018) concluded that the exposed rocks along the area are mapped to ages varying from Lower Eocene to Oligocene.
The following rock units are recognized from the base to top: El-Minia formation, limestone to massive limestone with shale intercalations up to 36 m thick, Samalout formation, gray shale and claystone with white numilitic limestone at shallow depths, then turning into fractured limestone in the lower parts with a thickness of about 40 m-107 m (Said 1981), the Qatrani formation, sandstone-gravel, and finally the Katkut formation composed of a series of clastic sediments (siltstone and claystone), limestone and gravel fragments, respectively. While the Pliocene deposits in the Kom Al-Shelul formation consist of a layer of sandstone topped by deposits of limestone with sand, clay and conglomerates. The formation of Wadi El Rayan (shallow limestone with shale intercalation and sandy shale) covers the Samalut formation and then the Minia formation (Said 1981). Yousif et al. (2018) stated that the general lineaments are North East-South West followed by the North West-South East directions. The structural lineaments are intensified to the west and northwest, and there are many faults and fractures lineament trends that affected the study area. Where the main lineament trend is NE-SW followed by NW-SE directions.

Hydrogeology
In general, the previous geological and topographical changes played a large role in shaping the subsurface hydrogeological conditions of the area. From the hydrogeological map ( Fig. 2, right panel), there are two main shallow aquifers, the fractured to semi-fractured limestone aquifer to the W direction and the Quaternary aquifer to the E direction of the area. Fractured limestone aquifer rocks trace the Eocene age; It consists mainly of limestone and marl, with intercalations of shale. This aquifer is characterized by heterogeneity in terms of hydrogeology and varies in the quantity and quality of water from one place to another. This varies depending on the types and density of fractures. Therefore, this aquifer is a moderately productive aquifer with small, medium to highly saline thickness of its water (RIGW 2015).
Al Temamy and Abu Risha (2016) report that there is vertical leakage from the lower Nubian sandstone aquifer and horizontal leakage from Nile river waters in the east into the Eocene aquifer in the study area through Quaternary sediments. In general, groundwater in the area is found in  Klitszch et al. 1987) and regional hydrogeological map (right panel) (RIGW 2015) of the study area a shallow fraction of Quaternary sediments and fractured Eocene limestone formation, and the resistivity values of the aquifer increase with depth due to the decrease in fracture density (Mahmoud and Kotb 2017). The shallow fractured limestone in the Samalut formation is one of the most important sources of the aquifer, especially west of El-Minia and Assiut, and its water quality is good (Shabana 2010;RIGW 2015;Al Temamy and Abu Risha 2016). The Quaternary aquifer consists of gravel, sand and clay intercalation, and it is considered a moderate and limited productive aquifer (RIGW 2015). It covers the limestone aquifer in the eastern parts of the study area (Fig. 2, right panel) and as a very thin dry layer in the west where the limestone plateau is.
There are three shallow drilled wells ~ 7 km south of the study area. The total drilling depths of these wells range from 24 to130 m and their discharge rate ranges from 50 to 100 m 3 /hour. The groundwater TDS ranges for these wells range from 812 to 4243 ppm and reach 5261 ppm in some locations, especially in wells more than 90 m deep. There are also two wells, well (1) located ~ 1.8 km NE, with a total depth of 150 m and consisting of 15 m of dry sand and gravel (Quaternary sediments), 30 m of dry limestone, 20 m of dry marly limestone, 35 m of clay and then the saturated fractured limestone (Samalut Formation, shallow aquifer) that was recorded at a depth of 100 m (Fig. 3, left panel). Well (2) is located 2.14 km east of the study area with a total depth of ~ 130 m and consists of 5 m dry sand and gravel, 10 m wet clay sand, 20 m wet clay, 20 m saturated clay sand (Quaternary shallow aquifer), 32 m of wet clay and then saturated fractured limestone at a depth of 97 m (Fig. 3, middle panel). The salinity measured in these wells ranges from 2560 to 4243 ppm. Accordingly, the limestone aquifer is under confined conditions.

Geoelectric methods
Research Institute for Groundwater (RIGW) staff implemented field measurements using the SYSCAl-R2 resistivity meter and application of Schlumberger array. In this study, 21 VES, IP, and SP sites were measured as shown in Fig. 3, right panel. Apparent resistivity data were measured (Ω.m) of 7 Vertical Electrical Soundings (VESes), with maximum current electrodes spacing (C 1 C 2 /2), ranging between 500 and 700 m. Also, the SP (mV) of the 7 SP stations and the chargeability (M) (mV/V %) of the 7 TDIPs were measured, using the same previous spacing. These measured soundings are distributed in Fig. 3, right panel. Most of orientations of VES1, VES2, VES3, VES6, and VES7 are parallel to the western limestone plateau (N-S direction) except VES4 and VES5 are perpendicular to the limestone plateau (W-E direction). The measured raw data of these soundings are shown in curves (Fig. 4) referring to the highly conductive zone which will be of interest in aquifer determination and in understanding the geologic and the hydrogeologic conditions of the subsurface sediments.

Depth (m)
We ll

Spontaneous polarization potential (SP) method and interpretation
The SP method is negative, i.e., the differences in the earth's natural potential between two poles at the earth's surface are measured. These measured potentials range from < millivolts (mV) to > 1 Volts. The + and − signs of the potential are an interesting factor in explaining and interpreting the SP anomaly. The intrinsic potential is generated by a number of natural sources such as electrochemical reactions in the ground and does not require the injection of electric currents into the ground as is the case in the IP and resistivity The measured raw SP curves (mV) (light blue curves), normal chargeability curves (mV/V%) (dark red curves), and apparent resistivity curves (Ω.m) (black curves) of the seven SP, seven TDIPs, and seven VESes (1 (A), 2 (B), 3 (C), 4 (D), 5 (E), 6 (F) and 7 (G)) along the study area with referring to the highly conductive zone methods and unlike the IP method. The SP method was used widely in studies of groundwater flow. In general, groundwater is a common factor in the generation of SP by the flow of water and by the interaction of water as a solvent and as an electrolyte for several metals. Accordingly, when the source of the SP is groundwater movement, the type of anomaly is + or − up to hundreds of mV. The main factor that influences and helps in rock potential is electrical conductivity/resistivity. The electrical conductivity that aids in this parameter depends on the porosity and movement of water in the pores, ion mobility, solution concentrations, bulk water chemistry, and the water/mineral interface (Lange and Kilty 1991; Aubert and Atangana 1996;Stevanovic and Dragisic 1998;Fagerlund and Heinson 2003). In this study, there is a seepage from surface reservoirs (Fig. 5) and recharge from the adjacent Quaternary aquifer from the eastern direction. SP fluctuations and signals are used to depict subsurface resistivity discrepancies before any current is injected (Revil et al. 2012). − SP values indicate an increased content of clay and water.
SP values were measured in this study when the distance between the potential electrodes (P 1 P 2 ) is stable at 1 m, 10 m and 50 m with changing C 1 C 2 electrodes. Accordingly, SP measurements were implemented through one side of the used Schlumberger array. Therefore, the resulting curves are in three segments, as shown in Fig. 4. The anomalies of SP are qualitatively explained by the shape of profile, polarity (+ or −), amplitude or contour style. The data were plotted against the electrode spacing and compared with the resistivity and chargeability data (Fig. 4), and these data were then plotted for comparison with the chargeability using the calculated depths (Fig. 7). Therefore, the SP data will be interpreted qualitatively based on the − or + SP anomalies to infer the expected locations of surface water seepage and determine the direction of groundwater flow into the aquifer.

DC-resistivity method and interpretation
Generally, when clay is present with other geological materials, the resistivity becomes low and very low if the salinity value is high. This method is used for measuring the resistivity (apparent) of the shallow and deep geological layers as an EC (DC) is passed through them. In this study, the quantitative interpretation of vertical electric sounding curves (7 VESCes) was based on Zohdy and Bisdorf (1989) technique for the primary model and Rinvert (1999) software for carrying out the forward model and inverse model. This interpretation was used to determine the true Earth's resistivity and the thickness of the recorded subsurface geological layers. These parameters will help in determining the geological, structural and hydrogeological settings, characteristics of dry and water-bearing strata with depths. From the inversion of these VESes, the forward model and inverse model ( Fig. 6) for VES4, were obtained. The rocks encountered within the aquifer and dry strata, the geological and hydrogeological setting of these strata along the area can be detected and identified. Therefore, all the previous findings may reflect the changes neither vertical nor lateral in the rock formation of the penetrated section by the EC. In this study, from the resistivity contrasts between subsurface sedimentary rocks, a distinction can be made between dry, tight and saturated sedimentary layers.

Time-domain induced polarization method and interpretation
In general, it is assumed that the polarization increases with decreasing the porosity as the pores compress or not exist like to the west of the study area due to the occurrence of the hard limestone. The IP response may depend on the water amount and the capacity of cation exchange (CEC) with respect to the clay minerals as indicated from some of the laboratory results (Cohen 1981). Accordingly, this method is interesting in hydrogeological and geological investigations. The phenomenon of polarization can result from interactions between clay minerals and groundwater and also a decrease in resistivity (Cohen 1981). Therefore, the chargeability is high and the resistivity is low in saturated clayey deposits such as fractured shaly limestone.
Generally, this method is used to measure the chargeability of geological materials as a potential due to a power (electric current) outage called induced polarization potential. The decay rate of this potential depends on the sediment and pore geometry and the percentage of water saturation in the pores. Therefore, clay and shale layers show higher charging capacity (chargeability) than sandy or pure sediments (Seidel and Lange 2007). Vacquier et al. (1957) found that the chargeability of the mixtures of sand-clay is proportional to the clay properties. Keller and Frischknecht Fig. 5 One of the surface water reservoirs in the eastern sites of the study area and exactly next to VES3, SPs3 and TDIPs3 (~ 15 m), where the water source is the Nile River water channels in the far eastern sites which were via pipes to the area to irrigate the cultivated lands (1966) found that a peak polarization response is due to different types of clays and is depending on their concentrations. Accordingly, the difference in the polarization intensity indicates that the content of clay changes from place to another in the sediments. IP is weak in low CEC compacted clays, and strong in deposits with high CEC of clay particles on the surface of the large grains or host sediments or rocks such as shaly limestone. Parkhomenko (1971) indicated that higher or lower clay contents will result in lower IP and increase IP as content of groundwater increases until the saturation level is reached after which IP decreases. Ogilvy and Kuzmina (1972) found that a high IP response occurs for optimal water content in shaly sands. Draskovits and Smith (1990) noticed that dry sand and gravel had low polarizability and high resistivity as in shallow surface and dry sediments along the study area and that max polarization was associated with clay sediments with medium apparent resistivity while clay deposits with low values of resistivity and polarization were indicated as it seems to east of the area where the Quaternary sediments. Accordingly, the IP method is used to predict clay content in rocks and host aquifers.
The TDIP process measures the voltage decay caused by turning off the excited current oscillation again in one or several intervals of time windows. The decay response is used to study the IP (chargeability) of sediments. This chargeability is determined as the secondary potential to the primary potential ratio of the sent current. The magnitude of the chargeability is affected by the spaces of the time gate and the duration of the current pulses on and off times (Seigel 1959;Magnusson et al. 2010;Sumner 2012). The decay curve is a TDIP target because it is a property of subsurface materials according to the first magnitude, slope, and relaxation time. The V ip signal is usually integrated over n time windows or gates, for the computation of chargeability (M).
In this study, the measured chargeability (IP effect) is raw chargeability (MR) and considered as normalized chargeability (MN), due to the used period is 2000 ms and it is expressed, as shown in the following formula (1) (Schön 2015;Slater and Lesmes 2002): where; V DC [V] is the potential which is used to calculate the DC-resistivity, V ip : the intrinsic or secondary potential [mV], and t i and t i+1 : the open and close times [s] for the gate over which signal is integrated.

Interpretation
Interpretation of the IP can be qualitative using the observed data ( Fig. 7) and is considered more complex than resistivity or by inversion using iterative algorithms similar to resistivity. This inversion was implemented using Zohdy and Bisdorf (1999) technique in the form of calculated data, as shown in Fig. 7, compared to the observed SP data and Fig. 8 compared to the calculated resistivity data. The IP interpretation of subsurface geological layers is complicated by the heterogeneity of sediments and rocks at a small scale. Moreover, the mineral structure, chemical environment, texture and many other factors demonstrate the difficulty of polarization mechanisms (Revil and Florsch 2010;Weller et al. 2013;Revil et al. 2015). Therefore, the sediments of high clay content have high values of chargeability. The tight/hard formations may have chargeability when they include high content of shale or clay. Sediments that have a uniform sand and gravel particle size also produce less chargeability (Alabi et al. 2010). Variation in natural chargeability determines the potential of permeable rocks that contain or transport groundwater from pure deposits (non-shaly deposits). Accordingly, the measured and calculated chargeability values for all TDIPs with depth were used to separate shale/clay and non-clay/rock sediments, such as shale limestone (fractured, tight or compact, and pure limestone), clayey sand or sandy clay, and sand and gravel along this area. This separation of these deposits helps in understanding the geological sedimentation system, the hydrogeological framework, and the interaction between the different aquifers if they are established. Roy and Elliot (1980) observed that the simultaneous decrease in resistivity and IP response is indicative of saltwater zones but only decrease in resistivity is indicative of freshwater zones. Accordingly, as in the resistivity and chargeability field curves (Fig. 4) especially in the highly conductive region and from the comparison of the calculated resistivity and chargeability curves, IP helped in confirming the occurrence of the aquifer and this aquifer has an average saline water content and includes shale/clay. Keller and Frischknecht (1966) and Ogilvy and Kuzmina (1972) reported that, IP is low in clean gravel or pure shale but reaches a maximum value in some intermediate grain sizes, from conducting some studies on the effect of grain size parameters on the IP response. Kiberu (2002) reported that IP response relies on content of groundwater and clay, capacity of cation exchange and that an IP response is possible when soils contain some water rather than completely dry out and are highly dependent on water amount. Accordingly, the chargeability is high against the water-bearing layers as shown in Fig. 4. Barker (1990) reported that the chargeability did not exceed 500 Ms due to the low salinity, according to which the chargeability will increase with increasing salinity of the groundwater up to 500 ppm. Leroux et al. (2010) stated that high normalized chargeability mainly appeared in saturated or semi-saturated materials and that the composition and salinity of the electrolyte (groundwater) have an effect on the mode of occurrence of this max, as shown in Fig. 4. The high IP value decreased and shifted slightly towards low water content (0.2 < Sw% < 0.6) for clean sands (Iliceto et al. 1982) and almost linearly and was approaching zero at very small saturation (Fraser and Ward 1965).
In general, the conductivity of saturated pores will rise because the concentration of ions. The water conductivity in pores increases with clay ionization and surface conductivity (Keller and Frischknecht 1966). Therefore, the host soil conductivity will increase. Due to the relationship between the chargeability and the water content within the pores, the conductivity depends on the availability of ions for ion exchange, which leads to an IP response. Accordingly, if the water conductivity decreases, the mobile ions will decrease due to their close attachment to the clay surface and the IP response will also decrease (Keller and Frischknecht 1966). This would be useful in predicting TDS of a type where the chargeability decreases or increases in the groundwater zone, and the TDS is expected to be both low and high.
Valeriya and Leonie (2011) observed that chargeability is based on the applied electric current magnitude and increases when the electric current is reduced. This case occurs in sediments including pores with different surface areas. The potential difference amplitude (voltage) of these sediments depends not only on fluids, porosity, and tortuosity. Accordingly, the observed chargeability is proportional to the porosity. Accordingly, the chargeability is higher for disseminated shale/clay, such as shaly limestone, than for massive shale/clay and depends on the clay and mineral particle concentration. Also, the chargeability value increases with decreasing porosity as in the case of tight limestone and fractured shaly limestone or when the effective porosity is very low such as in the presence of clay or shale and varies with the amount of water in the pores. The chargeability response also depends on the current strength, which increases with increasing water salinity, clay content, and frequency of current.

Comparison of the observed and calculated chargeability and SP data
From the comparison of the spontaneous polarization potentials and the observed and calculated chargeability values with depth, as shown in Fig. 7, the following was found; An increase in SP values or positive ( +) SP indicates a decrease in the shale deposits and increase in clean deposits while decreasing SP values or a negative (-) SP indicates an increase in shale deposits and deposits of high amount of water. These indicators can be seen when the SP readings are opposite to the IP readings at some depths. Also, chargeability values with depth helped predict the increase or decrease in shale content with depth. Depths with low SP (-) values indicate low to medium clay and shale content with saturated clean sediment and increased effective porosity and groundwater flow. When there is a slight increase in SP may reflect a decrease in water salinity in which case it deviates in the same direction (not opposite) of the chargeability readings as for IPs3, IPs5 and IPs7. In Fig. 7, there is a layer of shale or clay in the upper depths and shale or clay intercalation in the lower depths. This may indicate the occurrence of clay of Quaternary with sand in the layer or intercalations at shallow depths and the occurrence of shaly limestone at deeper depths.

Comparison between calculated apparent resistivity and calculated chargeability
The calculated data for either resistivity or chargeability were implemented using Zohdy and Bisdorf (1999) technique (Fig. 8). The resistivity and chargeability data were compared with each other to separate shale/clay from pure sediment and high-water content zone, in addition to clarifying the expected aquifer zone, determining the depth of groundwater and assessing the expected salinity. Accordingly, this comparison indicated the presence of an aquifer variable in shale content and salinity in addition to the depth of groundwater corresponding to the depth of groundwater measured in the field.

2D visualizing profiles of the true resistivity, calculated chargeability and SP
The two-dimensional (2D) visualizing profiles of the true resistivity, chargeability and SP were mapped, as shown in Figs. 9 and 10, to show the electrical parameters of the subsurface recording layers and their geological and hydrogeological implications. These shallow geoelectrical studies demonstrated the thickness of surface sediments with high electrical conductivity, which includes and studies the shallow (upper) aquifer. These profiles showed the depths of the shaly and non shaly sediments, the main water-bearing layers, the depth of the groundwater, the qualitatively expected salinity, the direction of the expected groundwater flow, and the sources of surface seepage.
The 2D hydro-resistivity profiles (Figs. 9 and 10) showed that the shallow subsurface geological layers were divided into four different geoelectric layers in their true electrical resistivity values and sediments thickness. Where, in general, their sediments are composed of calcareous sand and gravel with clay intercalation as surface sediments, hard limestone to marly limestone, clay, sand with clay intercalation to fractured limestone with shale intercalation (shallow aquifer) and semi fractured to dense limestone. The details of the electrical properties of these sedimentary sequences are as follows: -The first (surface) geoelectric layer (L1): It varies in composition from one place to another. It consists of sand and gravel with intercalations of clay and silts. Its electrical resistivity ranges from 12.5 Ω.m, when clay increases, and 2000 Ω.m. Its thickness ranges from 1 to 3.5 m. It in other places, especially in the northwest, changes into calcareous limestone with sand with an electrical resistivity ranging from 44 to 670 Ω.m and thickness ranges from 1 to19 m. This layer covers the hard limestone and marl layer, which range its resistivity between 6 Ω.m, when increasing the marl, and 738 Ω.m with thickness of 3-83 m in the rest of the area.
-The second geoelectric layer (L2): is a highly conductive layer because it consists of wet clay that has a resistivity ranging between 0.5 Ω.m and 1.7 Ω.m. Its thickness ranging from 21 to104 m. -The third geoelectric layer (L3): It is also a highly conductive layer because it has a resistivity ranging from 15 to 75.5 Ω.m. It is expected to be made up of fractured limestone with intercalations of the shale to the west and transformed into clayey sand to sandy clay in the east. This layer represents a shallow groundwater aquifer and its thickness ranges from 34 to 188.5 m. -The fourth geoelectric layer (L4): it has a resistivity ranging between 174 Ω.m and 2402 Ω.m. It is expected to consist of semi-fractured limestone with shale intercalation to dense limestone. It is considered an extension of the aquifer above but is not at all sites. The depth to the top of this layer is between 124 and 194 m.

3D visualizing models of the resistivity, chargeability and spontaneous polarization
These visualizing profiles display the distribution of these physical parameters of the subsurface materials in 3D to distinguish between these materials and their geological and hydrogeological properties. These characteristics include lithologic description, structural identification, dry and saturated strata, as well as determination of the location of seepage and expected groundwater flow. Therefore, this model is designed using the interpreted resistivity values to show the appearing and distribution of geologic materials whether they are dry or saturated as well as the expected error.

3D visualizing model of the true resistivity
Generally, the resulted max value of resistivity was 750 Ω.m while the minimum value was 1.5 Ω.m along the study area. The 3D visualizing models of the values from 750 to 200 Ω.m, from 200 to 100 Ω.m, from 100 to 60 Ω.m, from 60 to 30 Ω.m, from 30 to 10 Ω.m and from 10 to 1.5 Ω.m are shown in Figs. 11, 12, 13. Each of these ranges of resistivity values are referring to appear and distribute of different subsurface dry or saturated sediments, as well as shaly and non-shaly sediments along the area of study as follows: -The resistivity ranges from 750 to 200 Ω.m refer to the appearance and distribution of the dense limestone especially to the west at shallow and deep depths (Fig. 11, left panel). Whereas the ranges from 200 to 100 Ω.m indicate the appearing of dense to semi-fractured limestone with shale intercalation, as shown in Fig. 11, right panel, also to the west, at shallow depths, to the east at deep depths and appeared as wedge to the east at shallow depths. These values showed a fault at the deep depths in the limestone rock with its general down-throw direction towards the east where the Nile valley deposits. -The resistivity ranges from 100 to 60 Ω.m refer to the appearing and distribution of the fractured limestone aquifer (Fig. 12, left panel) especially to the west where the limestone plateau. While the values from 60 to 30 Ω.m refer to fractured limestone with shale intercalation and appearing of Quaternary sand and gravel aquifer (Fig. 12, right panel), especially towards the east of the study area. -The resistivity ranges from 30 to10 Ω.m refer to the main locations of appearing and distribution of Quaternary sand and gravel aquifer with clay intercalation especially towards the east where the Nile valley sediments (Fig. 13, left panel), as well as the wetted surface sediments. While the values from 10 to1.5 Ω.m indicate the appearance and distribution of the clay and shale sediments along the study area (Fig. 13, right panel) which looks more eastward.

3D visualizing model of lithological and structural setting from the true resistivity
This model is designed to show the subsurface rock registration from bottom to top and the expected structural setting along the study area depending on the resistivity interpreted values. This model also shows the sedimentation system of the geological layers and the change in the geological facies. Accordingly, it will help in understanding the distribution of sediments by referring to hard limestone, fractured limestone, shaly limestone, clayey sand, sandy clay, clay, and sand with gravel and their hydrogeological implications. These hydrogeological implications include the groundwater aquifer/aquifers and their lithology. Where the recorded aquifers in this area are classified into fractured limestone of the Eocene age to fractured shaly limestone to the west and Quaternary sands with clay intercalation to the east covering the fractured limestone oriented to the east. Also, this model helped to understand the structural setting of this area and it appeared that a fault occurred (inferred fault). This fault is directed to NE-SW and its zone has been filled with porous, highly permeable sediments, as shown in Figs. 14, A C B Fig. 9 The hydro-resistivity profile A-A (A) shows the recorded geoelectric layers; the 1st geoelectric layer (L1), the 2nd geoelectric layer (L2), the 3rd geoelectric layer (L3) (shallow aquifer), and the 4th geoelectric layer (L4), as well as the depth to groundwater, the 2D hydro-

3D visualizing model of chargeability and its lithological signification
According to Fraser and Ward (1965), Keller and Frischknecht (1966), Cohen (1981), Draskovits and Smith (1990), Barker (1990), Kiberu (2002), and Leroux et al. (2010), the polarization increased with decreasing porosity west of the study area where the limestone is tight. The chargeability was high and the resistivity was low in saturated clay deposits such as fractured limestone. Also, the chargeability was high for disseminated clays/shale sediments such as in the shale limestone. It increased in the zones of fresh water along the study area and also increased with decreased porosity as in tight limestone and in fractured limestone when the shale content is high  and varies with the amount of water in the pores along the area. Accordingly, the subsurface geological layers can be visualized in a 3D model according to normal chargeability values, as shown in Figs. 19,20,21. The 3D visualization model for chargeability values from 476-150 mV/V% refers to the dense limestone with shale intercalation (Fig. 19, left  panel). Whereas, the chargeability value of 100 mV/V% refers to the dense limestone to semi-fractured limestone with shale intercalation (Fig. 19, right panel). All these deposits appeared to the west and northwest where is the limestone plateau.
The chargeability values 60 mV/V% and 30 mV/V% refer to the fractured limestone with shale intercalation (Fig. 20, left panel) and to fractured limestone with minor shale to sandy clay and clayey sand (Fig. 20, right panel). Where the limestone appears to the west and northwest in the shallow depths with large thickness and transfers to a small thickness to the east and appears in the deeper depths. This Fig. 20 refers, in general, to the change of facies of sediments from Eocene limestone with shale intercalation, to the west, to sandy clay and clayey sand of the Quaternary age to the east. Also, the low values of chargeability 60-30 mV/V% indicate the fractured limestone aquifer and Quaternary aquifer along the area.
Also, the value of chargeability 20 mV/V%, shown in Fig. 21, left panel, refers to the sandy clay to clayey sand of the Quaternary aquifer which are more concentrated and appear to the east of the area of study. The former sediments are then covered with dry highly permeable sand and gravel of the Quaternary sediments which are generally thick to the east and thin to the west where the limestone plateau. The chargeability  values of these sediments decreased to 0.2 mV/V% (Fig. 21, right panel). Accordingly, the previous chargeability values assisted in separating the sediments from the west to the east, in determining the expected zones and depths of high groundwater content (Fig. 21, right panel), in separating the fractured Eocene limestone aquifer and Quaternary aquifer, in separating the depths of high and low shale content, as well as the pure sediments (Fig. 21, right panel) and in general helped to show the sedimentary facies of the subsurface geological layers from the west to the east along the area.

3D visualizing model of SP and detection of seepage and flow of water
According to Tripathi and Fryar (2016) who performed SP measurements along the electrical resistivity profiles, they found that a -SP (mV) anomaly is consistent with a low electrical resistivity anomaly for most of the profiles, but few are not similar and not comparable. The low-resistivity anomalies and corresponding -SP (mV) anomalies encountered water-filled conduits, although clay/shale filled voids. Also, they concluded that in most cases, the anomalies of low resistivity values are reflected as -SP anomalies and amount of water decrease. Therefore, under wet conditions, the anomaly of -SP resulted due to groundwater flow. This anomaly was masked by the + SP response due to infiltration. The SP is a highly sensitive method in measuring of streaming potential, due to groundwater flow (Lange and Kilty 1991; Aubert and Atangana 1996;Stevanovic and Dragisic 1998;Fagerlund and Heinson 2003). Bumpus and Kruse (2014) stated that a different polarity of SP can result from different flow rates.  Accordingly, in this study, there is a seepage from surface reservoirs and expecting a feeding from adjacent aquifers. According to the previous studies, the negative values (-mV) may give indication about the increased content of clay/shale, which is intercalated with other sediments such as the fractured shaly limestone (Eocene aquifer) and clayey sand or sandy clay. Also, decreasing negative values may reflect the occurrence of groundwater and its source and flow direction as recorded along the study area. Also, the SP values can be visualized in a 3D model, as shown in Figs. 22,23,24. Where high SP values of 89-0 mV indicate dry sediments with their permeability expected to be high (Fig. 22, left panel). While values from 0 to -30 mV indicate wet sediments with low or equal permeability expected, as well as low clay content as in dense to semi-fractured limestone, as shown in the west of the study area where the limestone plateau (Fig. 22, right panel). Medium Values from -30 to -50 mV indicate depths with low groundwater flow with the expectation of zones with medium porosity such as fractured limestone and sandy clay (Fig. 23, left panel). While medium to low values from (-50) to (-75) (-mV) are expected to indicate depths with medium groundwater flow and medium to high porosity, as well as the expected sources of seepage (Fig. 23, right panel).
SP low values from -75--100 mV indicate depths with high groundwater flow and accumulation (Fig. 24, left panel) and also a very low value at -150 mV indicate depths with high groundwater flow. In general, the above values refer to high groundwater flow, high groundwater accumulation depths, and high permeability depths, as well as the location of the main vertical sources of seepage from the surface reservoir. In the end, SP values helped classify high, medium and low groundwater flow depths, determine the expected direction of seepage, predict high, medium and low  permeability sediments, as well as predict high, medium and low groundwater accumulation depths.
Also, according to the previous results from the resistivity, chargeability and SP values of the recorded sediments and the role of these values in separating the dry and saturated sediments (aquifers) and facies of these sediments and in delineating the fault zone with the expected sediments which fill this fault, as well as the type of covered sediments. Also, according to previous results from the resistivity, chargeability and SP values of the recorded sediments, these values had an important role in separating the shallow and deep dry and saturated sediments (aquifers) and in defining the fault zone with the expected sediments that would fill this fault, as well as the type of covered sediments. It can be also expected from these results that the source of surface water seepage into the subsurface porous sediments (aquifer) was through the surface of highly permeable sediments (Fig. 25, left panel) then vertically through the fault zone which was filled with high porous sediments and locates between the Eocene fractured limestone and Quaternary   (Fig. 25, middle panel), in addition to the very low values of SP (-mV) down to -150 mV indicates the expected locations of the source of surface water seepage, depths of high water accumulation and direction of water at deeper depths (Fig. 25, right panel) where indicated for the same location and depths detected in the fault zone (Fig. 25, middle and right panels). So, finally, very low values of negative -SP indicate high groundwater content, high to medium clay content, while medium to low values of -SP indicate low water content and low shale content, fractured limestone to semi-fractured limestone, as well as low to medium in porosity of the sediments (such as secondary porosity of fractured limestone).

Conclusion
In this study, SP, DC-resistivity and DC-TDIP methods were used at 21 sites to determine groundwater conditions and visualize the aquifer sediments, surface water seepage and groundwater flow in 2D and 3D. The resistivity results indicated the existence of Quaternary and Tertiary deposits. A comparison of resistivity and chargeability indicated a shallow aquifer consists of Eocene and Quaternary deposits and started from a depth of 52 to 135m and a thickness of 34-188m. The chargeability distinguished between shaly and non-shaly sediments and helped in  with referring to the expected locations and direction of seepage and the depths of high groundwater flow and groundwater accumulation (right panel) along the study area aquifer exploration by its value was high opposite to water-bearing layers and the shale sediments and increased with salinity increasing. It sometimes increases with less effective porosity sediments due to increased shale and it was varying with the water amount and low in dry sand and gravel. From Comparison of SP and chargeability, an increase in SP indicated a decrease in shale sediments and an increase in clean sediments while a decrease in SP indicated an increase in shale deposits, clean deposits with high water amounts, and an increase in effective porosity and groundwater flow. If there is a slight increase in SP, this may indicate a decrease in water salinity. The 3D resistivity model helped in understanding the distribution of sediments and their geological and hydrogeological implications. The 3D chargeability model showed that the polarization increased with decreasing porosity and it was medium in the fractured limestone and shaly limestone and in fresh water zones. Chargeability values of 60-20 mV/V% indicated the presence of the Eocene aquifer then the Quaternary aquifer. The 3D SP model with -SP values may indicate increased clay/ shale content and reflect groundwater occurrence, source and flow direction. SP values -30 to -50 mV, -50 to -75 mV and -75 to -100 mV and -100 to -150 mV indicated low groundwater flow depths with medium permeability, medium groundwater flow depths with medium to high permeability and sources of seepage, depths of high groundwater flow and accumulation then very high groundwater flow depths, respectively.
Finally, the results showed that these methods had an important role in separating the shallow and deep dry and saturated sediments, determining the water seepage source and the depths of different groundwater accumulations, classifying the depths of different groundwater flow and predicting the sediments of different permeability. The IP response depended highly on the water amount and shale content and it assisted the resistivity method in groundwater exploration, differentiation between the groundwater/shales and shaly/ pure sediments. The SP assisted in delineating the surface water seepage and groundwater flow direction and in predicting the qualitative permeability of sediments.