Environ Earth Sci (2017)76:511 DOI 10.1007/s12665-017-6840-1

ORIGINAL ARTICLE

GIS-based multi-criteria analysis and vulnerability method for the potential groundwater recharge delineation, case study of Manouba phreatic aquifer, NE Tunisia S. Saidi1,2 • S. Hosni1 • H. Mannai1 • F. Jelassi3 • S. Bouri2,4 • B. Anselme5

Received: 28 August 2016 / Accepted: 17 July 2017  Springer-Verlag GmbH Germany 2017

Abstract Groundwater resources in semi-arid climate like Manouba region (North of Tunisia) have experienced deterioration of available water resources in terms of both quantity and quality. Managers and water engineers developed techniques to compensate for the high exploitation and remedy the poor quality of groundwater resources. One of the most important solutions is to focus on the recharge capacity of the aquifer. This paper aims to locate potential groundwater recharge areas using the multi-criteria approach. Delineation of groundwater potential recharge zones relies on physical parameters and groundwater vulnerability. Geographical information system can provide tools for handling multidisciplinary data used for each parameter. In this work, weighting coefficients attributed for parameters are assigned based on local conditions in the region, using the analytic hierarchy process by applying the intelligent decision system software. Results show that most of the study area has a groundwater recharge rate potentially moderate to high. Only 24% of the study area have low groundwater recharge rate located especially in Sebkhat Essijoumi and its vicinity, that mainly match to the least permeable and hilly areas. Considering both the & S. Saidi [email protected] 1

Faculty of Sciences of Tunis, University of Tunis El Manar, 2092 Manar II-Tunis, Tunisia

2

Water Energy and Environment Laboratory ENI-Sfax, Sfax, Tunisia

3

Agricultural Development Office of Manouba (CRDAManouba), Manouba, Tunisia

4

Faculty of Sciences of Sfax, University of Sfax, Sfax, Tunisia

5

PRODIG Laboratory, University of Sorbonne, Paris1, Paris, France

vulnerability to pollution and the potential groundwater recharge capacity, the most suitable recharge sites are located in the North West of the study area. Keywords Potential groundwater recharge  Groundwater management  Manouba phreatic aquifer  IDS  GIS-AHP method  Groundwater sensibility

Introduction Groundwater is considered a key factor in the development of urban and rural regions with increasing needs. The situation became critical in areas with semi-arid climate and high exploitation like Manouba region. In fact, its exploitation by pumping is estimated at 1.093 Mm3/an (Hosni and Mannai 2014). Generally, groundwater is less vulnerable to pollution than surface water, which makes it a valuable natural resource. It is a heavily exploited resource due to intensive pumping, which can consequently lead to a rapidly declining water table. So in order to remediate to this overexploitation and to control water resources problem, it is necessary to pump groundwater in specific areas. In fact, potential groundwater recharge mapping is one of the most important aspects of groundwater studies that could assist water resource managers to have better exploitation and management plans. This issue has become a critical one for hydrologists, land managers and decision makers who are responsible for water resources for several decades. More recently, a lot of models have been applied for assessing groundwater potential mapping; for example, the indexing models which are applied in many studies (Dar et al. 2010; Madrucci et al. 2008; Nag et al. 2012; Prasad et al. 2008). In addition to other ones which are: probabilistic

123

511

Page 2 of 20

models such as multi-criteria decision analysis (MCDA) (Malczewski 2004; Gupta and Srivastava 2010; Murthy and Mamo 2009; Rahman et al. 2012; Feizizadeh et al. 2014), weights-of-evidence (Corsini et al. 2009; Lee et al. 2012), frequency ratio (FR) (Oh et al. 2011), fuzzy logic (Shahid et al. 2002; Ghayoumian et al. 2007) and analytic hierarchy process (AHP) (Chowdhury et al. 2009; Pradhan 2013) that have been used for groundwater potential mapping. AHP is one of the promising methods, developed by Saaty (1977), and based on individual criterions through quantitative analysis. It is a structured technique for organizing and analyzing complex decisions, and it is used worldwide in a large variety of decision situations (Vaidya and Kumar 2006). Hence, it was used by previous research studies in many topics of groundwater management such as the vulnerability assessment in suitability analysis (An et al. 2012) and in recharge assessment (Abdalla 2012; Senanayake et al. 2015; Zaidi et al. 2015; Herrmann et al. 2016). Groundwater recharge is influenced by many parameters. So in order to combine, analyze and handle the wide range of parameters, the use of GIS-based multi-criteria decision analysis (MCDA) is particularly attractive. GIS has been widely used in studies on water management (Babiker et al. 2005; Raj Pathak et al. 2008; Abdalla 2012; Rahman et al. 2012; Boughariou et al. 2014; Machiwal and Jha 2014; Zaidi et al. 2015). In this context, the aim of the present work is to provide a simple methodology to perform a groundwater potential recharge delineation map. The main difference between this research and the approaches described in the aforementioned publications is that the delineation of potential recharge zones considers not only recharge factors but also vulnerability to pollution. Another specific objective of this study is to establish an innovative methodology to calculate weights of each parameter used in potential groundwater recharge using GIS-AHP method and intelligent decision system (IDS), introduced by Xu and Yang (2001). The use of MCDA and GIS-AHP methods in the delineation of potential groundwater sites can reduce the cost of recharge operations and water management studies since it can help water mangers to choose the appropriate sites.

Environ Earth Sci (2017)76:511

Administratively, it is bounded by Tunis and Ariana in the East, Manouba in the West, Mohamedia and Fouchana in the South. According to the formula of Emberger, the study area experiences semi-arid climate. In fact, it is characterized by an average annual rainfall of 495 mm (CRDA Manouba 2014) and an average annual temperature of 20 C. The annual potential evapotranspiration is 1369 mm (Trabelsi 2013). Stratigraphic and geological settings The Manouba plain is an endorheic basin bordered by several mountain, in the North Jebel Ammar and in the South Jebel Sidi Salah and Jebel Nadhour (Added et al. 1995), which are made up of a sedimentary sections aged late Cretaceous to Quaternary (Figs. 1, 2a, b). This succession includes from bottom to top, the Senonian calcareous facies composed of marls and limestone alternations, the Eocene limestones and marls, the Oligocene sandstones, the Mio-Pliocene continental sands and siltstones, and the Quaternary clays and sands alluvial deposits which cover largely most of the plain area (Pimienta 1959; Pini 1971; Ennabli 1980; Added et al. 1995; Azizi et al. 2015) (Fig. 2b). Tectonic setting Manouba Essijoumi is located in the Greater Tunis part of Tunisian Atlas (the foreland basin of the Alpine Tunisian chain). It is an area of variable tectonic styles, essentially consisting of NE–SW trending folds (Pique´ et al. 2002). In the North of the study area, the Jebel Ammar anticline is a complex NE–SW anticline structure formed by a Mesozoic aged series. This structure is limited on the East and the West by two NW–SE major normal faults (Azizi et al. 2015). The South of Manouba graben is limited by a dextral strike-slip fault, characterized by high seismic activity (Baccara 2003). This fault has a significant normal component marked by a thick Mio-Plio-Quaternary series in the graben structure of Manouba plain (Azizi et al. 2015). Hydrology and hydrogeology

Study area characteristics General settings The Manouba phreatic aquifer is located in Northeast of Tunisia and the western part of Tunis region. Geographically, the study area lies between 590640 N to 609324 N latitude and 4057350 E to 4081416 E longitudes and occupies 230 km2.

123

Hydrologically, the study area is characterized by Wadi Gueriana which represents the most important river. It is subdivided into many affluents which have the sole collector Sebkhat Essijoumi with an area of 30 km2 and an average topography of 10 m (Zebidi and Chaumont 1962). The study area occurs as a set of complex aquiferous systems made up of three superimposed aquifers displaying a total thickness of 40, 40–80 and more than 80 m, respectively. These aquifers are lodged in Quaternary

Environ Earth Sci (2017)76:511

Page 3 of 20

511

Fig. 1 Location of the study area

continental materials which show large lateral facies variations extending from sands and clays, westward in the Manouba area, to reddish sands, clays and recent alluvia in the Fouchana area located in Southwest (Drogue 1966; Hechmi 1989; Added et al. 1995). The phreatic aquifer of Manouba aim of this study is lodged in the Plio-Quaternary sediments and exploited by 979 wells (Habboubi 2014). Previous studies showed that water consumption for irrigation during summer has caused groundwater level drawdown and a subsequent aquifer overexploitation.

Materials and methods Potential recharge sites identification Selection of potential groundwater recharge sites is one of the basic decisions in water resource management and especially in artificial recharge of aquifers. To take more efficient decisions on water resource management, it is necessary to consider potential

groundwater recharge sites: physical characteristics influencing the recharge capacity of the aquifer and parameters expressing the vulnerability to pollution of the zone (Eq. 1): PRS ¼ PR þ AV:

ð1Þ

PRS potential groundwater recharge site; PR potential groundwater recharge index; AV aquifer vulnerability. In order to combine PR and AV resulting maps, a simple summation of raster maps by executing a single Map Algebra expression using Python syntax in a calculator interface of ArcGIS software is performed. So in this study, two considerations should be taken into account for the selection of potential recharge sites, the vulnerability to pollution and the potential recharge capacity of the aquifer. Potential groundwater recharge assessment To locate the potential recharge areas, an Analytic Hierarchy Process (AHP) combined with a GIS was used. The methodology involved the following major steps:

123

511

Page 4 of 20

Environ Earth Sci (2017)76:511

Fig. 2 Geological context of the study area; a geological map of the Manouba Essijoumi basin. b Synthetic log of the Mesozoic and Cenozoic series in the Greater Tunis area (Azizi et al. 2015)

1. 2.

Select criteria for potential recharge assessment; Develop a decision hierarchy structure and identify priorities (weights) for each decision criterion a pairwise comparison matrix; Determine the vulnerability to pollution of the aquifer on the basis on such criteria; Combine the potential groundwater recharge map and the vulnerability to pollution one to delineate the most potential groundwater recharge sites.

important parameters influencing potential groundwater recharge are: slope (S), permeability (P), land use (Lu), rainfall (R), drainage density (Dd) and fault density (Fd). The potential groundwater recharge map is a result of overlaying of the six parameters according to their influence and impact on water infiltration. Technically, the potential groundwater recharge index (PRI) is the sum of products of normalized weights (w) and ratings (R) of each criterion, using the following formula:

Selection of criteria/parameters affecting potential groundwater recharge (Fig. 5) Several previous research studies have shown the benefit of multi-criteria approach for estimating potential groundwater recharge areas, such as Andres (2004), Waichler (2005), Vaidya and Kumar (2006), Haouchine et al. (2011), Harrison et al. (2011), Huang et al. (2011). Based on these previous works and after screening references and studies (Ghayoumian et al. 2007; Abdalla 2012; Adiat et al. 2012; Boughariou et al. 2014), the most

PR index ¼ Sw  Sr þ Fdw  Fdr þ Ddw  Ddr þ Pw  Pr þ Luw  Lur þ Rw  Rr ; ð2Þ

3. 4.

123

with w and r indexes representing weights and rates for each criterion. Weights calculation using AHP method and IDS software The major drawback of multi-criteria methods is the subjectivity of the determination of the rating scale and the weighting coefficients (Saidi et al. 2011). In this study, the Analytic Hierarchy Process (AHP), as developed by Saaty

Environ Earth Sci (2017)76:511

Page 5 of 20

511

Fig. 2 continued

(1977), is adopted to weight assessment. It is one of the expert systems that deal with imprecision of results by assigning a weighting to each criteria (Kim et al. 2012) and by reducing complex decision to a series of one-on-one comparisons and then synthesizes the final results (Ting and Cho 2008). The reasons to use AHP method are its flexibility. It can be integrated with different techniques such as Intelligent Decision System (IDS) based on linear programming that will be used in this study. It represents a

windows-based software package that has been developed on the basis of the Evidential Reasoning (ER) approach (Xu and Yang 2001). It is used in a recent development in handling hybrid Multiple Criteria Decision Analysis (MCDA) problems of both a quantitative and qualitative data with uncertainties (Yang and Xu 2004). Recently, IDS is recognized in many projects since it can help to pre-quantify the weights of each parameter in multi-criteria assessment, like the case of the present study.

123

511

Page 6 of 20

In the IDS interface, the user can directly define certain qualitative characteristics such as name, number of measurement grades to assess quality of parameters, as well as a short description for each measurement grade. Generally, the grades vary from very low level designed by the worst (rate = 1) to very high level or the best (rate = 9). Furthermore, the user can provide comments on why the assessment is given this way. Such an assessment process is referred to as an evidence-based mapping process. The weights are assessed based on both the spatial variation of the criteria (different classes of each potential groundwater recharge mapped in GIS environment) and its influence on potential groundwater recharge (rates attributed to each class). This approach is designed to improve the objectivity and accuracy of the inherent subjective process. After that, rates are assigned for all parameters and they are classified according to their potential for groundwater recharge and on the basis of similar studies (Ghayoumian et al. 2007; Abdalla 2012; Adiat et al. 2012; Boughariou et al. 2014) and after discussion with local authorities. After assigning rates, IDS interface allows to estimate the weight for each parameter (alternative) as already cited on the basis of rates and the area occupied by each class (or rate). For that, two approaches can be used in IDS to support weight assignment to each criterion: The first one is the visual assignment, whereas the second one is based on pairwise comparison of criteria using Saaty matrix. In this study, the second method is chosen because it consists in the selection of the attribute and proceeds to a comparison and if the result displayed is satisfying, a confirmation of the comparison should be made. The same process is repeated for all attributes (slope, permeability, rainfall distribution, drainage density, land use, fault density). The resulting values were rescaled linearly into [0–1] interval in order to show the weight for each parameter. Processing parameters •

Slope (S)

Slope is considered as an important parameter affecting runoff and infiltration, a steeper slope leading to less infiltration and increased runoff. In this study, the slope map was generated from Shuttle Radar Topography Mission (SRTM) topographic data, following standard GIS routines using Global mapper and ArcGIS software. In fact, the contours are generated from SRTM image using generate contours command of global mapper which allows the user to generate equally spaced contour lines from the SRTM. The contour interval and units as well as the grid spacing to use are fixed in the contour options panel. Then, the contours are imported to ArcMap interface

123

Environ Earth Sci (2017)76:511

of ArcGIS, and using Slope function, slope map can be generated (Table 1; Figs. 3, 4). •

Permeability (P)

Groundwater storage is controlled by the porosity and permeability of the aquifer’s lithology. In fact, a higher porosity contributes to higher groundwater storage, and higher permeability leads to increased groundwater yields. In this case, the permeability is determined by the lithology of geological formations. According to Jang et al. (2013), the condition of soil textures is closely related to subsoil infiltration and unsaturated soil percolation and it represents the most important factor to dominate aquifer recharge. The lithology map is derived from the digitalization of geological map and cross sections of CRDA (Jelassi 2012). The classification is based on conductivity values from the literature (Castany 1982; Ge´raud 2000) (Table 1) in addition to CRDA permeability index attributes to each lithological nature. Thereby, the less permeable grounds with a fine or non-fractured lithology such as clay are ranked by the lowest number, while the highest rank is attributed to sand. •

Rainfall distribution (R)

Naturally, groundwater recharge increases with the amount of precipitation. The rainfall map is the result of interpolation of rainfall data. These data are collected at each of the six climatological weather stations included into the study area, operated by meteorological and agricultural offices and represent a mean of 63 years (1950–2013), and the newest data are these of 2013 hydrologic year (Figs. 3, 5; Table 2); (CRDA 2014; INM 2014) (Tables 1, 4). Inverse distance weighting method of ArcGIS software is used for interpolation. •

Land use (Lu)

The high spatial resolution of satellites images reveals very fine details in urban areas and greatly facilitates the classification and the extraction of urban-related features such as buildings (Jin and Davis 2005; Benediktsson et al. 2003; Guindon 2000). Two landsat thematic mapper (TM) images acquired on April 28, 2014, with a 30-m spatial resolution, are used, and its processing was carried out applying the ENVI image analysis (Table 1; Figs. 3, 5). In this case, ENVI software is used for image classification adopting the following steps: – –

Preprocessing by making an atmospheric calibration and removing the disturbances; Making supervised classification based on maximum likelihood algorithm was applied to the images to classify land cover. In fact, in this classification method

Environ Earth Sci (2017)76:511

Page 7 of 20

511

Table 1 Data source and mode of processing of each potential recharge parameter Parameter

Data source

Mode of processing

Slope (S)

Topographic maps (sheets of 1/50,000)

Digitalization/3D analyst ArcGIS

Permeability (P)

Geological maps (sheets of 1/50,000), CRDA reports (Jelassi 2012)

Digitalization/classification Interpolation/classification

Rainfall (R) distribution

Rainfall data (INM 2014; CRDA 2014)

Land use (Lu)

Landsat 8 imagery

Classification in ENVI

Drainage density (Dd)

SRTM data, topographic maps (sheets of 1/50,000)

Stream generation with Arc Hydro Tools/density line calculation/validation

Fault density (Fd)

SRTM data, geological map

SRTM filter, analysis and classification/density line calculation/validation

Fig. 3 Flowchart describing the methodology adopted in the potential recharge delimitation zones

Bibliography research assessment

Selection of the most important potential recharge parameters

-Screen the references -assess the remaining studies -select the principal parameters influencing potential recharge.

Raw data (xls data, scanned images and plans…)

Data treatment (interpolation, digitalisation…)

Thematic layers (maps of different potential groundwater recharge parameters)

Definition of criteria score

Criteria layer 1

Criteria layer 2

Criteria layer n

Comparison of all factors using parwaise comparison

Generation of weights of different factors Superposition Potential groundwater recharge index calculation by the use of raster calculator

Vulnerability assessment

Analysis of quantitative and qualitative data

Favourable Potential groundwater recharges sites

sample land cover called ‘‘training sites’’ is used to identify the land cover classes in the entire image. These training sites serve in the classification and the validation of the map.

Result of the classification process is completed with the GIS database that includes the limits for each class. •

Drainage density (Dd)

123

511

Page 8 of 20

Environ Earth Sci (2017)76:511

Creation of the Stream network

Creation of the lineaments or faults

SRTM

Acquire a DEM

Acquire a DEM

Pre-processing (smoothing filter using pixel 30*30 pixels)

Set Data Frame Coordinate System to the DEM

Analyzing the smoothness/roughness by applying Linear filters in all main directions (N; NW; W; SW)

Fill Holes or skins in the DEM Flow Direction

Classification using Imagine Objective (ERDAS)

Stream definition

Differentiation of lineaments after applying the appropriate algorithm and comparison to geological and structural maps

Validation by comparison to the wadis digitalized from topographic map.

Fig. 4 Generation process of stream network and faults from SRTM data

Drainage density is one of the most commonly used parameters for the assessment of groundwater recharge as it is directly linked to the relationship between runoff and infiltration (Boughariou et al. 2014). In order to extract the stream network from SRTM image, DEM is modified and corrected using Arc Hydro Tools extension of GIS tools. After that, some operations are utilized using Arc Hydro Tools: flow direction, stream definition after fill skins (Fig. 4). The drainage density map is created using line density function of ArcGIS. It is the result of dividing stream length by the study zone area. In this case, natural stream network was derived from a digital elevation model produced from SRTM data (Table 1; Figs. 1, 4). •

Fault density (Fd)

Tectonic lineaments were extracted from SRTM data and geological map of the study area. By using digital elevation models (DEM), extracted lineaments rely solely on elevation information (Fig. 4). The first step is the preprocessing by applying a smoothing filter before further analysis. After that, a filter of 30 9 30 pixels and using different directions is applied. The second step is to use the classification of Imagine Objective (ERDAS). This classification provides important information on surface and subsurface fracture systems that may control the movement and storage of groundwater (Pradeep 1998; Sreedhar et al. 2009). Using line density of GIS, the fault density map was produced in the same way as that for stream network. And validation is made by

123

comparison with the stream network digitalized from topographic map (Table 1; Fig. 4). Vulnerability to pollution assessment A DRASTIC model applied in a GIS environment was used to evaluate the vulnerability of the Manouba phreatic aquifer. It was developed by the US Environmental Protection Agency (EPA) to evaluate groundwater pollution potential for the entire USA (Aller et al. 1987). The DRASTIC method that relies on hydrogeological parameters provides an approach to evaluate an area based on known conditions without the need for extensive sitespecific pollution data. It relies on the hydrogeological parameters that affect and control the transport of contaminants from the surface to the groundwater. The acronym DRASTIC stands for the seven parameters used in the model which are (Table 1): •

•

Depth to water The depth to water table is obtained by subtracting the water table level from the elevation of the well. Therefore, an exact interpolation scheme is appropriate for generating a smooth surface representation for the high degree of spatial continuity of the groundwater surface in an aquifer. The inverse distance weighting interpolation technique is used. Net Recharge is evaluated using water table fluctuation method which considers that recharge occurring between times t0 and tj is the product of both the specific yield (Sy) and the peak water level rise

Environ Earth Sci (2017)76:511

Page 9 of 20

511

Fig. 5 Recharge parameters classification

123

511

Page 10 of 20

Environ Earth Sci (2017)76:511

Table 2 Interannual rainfall for the period (1950–2013) (CRDA Manouba 2014) Month

September

October

November

December

January

February

March

April

May

June

July

Augest

Year

Manouba

40.4

57.0

58.3

71.1

68.2

53.5

48.7

53.5

17.2

13.9

2.6

10.7

495

Mornaguia

47.1

66.5

68.0

82.9

79.6

62.4

56.8

62.3

20.0

16.2

3.0

12.5

577

Borj Chakir

36.5

51.6

52.8

64.4

61.8

48.4

44.1

48.4

15.5

12.6

2.3

9.7

448

Tunis Manouba

37.1

52.4

53.6

65.4

62.7

49.2

44.8

49.2

15.8

12.8

2.4

9.9

455

Mnihla

38.5

54.4

55.6

67.8

65.1

51.0

46.5

51.0

16.4

13.3

2.5

10.2

472

Mean

39.9

56.4

57.7

70.3

67.5

52.9

48.2

52.9

17.0

13.7

2.5

10.6

489

attributed to the recharge period DH (tj) according to the following equation (Sophocleous 1991)

R tj ¼ Sy  DH tj ð3Þ

•

•

•

•

where Sy = f - Sr; f is porosity and Sr is specific retention. Net recharge can increase with the permeability of vadose zone and the soil texture. The aquifer media (A) and the impact of vadose zone (I) are obtained using a subsurface geology map and geological sections of the Manouba Essijoumi aquifer. The impact of the vadose zone controls the path length and routing, thus affecting the time available for attenuation and the quantity of material encountered. The materials at the top of the vadose zone also exert an influence on soil development (Aller et al. 1987). Soil media It considers the uppermost part of the vadose zone, and it influences the pollution potential. A soil map, for the study area, is obtained from CRDA of Manouba (CRDA and classified according to Aller et al. (1987) rating. Topography the same vector layer extracted from SRTM image for potential recharge assessment is used. However, in its classification, rates are attributed according to their influence on pollutant infiltration on the basis of Aller et al. (1987) classification (Table 4; Fig. 4). Hydraulic conductivity It refers to the ability of the aquifer materials to transmit water, which in turn, controls the rate at which groundwater will flow under a given hydraulic gradient (Aller et al. 1987). Hydraulic conductivity is measured by aquifer pumping tests in wells (Jelassi 2012). Then, the relative map is established by interpolation and it is divided and reclassified into ranges according to their influence on pollution potential.

The model yields a numerical index that is derived from ratings and weights assigned to the seven model parameters. The significant media types or classes of each parameter represent the ranges, which are rated from 1 to

123

10 based on their relative effect on the aquifer vulnerability. Each parameter is then assigned a weight, ranging from 1 to 5 reflecting their relative importance. The DRASTIC Index is then computed applying a linear combination of all factors according to the following equation: VI ¼ Dr Dw þ Rr Rw þ Ar Aw þ Sr Sw þ Tr Tw þ Ir Iw þ C r C w

ð4Þ

where VI is the DRASTIC Index (vulnerability), D the depth to groundwater, R the net recharge, A the aquifer media, S the soil type, T the topography (surface slope), I the impact of the vadose zone type, C the hydraulic conductivity, w the weight of each factor and r the rating associated. The same methodology used in the weights calculation of potential groundwater recharge parameters is used for vulnerability parameters.

Results and discussion Potential groundwater recharge estimation Spatial variation and classification of potential recharge parameters In order to delineate the potential groundwater recharge zones in the study area, the most important parameters are mapped: slope (S), rainfall distribution (R), drainage density (Dd), fault density (Fd), permeability (P), drainage density (Dd) and land use (Lu) (Table 2; Fig. 5). After mapping the thematic maps of different parameters, the input layers were ranked according to Table 2. Hence, the ranked layers are defined relying on their relative importance to control groundwater potential groundwater recharge according to this classification: very low (1), low (3), medium (5), high (7) and very high (9). The slopes were classified into three categories with respect to their hydrologic capacities (Fig. 3). About 70% of the study area is classified as being relatively flat

Environ Earth Sci (2017)76:511

Page 11 of 20

511

Table 3 Assigned ratings (R) and normalized weights calculated for potential recharge parameters Influencing parameters

Category (classes)

Potentiality for groundwater recharge

Fault density (Fd) (km/ km2)

0–1 1–2

Slope (S) (%)

Rainfall (R) (mm)

Land Use (Lu)

Rating (R)

Weight (W) (%)

Normalized weight

Very low

1

20

0.073

Low

3

3–3.43

Medium

5

0–2

Very High (flat)

9

85

0.31

2–4

High (undulating)

7

4–11

Medium (rolling)

5

380–435

Low

3

44

0.16

435–460

Medium

5

460–510

High

7

Water surface (Sebkhat)

Very low

1

44

0.16

Urban areas Bare soil

Low Medium

3 5 38

0.14

42

0.15

Irrigated areas Permeability (P)

Drainage density (Dd)

High

7

Low

3

Medium

5

High

7

0–0.5

Low

3

0.5–1.5

Medium

5

(0–2%), which is highly favorable to infiltration and therefore for groundwater recharge. A rate of 9 is assigned to this slope class. The rest of the study area is occupied by two classes of slope which are (2–4%) and (7–11%) considered as moderate to steep, with respective assigned rates of 7 and 5. Soil permeability determined from the geological data is divided into three classes; most of the area has a low-tomoderate permeability with respective assigned rates of 3 and 5 (Fig. 5). The last class that covers about 10% of the area is very favorable for groundwater recharge (rate = 7). The corresponding area is located mainly south of the study area, near the localities of Chafrou and Mornag and corresponds to very permeable sand formations (Fig. 5). The rainfall distribution map shows a minimum of 380 mm and up to 510 mm in three classes (rates 3, 5 and 7). The most favorable area with high rainfall occupies half the surface of Soukra and Morang regions in the north and east (Fig. 5). Considering faults density parameter, three categories are distinguished with rates of 1, 3 and 5, respectively, for high density (overlap structures and faults), moderate density (faults) and low density classes. The high density class, which occupies a small percentage of the total area, is a potential location of groundwater infiltration (Fig. 5). In the study area, four classes of land use are determined: urban areas, water surface, bare soil and irrigated areas. Low rates of 1 and 3 are assigned to water surface

(Sebkhat) and urban areas, whereas rates of 5 and 7 are, respectively, assigned to bare soil and irrigated areas (Fig. 5). The drainage density is divided into two classes: a high density class corresponds to areas dug by large channel beds (rate of 5) and a low density class coincides with the dry channel network (rate of 3) (Fig. 5). Determining the weight of each criteria The comparisons ratings result of AHP process is on a scale of 1–9 (Table 3; Fig. 6a). In the AHP, the pairwise comparisons of all the criteria/parameters were taken as the inputs, while relative weights of the parameters were the outputs. The final weights are generated after comparison (in percent) (Fig. 6b): The last step is to normalize the weights to minimize the impact of inconsistencies in the ratio (Table 3). Table 3 shows that normalized weights values range from 0.073 to 0.31, thereby indicating that the most important criterion for potential groundwater recharge is the slope, while faults density has much less influence. Potential groundwater recharge evaluation The potential groundwater recharge calculation according to the (Eq 2) shows that the most favorable areas extend over 15% of the study area. The most important criteria in

123

511

Page 12 of 20

Environ Earth Sci (2017)76:511

Fig. 6 Weights generation approach using IDS software: a pairwise comparison, b weighted generated in %

aquifer recharge ranked from most to less influential being the slope (85%), rainfall and land use ([44%), the drainage density (42%) and soil permeability (38%). The fault density criterion has the lowest percentage (20%) and

123

therefore is the one having the least influence on groundwater recharge (Fig. 7). In fact, in the area there is neither a high density of faults nor a high diversity of fault types.

Environ Earth Sci (2017)76:511

Page 13 of 20

511

Fig. 7 Potential groundwater recharge map

Results show that 24.24% of the study area has a low capacity for groundwater recharge. This corresponds to steep slopes, with low rainfall and the low permeability soil. The rest of the area is considered to have moderate to high capacity for groundwater, corresponding to relatively flat areas associated with heavy rainfall and infiltration potentialities, as well as irrigated areas (Fig. 7).

Some areas of high potential groundwater recharge are probably associated with semi-urban and rural settlements, which promote recharge capacity such as artificial recharge in the vicinity of hydrographic network (principal affluent of wadis Gueriana and Chafrou (Fig. 1). Also the high potential groundwater recharge zones of the study area have key roles in choosing the most suitable zones for wells and therefore

123

511

Page 14 of 20

Fig. 8 Spatial variation of vulnerability parameters

123

Environ Earth Sci (2017)76:511

Weight = 2 Weight = 5

7

Depth of groundwater (D) map as a result of interpolation of groundwater depth measures indicates that 45% of the groundwater basin has depth range between 2 and 9 m (rating values of 9 and 10 according to DRASTIC) which is associated with high vulnerability to pollution. The depth of groundwater of the rest of the study area varies between 9 and 15 m which provide relatively lower vulnerability to pollution (Fig. 8; Table 4). The net recharge parameter shows that the whole of the study area is between 17 and 35 mm/year. The lower net recharge is contributed by low average annual rainfall of 495 mm for the Manouba phreatic aquifer (Table 4). Sand and gravel aquifer media types dominate most of the aquifer (65%).This makes the majority of the shallow aquifer to be highly susceptible to contamination according to DRASTIC model assumptions (Fig. 8). Silt and sand soil types occupy most part of Manouba plain are calcimorphic associated with vertisols, modal Aeolian and halomorphic soils. Ratings were assigned on different types of soil according to their infiltration capacity which depends on soil texture. Topography map displays a gentle slope for the majority of the study area. A high rate is assigned to these areas because vulnerability decreases with the slope. The impact of the vadose zone (I) is expressed by the lithology of the unsaturated zone. The vadose zone is ranged from clay, sand to calcareous lithology. Hydraulic conductivity of the study area shows maximum of 4.3 9 10-4 m/s in the northwest and increase toward the central part of the study area (Fig. 8; Table 4).

8 Slightly sandy clay

Limestone, sandy clay

Weight = 3 10 to 4 9 10

Weight = 3

3 9 10

8 2 9 10-4 to 3 9 10-4 1 15–60

Vulnerability index calculation

Weight = 1

6 1 9 10-4 to 2 9 10-4 3 10–15

-4

4 -2 9 10-5 to 1 9 10-4 5 5–10

-4

5

7 Halomorphic soil

Modal aeolian soil 5

6 Calcareous marl, limestone, sand and clay

Limestone crust and gypsum

Spatial variation of vulnerability parameters

6

4 2 -5 9 10-5 to -2 9 10-5 9 3–5

1

Calcareous, gravel

Conglomerate

2 Calcimorphic and vertisols

Vulnerability assessment

5 9–15

R rate

7 4.5–9

Weight = 5

9 2–4.5

Weight = 4

0.017–0.035

1

0–3

511

can considerably reduce the cost of well drilling by minimizing the failure of obtaining suitable well sites.

-9.52 9 10 to -5 9 10-5

-5

R R R

Page 15 of 20

10

Clay and calcareous

3

Clay, dune, dolomites, sandstones and evaporites

4

R Soil type R

Lithology class

R

Lithology class

R

Soil type (S) Impact of the vadose zone (I) Aquifer lithology (A) Hydraulic conductivity (m/s) (C) Topography/ slope % (T) Net recharge (m) (R) Depth to groundwater (m)

Table 4 Rates affected for each vulnerability parameter according to DRASTIC method and using the Aller et al. (1987) classification in Manouba phreatic aquifer

Environ Earth Sci (2017)76:511

Using the weights and rates of DRASTIC method and verified by IDS software the same as the method used in weight calculation of recharge parameters, the DRASTIC Index is calculated by affecting each parameter by the corresponding weight. The resulting DRASTIC values lay between 69 and 159 (Fig. 9). This range on the basis of the above classification (Engel et al. 1996) as: (1) 69–100 which is assigned a low vulnerability, (2) 100–141, is represented by a moderate vulnerability and (3) 141–159, which is assigned high vulnerability (Fig. 9). The results of intrinsic vulnerability (Fig. 9) indicate that about 34% of the total groundwater basin area is occupied by high vulnerability zones.

123

511

Page 16 of 20

Environ Earth Sci (2017)76:511

Fig. 9 Vulnerability to pollution of the Manouba phreatic aquifer

Also moderate vulnerability zones were found to occupy about 41% of the total groundwater basin. Large area of Northern West and the South of Manouba phreatic aquifer is occupied by moderately and high vulnerability zones. This is due to the high permeability, the flat topography and the low depth of groundwater. The low permeability and the emplacement of urban zones and settlements in the behavior of Sebkhat Essijoumi and in the Extreme Northeast of the region make it low vulnerable to pollution (Fig. 9). Results of the vulnerability assessment are validated by means of nitrate monitoring of the Manouba

123

Essijoumi phreatic aquifer. Nitrate concentrations vary from a minimum of 4 and a maximum of 186 (Habboubi 2014). The comparison between vulnerability map and nitrate concentration shows clear coincidence, and high concentrations are situated in high vulnerability zones. Delineation of potential groundwater recharge sites Results show that the extreme northern part of the study area is favorable to recharge (30%) and the most part of the groundwater basin is moderate favorable to recharge. This

Environ Earth Sci (2017)76:511

Page 17 of 20

511

Fig. 10 Potential groundwater recharge sites selection using quantity and quality parameters

is due to the fact that most part of the basin is classed as moderate vulnerable and moderate favorable to recharge. However, low favorable zones such as Sebkhat Essijoumi and its behavior (Sidi Thabet) correspond with low vulnerability zones. Sebkhat Essijoumi can not represent a promote zone for recharge despite its low vulnerability, which is due to low permeability of its soil and vadose zone (Figs. 5, 9, 10). Considering the quantitative (potential groundwater recharge) and qualitative factors (vulnerability to pollution), the least exposed recharge areas to the vulnerability to pollution are those of the extreme Southwest

(Mohamedia and Kabbaria) and Northwest (El Mnihla). These areas are of great concern to decision makers in terms of water resource management (Figs. 7, 9, 10). Discussion of the results The model used in this study allows us to identify areas that have a high potential for groundwater recharge. Despite the subjectivity of the approach and the availability of thematic data, their accuracy level and the local conditions play an important role in such groundwater management endeavors.

123

511

Page 18 of 20

To validate the results, there is no direct method for verifying potential groundwater recharge delineation and results are indirectly verified by comparison with the water budget, by the field results (for land use parameter) as well as by the local administrative authorities. The fault density is validated by comparison with structural studies undertaken on the study area. In vulnerability assessment, only intrinsic parameters are used in this study, which can constitute a serious weakness in the application. Hence, the existence of potentially polluting activities should be considered in this assessment.

Conclusion Integration of geospatial technology with the Analytic Hierarchy Process seems to be a very relevant tool for identifying potential areas of groundwater recharge in Manouba phreatic aquifer. Results show that around 18% of the total area was identified as high potential zone for groundwater recharge corresponding to the extreme Northwest and Southwest of the study area (Fig. 10). However, hilly terrains in the extreme Northwest and around the Sebkhat Essijoumi are considered unsuitable areas for groundwater recharge processes and especially areas corresponding to a high vulnerability degree. These areas are a priority for building artificial groundwater recharge structures to store rainwater and reduce surface runoff. It is obvious that integrated approach of GIS and AHP developed in this study is simple, spatial and flexible. It can provide the appropriate platform for analysis of different datasets for decision making is not concerned only with mapping and planning of groundwater resources but also with the management of groundwater resources. Indeed, this process reveals also to decision makers how changes in criteria weights affect evaluation outcomes spatially and quantitatively. The potential groundwater recharge delineation maps can be used to plan the implantation of future wells and hydraulic structures on these high potential areas. However, this study should be completed by future projections of the potential groundwater recharge delineation in function of the climatic changes including precipitation, temperature and land use changes. Furthermore, it needs to be extended in further research studies such as developing of web software solutions to publish the maps established in this study. Also, the AHPGIS methodology is developed here as a web decision support system.

123

Environ Earth Sci (2017)76:511 Acknowledgements The authors thank Ms. Yousra SAIDI for her translation help. They are also grateful to the Office of Agricultural Development of Manouba region for its support in this work.

References Abdalla F (2012) Mapping of groundwater prospective zones using remote sensing and GIS techniques: a case study from the Central Eastern Desert, Egypt. J Afr Earth Sci 70:8–17. doi:10. 1016/j.jafrearsci.2012.05.003 Added A, Ben Mammou A, Fernex F, Rezzoug S, Bernat M (1995) Distribution of uranium and radium isotopes in an aquifer of a semi-arid region (Manouba-Essijoumi, Northern Tunisia). J Environ Radioact 82(2005):371–381 Adiat KAN, Nawawi MNM, Abdullah K (2012) Assessing the accuracy of GIS-based elementary multi criteria decision analysis as a spatial prediction tool—a case of predicting potential zones of sustainable groundwater resources. J Hydrol 440–441:75–89. doi:10.1016/j.jhydrol.2012.03.028 Aller L, Bennet T, Lehr JH, Petty RJ, Hacket G (1987) DRASTIC: a standardized system for evaluating ground water pollution potential using hydrogeologic settings. US Environmental Protection Agency Report (EPA/600/2-87/035), Robert S. Kerr Environmental Research Laboratory An Y, Wang Y, Zhang H, Wu X (2012) GIS-based suitability assessment for shallow groundwater development in Zhangye Basin, 2011 international conference on environmental science and engineering (ICESE 2011). Proc Environ Sci 12:1397–1403 Andres AS (2004) RI66 Ground-water recharge potential mapping in Kent and Sussex Counties, Delaware. Delaware Geological Survey Report of Investigation No. 66. http://www.dgs.udel.edu/ publications/ri66-ground-water-recharge-potential-mapping-kentand-sussex-counties-delaware Azizi R, Kadri A, Chihi L (2015) Neogene tectonic evolution in Northern Tunisia: case of Chaouat-Mannouba area. Palaeoseismic event associated. Arab J Geosci 8:8911–8925. doi:10.1007/ s12517-015-1789-4 Babiker IS, Mohammed MAA, HiyamaT Kato K (2005) A GIS-based 337 DRATIC model for assessing Aquifer vulnerability in Kakamigahara Heights, 338 Gifu Prefecture, central Japan. Sci Total Environ 345:127–140. doi:10.1016/j.scitotenv.2004.11.005 Baccara K (2003) Etude tectonique et se´ismotectonique: les deformations anciennes et re´centes associe´es au Jbel Ammar. DEA, Faculte´ des sciences de Tunis Benediktsson J, Pesaresi AM, Arnason K (2003) Classification and feature extraction for remote sensing images from urban areas based on morphological transformations. IEEE Trans Geosci Remote Sens 41:1940–1949 Boughariou E, Saidi S, Barkaoui AE, Khanfir H, Zarehloul Y, Bouri S (2014) Mapping recharge potential zones and natural recharge calculation: study case in Sfax region. Arab J Geosci 8:5203–5221. doi:10.1007/s12517-014-1512-x Castany G (1982) Principes et me´thodes de l’hydroge´ologie. Universite´ de Pierre et Marie Crue (Paris VI), Paris, p 233 Chowdhury A, Jha MK, Chowdary VM, Mal BC (2009) Integrated remote sensing and GIS-based approach for assessing groundwater potential in West Medinipur District, West Bengal, India. Int J Remote Sens 30(1):231–250 Corsini A, Cervi F, Ronchetti F (2009) Weight of evidence and artificial neural networks for potential groundwater spring mapping: an application to the Mt. Modino area (Northern Apennines, Italy). Geomorphology 111(1):79–87 CRDA (2014) Weather data of Manouba. Unpublished report, CRDA Manouba (in French)

Environ Earth Sci (2017)76:511 Dar IA, Sankar K, Dar MA (2010) Remote sensing technology and geographic information system modeling: an integrated approach towards the mapping of groundwater potential zones in Hardrock terrain, Mamundiyar basin. J Hydrol 394:285–295 Drogue C (1966) Note sur la nappe phre´atique de Manouba. Rapport interne, Bureau d’Inventaires des Ressources Hydraulique (BIRH), Tunis Engel BA, Navulur KCS, Cooper BS, Hahn L (1996) Estimating groundwater vulnerability to non-point source pollution from nitrates and pesticides on a regional scale. Int Assoc Hydrogeol Sci Publ 235:521–526 Ennabli M (1980) Etude hydroge´ologique des aquife`res du Nord-Est de la Tunisie, pour une gestion inte´gre´e des ressources en eau. The`se d’Etat, Universite´ de Nice Feizizadeh B, Jankowski P, Blaschke T (2014) A GIS based spatiallyexplicit sensitivity and uncertainty analysis approach for multicriteria decision analysis. Comput Geosci 64:81–95. doi:10. 1016/j.cageo.2013.11.009 Ge´raud Y (2000) Rock permeability and Darcy law. Benedict Urgelli, University of Louis Pasteur, Strasbourg, Surface Geochemistry Center. http://planet-terre.ens-lyon.fr/article/permeabilite-desroches.xml Ghayoumian J, Mohseni Saravi M, Feiznia S, Nouri B, Malekian A (2007) Application of GIS techniques to determine areas most suitable for artificial groundwater recharge in a coastal aquifer in southern Iran. J Asian Earth Sci 30(2):364–374. doi:10.1016/j. jseaes.2006.11.002 Guindon B (2000) A framework for the development and assessment of object recognition modules from high-resolution satellite images. Can J Remote Sens 26:334–348. doi:10.1080/07038992. 2000.10874784 Gupta M, Srivastava PK (2010) Integrating GIS and remote sensing for identification of groundwater potential zones in the hilly terrain of Pavagarh, Gujarat, India. Water Int 35(2):233–245 Habboubi A (2014) Study of groundwater contamination of groundwater resource and its management proposals through GIS tools: case of the Manouba aquifer. Master report (MPGTA), Tunis El Manar University, Faculty of Sciences of Tunis (in French) Haouchine A, Boudoukha A, Haouchine FZ, Nedjai R (2011) Potential recharge mapping of the aquifer in arid climate. Eur J 45:1–13 (in French) Harrison EM, Saroj KS, Kalanithy V (2011) Multi-criteria decision analysis: a strategic planning tool for water loss management. Water Resour Manage 25:3947–3969. doi:10.1007/s11269-0119896-9 Hechmi H (1989) Carte des ressources en souterraine de la Tunisie. Ech. 1/200 000, Feuille de Tunis. Direction Ge´ne´rale des Ressources en Eau (DGRE) Herrmann F, Baghdadi N, Blaschek M, Deidda R, Duttmann R, La Jeunesse I, Sellami H, Vereecken H, Wendland F (2016) Simulation of future groundwater recharge using a climate model ensemble and SAR-image based soil parameter distributions—a case study in an intensively-used Mediterranean catchment. Sci Total Environ 543 Part B:889–905. doi:10. 1016/j.scitotenv.2015.07.036 Hosni S, Mannai H (2014) Mapping of suitable areas for recharge using GIS-multidisciplinary data: case of Tunis groundwater. End of Studies Project (ESP-LAGTE), Tunis El Manar University, Faculty of Sciences of Tunis (in French) Huang IB, Keisler J, Linkov I (2011) Multi-criteria decision analysis in environmental sciences: ten years of applications and trends. Sci Total Environ 409:3578–3594. doi:10.1016/j.scitotenv.2011. 06.022 INM (2014) Weather data of Manouba. Unpublished report, Tunis (in French)

Page 19 of 20

511

Jang CS, Chen SK, Kuo YM (2013) Applying indicator-based geostatistical approaches to determine potential zones of groundwater recharge based on borehole data. CATENA 101(2013):178–187 Jelassi F (2012) Ressources en eaux du gouvernorat de Manouba, C.R.D.A. Manouba. Inedit report Jin X, Davis CH (2005) An integrated system for automatic road mapping from high-resolution multispectral satellite imagery by information fusion. Int J Inf Fusion 6(4):257–273. doi:10.1016/j. inffus.2004.06.003 Kim SM, Choi Y, Suh J, Oh S, Park HD, Ho Yoon S, Go WR (2012) ArcMine: a GIS extension to support mine reclamation planning. Comput Geosci 46:84–95. doi:10.1016/j.cageo.2012.04.007 Lee S, Kim YS, Oh HJ (2012) Application of a weights-of-evidence method and GIS to regional groundwater productivity potential mapping. J Environ Manage 96(1):91–105 Machiwal D, Jha MK (2014) GIS-based water balance modeling for estimating regional specific yield and distributed recharge in data-scarce hard-rock regions. J Hydro-Environ Res 9(4):554–568. doi:10.1016/j.jher.2014.07.004 Madrucci V, Taioli F, Cesar de Araujo C (2008) Groundwater favourability map using GIS multi criteria data analysis on crystalline terrain, Sao Paulo State, Brazil. J Hydrol 357:153–173 Malczewski J (2004) GIS-based land-use suitability analysis: a critical overview. Prog Plann 62(1):3–65. doi:10.1016/j.progress. 2003.09.002 Murthy KSR, Mamo AG (2009) Multi-criteria decision evaluation in groundwater zones identification in Moyale-Teltele sub-basin, South Ethiopia. Int J Remote Sens 30(11):2729–2740 Nag A, Ghosh S, Biswas S, Sarkar D, Sarkar PP (2012) An image Steganography technique using X-box mapping. In: 2012 International conference on advances in engineering, science and management (ICAESM), pp 709–713. IEEE Oh HJ, Kim YS, Choi JK, Park E, Lee S (2011) GIS mapping of regional probabilistic groundwater potential in the area of Pohang City, Korea. J Hydrol 399(3):158–172 Pimienta J (1959) Le cycle plioce`ne actuel dans les bassins paraliques de Tunis, vol 85. Me´moire de la Socie´te´ Ge´ologique de France, Paris, p 176 Pini S (1971) Carte ge´ologique de Tunisie au 1/50 000. Feuille de l’Ariana, Tunis Pique´ A, Tricart P, Guiraud R, Laville E, Bouaziz S, Amrhar M, Ouali R (2002) The Mesozoic-Cenozoic Atlas belt (North Africa): an overview. Geodin Acta 15:185–208 Pradeep R (1998) Remote sensing techniques to locate groundwater potential zones in upper Urmil River Basin, District Chhatarpur—Central India. J Indian Soc Remote Sens 26:135–147. doi:10.1007/BF03026671 Pradhan B (2013) A comparative study on the predictive ability of the decision tree, support vector machine and neuro-fuzzy models in landslide susceptibility mapping using GIS. Comput Geosci 51:350–365. doi:10.1016/j.cageo.2012.08.023 Prasad RK, Mondal NC, Banerjee P, Nandakumar MV, Singh VS (2008) Deciphering potential groundwater zone in hard rock through the application of GIS. Environ Geol 55(3):467–475 Rahman MA, Rusteberg B, Gogu RC, Lobo Ferreira JP, Sauter M (2012) A new spatial multicriteria decision support tool for site selection for implementation of managed aquifer recharge. J Environ Manage 99:61–75. doi:10.1016/j.jenvman.2012.01. 003 Raj Pathak D, Hiratsuka A, Awata I, Chen L (2008) Groundwater vulnerability assessment in shallow aquifer of Kathmandu Valley using GIS-based DRASTIC model. Earth Environ Sci 57(2):1569–1578. doi:10.1007/s00254-008-1432-8

123

511

Page 20 of 20

Saaty TL (1977) A scaling method for priorities in hierarchical structures. J Math Psychol 15:234–281. doi:10.1016/00222496(77)90033-5 Saidi S, Bouri S, Ben Dhia H, Anselme B (2011) Assessment of groundwater risk Using intrinsic vulnerability and hazard mapping: application to Souassi aquifer, Tunisian Sahel. Agric Water Manag 98:1671–1682. doi:10.1016/j.agwat.2011.06.005 Senanayake IP, Dissanayake DMDOK, Mayadunna BB, Weerasekera WL (2015) An approach to delineate groundwater recharge potential sites in Ambalantota, Sri Lanka using GIS techniques. Geosci Front 7(1):115–124. doi:10.1016/j.gsf.2015.03.002 (article in press) Shahid S, Nath SK, Kamal ASMM (2002) GIS integration of remote sensing and topographic data using fuzzy logic for ground water assessment in Midnapur District, India. Geocarto Int 17(3):69–74 Sophocleous MA (1991) Combining the soil water balance and waterlevel fluctuation methods to estimate natural groundwater recharge: practical aspects. J Hydrol 24:229–241 Sreedhar G, Vijaya Kumar GT, Murali Krishna IV, Ercan K, Cu¨neyd DM (2009) Mapping of groundwater potential zones in the Musi basin using remote sensing data and GIS. Adv Eng Softw 40:506–518. doi:10.1016/j.advengsoft.2008.10.001 Ting SC, Cho DI (2008) An integrated approach for supplier selection and purchasing decisions. Supply Chain Manag Int J 13:116–127 Trabelsi S (2013) Hydrogeological and structural studies of the Tebourba—Manouba, reservoirs, eastern North of Tunis. Master thesis, Faculty of Sciences Bizerte (in French)

123

Environ Earth Sci (2017)76:511 Vaidya OS, Kumar S (2006) Analytic hierarchy process: an overview of applications. Eur J Oper Res 169:1–29 Waichler SR (2005) Potential groundwater recharge from the infiltration of surface runoff in cold and dry creeks, Phase 2, Pacific Northwest National Laboratory Richland, Washington 99352, PNNL-15533. http://www.pnl.gov/main/publications/ external/technical_reports/PNNL-15533.pdf Xu DL, Yang JB (2001) Introduction to multicriteria decision making and the evidential reasoning approach. Manchester School of Management, University of Manchester Institute and Technology. https://php.portals.mbs.ac.uk/Portals/49/docs/jyang/XuYang_ MSM_WorkingPaperFinal.pdf Yang JB, Xu DL (2004) Intelligent decision system for supplier assessment, DSS 2004 Italy. https://www.researchgate.net/pub lication/228570159_Intelligent_decision_system_for_supplier_ assessment Zaidi FK, Nazzal Y, Izrar A, Naeem MM, Jafri K (2015) Identification of potential artificial groundwater recharge zones in Northwestern Saudi Arabia using GIS and Boolean logic. J Afr Earth Sci 111:156–169. doi:10.1016/j.jafrearsci.2015.07.008 Zebidi H, Chaumont M (1962) Etude hydroge´ologique de la bordure Est de Lac Se´joumi. Bureau d’Inventaire des Ressources Hydrauliques (BIRH), Tunis