In this study, the groundwater potential zones in Baringo County (Kenya) were identified using an integrated approach of Remote Sensing-Geographic Information System (RS-GIS) and Analytical Hierarchy Process (AHP) techniques, and thematic layers of rainfall, slope, lithology, soil type, land use, drainage density and lineament density. The data were validated using borehole yield and fluoride data obtained from boreholes in the area. Using RS-GIS-based methods to identify potential groundwater zones is more convenient and less expensive compared with the conventional methods. Rainfall, lithology, soils and land use/land cover were found to be the most significant factors influencing groundwater with weighted values of 31, 24, 18 and 10% respectively. The very good, good and poor potential zones covered 0.39, 65.33 and 34.27% area coverage of the study area respectively. Low yield boreholes were found in the poor groundwater potential zones and vice versa. Moderate yield boreholes were found in the good groundwater potential zone. The majority of the boreholes with acceptable fluoride concentration was found in the good groundwater potential zones, while those with high fluoride concentration levels were found in the poor groundwater potential zone. Generally, from this study, it can be concluded that remote sensing and GIS in conjunction with AHP are important tools in monitoring and evaluating groundwater resource potential areas.

  • There is overreliance on groundwater.

  • Conventional groundwater exploration methods are expensive in developing countries.

  • AHP was used to develop groundwater potential map using catchment properties and validated using an observed water quality parameter.

  • Rainfall, lithology, soils and land use/land cover were found to be the most significant factors influencing groundwater with weighted values of 31%, 24%, 18% and 10%, respectively.

  • The very good, good and poor potential zones covered 0.39%, 65.33% and 34.27% area coverage of the study area.

Graphical Abstract

Graphical Abstract
Graphical Abstract

Water is a basic requirement of all life on Earth. The increase in population and urbanization, among other factors, has necessitated growth in the agricultural and industrial sectors which demand more fresh water from groundwater. This is because of its continuous availability in large quantities and in most cases better quality.

Groundwater is a part of the hydrologic cycle. Precipitation that lands on the ground surface infiltrates into the sub-surface. The infiltrated water flows through the soil until it reaches rock material that is saturated and forms groundwater recharge. Groundwater pollution often results from improper disposal of wastes on land. Major sources include industrial and household chemicals and garbage, landfills, leaking underground oil storage tanks and pipelines, sewage sludge and septic systems.

According to Foster (1998), although groundwater is only estimated to be about 22% of the available freshwater it constitutes 97% of the freshwater used for human consumption. This has been motivated by inadequate quantity and quality of surface water. This has made groundwater to be an important source of water supply for both urban and rural areas in many countries (Todd & Mays 2005).

In many developing countries, high population growth and water scarcity are the two common challenges facing planners and decision-makers in the development of water resources. Continued focus on groundwater resources has increased in the recent past, leading to its over-extraction. This could eventually cause problems that include lowered groundwater levels, depletion of groundwater, water pollution and deterioration in quality. To ensure sustainability, it is necessary to undertake proper planning and investigations and apply novel groundwater exploitation techniques.

Conventional groundwater exploration methods, however, are expensive, time consuming and present many problems such as scarcity of existing data (Shaban 2010). Development of groundwater resources calls for studies and surveys that identify areas with high potential for exploration (Israil et al. 2006; Jha et al. 2010). Groundwater mapping, a graphical representation of occurrence and distribution of groundwater within a geographical region, is a promising GIS and remote sensing approach. It endeavors to provide information on availability of groundwater by serving as an efficient tool for detailed ground-based hydrogeological surveys, which ultimately lead to the selection of suitable sites for boreholes. A basis for understanding the relationship between groundwater, geological and hydrological environment is also provided through groundwater mapping.

A review of literature reveals that researchers have been using different methods to delineate the groundwater potential zones and its mapping. Some researchers have applied probabilistic models such as frequency ratio (Razandi et al. 2015), multi-criteria decision analysis (Pradhan 2009), weights of evidence (Lee et al. 2012; Pourghasemi & Beheshtirad 2015), logistic regression (Pourtaghi & Pourghasemi 2014), evidential belief function (Nampak et al. 2014), certainty factor (Razandi et al. 2015), decision tree (Chenini & Mammou 2010), artificial neural network model (Lee et al. 2012), Shannon's entropy (Naghibi et al. 2015) and machine learning techniques such as random forest and maximum entropy (Rahmati et al. 2016).

Effectiveness in spatial data and decision-making in areas including engineering and environment has been demonstrated with the advancement in technology and geospatial tools. These tools enable analysis of large spatial, temporal and spectral data as well as understanding of complex features through simulation modeling (Stafford et al. 2008; Gitas et al. 2014; Pinto et al. 2015).

Remote sensing and GIS constitute a powerful tool that can be used for fast estimation of natural resources. The method is cost effective and can effectively be used for groundwater exploration (Jha et al. 2007).

Several studies have been carried out on this aspect which reiterates the use of remote sensing and GIS techniques for mapping groundwater potential zones in different parts of the world (Jaiswal et al. 2003; Solomon & Quiel 2006; Madrucci et al. 2008; Prasad et al. 2008; Yeh et al. 2009; Dar et al. 2010; Saha et al. 2010).

The worldwide scarcity of surface water (particularly in developing countries) has led to dependency on groundwater resources for industrial, irrigation, household, livestock among other uses. The increasing population growth is likely to worsen the water shortage particularly when adverse impacts of climate change expected in the near future are considered (Al-Bakri et al. 2013). Without developing alternative water resources and instituting novel management strategies, the expected annual water deficit will continue to rise.

Large numbers of boreholes have failed to achieve the desired high yield owing to inadequate groundwater investigations due to insufficient funds. This situation is equally depicted in Baringo County, Kenya. This study aimed to derive a groundwater potential map which would provide a simpler way to locate groundwater potential zones, hence saving on cost, time and other resources. The specific objectives of the study are: (i) to generate thematic maps for rainfall, soil, lithology, drainage density, lineament density, slope and land use/cover; (ii) to derive rank and weightage wise thematic maps; (iii) to delineate the groundwater potential zones; (iv) to assess the water quality (Fluoride concentrations) of the study area; and (v) to validate the groundwater potential zones.

AHP and GIS techniques were used for delineating the groundwater potential zones map. The resulting groundwater potential map was validated using measured yield and fluoride levels data. The Analytical Hierarchy Process is useful for reducing complex decisions to a series of pairwise comparisons and then synthesizing the results (Mallick et al. 2014; Rahmati et al. 2014). Additionally, the AHP tool is a suitable technique for evaluating the consistency of the result, consequently reducing the bias in the decision-making process (Sharma et al. 2012).

Study area

Baringo County lies between latitudes 00°13″ South and 1°40″ North and longitudes 35°36″ and 36°30″ East. The County has six sub-counties: Baringo Central, Baringo South, Baringo North, Eldama Ravine, Mogotio and Tiaty, as illustrated in Figure 1. The County covers a total area of 11,015 km2, with about 165 km2 being under water (Lake Baringo, Lake Bogoria and Lake Kamnarok). The climate of Baringo varies from humid highlands to arid lowlands, while some regions are between these extremes (Odada et al. 2006). The County covers a range of climatic zones, from semi-arid, arid, and very arid through semi-humid and sub-humid, to a small portion in the humid zone. The mean annual rainfalls in these zones are 450–900 mm (semi-arid), 800–1,400 mm (semi-humid), 1,000–1,600 mm (sub-humid) and 1,100–2,700 mm (humid).
Figure 1

Map of Baringo County, Kenya.

Figure 1

Map of Baringo County, Kenya.

Close modal

The methodology adopted for this study consisted of three major phases: pre-fieldwork, fieldwork and post fieldwork activities.

Pre-field and fieldwork activities

The pre-fieldwork activities included the review of previous studies on the subject and reconnaissance of the study area. Collection of various data such as station rainfall, borehole yields and fluoride levels was also conducted in the fieldwork activities. Geological and structural investigations were conducted in the fieldwork for identifying the major geological units and structural configurations in the study area. As a result, geological map of the area was produced. Hydrogeological field investigation was done by giving more attention on differentiating the sediments and rock units of groundwater significance such as the degree of fracturing of the rock units, space between fractures and opening space of fractures in the rocks.

Post-fieldwork activities

In the post-fieldwork, data processing and analysis of primary and secondary data obtained from pre-field and fieldwork was the main activity. Some of the important activities were: delineation of the study area by extracting the catchment from topographic maps and digital elevation model (DEM) images using GIS software. GPS readings from different locations of the study area were used for further ground checking. After delineating the study area, thematic maps for each variable were generated. Weighted value estimation for each thematic map was computed since all these factors have different influencing values to groundwater occurrence. The weighted value estimation was conducted using the multi-criteria evaluation decision analysis. Finally, these thematic maps were overlain using the GIS software to develop the groundwater potential map. There was need for further validation to check whether the selected suitability classes were correct or not. Accordingly, point data on bore hole yields collected across the study area were used for accuracy assessment of the groundwater potential map.

Generation of thematic maps

Development of thematic layers involved digital image processing of remote sensed data and digitization of existing maps and field data for extraction of pertinent information. The existing lithology map from Kenya Survey was scanned, rectified and digitized in GIS software to prepare the thematic layer of lithology. The soil map from Kenya Soil and Terrain (KENSOTER) database was also scanned, rectified and then digitized using GIS software to obtain the soil thematic map. The slope map was generated from a DEM using the spatial analyst tool. The slopes were determined by applying partial derivatives in x and y direction by fitting surfaces to neighborhoods of pixel and by linear filtering.

A drainage network map was generated from the data on Kenyan rivers. Subsequently, the drainage density map was prepared from the drainage map. The rainfall map was generated by taking the average annual rainfall amount from the rainfall stations. These amounts of rainfall were interpolated using GIS to generate the areal rainfall map for the whole study area.

The land-use/land-cover map was generated from Landsat imagery. Supervised classification was performed using Bayesian Maximum Likelihood Classifier (MLC). After the classification, accuracy assessment of the land-use map was substantiated by correlating ground truth information.

The lineament density map was prepared by sub-dividing the study area into a number of grids of dimension 1 × 1 km. Density of the lineaments of a single grid was obtained from the values of the ratio of total length of the lineaments in a single grid (L) and the area of that single grid (A). By calculating the value of L/A (density) for each grid, the value was plotted at the center of each grid. The lineament density map was prepared by joining the fields of equal density values.

These thematic maps were converted into raster format. Raster classification was performed for them to demarcate their classes. Table 1 shows a summary of the data sources and resolutions.

Table 1

Data, their sources and some attributes

Data typeSourceResolution/period
Borehole locations, their fluoride concentration and yields Central Rift Valley Water Works Development Agency 2018–2020 
Land use/land cover United States Geological Survey (USGS) satellite image 30 m × 30 m, 2020 
Soil Kenya Soil and Terrain (KENSOTER) database 2018 
Slope United States Geological Survey website, digital elevation model (DEM) 30 m × 30 m, 2020 
Drainage density Basins and sub-basins in Kenya database 2018 
Lineament density United States Geological Survey website 2018–2020 
Lithology Survey of Kenya 2018 
Rainfall Kenya Meteorological Department 2010–2020 
Data typeSourceResolution/period
Borehole locations, their fluoride concentration and yields Central Rift Valley Water Works Development Agency 2018–2020 
Land use/land cover United States Geological Survey (USGS) satellite image 30 m × 30 m, 2020 
Soil Kenya Soil and Terrain (KENSOTER) database 2018 
Slope United States Geological Survey website, digital elevation model (DEM) 30 m × 30 m, 2020 
Drainage density Basins and sub-basins in Kenya database 2018 
Lineament density United States Geological Survey website 2018–2020 
Lithology Survey of Kenya 2018 
Rainfall Kenya Meteorological Department 2010–2020 

Derivation of rank and weightage wise thematic maps

The multi-criteria decision analysis (MCDA) is a generally recognized and highly appropriate technique for complex decision-making problems. The Analytical Hierarchy Process helps to find out the weight of criterion by pairwise comparison and is based on the judgment of common people to derive priority weights (Saaty & Vargas 2001; Das & Pal 2019).

Other approaches such as multi-attribute value theory (MAVT) and multi-attribute utility theory (MAUT) have been used in derivation of ranks and decision-making. The MAVT and MAUT approaches produce more inclusive information (assessment of the relative importance of each criterion as well as value function characterization with the level of satisfaction by each alternative under each criterion) than AHP. However, these methods are more difficult to use. The AHP approach is easier, flexible and requires less skills than MAVT and MAUT (Ananda & Herath 2009). In most MAVT and MAUT studies, the decision-makers are experts, except in valuation studies where the inputs are provided by the common people. It is feasible to use MAVT and MAUT by innovative adaptations such as the use of utility indices. However, Ananda & Herath (2009) reported that MAUT and MAVT approaches have not been widely documented.

A pairwise comparison matrix (PCM) for the thematic maps was calculated based on Saaty's 1–9 scale weights as described in Ouma & Tateishi (2014) and summarized in Table 2. The nine points are chosen because psychologists conclude that nine objects are the most that an individual can simultaneously compare and consistently rank. Pairwise judgments are made based on the best information available and the decision-maker's knowledge and experience. Moreover, there are other scale methods such as the 5-point, 7-point, and 10-point scales.

Table 2

Nine-point intensity of importance scale (after Ouma & Tateishi 2014)

Intensity of importanceDefinitionDescription
Equally important Two factors contribute equally to the objective. 
Moderately more important Experience and judgment slightly favor one over the other. 
Strongly more important Experience and judgment slightly favor one over the other. 
Very strongly more important Experience and judgment slightly favor one over the other. Its importance is demonstrated in practice. 
Extremely more important The evidence favoring one over the other is of the highest possible validity. 
Intensity of importanceDefinitionDescription
Equally important Two factors contribute equally to the objective. 
Moderately more important Experience and judgment slightly favor one over the other. 
Strongly more important Experience and judgment slightly favor one over the other. 
Very strongly more important Experience and judgment slightly favor one over the other. Its importance is demonstrated in practice. 
Extremely more important The evidence favoring one over the other is of the highest possible validity. 

Intermediate values (2, 4, 6, and 8) are applied when a compromise is needed. If an element i has one of the above numbers assigned to it when compared with element j, then j has the reciprocal value when compared with i. If the activities are very close, e.g. ratios 1.1–1.9, there may be difficulty in assigning the best value. However, when compared with other contrasting activities the size of the small numbers would not be too noticeable, yet they can still indicate the relative importance of the activities.

The process of AHP can be summarized in four steps: construct the decision hierarchy; determine the relative importance of attributes and sub-attributes; evaluate each alternative and calculate its overall weight in regard to each attribute, and check the consistency of the subjective evaluations. In the first step, the decision is decomposed into its independent elements. Secondly, the user is asked to subjectively evaluate pairs of attributes on a nine-point scale. In the third stage, a weight is calculated for each attribute (and sub-attribute), based on the pairwise comparisons. Since judgments are given subjectively by the user, the logical consistency of these evaluations is tested in the last stage. The ultimate outcome of the AHP is a relative score for each decision alternative, which can be used in the subsequent decision-making process.

The weights were assigned according to their potential for groundwater recharge. High normalized weight values were assigned to the themes of very good groundwater potential, while the low weight values were assigned to poor groundwater potential themes.

The attributes of each of the thematic map were assigned a rank of 1–5 (1, 2, 3, 4, 5 denoted very high, high, moderate, low, and very low prospects for groundwater respectively).

Delineation of the groundwater potential zones

All the maps were converted into raster format and georeferenced to common reference point in the Universal Transverse Mercator plane coordinate system. The seven different thematic maps were then integrated to generate the groundwater potential map for the study area. The ground water potential index was computed as suggested by Hapuarachchi et al. (2011). Thus:
formula
where: G, Lu, D, Sl, S, L and R represents geology, land use/land cover, drainage density, slope, soil type, lineament density and rainfall, respectively

w is the normalized weight of each thematic map.

wi is the normalized weight of each class.

Assessment of the fluoride concentration

A map of fluoride concentration levels for the existing borehole data was generated. Samples of water pumped from three selected boreholes were collected in clean labelled plastic bottles and stored in a lightproof insulated box. The samples were tested for fluoride levels using photometer procedures on drinking water. This test was done within 6 hours from the time of sampling. The colorimetric method was used. The results were then compared with the WHO standards on permissible levels (WHO 2002).

Validation of the groundwater potential map

The delineated groundwater potential zone map was verified using yield and fluoride concentration data of 85 boreholes from Central Rift Valley Water Works Development Agency (CRVWWDA). This Agency is mandated by the Government of Kenya to provide quality water services and therefore has all records on boreholes. Three boreholes used to validate the existing data collected from CRVWWDA and were named borehole 1, 2 and 3. The borehole yield points and their fluoride concentrations were overlain onto the final groundwater potential map to check the accuracy of the study in the various groundwater potential zones.

A summary of the techniques and methods used in the generation of the groundwater potential map is illustrated in the Figure 2.
Figure 2

Flow diagram for the materials and methods.

Figure 2

Flow diagram for the materials and methods.

Close modal

Ranking criteria of the groundwater mapping

The ranking and prioritization process is the main purpose of AHP based multi-criteria decision-making. When making decisions, hydrologists and engineers frequently use experiential judgments from the public who are the end-users. In this study, to determine the objectives and formulate the decision-making process, 12 experts comprising of four hydrologists, four engineers and four end-users were asked to give their assessments and judgments regarding the variables related to groundwater and their significances in terms of weights, out of the seven factors analyzed. Experts or decision-makers comprise those with the technical skills for solving a given problem, while end-users are the public who are affected by the phenomenon and for this case study comprised of representatives from the community leaders. Each of the expert participants assigned weights to the objective factors in three steps, with each step using a different approach comprising of the following steps:

Step 1: Assign each objective factor a percentage to indicate the weight;

Step 2: Use step 1 to indicate the lowest importance, and assume the importance scale among the objectives is linear;

Step 3: The importance of objectives should be ranked using a 1–5 scale, with 1 representing the most important and 5 representing the least important.

Table 3 shows the PCM.

Table 3

Pairwise comparison matrix for the AHP process

Groundwater potential influencing factors
RGSLuSlLD
Rainfall, R 1.00 2.00 4.00 3.00 5.00 6.00 8.00 
Geology, G 0.5 1.00 3.00 2.00 5.00 6.00 7.00 
Soils, S 0.25 0.33 1.00 2.00 3.00 7.00 8.00 
Land use/land cover, Lu 0.33 0.50 0.50 1.00 7.00 5.00 8.00 
Slope, Sl 0.20 0.20 0.33 0.14 1.00 0.33 4.00 
Lineament density, L 0.17 0.17 0.14 0.20 3.00 1.00 2.00 
Drainage density, D 0.13 0.14 0.13 0.13 0.25 0.50 1.00 
Groundwater potential influencing factors
RGSLuSlLD
Rainfall, R 1.00 2.00 4.00 3.00 5.00 6.00 8.00 
Geology, G 0.5 1.00 3.00 2.00 5.00 6.00 7.00 
Soils, S 0.25 0.33 1.00 2.00 3.00 7.00 8.00 
Land use/land cover, Lu 0.33 0.50 0.50 1.00 7.00 5.00 8.00 
Slope, Sl 0.20 0.20 0.33 0.14 1.00 0.33 4.00 
Lineament density, L 0.17 0.17 0.14 0.20 3.00 1.00 2.00 
Drainage density, D 0.13 0.14 0.13 0.13 0.25 0.50 1.00 

The relative weight matrix and normalized weights were determined based on PCM, to determine the percentage of impact of the themes as indicated in Table 4.

Table 4

Normalized weights for thematic layers

Groundwater potential influencing factors
Normalized weightPercentage influence
RGSLuSlLD
Rainfall, R 0.39 0.46 0.44 0.35 0.21 0.23 0.21 0.31 31 
Geology, G 0.19 0.23 0.33 0.24 0.21 0.23 0.18 0.24 24 
Soils, S 0.10 0.08 0.11 0.24 0.12 0.27 0.21 0.18 18 
Land use/land cover, Lu 0.13 0.0.12 0.05 0.12 0.29 0.19 0.21 0.10 10 
Slope 0.08 0.05 0.04 0.02 0.04 0.01 0.11 0.06 
Lineament density, L 0.07 0.04 0.02 0.03 0.12 0.04 0.05 0.06 
Drainage density, D 0.05 0.03 0.01 0.02 0.01 0.02 0.03 0.04 
Groundwater potential influencing factors
Normalized weightPercentage influence
RGSLuSlLD
Rainfall, R 0.39 0.46 0.44 0.35 0.21 0.23 0.21 0.31 31 
Geology, G 0.19 0.23 0.33 0.24 0.21 0.23 0.18 0.24 24 
Soils, S 0.10 0.08 0.11 0.24 0.12 0.27 0.21 0.18 18 
Land use/land cover, Lu 0.13 0.0.12 0.05 0.12 0.29 0.19 0.21 0.10 10 
Slope 0.08 0.05 0.04 0.02 0.04 0.01 0.11 0.06 
Lineament density, L 0.07 0.04 0.02 0.03 0.12 0.04 0.05 0.06 
Drainage density, D 0.05 0.03 0.01 0.02 0.01 0.02 0.03 0.04 

CI = 0.1, n = 7, RI = 1.32, λmax = 1.6, Consistency ratio, CR = 0.076 < 0.1.

The principle is to check that the consistency ratio (CR) is not equal to 0.1. This indicates a satisfactory reciprocal matrix. When the ratio is above 0.1, it indicates a change of the PCM thus corresponding weights should be re-examined to prevent inconsistency (Rajasekhar et al. 2019).

CR is calculated as:
formula
where CI is the consistency index, and RI is the random index and is dependent on the order of the matrix as shown in Table 5 (Ouma & Tateishi 2014).
Table 5

Random index (RI) used to compute consistency ratios (CR)

N 10 
RI 0.58 0.90 1.12 1.24 1.32 1.41 1.45 1.49 
N 10 
RI 0.58 0.90 1.12 1.24 1.32 1.41 1.45 1.49 
CI is calculated as:
formula
where;

λmax is a principal eigenvalue;

n is the number of factors.

Therefore, CR is calculated as:
formula

The obtained CI is lower than the threshold value of 0.1 and indicates a high level of consistency in the pairwise judgments, implying that the determined weights are acceptable. The computed eigenvector is used as a coefficient for the respective factor maps to be combined in the weighted overlay.

Weighting and ranking of the groundwater influencing factors

The PCM and the factor maps are used in this step. The principal eigenvector of the pairwise comparison matrix is figured out to produce a best fit to the weight set. Weight values represent the priorities which are absolute numbers between zero and one. Using a weighted linear combination, it implies that the weights sum to 1.

A summary of the groundwater potential variables showing the various factors, their respective weights and how they are ranked according to their influence to groundwater is presented in Table 6.

Table 6

Weighted groundwater potential ranking

Groundwater influencing factorRelative weightClassesRanking
Land use/land cover 0.31 Water bodies 
Agricultural land 
Forest cover 
Fallow land 
Urban areas 
Rainfall (mm/annum) 0.24 1,183–1,418 
1,003–1,183 
832–1,003 
678–832 
482–678 
Slope (%) 0.18 <3 
3–8 
8–14 
14–22 
>22 
Geology 0.1 Sedimentary rocks 
Igneous rocks 
Metamorphic rocks 
Soil type 0.6 Sandy 
Loamy 
Clayey 
Very clayey 
Lineament density 0.6 0–1.55*10−4 
1.56*10−4–3.21*10−4 
3.22*10−4–5.16*10−4 
5.17*10−4–7.71*10−4 
7.72*10−4–1.28*10−3 
Drainage density 0.4 6.9–28.3 
28.4–38.4 
38.5–47.6 
47.7–57.7 
57.8–82.6 
Groundwater influencing factorRelative weightClassesRanking
Land use/land cover 0.31 Water bodies 
Agricultural land 
Forest cover 
Fallow land 
Urban areas 
Rainfall (mm/annum) 0.24 1,183–1,418 
1,003–1,183 
832–1,003 
678–832 
482–678 
Slope (%) 0.18 <3 
3–8 
8–14 
14–22 
>22 
Geology 0.1 Sedimentary rocks 
Igneous rocks 
Metamorphic rocks 
Soil type 0.6 Sandy 
Loamy 
Clayey 
Very clayey 
Lineament density 0.6 0–1.55*10−4 
1.56*10−4–3.21*10−4 
3.22*10−4–5.16*10−4 
5.17*10−4–7.71*10−4 
7.72*10−4–1.28*10−3 
Drainage density 0.4 6.9–28.3 
28.4–38.4 
38.5–47.6 
47.7–57.7 
57.8–82.6 

A higher weight value of the factors represents more priority or more impact than others within the study. From the factor weights found for this study area, it is clear that rainfall and geology have the highest weights, implying that they have more influence on groundwater in the area compared with the other factors.

Evaluating physical and environmental factors influencing groundwater potential

Soil

Soil properties influence the relationship between runoff and infiltration rates which in turn control the degree of permeability that determines the groundwater potential (Tesfaye 2010). Soil texture is the most important component and characteristic of soils. Soil texture has a great impact on groundwater because sandy soils absorb water fast and generate little runoff.

The soil in the study area was classified using the Unified Soil Classification System (USCS), on the basis of infiltration capacity. Four soil categories namely sandy, loamy, clay and very clayey were identified. Table 7 shows the area covered by different types of soil.

Table 7

Area coverage for soil classes

Soil typeRankInfluence to groundwaterArea covered in Km2% area
Sandy Very good 495.846 5.55 
Loamy Good 946.5183 10.60 
Clay Poor 3,454.6473 38.69 
Very clayey Very poor 3,884.4819 43.50 
Soil typeRankInfluence to groundwaterArea covered in Km2% area
Sandy Very good 495.846 5.55 
Loamy Good 946.5183 10.60 
Clay Poor 3,454.6473 38.69 
Very clayey Very poor 3,884.4819 43.50 

The very clayey soils occupy most parts of the County with a 43.5% area coverage. These soils are experienced around Lake Baringo and the western parts of the County (Figure 3). The clay soils occupy the second largest area (38.69%) in the study area. These soils are mostly situated in the eastern parts of the study area. The sandy and loamy soils occupy the least parts of Baringo with an area coverage of 5.55 and 10.6%, respectively.
Figure 3

Ranked soil map.

Figure 3

Ranked soil map.

Close modal

Considering recharge from rainfall, this implies that just a small area of the County (about 16%) has a high potential for groundwater due to the expected higher infiltration rates for the sandy and loamy soils. The remaining regions, about (84%), have low groundwater potential as they are covered by the clay and very clayey types of soil, which have lower groundwater infiltration rates (Tesfaye 2010). The ranked soil map is shown in Figure 3.

Land use/land cover

Land use/land cover plays a vital role in groundwater prospecting. Land use and land cover affect the rate of recharge, runoff, groundwater abstraction and evapotranspiration. The rate of infiltration is affected by the density of vegetation cover (Singh 2014). When the ground surface is covered with dense forests, the infiltration rate is higher while the runoff rate becomes lower. Land use/land cover gives essential information on infiltration, soil moisture, groundwater and surface water (Ibrahim & Ahmed 2016).

Supervised classification was performed to classify the types of land use/land cover and five classes were identified as shown in Table 8. The ranks of the various land use land cover classes in relation to groundwater potential was also developed as shown in Figure 4. The land use/land cover (LULC) classes like forest cover and agriculture land hold substantially high proportion of water than urban areas and fallow land (Rajaveni et al. 2017).
Table 8

Area coverage for land-use/land-cover classes

LULC classesRankInfluence to groundwaterArea covered in km2% area
Water bodies Very good 894.7827 8.12 
Agricultural land Good 2,187.7182 19.86 
Forest cover (shrubs and bush lands) Moderate 432.8487 3.93 
Fallow land Poor 7,375.2462 66.97 
Urban areas Very poor 122.7105 1.11 
LULC classesRankInfluence to groundwaterArea covered in km2% area
Water bodies Very good 894.7827 8.12 
Agricultural land Good 2,187.7182 19.86 
Forest cover (shrubs and bush lands) Moderate 432.8487 3.93 
Fallow land Poor 7,375.2462 66.97 
Urban areas Very poor 122.7105 1.11 
Figure 4

Ranked land-use/land-cover map.

Figure 4

Ranked land-use/land-cover map.

Close modal

Approximately 4% of the total area is covered with forest cover. Agricultural land and water bodies occupy approximately 20 and 8% of the study area, respectively. Fallow lands occupy around 67% of the region. The remaining 1% of the study area is covered by urban areas.

The areas around water bodies and under agriculture, constituting 28%, are likely to have a higher potential for groundwater. About 68% of the County is likely to have a low potential for groundwater since they are covered by fallow lands and urban areas.

Slope

Slope plays an important role in governing the stability of a terrain. Slope influences the direction and amount of surface runoff or subsurface drainage reaching a site. Slope has a dominant effect on the contribution of rainfall to stream flow. It controls the duration of overland flow, infiltration and subsurface flow.

Flat and gentle sloping areas promote infiltration and groundwater recharge, while steep sloping grounds encourage runoff and little infiltration (Ettazarini 2007). Groundwater potential is expected to be greater in the flat and gentle sloping area (Solomon 2003). Flatter topography therefore gives more prospect for groundwater accumulation.

The area coverage of each class of slope in the County is presented in Table 9.

Table 9

Area coverage for slope classes

Slope classes in %RankInfluence to groundwaterArea covered in km2% area
<3 Very good 183.2805 1.76 
3–8 Good 5,611.4172 53.87 
8–14 Moderate 703.8288 6.76 
14–22 Poor 1,399.797 13.44 
>22 Very poor 2,519.0865 24.18 
Slope classes in %RankInfluence to groundwaterArea covered in km2% area
<3 Very good 183.2805 1.76 
3–8 Good 5,611.4172 53.87 
8–14 Moderate 703.8288 6.76 
14–22 Poor 1,399.797 13.44 
>22 Very poor 2,519.0865 24.18 

Very gentle slopes (<3%) are experienced mostly in the areas covered with water bodies. This implies that areas around these regions possibly have a higher groundwater potential since most water is infiltrated due to the gentle slopes. Baringo County has 24.18% of its area with a steep slope of >22%. With regards to the groundwater potential, these locations could have a lower groundwater potential since steep areas have a lower capability of holding rainfall that would facilitate groundwater recharge (Sisay 2007).

Figure 5 presents the ranked slope map.
Figure 5

Ranked slope map.

Figure 5

Ranked slope map.

Close modal

Lineament density

Lineaments are natural, linear surface elements, interpreted from satellite imagery. They are features with secondary permeability and influence movement and storage of groundwater (Subba et al. 2001).

The lineament density classes ranged from 0 to1.28 × 10−3 m/m. The highest satellite lineament density class (7.72 × 10−4–1.28 × 10−3 m/m) was observed mostly in the central parts of Baringo, as presented in Figure 6. The lowest lineament density class ranging from 0 to 1.55 × 10−4 occupied 33.82% of the study area. According to Al-Abadi & Al-Shamma'a (2014), high lineament density indicates high porosity, thus representing a zone with high groundwater potential. It was therefore concluded that these regions with the lowest lineament density class were of low susceptibility to groundwater potential (Bhuvaneswaran et al. 2015).
Figure 6

Ranked lineament density map.

Figure 6

Ranked lineament density map.

Close modal
Figure 7

Drainage network map.

Figure 7

Drainage network map.

Close modal

Table 10 shows the areas covered by each lineament density class in the study area.

Table 10

Area coverage for lineament density classes

Lineament density classesRankInfluence to groundwaterArea covered in km2% area
0–1.55 × 10−4 Very poor 3,725.0379 33.82 
1.56 × 10−4–3.21 × 10−4 Poor 2,897.8479 26.31 
3.22 × 10−4–5.16 × 10−4 Moderate 2,497.2579 22.68 
5.17 × 10−4–7.71 × 10−4 Good 1,356.4386 12.32 
7.72 × 10−4–1.28 × 10−3 Very good 536.6673 4.87 
Lineament density classesRankInfluence to groundwaterArea covered in km2% area
0–1.55 × 10−4 Very poor 3,725.0379 33.82 
1.56 × 10−4–3.21 × 10−4 Poor 2,897.8479 26.31 
3.22 × 10−4–5.16 × 10−4 Moderate 2,497.2579 22.68 
5.17 × 10−4–7.71 × 10−4 Good 1,356.4386 12.32 
7.72 × 10−4–1.28 × 10−3 Very good 536.6673 4.87 

Drainage density

Drainage density is the closeness of spacing of stream networks. Groundwater potential is poor in areas with high drainage density since water is lost in the form of runoff (Terzer et al. 2013). On the other hand, areas with low drainage density enhance infiltration (due to a higher concentration time) and therefore have higher potential for groundwater (Murasingh 2014).

In the study area, five main drainage density classes were identified. The values of the drainage density varied from 6.9 to 82.6 m/m. Table 11 shows the drainage density classes in Baringo County. The drainage pattern in the south of the study area was found to be dendritic while that in the northern part sub-parallel as shown in Figure 7. The dendritic drainage pattern had a high influence to groundwater potential while the sub-parallel had a low influence to groundwater potential.

Table 11

Area coverage for drainage density classes

Drainage density classesRankInfluence to groundwaterArea covered in km2% area
6.9–28.3 Very good 837.642 8.10 
28.4–38.4 Good 1,351.60 13.07 
38.5–47.6 Moderate 2,021.44 19.54 
47.7–57.7 Poor 2,703.53 26.13 
57.8–82.6 Very poor 3,430.36 33.16 
Drainage density classesRankInfluence to groundwaterArea covered in km2% area
6.9–28.3 Very good 837.642 8.10 
28.4–38.4 Good 1,351.60 13.07 
38.5–47.6 Moderate 2,021.44 19.54 
47.7–57.7 Poor 2,703.53 26.13 
57.8–82.6 Very poor 3,430.36 33.16 

Areas classified with very high drainage density were found near the western border of the County and some central parts of the study area as presented in Figure 8. Groundwater potential is poor in areas with very high drainage density since water is lost majorly in the form of runoff (Hussein et al. 2017). Areas with low drainage density depicted comparatively higher infiltration due to lower runoff potential. Therefore, the low drainage density areas in the study area had a high groundwater potential while high drainage density regions had low groundwater potential.
Figure 8

Ranked drainage density map.

Figure 8

Ranked drainage density map.

Close modal

Lithology

Lithology is another factor controlling the quantity and quality of groundwater occurrence in a given area (Ayazi et al. 2010). Geologic setting plays a vital role in the occurrence and distribution of groundwater in any terrain (Yeh et al. 2016). Lithology influences the porosity and permeability of aquifer rocks. Higher porosity contributes to higher groundwater storage as higher permeability contributes to higher groundwater yields (Adiat et al. 2012).

The several rock types in the study area were classified into igneous, metamorphic and sedimentary rocks as indicated in Table 12. Their respective area coverages are presented in Table 13.

Table 12

Classification of rock types

Sedimentary rocks-rank 1Igneous rocks-rank 2Metamorphic rocks-rank 3
Eolian-unconsolidated rock
Sandstone, greywacke, arkose
Fluvial
Andesite
Lacustrine unconsolidated rock 
Igneous rock
Intermediate igneous rock
Pyroclastic unconsolidated
Andesite-trachite-phonolite rock
Basalt
Granite 
Gneiss-migmatite
Organic unconsolidated rock quartzite 
Sedimentary rocks-rank 1Igneous rocks-rank 2Metamorphic rocks-rank 3
Eolian-unconsolidated rock
Sandstone, greywacke, arkose
Fluvial
Andesite
Lacustrine unconsolidated rock 
Igneous rock
Intermediate igneous rock
Pyroclastic unconsolidated
Andesite-trachite-phonolite rock
Basalt
Granite 
Gneiss-migmatite
Organic unconsolidated rock quartzite 
Table 13

Area coverage for lithology classes

Rock typeRankInfluence to groundwaterArea covered in km2% area
Sedimentary rocks Good 1,917.03 17.41 
Igneous rocks Moderate 8,669.55 78.72 
Metamorphic rocks Poor 246.05 2.23 
Rock typeRankInfluence to groundwaterArea covered in km2% area
Sedimentary rocks Good 1,917.03 17.41 
Igneous rocks Moderate 8,669.55 78.72 
Metamorphic rocks Poor 246.05 2.23 

Igneous, sedimentary and metamorphic rocks covered about 78.72, 17.41 and 2.23% of the County. Metamorphic rocks were sparingly found in the south eastern and north western regions of the County as presented in Figure 9.
Figure 9

Ranked lithology map.

Figure 9

Ranked lithology map.

Close modal

Characteristics such as weathering, types of rocks, origin and occurrence, were given importance while assigning the weight. According to the rock characteristics, higher weights were assigned for sandstone, fluvial and andesite rocks. Moderate and low weights were assigned for basalt, migmatite, and quartzite rocks. Unconsolidated sedimentary and fractured crystalline rocks are more favorable for groundwater movement and storage than massive type of rocks (Balamurugan et al. 2016). The sedimentary rocks were taken to have the highest prospects for groundwater potential. It was therefore projected that the central and some parts of North West areas of the study area were likely to have higher groundwater prospects.

Rainfall

Rainfall is the major water source in the hydrological cycle and the most dominant influencing factor in the groundwater of an area. Infiltration depends on the intensity and duration of rainfall. High intensity and short duration rain results into less infiltration and more surface runoff. Conversely, low intensity and long duration rain causes high infiltration (Ibrahim & Ahmed 2016).

A mean annual rainfall for ten years (2010–2020) was considered and interpolated using Inverse Distance Weighting (IDW) to create a continuous raster rainfall data for the study area. The resulting raster layer was finally reclassified into the five classes using an equal interval. The reclassified rainfall data were given a value 1 for the highest rainfall and 5 for the least rainfall (Table 14).

Table 14

Area coverage for the rainfall classes

Rainfall amount (mm/annum)RankInfluence to groundwaterArea covered in km2% area
1,183–1,418 Very good 553.3 5.02 
1,003–1,183 Good 1,186.3 10.77 
832–1,003 Moderate 2,060.8 18.43 
678–832 Poor 2,648.4 24.04 
482–678 Very poor 4,568.8 41.47 
Rainfall amount (mm/annum)RankInfluence to groundwaterArea covered in km2% area
1,183–1,418 Very good 553.3 5.02 
1,003–1,183 Good 1,186.3 10.77 
832–1,003 Moderate 2,060.8 18.43 
678–832 Poor 2,648.4 24.04 
482–678 Very poor 4,568.8 41.47 

The study area experienced the highest rainfall category (1,183–148 mm/annum) in the smallest percentage of its area (5.02%), while it experienced the lowest rainfall category (472–678 mm/annum) in major parts of its area (41.47%). Figure 10 presents the ranked rainfall map for Baringo.
Figure 10

Ranked rainfall map.

Figure 10

Ranked rainfall map.

Close modal

Rainfall being the major source of groundwater recharge, it determines the amount of water that would be available to percolate into the groundwater system (Hussein et al. 2017). Therefore, rainfall controls groundwater potential in an area. From Figure 10, areas with high amount of rainfall were considered to result in higher groundwater potential compared with areas with low annual rainfall.

Groundwater potential zone map and validation

The groundwater potential map of the study area was generated by integrating seven thematic layers using weighted overlay method of GIS software. The GIS-based multi-criteria evaluation based on Saaty's AHP was used to compute the classes, weights and ranks for thematic layers. The consistency ratio value for all the thematic maps was found to be less than 0.1 (Table 4). Therefore, the judgments of the pairwise comparison within each thematic layer were acceptable (Saaty 1980).

Rainfall, lithology, soils and land use/land cover were found to be the most significant factors influencing groundwater with weighted values of 31, 24, 18 and 10%, respectively as depicted in Table 4.

The final groundwater potential map was developed using weighted linear combination of the thematic layers and reclassified into three classes with ranks 1, 2 and 3. With reference to the scale of importance in ranking, the respective classes were very good, good and poor potential zones as shown in Table 15 and Figure 11.
Table 15

Groundwater potential zone classes

Groundwater potential zoneRankArea covered in km2% area
Very good 42.688 0.39 
Good 7,011.93 65.33 
Poor 3,678 34.27 
Groundwater potential zoneRankArea covered in km2% area
Very good 42.688 0.39 
Good 7,011.93 65.33 
Poor 3,678 34.27 
Table 16

Existing and measured yield and fluoride levels

BoreholeParameterExisting dataField data collected
Yield (m3/h) 4.5 3.3 
Fluoride (mg/L) 1.2 1.2 
Yield (m3/h) 10 8.6 
Fluoride (mg/L) 0.55 0.55 
Yield (m3/h) 13 12 
Fluoride (mg/L) 1.5 1.5 
BoreholeParameterExisting dataField data collected
Yield (m3/h) 4.5 3.3 
Fluoride (mg/L) 1.2 1.2 
Yield (m3/h) 10 8.6 
Fluoride (mg/L) 0.55 0.55 
Yield (m3/h) 13 12 
Fluoride (mg/L) 1.5 1.5 
Figure 11

Groundwater potential zone map.

Figure 11

Groundwater potential zone map.

Close modal

Baringo County had about 0.39% of its area having very good groundwater potential zones. Most of its area (65.33%) had good groundwater potential, while the poor groundwater potential zones covered about 34.27%. In general, areas classified as very good and good in the groundwater map accounted for a large areal coverage. Hence, it can be inferred that the study area had high groundwater potential suitability in most of its area. However, this may change when the quality and yield of the groundwater at specific locations are considered.

Validation of groundwater potential with borehole yield

The groundwater potential map of the study area was validated by the yield data collected from 85 selected boreholes shown in Figure 12. Additionally, three boreholes were sampled to validate the existing data collected by the CRVWWDA, and the actual data in the field. The data collected from the three boreholes are shown in Table 16.
Figure 12

Borehole locations and their yields map.

Figure 12

Borehole locations and their yields map.

Close modal

There was no change in flouride levels in comparison with the existing data for the sampled boreholes. However, the yield had reduced. The reduction in yield could have been as a result of seasonal variation (rainfall pattern change) between initial drilling of these boreholes and when the field pumping test was carried out. The reduction in yield could also be due to overexploitation of the aquifer or silt deposition which causes closure of perforations in the borehole screens.

From the validated groundwater potential map, the low yield boreholes (1–4 m3/h) were found in the poor groundwater potential zones while moderate yield boreholes (5–18 m3/h) were found in the good groundwater potential zone. Additionally, the majority of the high yield boreholes (19–28 m3/h) was lying in the good groundwater potential zones as presented in Figure 13. The groundwater potential map was hence proven to be satisfactory.
Figure 13

Borehole yields overlaid on the groundwater potential map.

Figure 13

Borehole yields overlaid on the groundwater potential map.

Close modal

Validation of groundwater potential zone map with fluoride concentration levels

With reference to the acceptable fluoride concentration of 3 mg/L in Kenya, the majority of the boreholes in the very good and good groundwater potential zones were within this limit as presented in Figure 14. However, 7% of the boreholes exceeded the threshold. One-half of these boreholes was within the poor groundwater potential zones while the other half was within the good groundwater potential zones. The area coverage of the good groundwater potential, with acceptable fluoride concentration, was estimated to be about 65%.
Figure 14

Borehole fluoride concentration overlaid on groundwater potential map.

Figure 14

Borehole fluoride concentration overlaid on groundwater potential map.

Close modal

Further discussion of results

Mapping of groundwater potential zones as carried out in this study is important for decision-makers for planning and management activities. The groundwater potential assessment applied in this study consists of two phases. Firstly, the effective factors influencing groundwater are determined. Secondly, AHP in conjunction with GIS is applied, and these approaches are evaluated in finding the groundwater potential areas. From the study results, it is observed that AHP offers a flexible, step-by-step and transparent way of analyzing complex problems in a MCDA environment (based on experts and end-user preferences, knowledge, and judgments). In a typical MCDA situation, in which multiple criteria and different fields are involved, there is a need to consider multiple stakeholders and wide-ranging expertise. Owing to its ability to readily incorporate multiple judgments, AHP and its combination with other tools such as GIS offer a solution to multi-dimensional and multi-parametric complex problems. Indeed, the application of AHP as a decision support tool or weighting method for GIS-based MCDA marks its potential usefulness to wider applications in which multi-parametric geospatial analysis is involved.

Furthermore, the AHP-GIS-based approach to groundwater potential assessment as applied in this study is seen as a relatively inexpensive, easy to use, and more importantly, allows interactive use by groundwater managers for continuing improvement. This is because the AHP rating indices and numerical results can be used as a reference for groundwater potential monitoring.

However, AHP has some challenges. Data used for assessment of groundwater are derived from different sources, in different formats, periods and resolutions. Therefore, it may be difficult to standardize the datasets for assessing groundwater. This however may not significantly affect the quality of ranking the groundwater influencing factors because they are weighted differently. Review-wise, researchers have still not reached a consensus on some issues related to the implementation of AHP as a weighting method for GIS-based MCDA. The areas of contention include: (i) the method for capturing experts' opinions using the pairwise comparison method; (ii) the method for aggregating individual expert ratings (in cases in which consensus ratings are not used); and (iii) the method for standardizing the criteria or factors involved in the analysis. Nevertheless, the AHP method has received considerable attention because it also places greater emphasis on the structure of the preferences of the decision-makers.

This paper presents an empirical approach for mapping groundwater potential through the integration of AHP and GIS techniques. The groundwater potential map developed in this study is very useful for planners, policy/decision-makers, researchers and engineers seeking suitable locations to drill boreholes.

The groundwater potential map has been derived using a multi-parametric approach that combines physical and socio-economic factors. Rainfall, lithology, soils and land use/land cover were found to be the most significant factors influencing groundwater with weighted values of 31, 24, 18 and 10%, respectively.

A consistency ratio of 0.076 was determined from the judgment process. This validates the reliability of the proposed approach and results. The groundwater potential map was demarcated into three zones, namely; Very good, good and poor. The verification of the groundwater potential map using borehole discharge and fluoride concentration was satisfying.

The results of this study confirm that the integration of AHP and GIS techniques provides a powerful tool for decision-making procedures in groundwater potential mapping, as it allows a coherent and efficient use of spatial data. Generally, the case study results show that the GIS-AHP based category model is effective in the groundwater potential zonation.

Regulation and control of groundwater development is required to be made operational for not only protecting the environment but also to ensure equity in sharing groundwater. Awareness campaigns for ground water conservation should be taken up on a large scale since the groundwater resource is becoming scarcer. Artificial groundwater recharge systems should also be constructed to augment the poor groundwater potential zones. For further studies, research efforts could be focused on how AHP can be combined with other techniques such as fuzzy logic, as suggested by Boroushaki & Malczewski (2010).

The authors appreciate Tanzania Water Partnership in collaboration with Moi University (Kenya), Makerere University (Uganda), and University of Dar-es Salaam (Tanzania), under the project- Building Capacity in Water Engineering for Addressing Sustainable Development Goals in East Africa (CAWESDEA), for presenting financial support and guidance. The support and supervision from CRVWWDA and the Civil and Structural Department of Moi University is also acknowledged.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.

Al-Abadi
A. M.
&
Al-Shamma'a
A.
2014
Groundwater potential mapping of the major aquifer in Northeastern Missan Governorate, South of Iraq by using analytical hierarchy process and GIS
.
J. Environ. Earth Sci.
10
,
125
149
.
Al-Bakri
J. T.
,
Salahat
M.
,
Suleiman
A.
,
Suifan
M.
,
Hamdan
M. R.
,
Khresat
S.
&
Kandakji
T.
2013
Impact of climate and land use changes on water and food security in Jordan. Implications for transcending the tragedy of the commons
.
Sustainability
5
(
2
),
724
748
.
Ayazi
M.
,
Pirasteh
S.
,
Rizvi
S.
,
Safari
H.
,
Ramli
F. M.
,
Pradhan
B.
&
Rizvi
S. M.
2010
Using ERS-1 synthetic aperture radar for flood delineation, Bhuj Taluk, Kuchch District Gujarat, India
.
Int. Geoinformatics Res. Dev. J.
1
,
13
22
.
Balamurugan
G.
,
Seshan
K.
&
Bera
S.
2016
Frequency ratio model for groundwater potential mapping and its sustainable management in cold desert, India
.
J. King Saud Univ. Sci.
3
.
https://doi.org/10.1016/j.jksus
.
Bhuvaneswaran
C.
,
Ganesh
A.
&
Nevedita
S.
2015
Spatial analysis of groundwater potential zones using remote sensing, GIS and MIF techniques in upper Odai sub-watershed, Nandiyar, Cauvery basin, Tamil Nadu
.
Int. J. Curr. Res.
7
(
09
),
20765
20774
.
Foster, S. 1998 Groundwater: assessing vulnerability and promoting protection of a threatened resource. Proceedings of the 8th Stockholm Water Symposium, Sweden, 10–13 August 1998, pp. 79–90.
Gitas
I. Z.
,
Ayanz
J. S. M.
,
Chuvieco
B. D. E.
&
Camia
A.
2014
Advances in remote sensing and GIS applications in support of forest fire management
.
Int. J. Wildland Fire
23
,
603
605
.
Hapuarachchi
H. A. P.
,
Wang
Q. J.
&
Pagano
T. C.
2011
A review of advances in flash flood forecasting
.
Hydrol. Process.
25
,
2771
2784
.
Hussein
A. A.
,
Govindu
V.
&
Nigusse
A. G. M.
2017
Evaluation of groundwater potential using geospatial techniques
.
Appl. Water Sci.
7
(
5
),
2447
2461
.
Ibrahim
B. K.
&
Ahmed
S. A.
2016
Geospatial technology for delineating groundwater potential zones in Doddahalla watershed of Chitradurga district, India
.
Egypt. J. Remote Sens. Sp. Sci.
19
,
223
234
.
Mallick
J.
,
Singh
C. K.
,
Al-Wadi
H.
,
Ahmed
M.
,
Rahman
A.
,
Shashtri
S.
&
Mukherjee
S.
2014
Geospatial and geostatistical approach for groundwater potential zone delineation
.
Hydrol. Processes
.
https://doi.org/10.1002/hyp.101
.
Murasingh
S.
2014
Analysis of Groundwater Potential Zones Using Electrical Resistivity, Remote Sensing & GIS Techniques in a Typical Mine Area of Odisha
.
Doctoral Dissertation
,
National Institute of Technology
,
Rourkela
.
Naghibi
S. A.
,
Pourghasemi
H. R.
,
Pourtaghi
Z. S.
&
Rezaei
A.
2015
Groundwater qanat potential mapping using frequency ratio and Shannon's entropy models in the Moghan watershed, Iran
.
Earth Sci. Inf.
8
,
171
186
.
Odada
E. O.
,
Onyando
J. O.
&
Obudho
P. A.
2006
Lake Baringo: addressing threatened biodiversity and livelihoods
.
Lakes and Reservoirs: Res. Manage.
11
(
4
),
287
299
.
https://doi.org/10.1111/j.1440-1770.2006.00309.x
.
Pinto
D.
,
Shrestha
S.
,
Babel
M. S.
&
Ninsawat
S.
2015
Delineation of groundwater potential zones in the Comoro watershed, Timor Leste using GIS, remote sensing and analytic hierarchy process (AHP) technique
.
Appl. Water Sci.
https://doi.org/10.1007/s13201-015-0270-6
.
Prasad
R. K.
,
Mondal
N. C.
,
Banerjee
P.
,
Nandakumar
M. V.
&
Singh
V. S.
2008
Deciphering potential groundwater zone in hard rock through the application of GIS
.
Environ. Geol.
55
,
467
475
.
Rahmati
O.
,
Nazari
S. A.
,
Mahdavi
M.
,
Pourghasemi
H. R.
&
Zeinivand
H.
2014
Groundwater potential mapping at Kurdistan region of Iran using analytic hierarchy process and GIS
.
Arab. J. Geosci.
https://doi.org/10.1007/s12517-014-1668-4
.
Rajasekhar
M.
,
Sudarsana Raju
G.
,
Sreenivasulu
Y.
&
Siddi Raju
R.
2019
Delineation of groundwater potential zones in semi-arid region of Jilledubanderu river basin, Anantapur District, Andhra Pradesh, India using fuzzy logic, AHP and integrated fuzzy-AHP approaches
.
HydroResearch
2019
(
2
),
97
108
.
doi:10.1016/j.hydres.2019.11.006
.
Rajaveni
S. P.
,
Brindha
K.
&
Elango
L.
2017
Geological and geomorphological controls on groundwater occurrence in a hard rock region
.
Appl. Water Sci.
7
,
1377
1389
.
Saaty
T. L.
1980
The Analytic Hierarchy Process
.
McGraw-Hill
,
New York
.
Saaty
T. L.
&
Vargas
L. G.
2001
Models, Methods, Concepts and Applications on the Analytical Hierarchy Process
.
Kluwer's International Series, Kluwer Academic Publishers
,
Massachusetts, USA
.
Shaban
A.
2010
Support of space technologies for groundwater in Lebanon
.
J. Water Resour. Prot.
37
,
69
77
.
Sharma
M. P.
,
Kujur
A.
&
Sharma
U.
2012
Identification of groundwater prospecting zones using remote sensing and GIS techniques in and around Gola block, Ramgargh district, Jharkhand India
.
Int. J. Sci. Eng. Res.
3
(
3
),
1
6
.
Sisay
L.
2007
Application of Remote Sensing and GIS for Groundwater Potential Zone Mapping in Northern Ada'a Plain (Modjo Catchment)
.
University/Publisher Addis Ababa University, Addis Ababa
.
Solomon
S.
2003
Groundwater Study Using Remote Sensing and Geographic Information Systems (GIS) in the Central Highlands of Eritrea
.
Doctoral Dissertation
,
Environmental and Natural Resources Information Systems, Royal Institute of Technology, Stockholm
.
Stafford
K. W.
,
Rosales-Lagarde
L.
&
Boston
P. J.
2008
Castile evaporate karst potential map of the Gypsum Plain, Eddy County, New Mexico and Culberson County, Texas: a GIS methodological comparison
.
J. Cave Karst Stud.
70
(
1
),
35
46
.
Subba
R. N.
,
Chakradhar
G. K. J.
&
Srinivas
V.
2001
Identification of groundwater potential zones using remote sensing techniques in and around Gunur town, Andhra Pradesh, India
.
J. Indian Soc. Remote Sens.
29
,
69
78
.
Terzer, S., Wassenaar, L. I., Araguás-Araguás, L. J. & Aggarwal, P. K. 2013 Global isoscapes for δ18O and δ2H in precipitation: improved prediction using regionalized climatic regression models, Hydrol. Earth Syst. Sci. 17, 4713–4728.
Tesfaye
T.
2010
Ground water potential evaluation based on integrated GIS and RS techniques in Bilate river catchment, South rift valley of Ethiopia
.
Am. Sci. Res. J. Eng. Technol. Sci. (ASRJETS)
.
ISSN (Print) 2313-4410, ISSN (Online) 2313-4402. Global Society of Scientific Research and Researchers. Available from: http://asrjetsjournal.org
Todd
D. K.
&
Mays
L. W.
2005
Groundwater Hydrology
, 3rd edn.
Wiley
,
New York
, p.
636
.
WHO
2002
Fluorides. Environmental Health Criteria
.
World Health Organization
,
Geneva
.
Yeh
H. F.
,
Lee
C. H.
,
Hsu
K. C.
&
Chang
P. H.
2009
GIS for the assessment of the groundwater recharge potential zone
.
Environ. Geol.
58
,
185
195
.
Yeh
H. F.
,
Cheng
Y. S.
,
Lin
H. I.
&
Lee
C. H.
2016
Mapping groundwater recharge potential zone using a GIS approach in Hualian River, Taiwan
.
Sustainable Environ. Res.
26
,
33
43
.

Author notes

Deceased.

This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/).