Journal of Hydrology: Regional Studies 28 (2020) 100676

Contents lists available at ScienceDirect

Journal of Hydrology: Regional Studies

journal homepage: www.elsevier.com/locate/ejrh

Identification of Groundwater Potential Zones Using Proxy Data:

T
Case study of Megech Watershed, Ethiopia
Kibrit Gedam Berhanua, Samuel Dagalo Hatiyeb,*
a
University of Gondar, Institute of Technology, Gondar, Ethiopia
b
Arba Minch University, Arba Minch Water Technology Institute, Water Resources Research Center, Arba Minch, Ethiopia

ARTICLE INFO ABSTRACT

Keywords: Study Region: Part of the Lake Tana sub-basin of the Blue Nile, Ethiopia
AHP Study focus: This study is aimed to identify and locate the groundwater potential zones of Megech
GIS and remote sensing watershed using geographic information system (GIS) and remote sensing. Lithology, lineament
Groundwater potential density, slope, geomorphology, soil, land use and land cover, rainfall, drainage density and
Overlay analysis
elevation were used as proxy data to map the groundwater potential zones of the watershed.
Proxy data
Weights were assigned for each thematic layer and thematic layer classes using the PriEsT
Analytical Hierarchy Process (AHP) tool. Suitably ArcGIS overlay analysis was conducted to
obtain the groundwater potential map. Geology, lineament density, slope and geomorphology
were found to be the dominant and sensitive factors for groundwater zoning as portrayed in the
map removal analysis.
New hydrological insights for the region: Three groundwater potential zones were identified in the
study area namely low, moderate and high. These groundwater potential zones were compared
with an independent set of groundwater inventory data to validate the results. The validation
analysis showed 78.5 % agreement between the groundwater inventory data and the qualitative
results. Use of proxy data can serve as a credible source of information for groundwater potential
zoning and could greatly help in the groundwater development and management of a region.

1. Introduction

Water is the primary source of all life and it should be available sufficiently for all the required demands like domestic, agri-
culture, industrial, recreational and ecological demands. Groundwater is one of the most valuable natural resources supporting
human health and economic development. Because of its continuous availability and excellent natural quality, groundwater (GW)
becomes an important source of water supply in many urban and rural areas of the world (Todd and Mays, 2005). In Ethiopia, most
towns and villages get their domestic water supply from groundwater sources through developed springs, bore wells and shallow
wells. According to the International Atomic Energy Agency (IAEA) (2013), over 70 % of Ethiopia’s water supply comes from
groundwater and only 34 % of the population has access to an improved water supply. This indicates that there is a large demand for
groundwater development and utilization in the country. However, there is lack of in-depth understanding of the groundwater
potential of Ethiopia. There is high disparity between the estimates of groundwater reserve in Ethiopia. Earlier reports show that the
annual replenishable groundwater reserve of Ethiopia was around 2.6 Bm3 (MoWR, 2002). In the very recent past, however, this
figure was deemed to be underestimated. Studies show that the groundwater storage in areas like Adda Becho (central Ethiopia);

⁎
Corresponding author.
E-mail address: samueldagalo@gmail.com (S.D. Hatiye).

https://doi.org/10.1016/j.ejrh.2020.100676
Received 8 January 2020; Received in revised form 21 February 2020; Accepted 24 February 2020
Available online 29 February 2020
2214-5818/ © 2020 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/BY-NC-ND/4.0/).
K.G. Berhanu and S.D. Hatiye Journal of Hydrology: Regional Studies 28 (2020) 100676

Kabo Granna Vallay and Raya (Northern Ethiopia) alone is significantly more than the earlier estimation of 2.6Bm3(MoWR and GW-
MATE, 2011). Best estimates of the usable groundwater potential of the country falls between 12 and 30Bm3 (AGWATER, 2012;
Belete et al., 2014). Yet other studies show a different figure of groundwater potential in Ethiopia (40Bm3) (Haile et al., 2019).
Moreover, there were very limited attempts of studies on groundwater potentials of smaller catchments. Therefore, there is a pressing
need for the study and development of groundwater resources of the nation so that many, especially, the rural population have access
to an improved water supply.
On the other hand, there are also very limited studies on the use of groundwater for irrigation purpose in Ethiopia. This may be
attributed to lack of understanding of the available groundwater resource and need of large initial investment for groundwater
development for irrigation in the country. Groundwater potential investigation may, therefore, be an indispensable approach to
ensure the sustainable use of groundwater in the nation for various demands including domestic, agricultural and industrial purposes.
However, extensive use of groundwater based on only limited information may cause unsustainable use of the resource. In Ethiopian
condition, subsurface studies are often carried out when there arises a need for local specific developments of groundwater exists
including borehole, spring, shallow or hand dug wells for domestic water supply. Hussein et al.(2017) reported that the issue of
unsustainable groundwater utilization is becoming an evident problem and the key concern for many developing countries. One of
these problems is the absence of updated spatial information on the quantity and distribution of groundwater resource. Nowadays,
there is a strong need to use groundwater for the socio-economic development in Ethiopia, especially for rural areas. This indeed
requires the knowledge of groundwater resource potential assessment before using and managing it (Tesfaye, 2015).
There are several methods available for investigation of groundwater exploration or its availability in sub-surface environment.
They range from the oldest traditional water dowsing methods to the recent electromagnetic resonance (EMR) technologies. In
general, all the methods can be categorized into two broad classes: surface and sub-surface. The surface methods of groundwater
exploration or potential assessment are the least costly and easy to undertake. These methods include: esoteric, geomorphologic,
geological, soil and micro-biological, remote sensing techniques and surface geophysical methods. The sub-surface methods mainly
consist of test drilling of boreholes and geophysical logging techniques. Although the sub-surface methods are accurate for
groundwater exploration, they incur large investment since drilling, completing and development of wells may be needed for the
effective application of these methods. Therefore, it is a usual practice to undertake thorough surface investigation methods for
locating potential groundwater sources.
Remote sensing technique integrated with GIS is becoming a powerful tool for the identification and mapping of groundwater
potential zones in time and cost effective manner (Moore et al., 1991; Krishnamurty et al., 2000; Jha et al., 2012; Arkoprovo et al.,
2012; Hammuori et al., 2012; Lee et al., 2012; Davoodi et al., 2015; Das and Pardeshi, 2018; Andualem and Demeke, 2019). GIS and
remote sensing approaches often use spatial data to analyze and characterize required information. In fact, for studying sub-surface
phenomena or object, the geographic information system utilizes some proxy data from the ground surface. Geology or lithology,
geomorphology or landforms, drainage density, rainfall, geological structures or lineaments, slope, land use/land cover (LULC) and
soil characteristics are often the groundwater controlling parameters used to identify groundwater potential zones in a given area of
interest. These data were used as proxy data for groundwater potential mapping in this study. Use of such data for assessing
groundwater potential for Ethiopian conditions is immensely useful where there are very limited direct observations of groundwater,
particularly for smaller catchments. The Megech watershed is not exceptional to this limitation and there were no earlier studies on
the groundwater potential study in the watershed. Therefore, the aim of this study was to employ GIS and RS techniques to identify
and map the groundwater potential zones using the groundwater proxy data in the Megech watershed. The study also attempted to
verify the validity of the qualitative analysis obtained from the GIS and remote sensing techniques. The identified and organized
groundwater potential maps may provide information about productive well location in the study area. The results of the analysis
may also be used as geo-database which may enable decision makers and responsible stakeholders to properly use and manage
groundwater resources in the study area.

1.1. Description of the study area

Megech watershed is located in the North west part of Ethiopia (Amhara National Regional State) running between 12˚16′55′' N to
12˚45 40′' N latitude and 37˚18′7′' E to 37˚37′ 26′' E longitude with an approximate altitude ranging between 1755 m and 2983 m
average mean sea level (a.m.s.l.). The watershed covers a total surface area of 689.3km2. According to the meteorological data from
1995 to 2016, the annual rainfall of the study area ranges between 980 mm and 1140 mm. The monthly maximum temperature is
ranging between 18.4 °C and 29.2 °C while the monthly minimum temperature is falling between 8.3 °C and 13.1 °C. The area
generally falls in between Dega (cool and humid) and Woinedega (cool sub-humid) agro-climatic regions. The land use of the area is
dominantly agricultural followed by woody and shrub lands.
The Megech watershed has large irrigation potential and currently there is a construction of large irrigation scheme underway on
the Megech River. The project is also aimed to supply water for the nearby towns and villages from a storage dam. Upon completion,
the dam may enable to develop 17,000 ha of land around Lake Tana besides securing water supply for the Gondar town (the Zonal
capital).
The Megech River originates from the Semen Mountains and then flows to southern direction and terminates into Lake Tana (the
source of Blue Nile River). It is one of the main streams flowing into Lake Tana from the Northern part of Ethiopia. The northern part
of Megech watershed is characterized by a mountainous, wedge shaped steep slopes whereas the lower part, around Lake Tana, is
characterized by flat low-lying land with poor drainage conditions (WWDSE and TAHAL GROUP, 2008). According to the Ethiopian
geological survey of Ethiopia (GSE, 2011), the area is possesed by upper lava flows, basalts, trachiytes and lacustrine sedments. The

2
K.G. Berhanu and S.D. Hatiye Journal of Hydrology: Regional Studies 28 (2020) 100676

Fig. 1. Location map of Megech Watershed.

population in the watershed is mainly dependent on the subsistence agriculture for the livilehood income and potable water supplies
mainly come from groundwater sources. Fig. 1 shows the location map of Megech watershed.

2. Materials and methods

2.1. Preparation of input database

Input data for groundwater potential mapping of the study area was determined through field survey, analysis of available
information/literature and combined professional expert judgment (Sar et al., 2015). Based on the available information/ literature,
nine groundwater controlling factors were identified as proxy data. These were rainfall, elevation, slope, drainage density, geo-
morphology, geology/lithology, lineament density, land use/cover and soil type. These data are necessary for the analysis and
groundwater inventory data for comparison purpose were also collected from different and reliable sources. The type of data, sources
and the functions the data are presented in Table 1.
The long term average point rainfall values were used to generate areal rainfall map. Theissen polygon approach was used to
interpolate the areal rainfall map. Seven stations data were employed (Table 2) for the preparation of areal rainfall of the study area.

Table 1
The data collected, source and its purpose.
No. Data collected Sources of data Function of the data

1 Long term annual rainfall (from1995 National Meteorological Agency (NMA) Areal rainfall map
to 2016)
2 †SRTM DEM 30 Acquired on 27 January 2018 from https:// Elevation, drainage density, slope and geomorphological
earthexplorer.usgs.gov/ map
3 Existing maps GSE, MoWIE and ADSWE Geology, soil and LULC thematic maps
4 Landsat 8 OLI/TIRS Acquired on 27 January 2018 from https:// Lineament map
landsatlook.usgs.gov/
5 Existing wells/springs ADSWE, WWDE, TEWWE, GWSS and MoWIE For comparison of GIS and RS results with the GW
inventory data
6 Sample photos Field reconnaissance For ground control points

†SRTM = Shuttle Radar Topography Mission; ADSWE = Amhara Design and Supervision Enterprise; WWDE = Water Well Drilling Enterprise;
TEWWE = Tis Esat Water Works Enterprise; GSE = Geological Survey of Ethiopia; MoWIE = Ministry of Water, Irrigation and Energy;
GWSS = Gondar Water Supply service; GW = groundwater; DEM = Digital elevation model; OLI= Operational Land Imager; TIRS=Thermal
Infrared Sensor.

3
K.G. Berhanu and S.D. Hatiye Journal of Hydrology: Regional Studies 28 (2020) 100676

Table 2
The meteorological stations and mean annual rainfall in and around Megech watershed.
Station Name Latitude(˚) Longitude(˚) Elevation(m) Mean Annual Rainfall(mm)

Chuahit 12.34 37.23 1925 979.8

Aymba 12.53 37.30 2038 1055.5
Gorgora 12.24 37.29 1816 1114.8
Gondar 12.52 37.43 1973 1102.6
Shembekit 12.67 37.50 2403 1140.2
Amba Giorgis 12.77 37.62 2948 1117.6
Maksegnit 12.39 37.56 1912 1118.5

The data for elevation was acquired from the Shuttle Radar Topography Mission Digital Elevation Model (SRTM DEM) 30 m x 30 m
resolution data for its better accuracy both in its vertical and horizontal resolution (Odutola et al., 2012). Then the elevation map of
the Megech watershed was mapped using Arc GIS. Further, the elevation data was used to generate the slope raster data in each grid
cell. In the elevation raster, slope was measured by the identification of maximum rate of change in value from each cell to
neighboring cells (Waikar and Nilawar, 2014). The slope classes in the watershed were identified using slope spatial analyst tool in
the GIS package.
On the other hand, the data related to drainage density were generated indirectly from the watershed slope data. Drainage density
is defined as the closeness of spacing of stream channels (Sar et. al. 2015). According to Rahmati et al. (2014), drainage density (Dd) is
the measure of the total length of the stream segment of all orders per unit area which may be given as:
n
Di
Dd =
i= 1
A (1)

where, ΣDi is the total length of all streams in stream order i (km) and A is the area of the watershed(km2). The drainage line
delineation process was undertaken considering the stream lengths of 1000 units (“value > 1000″) and then the drainage density was
computed using Eq. (1). The geomorphological map of the study area was prepared following the Hammond landform classification
technique as suggested by Skye (2016). The geology or lithology thematic map of the study area was prepared by merging the two
recent (2011 and 2013 years) geologic maps of North Gondar and Yifag, published by the Geological Survey of Ethiopia (GSE, 2013).
The thematic layer for lineament density (Ld) can be defined as the total length of all recorded lineaments divided by the area of
the catchment under consideration which is given as (Edet et al., 1998):
n
Li
Ld =
i= 1
A (2)

where, Li is the length of the ith lineament and ΣLi is the total length of all lineaments in kilometer and A is the area of the grid in
square kilometers. Lineaments can be extracted from satellite images either by automatic or manual extraction methods. In this study,
automatic lineament extraction method was used due to its effectiveness and user friendliness compared to the manual extraction
(Weerasekera et al., 2014). The lineaments of the study area were extracted from the Landsat 8 OLI/TIRS image (path 170, row 51
with zero cloud cover) using automated processing LINE tool in PCI Geomatica 2017. The use of LINE tool in PCI Geomatica was
witnessed by Umikaltuma and Felix (2012). Out of the 11 bands of the Landsat 8 OLI/TIRS images, the lineaments were extracted
from band 5 due to its quality and the highest number of lineaments delineated. After the lineaments were extracted, further
processes of editing the watershed divide line and road features were done in GIS to ensure the quality of the extracted lineaments.
The recent (2014) land use/land cover (LULC) map of the study area reproduced from the LULC map of Tana sub basin earlier
prepared by Amhara design and supervision works enterprise was used. The soil map of Megech watershed was modified from the
FAO (2003) soil map of Ethiopia which was collected from MoWIE (Table 3).

Table 3
Soil group and fractions.
FAO Soil Class Sand % Silt % Clay % %age area

Calcic cambisol 81.60 6.75 11.75 24

Orthic luvisols 73.95 9.40 16.65 16
Eutric regosols 69.95 15.15 14.90 13
Eutric fluvisols 68.90 13.45 17.70 9
Calcic fluvisols 54.25 22.85 22.90 8
Calcic xerosols 43.05 33.15 23.85 7
Calcaric fluvisols 40.65 39.85 19.55 6
Eutric cambisols 39.05 34.65 26.35 5
Dystric nitosols 35.40 15.70 49.00 4
Chromic virtisols 21.60 24.00 54.35 3

(Modified from FAO (2003).

4
K.G. Berhanu and S.D. Hatiye Journal of Hydrology: Regional Studies 28 (2020) 100676

Table 4
Saaty’s (1995) Pair wise comparison scale.
Verbal judgment AHP numeric value (scale)

Extremely important 9
Very strongly to extremely important 8
Very strongly important 7
Strongly to very strongly important 6
Strongly important 5
Moderately to strongly important 4
Moderately important 3
Equally to moderately important 2
Equally important 1

3. The analytical hierarchy process (AHP)

The occurrence of groundwater in an area is influenced by several factors including the hydrological, geological, topographical,
environmental and climatic variables. In such a case, it is tempting to decide which factor or parameter is influencing more or less.
Among the several available methods (Yeh et al., 2016; Hussein et al., 2017; Singh et al., 2017; Das and Pardeshi, 2018; Razavi-
Termeh et al., 2019), the Analytical Hierarchy Process (AHP) helps a lot to systematically identify the most influencing parameters.
The Analytical Hierarchy Process (AHP) was used in this study. AHP is mainly suitable for complex decisions which consist of the
comparison of decision elements which are difficult to quantify. It involves building a hierarchy of decision elements and then
making comparisons between each possible pair in each cluster as a matrix. The AHP approach relays on the hierarchal structuring of
the decision criteria which is incorporated in decision making problem. The hierarchy includes the knowledge, experience and
intuition of the decision maker (Anagnostopoulos et al., 2007). AHP uses the pairwise comparison matrix (Eq. 3) to evaluate the
hierarchy (for instance, to compare two alternatives Ai and Aj using a criterion) based on the fundamental scale of values ranging
from 1 to 9. The definition of the ranks according to Satty(1995) are given in Table 4.The pair wise matrix is given by (Saaty 1995):
K A1 A2 ...... An
A1 1 a12 ...... a1n
A= A2 1/a12 1 ...... a2n
: : : :
: : : :
An 1/a1n 1/a2n ...... 1 (3)

where A is a pair wise comparison matrix of alternatives Ai, i = 1, 2, 3…n with respect to criteria K.
AHP incorporates an effective technique for checking the consistency of the evaluations made by the decision maker when
building each of the pairwise comparison matrix A (Eq. (3)), involved in the process. Computation of a suitable consistency index (CI)
is then obtained by first computing the scalar x as the average of the elements of the vector whose jth element is the ratio of the jth
element of the vector A multiplied by, weight vector, to the corresponding element of the vector w. The consistency index may be
expressed as (Satty, 1995):
x m
CI =
m 1 (4)

where m is the number of evaluation criteria (thematic layers in the case of this study).
A perfectly consistent decision maker should always obtain CI = 0, but small values of inconsistency may be tolerated if the
consistency ratio (CR) < 0.1 (Saaty, 1990 and Malczewski, 1999). The consistency ratio may be defined as:
CI
CR =
RI (5)

where RI is the random index whose values for m number of evaluation criteria are given in Table 5 according to Saaty(1990).
Later on Saaty (1995) suggested different CR value for different consistent pair wise comparison matrix size. Recommended CR
value for a 3 × 3 matrix is less than 0.05, a 4 × 4 matrix is 0.09 and for larger matrices it is is 0.1 (Saaty 1995).

4. Priority estimation tool (PriEsT)

Priority estimation tool (PriEsT) was used to identify dominant groundwater occurrence and distribution factors in this study.

Table 5
Random index (RI) corresponding to the number of evaluation criteria (m).
m 1 2 3 4 5 6 7 8 9 10

RI 0 0 0.58 0.9 1.12 1.24 1.32 1.41 1.45 1.49

5
K.G. Berhanu and S.D. Hatiye Journal of Hydrology: Regional Studies 28 (2020) 100676

Unlike other decision support tools based on AHP, PriEsT has the ability to visualize the inconsistency among a given set of pairwise
comparison matrix (Siraj, et al. 2015). PriEsT also provides the consistency measures (for example, dissonance, congruence, etc.)
which are recommended for pairwise comparison based on decision support tools (Siraj et al., 2015). PriEsT enables to select the
different elicitation methods. In this study, Eigen vector (EV) method was used to derive a priority vector from a set of pairwise
judgments (Saaty, 1977).
There are two types of consistency measures in pair wise comparison matrix to arrive at a judgment. These are the ordinal and the
cardinal consistency measures. The judgments given by the decision maker in the pair wise comparison matrix A is said to be
ordinally consistent if A1 is preferred to A2 and A2 is preferred to A3, then A1 is preferred to A3 where A1, A2 and A3 are elements of
the decision criteria in comparison matrix. For example, lithology, lineament density and slope were some of the elements considered
in this study. On the other hand, the judgments are said to be cardinally consistent if aij = 1/aji and ai j= aikakj for all i,j and k (Siraj
et al., 2015).
Siraj (2011) proposed different measures to determine the ordinal as well as the cardinal consistency and to detect outlying
judgments. These measures are dissonance (ψ), congruence (θ) and logarithmic mean square (L). Dissonance (ψ) is proposed for
ordinal consistency measure while congruence (θ) is the cardinal consistency measure in pairwise comparison matrix. When the value
of ψ is zero, it reveals no latent violation of priorities in the pairwise comparison matrix (PCM) whereas when ψ is a non-zero value,
there is latent violation. L indicates whether the PCM is transitive or not. When L is equal to zero, PCM is transitive and when L is not
equal to zero, PCM is intransitive.
The dissonance and congruence inconsistency measures are given by (Siraj et al., 2015):
n
1
ij = | log(aij ) log(aik akj )|
n-2 k=1 (5)

n
1
ij = step ( log(aij aik akj )
n-2 k (6)

4.1. Groundwater Potential Index (GWPI)

After preparing all the thematic maps, overlay analysis was conducted in GIS to obtain the ground water potential index (GWPI).
GWPI can be defined as a dimensionless quantity that helps to predict the groundwater potential zones in an area. A weighted linear
combination method was used to estimate the GWPI (Malczewski (1999); Shekhar and Pandey (2014)) which may be expressed as:
m n
GWPI = (Wj × Xi)
j= 1 i= 1 (7)

where, Wj is the normalized weight of the j thematic layer, Xi is the rank/rate value of each class with respect to the j layer, m is the
total number of thematic layers and n is the total number of classes in each thematic layer.

4.2. Sensitivity analysis

Sensitivity analysis reveals how much each thematic map and the weights assigned for them influence the output map. The
sensitivity analysis was essentially conducted using the map removal analysis. According to Lodwick et al. (1990) the map removal
sensitivity analysis shows that which thematic map(s) is/are the most or the least influential on determining the groundwater
potential zone. Sensitivity analysis uses the index number, S, to measure the influence of each thematic layer. The sensitivity index, S,
was computed using:

GWPI GWPI'
N N'
S=
GWPI (8)

where GWPI and GWPI' are the output of groundwater potential map index of all the thematic layers and when one of the thematic
layers is removed, respectively. N is the number of the full thematic layers used to compute GWPI and N' is the number of thematic
layers used to compute GWPI'.

4.3. Validation process

The identified groundwater potential zones using the GIS and RS techniques were compared with the existing ground data from
boreholes, springs and hand dug wells in the study watershed to validate the credibility of the qualitative result. This would build
more confidence on the use of qualitative results such as the one obtained from GIS and RS analysis. The general framework for data
processing and analysis is shown in Fig. 2.

6
K.G. Berhanu and S.D. Hatiye Journal of Hydrology: Regional Studies 28 (2020) 100676

Fig. 2. The work flow and methods of the study.

4.4. Results and discussions

In the following sections the results of the analysis are presented for each of the factors controlling groundwater recharge and
thereby the groundwater potential.

4.5. Rainfall

Rainfall plays an important role in the hydrologic cycle and controls groundwater potential. Knowing the nature and char-
acteristics of rainfall may enable one to conceptualize and predict its effects on runoff, infiltration, and groundwater recharge (Ramu
et al. 2010). The possibility of groundwater recharge would be high at the place where the rainfall is high and is low where rainfall is
low (Zomlot et al. (2015); Kotchoni et al. (2019); Shakya et al. (2019); Wang et al. (2020)). It can also be noted that the groundwater
recharge phenomena is the result of the long term effect than incidental highs or lows, particularly for humid environments.
Therefore, it is more plausible to consider long term influence of rainfall than the short term precipitation values to determine the
influence of rainfall on groundwater recharge. Fig. 3 shows the areal average rainfall of the watershed.
The mean annual rainfall of the watershed ranges from 979.8 to 1140.2 mm. The highest rainfall (1140.2 mm) was recorded at
Shembekit station while the lowest (1102.6 mm) was recorded at Gondar stations, which are covering 46.0 % and 40.3 % of the study
area, respectively. This may be attributed to the orographic effect arising from altitude difference (Mezga et al., 2014). It is evident
that more rainfall may cause more recharge potential, although that potential is challenged by other constraining factors such as
slope, geology, land use/cover, drainage density, lineament density and others. Therefore, high recharge area does not necessarily
imply high groundwater potential areas (Kotchoni et al. (2019).

4.6. Elevation

Elevation or altitude can have an indirect and inverse effect on the groundwater potential of a given area. It relates mainly to the
rainfall occurrence and thereby the recharge. Therefore, high altitudes favor more recharge and ensure the availability groundwater
in lowland areas in a watershed. Mountainous areas are often favorable for recharge in deep seated confined aquifers situated at low
land areas (Duffy and Al-Hassan, 1988; Todd and Mays, 2005). In some instances, highland areas can also serve for storage of some

7
K.G. Berhanu and S.D. Hatiye Journal of Hydrology: Regional Studies 28 (2020) 100676

Fig. 3. Average annual areal rainfall map of Megech watershed.

recharge and then contribute to the groundwater potential of an area depicted mainly through the presence of springs.
Water tends to store at lower topography than at the higher topography (Ramu et al. 2010). Therefore, higher the elevation, lesser
the groundwater potential and vice versa. Hence elevation or altitude data is needed to be considered in the groundwater potential
studies. Elevation/altitude ranges were arbitrarily created for qualitative assignment of values as shown in Fig. 4. Evidently, the
watershed has higher elevation (2543−2978 m) at the northern part and lower elevation (1784−2004 m) at the southern part of the
study area bordering Lake Tana. Fig. 4 provides the topographic map of the Megech watershed which is included in the weighted
overlay analysis in this study.

4.7. Slope

The land surface slope is another factor which influences the hydrology of a given watershed. It mainly affects the surface runoff
process and hence partially determines the groundwater recharge of a watershed. The lower slope values indicate the flatter terrain
(gentle slope) and the higher slope values indicate steep and undulating terrain. The lower slope areas of the flat terrain allow rainfall
infiltration and percolation while the higher slope areas generate quick runoff from the terrain and hence offer little volume of water
for groundwater recharge. Within a watershed, the effect of topographic gradient has significant influence on the groundwater flux
than the groundwater depth and hence the hydraulic gradient (Grinevskii, 2014; Condon and Maxwell, 2015). This may be attributed
to scale issue where the effect of groundwater hydraulic gradient may be negligible but the effect of topographic gradient cannot be
disregarded. However, slope exposure is more influencing factor than the slope gradient for groundwater recharge and thereby for
the groundwater potential (Grinevskii, 2014; Rukundo and Dogan, 2019).
As can be depicted from Fig. 5, there are four classes of slopes identified in Megech watershed using slope spatial analyst tool in
the GIS package. There is indeed no hard and fast rule for slope classification for groundwater recharge or runoff generation.
However, slope classes were identified based on their level of importance to groundwater recharge in GIS (Yeh et al., 2016; Andualem
and Demeke, 2019). Four classes of slope were extracted ranging from plain/gentle slope areas to extremely steep slopes. The gentle
slopes (0˚- 7˚) is observed in the southern part of the watershed, near Lake Tana covering an area of nearly 47 % showing significant
area of the watershed available for groundwater recharge with respect to slope. On the other hand, areas having extremely steep
slopes (28˚-77˚) are located at the north and northeastern part of the watershed. The steepest slope portion covers a small area of
around 8%. The other two classifications of the slopes (7˚-16˚ (medium) and 16˚-28˚ (steep)), respectively cover approximately 31 %
and 14 % of the total watershed area (Fig. 5).

8
K.G. Berhanu and S.D. Hatiye Journal of Hydrology: Regional Studies 28 (2020) 100676

Fig. 4. Elevation map of the study area.

4.8. Drainage density

Drainage density is the total length of streams per unit area. It is one of the factors which play an important role in potential
groundwater zoning. High drainage density favors high runoff through the dissected channels and thus provides less opportunity to
infiltration into the ground and percolation thereafter. On the other hand, low drainage density areas have less surface runoff and
consequently favor more infiltration (Jenifer and Jha, 2017; Thomas and Duraisamy, 2017).
Although large drainage density depicts less groundwater recharge, it may come to happen that the stream beds and banks may
contribute to substantial amount of groundwater recharge which needs further assessment (Das and Pardeshi, 2018; Yeh et al., 2009).
Furthermore, in perennial and seasonal streams, the water contact time with the bed and banks is long that it provides good op-
portunity for percolation. Therefore, areas nearby drainage channels/water courses may have good groundwater potential (Todd and
Mays, 2005). Drainage density enables to visualize high groundwater recharge areas. The drainage density map of Megech watershed
(Fig. 6) is produced by employing the line density tool in ArcGIS (Das and Pardeshi, 2018).
The drainage density of the study area is ranging from 0.083 to 4.024 km/km2. The drainage densities of the Megech watershed
were classified into four classes as very low (0.083-0.995 km/km2), moderately low (0.995–1.536 km/km2), high (1.536–2.216 km/
km2) and very high (2.216–4.024 km/km2) as shown in Fig. 6. The lowest drainage density areas are mainly located on the
northwestern boundaries of the watershed (Fig. 6). On the other hand, the highest drainage density areas are scattered in the central
and southern portions of the watershed. This may be due to the presence of closely spaced stream networks or drainage lines in these
areas.

4.9. Geomorphology

Geomorphology is the study of earth structures and also depicts the various landforms related to the groundwater occurrence and
structural features (Ramu and Vinay, 2010). Geomorphology controls the subsurface movement of groundwater. It is one of the most
important features in evaluating the groundwater potential and prospect. Therefore, geomorphology can be utilized for management
of groundwater resources (Kumar and Kumar, 2011; Valliammai et al., 2013).
The Hammond landform classification technique revised by Skye (2016) was used to prepare the geomorphological map. There
are seven major land forms (sub-classes) identified which include flat or nearly flat plains, smooth plains with some local relief, plains
with high hills, low hills, high hills, low maintains and high mountains (Fig. 7). The flat or nearly flat plains are found mostly in the

9
K.G. Berhanu and S.D. Hatiye Journal of Hydrology: Regional Studies 28 (2020) 100676

Fig. 5. Slope map of the study area.

southern portions of the study area. On the other hand, the highly mountainous areas are observed at the northern end of the study
area (Fig. 7). Accordingly, flat or nearly flat plains comprise of 29 %, smooth plains with some relief equal to 28 %, plains with high
hills comprising 14 %, low hills equal to 13 %, high hills equal to 8%, low mountains equal to 5% and high mountains equal to 3% of
the total watershed.
High and low hills linear ridge landforms are considered as very low groundwater potential zones while flat and smooth plains
with local relief are high groundwater potential prospect zones. Valley plains and flat lands provide good opportunity for ground-
water recharge and show good groundwater potential zones (Ramaiah et al., 2012; Rajaveni et al., 2017).

4.10. Geology

Geology influences both the porosity and permeability of the aquifer material (Yazi et al., 2010; Chowdhury et al., 2010). Geology
or lithology is one of the groundwater controlling parameters considered in ground water studies which play a significant role in the
distribution and occurrence of groundwater. There are four major lithology types identified in the study area namely tertiary plateau
basalts and pyroclasts, tertiary upper basalts and trachyte, quaternary lacustrine sediments and tertiary upper lava flows (Fig. 8).
Tertiary plateau basalts and pyroclasts are the tertiary period lithological formations characterized by dark grey, mostly coarse
grained and sometimes fine to medium grained rock. At some places, the tertiary plateau basalts are occurring as massive, hard,
dense, and slightly weathered rocks. It consists of plagioclase phyric, olivine phyric and aphanitic basalts. Its thickness can reach up
to 800 m (GSE, 2013). The geological survey of Ethiopia (GSE) (2013) found out that the tertiary plateau basalts and pyroclasts are
extensive and moderately to low productive fissured aquifers categorized under the hydraulic properties: permeability, K > 100 km²,
transmissivity, T = 1–100 m²/d, unit flow rate, and q = 0.01–1 l/s.m and discharge Q = 0.5–5 l/s.
Tertiary upper basalts and trachytes are the tertiary period lithological formations which behave like aphanitic, dark grey, fine
grained formation materials and characterized by jointing, fracturing and spheroidal weathering. These are rarely columnar and
mainly massive in structure, dense (300 m thick) and exhibit compact flows. According to GSE (2013) tertiary upper basalts and
trachyte are extensive and highly productive fissured aquifer systems with the hydraulic properties whose permeability can go above

10
K.G. Berhanu and S.D. Hatiye Journal of Hydrology: Regional Studies 28 (2020) 100676

Fig. 6. Drainage density of the Megech watershed.

100 km², transmissivity varying from 100 to 500 m²/d, and well yields vary from 5 to 25 l/s.
Quaternary lacustrine sediments are the young lithological formations which have a property of thick brownish clayey soil and
light grey soil material, assumed to have developed from lake sediments. Quaternary lacustrine sediments have the highest water
bearing capacity among the lithological units of the study area. This may be witnessed by the borehole yield data falling in the range
of 20 l/s to 35 l/s from groundwater inventory data. These lithological units are categorized under the extensive and highly pro-
ductive porous aquifer systems which have the permeability values over 100km2, transmissivity values ranging between
100−500 m2/d and well yields can go up to 25 l/s and not less than 5 l/sec (GSE, 2013). These are mainly covering the lowland areas
of Dembia area in the neighborhood of Lake Tana (Fig. 8).
The tertiary upper lava flows are the Miocene age lithological formations, structurally characterized by massive, less columnar,
horizontal and deeply weathered faults (GSE, 2011). These comprise of mainly basalts and lesser scoriaceous basalts and scoria falls.
In view of groundwater bearing capacity, these are preferably more important than the basaltic formations consisting of pyroclasts
and trachyte. This is because the tertiary upper lava flows include scoriaceous basalts in which groundwater can occur (Seifu, 2013).
These geologic units are covering the highland areas of Lay Armachew and stretching to the plateau of Dembia. The tertiary upper
lava flows have the highest areal coverage (45.7 %) among the lithological units of the study area (Fig. 8).

4.11. Lineament density

A lineament may represent a fault, fracture and master joint, a long and linear geological formation, topographic linearity, valleys
or straight course of streams, boundaries between the different lithological units, vegetation cover or artificial objects such as road,
bridge etc. (Pradhan, 2009; Adiri et al., 2017). In the Megech watershed, there are few road networks. These road networks were
extracted and excluded during the map overlay process and hence only actual lineaments were considered. Pradhan (2009) also
stated that the presence of lineaments usually signify a permeable zone. Therefore, lineament density of an area can ultimately
explain the groundwater potential in the area.
Fig. 9 shows the lineament density of Megech watershed divided into four classes namely: 1st class (0-0.5 km/km2), 2nd class
(0.5–1.1 km/km2), 3rd class (1.1–1.9 km/km2), and 4th class (1.9–3.9 km/km2) (Fig. 9). It is evident that areas having higher
lineament density have higher groundwater potential and vice versa (Das et al. 2018; Andualem and Demeke, 2019). As can be
depicted in the lineament map of the study watershed, areas having highest groundwater prospect zones have lineament density
ranging between 1.9–3.9 km/km2 (Andualem and Demeke, 2019) covering an area of 45 % of the total land scape. The other classes
of lineament density: 1.1–1.19 km/km2, 0.5–1.1 km/km2 and 0-0.5 km/km2, respectively, cover nearly 26 %, 17 % and 12 % of the
overall watershed.

11
K.G. Berhanu and S.D. Hatiye Journal of Hydrology: Regional Studies 28 (2020) 100676

Fig. 7. Geomorphological map of Megech watershed.

4.12. Land use/cover

The surface covered by vegetation like forests and agriculture traps and holds the water in root of plants whereas the built-up and
rocky land use affects the recharge of groundwater by increasing runoff during rainfall (Shaban et al., 2006). Therefore, it is necessary
to study which features have covered the study area in order to carry out the groundwater potential studies. Fig. 10 shows the LULC
of the Megech watershed. Overall 7 main classes of LULC units were identified which are further detailed to 17 sub-classes. The study
area mainly comprises of cultivated land, grassland, woody shrub land, forest areas, water bodies and wetlands, residential areas and
few afro alpine vegetation cover. The dominant LULC units are the cultivated lands covering an area of about 73.28 % of the total
Megech watershed followed by settlement and residential areas (9.76 %). This may be attributed to the intensive agricultural practice
in the area for satisfying the growing food demand in the watershed. Built-up and rocky surfaces have less probability of groundwater
potential occurrence whereas the surfaces covered by vegetation like agricultural plants and forests have higher chance of
groundwater water probability (Singh et al., 2010; Fenta et al., 2015; Li et al., 2018). The priorities for LULC units were given
following Fenta et al. (2015) where water body > forest > shrub land > cultivated land > grassland > bare land for ground-
water occurrence suitability.
Land use and cover conditions affect the hydrologic cycle in many ways primarily by altering the surface runoff behavior and
thereby the groundwater recharge. In fact, the LULC also affects the other hydrologic processes such as evapotranspiration, tran-
spiration, infiltration, evaporation and interception. Due to population increase and other anthropogenic influences in many wa-
tersheds there is change of land use and cover from one form to the other (Scanlon et al., 2005; Kenji et al., 2009; Pan et al., 2011;
Owuor et al., 2016; Riley et al., 2019). Some land uses favor groundwater recharge and potential while others are bringing negative
consequence to groundwater recharge. For example, settlement and urban areas mainly generate the runoff process and hence have
reduced recharge (Pan et al., 2011; Owuor et al., 2016). Agricultural areas may or may not positively contribute to the groundwater
depending on the management procedures taken for soil and water conservation.
Extensive review on the effect of land use and land cover change on groundwater recharge and then the groundwater availability
was conducted by Owuor et al. (2016). Forests were observed to lower groundwater recharge rates compared to others while the
conversion of the other LULC to cropland or grassland increases groundwater recharge (Owuor et al. (2016)). However, this may not
be taken as a universal truth and site specific conditions may prevail in the groundwater recharge process (Yeh et al., 2009; Singh
et al., 2010; Pan et al., 2011; Li et al., 2018). On the other hand, forests and grasslands are reducing the runoff process and thereby

12
K.G. Berhanu and S.D. Hatiye Journal of Hydrology: Regional Studies 28 (2020) 100676

Fig. 8. Geological/lithological map of Megech watershed.

provide better opportunity for groundwater recharge. Wetlands and water bodies are the foremost contributors for groundwater
augmentation. In particular cases of irrigated areas, there may be an incidental recharge from the irrigation areas and this may
contribute to aquifer replenishment (Pan et al., 2011). The effects of the other land uses and cover conditions can be separately seen
since LULC not only controls the groundwater recharge alone but the other hydrological processes too.

4.13. Soil

Soil is an important factor for delineating the groundwater potential zones. The water holding capacity of an area depends upon
the soil types and their permeability (Kumar et al., 2016). The dominant soil types of the study area were eutric cambisols, chromic
vertisol and eutric regosols which covered 46.14 %, 18.40 % and 15.90 % of the total area of Megech watershed, respectively. The
remaining soil types contribute very limited portions. Fig. 11 illustrates the soil types of the watershed.
The permeability of the soil types depends up on their texture. Therefore, the identified soil types of the study area were slightly
modified from FAO (2003) for this study to relate it with the USDA soil textural class. Soil affects the process of groundwater recharge
besides its effect on the other hydrologic processes. Soil mainly affects the infiltration and percolation processes which ultimately
influence the groundwater recharge and then the groundwater potential of a given area (Anuraga et al., 2006; Kumar et al., 2016;
Rukundo and Dogan, 2019). Alluvial soils favor recharge and hence more potential for groundwater availability. On the other hand,
compacted soil environments and fine textured soils have very low permeability and hence favor surface runoff than sub-surface flow
(Rukundo and Dogan, 2019). Soil type may be more important than the land use for groundwater recharge particularly course
textured soils (Anuraga et al., 2006). However, specific studies on individual soil texture and comparative land use/cover conditions
were not studied in detail.

4.14. Weighted thematic maps and their classes

Weighting the thematic maps was carried out using PriEsT software package by employing the pairwise comparison (PC) method.
PriEsT was developed to support the Analytical Hierarchy Process in decision making and provides the opportunity to identify and
revise inconsistent judgments based on consistency measures (Siraj et al., 2015). The result of the analysis is shown in Table 6. The
dominant factors governing the groundwater potential in the watershed were determined using the weight vector. Three consistency
measures (dissonance, logarithmic mean square and congruence) were used to describe the consistency of the judgments.
As can be observed from PriEsT result for the judgments, CR value of less than 0.1 shows that the judgments are highly acceptable.

13
K.G. Berhanu and S.D. Hatiye Journal of Hydrology: Regional Studies 28 (2020) 100676

Fig. 9. Lineament density map of Megech watershed.

In addition to this, the judgments do not involve the intransitive judgments since the value of logarithmic residual mean square (L) is
equal to 0 and there are no latent priority during the pair wise comparison as the value of dissonance (ψ) is equal to 0. The priority
vector computed in PriEsT using Eigen vector (EV) method (Table 6) showed the weighted value of nearly 30 % for the lithology, 23
% for lineament density, 15 % for slope, 10 % for geomorphology, 7% for soil, 6% for LULC, 4% for rainfall, 3% for drainage density
and finally 2% for elevation.
According to the calculated priority vector, lithology is the most dominant groundwater governing parameter in the study area.
This is due to its capability in controlling groundwater potential as it can be inferred from the studies by GSE (2011; 2013) and the
borehole yields in the study area. On the other hand, the elevation was prioritized/ranked the last since it has the least influence
among the parameters controlling groundwater potentiality considered in the study. In the same fashion, each thematic map classes
were weighted and reclassified based on the weighted values of their classes. After the weightages were determined for each thematic
layer, GIS overlay Analysis was carried out.

4.15. GIS overlay analysis

All the thematic maps or layers were prepared in the reclassified raster format and were given the normalized weight in ac-
cordance with the groundwater potential controlling capability. Similarly, each thematic layer’s classes were given the normalized
rank or weight. Then overlay analysis was carried out using the weight vectors from PriEsT as described in the earlier section. The
GWPI was calculated using:

GWPI=0.3 × Lithr + 0.23 × Ldr + 0.15 × Slopr + 0.10 × Geomr+

0.07 × Soilr + 0.06 × LULCr + 0.04 × Rnflr + 0.03 × Ddr + 0.02×Elvnr (9)

where, Lithr = reclassified lithological map; Ldr = reclassified lineament density map, Slopr = reclassified slope map, Geomr =
reclassified geomorphological map, soilr = reclassified soil map, LULCr = reclassified land use/cover map, Rnflr = reclassified
rainfall map, Ddr = reclassified drainage density map, Elvnr = reclassified elevation map, and the numbers preceding the ab-
breviations are weights (in %).
GWPI is a dimensionless quantity computed considering the weights computed for each layer and sub-classes in each thematic
layer. After the overlay process has been completed, the groundwater potential zone map was classified as high (GWPI = 23.5–32.2),

14
K.G. Berhanu and S.D. Hatiye Journal of Hydrology: Regional Studies 28 (2020) 100676

Fig. 10. Land use and cover map (2014) of Megech watershed.

moderate (GWPI = 17.2–23.5) and low (GWPI = 9.7–17.2). The output map (Fig. 12) indicates the high groundwater potential zone
situated mostly near the Lake Tana (Dembia plain). This is mainly attributed to the availability of the high water bearing quaternary
lacustrine sediment aquifer in plain terrain. Few other patches of high groundwater potential zone areas are located where there are
high to very high (1.1–3.9 km/km2) lineament density with gentle slope (0˚ - 7˚) (Fig. 12). The high groundwater potential zone
covers an area of nearly 124 km2 (18 % of the total area of the Megech watershed).
The groundwater potential zone designated as moderate is mostly found in the northwest part of Megech watershed which covers
part of the Dembia plateau, Gondar town and Lay Armachew areas. The moderate groundwater potential zone is distributed along the
areas where there is tertiary upper lava flows (the second higher water bearing geological unit), and the moderately slopping areas (7˚
- 16˚), flat or nearly flat plains, plains with high hills and the eutric regosols soil type. A few portion of the moderate groundwater
potential zone is also scattered along the northeast direction where the lineament density is very high (1.9–3.9 km/km2). The
moderate groundwater potential zone covers relatively higher area (320.1 km2or 46 % of the Megech watershed).
Low groundwater potential areas are found along the northeast portion of the study area which covers Lay Armachew, Wogera
and Gondar zuria districts. This may be due to the fact that the area exhibits high distribution of mountainous geomorphology and
relatively less water bearing lithological formations (upper basalts and trachytes and plateau basalts and pyroclasts). On the other
hand, this groundwater potential zone is sparsely distributed along the northwest part of Megech watershed which includes Lay
Armachew, Gondar town and Dembia. The low groundwater potential zone covers nearly 245 km2 or 36 % of the study area. In
general, moderate to high groundwater potential zones are located in the North western and southern portions of the study area while
the northeastern portions of the watershed possess low groundwater potential. Possible groundwater development works for water
supply can take place in these areas while groundwater for extensive use such as irrigation may be considered in the Dembia plain in
the south of the watershed.

4.16. Sensitivity analysis of the groundwater potential zones map

Sensitivity analysis was conducted by the map removal procedure. In this procedure, one of the thematic layers was removed at a

15
K.G. Berhanu and S.D. Hatiye Journal of Hydrology: Regional Studies 28 (2020) 100676

Fig. 11. Soil map of Megech watershed.

Table 6
Weight vectors for thematic layers calculated using PriEsT.
Thematic layers †Lith Ld Slop Geom Soil LULC Rnf Dd Elvn weight

Lith 1 1 3 4 5 6 7 8 9 0.303
Ld 1 2 3 3 4 5 6 7 0.228
Slop 1 2 3 3 4 5 6 0.151
Geom 1 2 2 3 3 5 0.098
Soil 1 1 2 3 4 0.068
LULC 1 2 2 3 0.059
Rnf 1 2 2 0.040
Dd 1 2 0.031
Elvn 1 0.022

†Lith = lithology, Ld = lineament density, Slop = slope, Geom = Geomorphology, LULC = land use/cover, Rnf = rainfall, Dd = drainage density,
Elvn = elevation.

time and the other thematic layers involved in the processing of the output map. The computed sensitivity index (S) values are
indicated in Table 7. In this particular study, geology/ lithology was the first influential thematic layer (S = 2.8 %) to determine the
output map of the groundwater potential zones of Megech watershed. The lineament density and the slope were observed to be the
second and third influential thematic layers which account 1.41 % and 1.38 % sensitivity index, respectively. On the other hand, the
least influential factor for the output map determination was the drainage density thematic map(S = 0.69 %).
The sensitivity analysis confirmed that the influence of thematic layers to alter the output map is directly related to the weights
given to them. In this study, the map removal sensitivity analysis revealed that the generated output map was highly sensitive to the
thematic layers having high weight and vice versa.

16
K.G. Berhanu and S.D. Hatiye Journal of Hydrology: Regional Studies 28 (2020) 100676

Fig. 12. Groundwater potential zones and inventory data.

Table 7
Statistics of the sensitivity analysis index numbers.
Layer removed Sensitivity analysis index number, S (in %)

Min Max Sum Mean Sd

Lith 2.80 3.22 5.27 2.80 2.13

Ld 2.24 1.54 3.24 1.41 0.16
slop 0.22 1.09 4.60 1.38 2.13
Geom 0.99 1.21 1.48 0.94 0.37
Soil 1.02 1.32 0.62 1.18 1.34
LULC 1.15 0.78 0.99 0.79 0.37
Rnfl 0.05 1.19 0.28 0.72 1.28
Dd 0.27 1.26 1.16 0.69 0.77
Elvn 1.15 1.05 0.25 1.13 1.06

4.17. Validation of qualitative results

An effort was paid to collect existing data from different sources for validation purpose. A total of 135 existing groundwater
inventory data including boreholes, springs and hand dug wells were collected. For the purpose of comparison or evaluation of the
qualitative results of the groundwater potential zones, well and spring yield were selected as a better candidate than other existing
data. However, only 79 groundwater well points have the recorded yield data. The highest yield among the collected groundwater
data was 35 l/s with 286 m well depth. The least yield was observed from the springs having a yield of 0.01 l/s.
Although there is no general classification scheme, well yields can be grouped into some classification schemes considering site
specific conditions. For example, the aquifer yields in Tana and Beles sub basins were divided into four categories by SOGREAH
(2013) as: 0−3 l/s: low, 3−6 l/s:moderate,6−20 l/s:high and more than 20 l/s as very high groundwater potential zones. On the
other hand, GSE (2013) classified Gondar and West Gondar aquifer yields, including other aquifer properties as 0.5−1 l/s low, 1–5 l/

17
K.G. Berhanu and S.D. Hatiye Journal of Hydrology: Regional Studies 28 (2020) 100676

Table 8
Agreement between the groundwater points and the qualitative results.
Classes of GWPZ Yield(l/s) Total no of wells and springs exist in the GWPZ No of wells and springs fall in the GWPZ Agreement (%)

Low 0-1 5 4 80
Moderate 1-5 41 38 92.7
High >5 33 20 60.6
Sum 79 62 78.5

s moderate to low and 5−25 l/s high groundwater potential areas. In this study, since Gondar and West Gondar are part of the
Megech watershed and its aquifer properties, the GSE (2013) classification scheme was adopted with some modification. Areas where
well/spring yield lie between 0 and 1 l/s were categorized as low, 1 and 5 l/s as moderate and more than 5 l/s were classified as high
potential groundwater regions.
Based on this classification, out of the 33 boreholes having yields over 5 l/s, 20 boreholes (60.6 %) have higher yield values (range
between 18−35 l/s) are falling on the high groundwater potential zone (GWPZ). Most of these are found in Dembia plain. Out of the
41 boreholes yield data, 38 (92.7 %) of them are falling on the moderate GWPZ whereas out of the 5 low yield (0−1 l/s) boreholes
and springs data, 80 % is falling on the GWPZ classified as low (Fig. 12).
In general, the cross validation analysis revealed that 78.5 % (Table 8) of the groundwater inventory data (boreholes, hand dug wells and
springs) agree to the corresponding groundwater potential zone classifications from qualitative analysis. Besides comparing with ground-
water inventory data, Seifu (2013) also witnessed that there is high groundwater potential near the Lake Tana plains and low groundwater
potential around the northern part of the study area. This confirms that there is a very good agreement between the groundwater inventory
data and groundwater potential zones delineated using GIS and RS techniques. Plotting the inventory data over the groundwater potential
map was also conducted. It can be clearly seen that only few groundwater development works are appeared on the northeastern portion of
the watershed where the qualitative results revealed low groundwater potential. The better yielding wells and springs are located in the
southern and northwestern portions of the watershed where moderate to high groundwater potential zones prevail (Fig. 12).

5. Conclusion

In this study, an attempt was made to develop a spatial model for demarcating the groundwater potential zones making use of the
nine thematic layers as proxy data for groundwater. These were geology/lithology, lineament density, slope, geomorphology, soil,
LULC, rainfall, drainage density and elevation. The groundwater potential zones were produced by integrating these thematic layers
in ArcGIS overlay analysis. The identified GW potential zones are high, moderate and low covering 18 %, 46 % and 36 % of the
Megech watershed area, respectively.
The three distinct groundwater potential zones were also validated using existing groundwater inventory data and the result
portrayed nearly 78.5 % agreement between the groundwater inventory data and the qualitative groundwater potential analysis. It
can be concluded that GIS and RS together with AHP decision making techniques are found to be time and cost effective approaches
in groundwater potential mapping. The results from AHP based PriEsT software revealed that a weightage of 30 % for geology/
lithology, 23 % for lineament density, 15 % for slope, 10 % for geomorphology, 7% for soil, 6% for LULC, 4% for rainfall, 3% for
drainage density and 2% for elevation. Higher weightage of the contributing proxy parameters refers to high groundwater potential
and vice versa. It can be observed that the first four factors (geology, lineament density, slope and geomorphology) are dominantly
contributing to the groundwater potential of the study areas. This was also revealed using the map removal analysis procedure.
This study may provide stakeholders and decision makers important information on the groundwater potential of the Megech
watershed and the districts (“Woredas”) in the watershed. We recommend that this kind of studies is quite important for groundwater
development schemes, particularly in Ethiopian conditions where observed data on groundwater is not available. The developed
groundwater potential map can be an aid to develop, manage and wisely utilize the groundwater resource for community water
supply and small scale commercial farming activities.

CRediT authorship contribution statement

Kibrit Gedam Berhanu: Conceptualization, Data curation, Formal analysis, Methodology, Writing - original draft. Samuel
Dagalo Hatiye: Project administration, Supervision, Validation, Writing - review & editing.

Declaration of Competing Interest

There is no conflict of interest

Acknowledgements

The financial support of the Ministry of Education, the Government of Ethiopia is greatly acknowledged. The authors would also
like to extend gratitude to the GSE, MoWIE, NMA, ADSWE and WWDE for their support in rendering me the important data.

18
K.G. Berhanu and S.D. Hatiye Journal of Hydrology: Regional Studies 28 (2020) 100676

Appendix A. Supplementary data

Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.ejrh.2020.
100676.

References

Adiri, Z., El Harti, A., Jellouli, A., Lhissou, R., Maacha, L., et al., 2017. Comparison of Landsat-8, ASTER and Sentinel 1 satellite remote sensing data in automatic
lineaments extraction: a case study of Sidi Flah-Bouskour inlier. Moroccan Anti Atlas. Advances in Space Research 60, 2355–2367 2017.
AGWATER, 2012. Groundwater Management in Ethiopia, in Agricultural Water Management Learning and Discussion Brief. Accessed 15 December 2019 (4:22:25
PM.) Addis Ababa, Ethiopia. http://awmsolutions.iwmi.org.
Anagnostopoulos, K., Graziou, M., Vavastikos, A., 2007. Using the Fuzzy Analytical Heirarch Process for Selecting Wastewater Fasilities at a Prefecture Level. E.W.
Publications.
Andualem, T.G., Demeke, G.G., 2019. Groundwater potential assessment using GIS and remote sensing: study of guna Tana landscape, Upper Blue Nile Basin. Ethiopia.
Journal of Hydrology: Regional Studies 24, 1006102 2019.
Anuraga, T.S., Ruiz, L., Mohan Kumar, M.S., Sekhar, M., Leijnse, A., 2006. Estimating groundwater recharge using land use and soil data: a case study in South India.
Agric. Water Manege. 84, 65–76 2006.
Arkoprovo, B., Adarsa, J., Prakash, S., 2012. Delineation of groundwater potential zones using satellite remote sensing and geographic information system techniques:
a case study from Ganjam district, Orissa,India. Res. J. Recent Sci. 1 (9), 59–66.
Belete, B.K., Yilma, S., Assefa, M.M., 2014. Surface Water and Groundwater Resources of Ethiopia: Potentials and Challenges of Water Resources Development.
Chapter 6. Springer Science Publisherhttps://doi.org/10.1007/978-3-319-02720-3_6.
Chowdhury, A., Jha, M., Chowdary, V., 2010. Delineation of groundwater recharge zones and identification of artificial recharge sites in West Medinipur District, West
Bengal using RS, GIS and MCDM techniques. Environ. Earth Sci. 59 (6), 1209–1222.
Condon, L.E., Maxwell, R.M., 2015. Evaluating the relationship between topography and groundwater using outputs from a continental-scale integrated hydrology
model. Water Resour. Res. 51, 6602–6621. https://doi.org/10.1002/2014WR016774.
Das, S., Pardeshi, S.D., 2018. Integration of different influencing factors in GIS to delineate groundwater potential areas using IF and FR techniques: a study of Pravara
basin, Maharashtra. India. Appl Water Sci 8https://doi.org/10.1007/s13201-018-084. 197 (2018).
Davoodi, M., Rezaei, M., Pourghasemi, H., Pourtaghi, Z., Pradhan, B., 2015. Groundwater spring potential mapping using bivariate statistical model and GIS in the
Taleghan watershed Iran. Arab J Geosci. https://doi.org/10.1007/s12517-013-1161-5.
Duffy, C.J., Al-Hassan, S., 1988. Groundwater circulation in a closed desert basin: topographic scaling and climatic forcing. Water Resour. Res. 24, 1675–1688.
Edet, A., Okereke, C., Teme, S.C., Esu, E., 1998. Application of remote-sensing data to groundwater exploration: a case study of the Cross River State, southeastern
Nigeria. Hydrogeol. J. 6, 394–404.
FAO, 2003. Digital Soil Map of the World. Rome.
Fenta, A.A., Kifle, A., Gebreyohannes, T., Hailu, G., 2015. Spatial analysis of groundwater potential using remote sensing and GIS-based multi-criteria evaluation in
Raya Valley, northern Ethiopia. Hydrogeol. J. 23, 195–206. https://doi.org/10.1007/s10040-014-1198-x. 2015.
Grinevskii, S.O., 2014. The effect of topography on the formation of groundwater recharge. Moscow Univ. Geol. Bull. 69 (1), 47–52. https://doi.org/10.3103/
s0145875214010025.
GSE, 2011. Geology, Geochemistry and Gravity Survey of West Gonder and Gonder Map Sheets (unpublished).Addis Ababa, Ethiopa.. .
GSE, 2013. Integrated Hydrogeological and Hydrochemical Mapping of Yifag Map Sheet (ND 37-14) (unpublished).Addis Ababa, Ethiopia.. .
Haile, A.M., Molla, B.D., Tamiru, A.A., 2019. Review: groundwater resource potential and status of groundwater resource development in Ethiopia. Hydrogeol. J.
https://doi.org/10.1007/s10040-019-01928-x.
Hammuori, N., El-Naqa, A., Barakat, M., 2012. An integrated approach to groundwater exploration using remote sensing and geographic information system. J Water
Resour Prot 4 (9), 717–724. https://doi.org/10.4236/jwarp.2012.49081.
Hussein, A.A., Govindu, V., Nigusse, A.G.M., 2017. Evaluation of groundwater potential using geospatial techniques. Appl. Water Sci. 7 (5) 2447–246.
International Atomic Energy Agency (IAEA), 2013. Assessing and Managing Groundwater in Ethiopia. Retrieved from https://www.iaea.org/technicalcooperation
/documents/Success Stories/2013/ETH8007.pdf. .
Jenifer, M.A., Jha, M.K., 2017. Comparison of analytic hierarchy process, catastrophe and entropy techniques for evaluating groundwater prospect of hard-rock aquifer
systems. J. Hydrol. (Amst) 548, 605–624.
Jha, M., Chowdary, V., Chowdhury, A., 2012. Groundwater assessment in Salboni Block, West Bengal (India) using remote sensing, geographical information system
and multi-criteriadecision analysis techniques. Hydrogeol. J. 18, 1713–1728. https://doi.org/10.1007/s10040-010-0631-z.
Kenji, J., Atsushi, T., Othoman, A., Susumu, S., Ronny, B., 2009. Effects of land-use change on groundwater recharge model parameters. Hydrol. Sci. J. Des Sci. Hydrol.
54 (2), 300–315. https://doi.org/10.1623/hysj.54.2.300.
Kotchoni, D.O.V., Vouillamoz, J., Lawson, F.M.A., et al., 2019. Relationships between rainfall and groundwater recharge in seasonally humid Benin: a comparative
analysis of long-term hydrographs in sedimentary and crystalline aquifers. Hydrogeol. J. 27, 447–457. https://doi.org/10.1007/s10040-018-1806-2.
Krishnamurty, J., Mani, A., Jayaraman, V., Manivel, M., 2000. Groundwater resources development in hard rock terrain: an approach using remote sensing and GIS
techniques. Int. J. Appl. Earth Obs. Geoinf. 3 (3–4), 204–215. https://doi.org/10.1016/S0303-2434(00)85015-1.
Kumar, B., Kumar, U., 2011. Ground waterrecharge zonation mapping and modeling usingGeomatics techniques. Int. J. Environ. Sci. 1 (7), 1671.
Kumar, P., Herath, S., Avtar, R., Takeuchi, K., 2016. Mapping of groundwater potential zones in Killinochi area. Sustain. Water Resour. Manag. https://doi.org/10.
1007/s40899-016-0072-5.
Lee, S., Song, K., Kim, K., Park, I., 2012. Regional groundwater productivity potential mapping using a geographic information system (GIS) based artificial neural
network model. Hydrogeol. J. 20, 1511–1527. https://doi.org/10.1007/s10040-012 0894-7.
Li, S., Yang, H., Lacayo, M., Liu, J., Lei, G., 2018. Impacts of land-use and land-cover changes on water yield: a case study in jing-Jin-Ji. China. Sustainability 10, 960.
https://doi.org/10.3390/su10040960.
Lodwick, W., Monson, W., Svoboda, L., 1990. Attribute error and sensitivity analysis of map operations in geographical information systems: suitability analysis. Int J
Geogr Inf Syst 4 (4), 413–428.
Malczewski, J., 1999. GIS And Multicriteria Decision Analysis. Wiley,United States of America, pp. 177–192.
Mezga, K., Urbanc, J., Cerar, S., 2014. The isotope altitude effect reflected in groundwater: a case study from Slovenia. Isotopes Environ. Health Stud. 50 (1), 33–51.
https://doi.org/10.1080/10256016.2013.826213.
Moore, I., Grayson, R., Ladson, A., 1991. Digital terrain modelling: a review of hydrological,geomorphological, and biological applications. Hydrol. Process. 5 (1), 3.
https://doi.org/10.1002/hyp.3360050103.
MoWR, 2002. National Water Supply and Sanitation Master Plan: Interim Master Plan, DHV Consultants BV and DHV. Addis Ababa, Ethiopia.. .
MoWR, GW-MATE, 2011. Ethiopia: Strategic Framework for Managed Groundwater Development. Addis Ababa, Ethiopia.. .
Odutola, C.A., Beiping, W., Yao, Y.Z., 2012. Assessing vertical accuracy of SRTM ver 4.1 and ASTER GDEM ver 2 using differential GPS measurements – case study in
Ondo State. Nigeria. International Journal of Scientific & Engineering Research 4.
Owuor, S.O., Butterbach-Bahl, K., Guzha, A.C., et al., 2016. Groundwater recharge rates and surface runoff response to land use and land cover changes in semi-arid
environments. Ecol. Process. 5 (16). https://doi.org/10.1186/s13717-016-0060-68-7. 2016.
Pan, Y., Gong, H., Zhou, D., Li, X., Nakagoshi, N., 2011. Impact of land use change on groundwater recharge in Guishui River Basin. China. Chinese Geographical

19
K.G. Berhanu and S.D. Hatiye Journal of Hydrology: Regional Studies 28 (2020) 100676

Science 21 (6), 734–743. https://doi.org/10.1007/s11769-011-050.

Pradhan, B., 2009. Groundwater potential zonation for basaltic watersheds Groundwater potential zonation for basaltic watersheds. Cent Eur J Geosci 1 (1), 120–129.
Rahmati, O., Samani, A.N., Mahdavi, M., Pourghasemi, H.R., Zeinivand, H., 2014. Groundwater potential mapping at Kurdistan regionof Iran using analytic hierarchy
process and GIS. Arabian Journal of Geosciences, DOI. https://doi.org/10.1007/s12517-014-1668-4,ISSN1866-7511.
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. https://doi.org/10.1007/s13201-015-0327-6.
Ramaiah, S.N., Gopalakrishna, G.S., Vittala, S.S., Najeeb, K.M., 2012. Geomorphological mapping for identification of groundwater potential zones in hard rock areas
using geo-spatial information—a case study in Malur Taluk, Kolar District, Karnataka. India. Nat Environ Pollut Technol 11 (3), 369–376.
Ramu Mahalingam, B., Vinay, M., 2010. Identification of ground water potential zones using GIS and Remote sensing Techniques: a case study of Mysore taluk
-Karnataka. International Journal of Geomatics and Geosciences 5 (3) 2014.
Razavi-Termeh, S.V., Sadeghi-Niaraki, A., Soo-Mi, Choi, 2019. Groundwater potential mapping using an integrated ensemble of three bivariate statistical models with
random forest and logistic model tree models. Water 2019 (11). https://doi.org/10.3390/w11081596. 1596.
Riley, D., Mieno, T., Schoengold, K., Brozović, Ni., 2019. The impact of land cover on groundwater recharge in the High Plains: an application to the Conservation
Reserve Program. Sci. Total Environ. 696, 133871 2019.
Rukundo, E., Dogan, A., 2019. Dominant influencing factors of groundwater recharge spatial patterns in Ergene river Catchment. Turkey. Water 11, 653. https://doi.
org/10.3390/w11040653.
Saaty, T., 1977. A scaling method for priorities in heirarchical stractures. Jornal of mathematical Psychology 15 (3), 234–281.
Saaty, T., 1990. The Analytic Hierarchy Process. McGraw-Hill International, New York.
Sar, N., Khan, A., Chatterjee, S., Das, A., 2015. Hydrologic delineation of groundwater potential zones using geospatial technique for Keleghai river basin. India. Model.
Earth Syst. Environ. https://doi.org/10.1007/s40808-015-0024-3.
Satty, T., 1995. Decision Making for Leaders, 3rd ed. RWS Publications, Pittsburgh.
Scanlon, B.R., Reedy, R.C., Stonestrom, D.A., Prudic, D.E., Dennehy, K.F., 2005. Impact of land use and land cover change on groundwater recharge and quality in the
southwestern US. Golbal Change Biology 11 (10), 1577–1593. https://doi.org/10.1111/j.1365-2486.2005.01026.x.
Seifu, K., 2013. Groundwater in Ethiopia Features, Numbers and Opportunities. Springer Heidelberg, New York Dordrecht London. https://doi.org/10.1007/978-3-
642-30391-3. 978-3-642-30390-6.
Shaban, A., Khawlie, M., Abdallah, C., 2006. Use of remote sensing and GIS to determine recharge potential zone: the case of Occidental. Lebanon. Hydrogeol J. 14 (4),
433–443. https://doi.org/10.1007/s10040-005-0437-6.
Shakya, B.M., Nakamura, T., Shrestha, S.D., Nishida, K., 2014. Identifying the deep groundwater recharge processes in an intermountain basin using the hydro-
geochemical and water isotope characteristics. Nord. Hydrol. 50 (5), 1216–1229. https://doi.org/10.2166/nh.2019.164. 2019.
Shekhar, S., Pandey, A., 2014. Delineation of groundwater potential zone in hard rock terrain of India using remote sensing, geographical information system (GIS) and
analytic hierarchy process (AHP) techniques. Geocarto Int. https://doi.org/10.1080/10106049.2014.894584.
Singh, S.K., Singh, C.K., Mukherjee, S., 2010. Impact of land-use and land-cover change on groundwater quality in the Lower Shiwalik hills: a remote sensing and GIS
based approach. Cent. Eur. J. Geosci. • 2 (2), 124–131. https://doi.org/10.2478/v10085-010-0003-x.
Singh, L.K., Jha, M.K., Chowdary, V.M., 2017. Multi-criteria analysis and GIS modeling for identifying prospective water harvesting and artificial recharge sites for
sustainable water supply. J. Clean. Prod. 142, 1436–1456.
Siraj, S., 2011. Preference Elicitation From Pairwise Comparison in Multicriteria Decisin Making. PhD. The University of Manchester, Manchester,UK.
Siraj, S., Mikhailov, L., Keane, J., 2015. PriEsT:an interactive decision support tool to estimate priorities from pairwise comparison judgments. Int. Trans. Oper. Res.
22, 217–235.
Skye, C., 2016. Landforms I : Hammond. Retrived fromgis4geomorphology.com/Hammond –landforms/. .
SOGREAH, 2013. Consulting Survice for Detailed Groundwter Investigation and Monitoring in Tana and Beles Sub-basinsfinal Stage 1 Report-part 6: Hydrogiological
Survey.
Tesfaye, G.T., 2015. Ground water potential evaluation based on integrated GIS and remote sensing techniques, in bilate RiverCatchment: south Rift Valley of Ethiopia.
American Scientific Research Journal for Engineering 10, 85–120 Technology, and Sciences (ASRJETS).
Thomas, R., Duraisamy, V., 2017. Hydrological delineation of groundwater vulnerability of droughts in semi-arid areas of western Ahmednagar district. Egypt J
Remote Sens Space Sci. https://doi.org/10.1016/j.ejrs.2016.11.008.
Todd, D.K., Mays, L.W., 2005. Ground Water Hydrology (3rd ed.). John wiley& Sons,Inc., United States of America.
Umikaltuma, I., Felix, M., 2012. Lineament extraction using landsat 8 (OLI) in Gedo. Somalia. International Journal of Science and Research (IJSR) ISSN (Online):
2319-7064 Impact Factor (2012): 3.358.
Valliammai, A., Balathanduytham, K., Tamilmani, D., Mayilswami, C., 2013. Identification of potential recharge zone of the selected watershed using Remote sensing
and GIS. Int. J. Sci. Eng. Res. 4 (8), 611.
Waikar, M.L., Nilawar, Ap., 2014. Identification of groundwater potential zone using remote sensing and GIS technique. International Journal of Innovative Research
in Science,Engineering and Technology 3 ISSN: 2319-8753.
Wang, J., Huo, A., Zhang, X., et al., 2020. Prediction of the response of groundwater recharge to climate changes in Heihe River basin. China. Environ Earth Sci. 79, 13.
https://doi.org/10.1007/s12665-019-8752-8. 2020.
Weerasekera, W., Mayadunna, B.B., Senanayake, I., Dissanayake, D., 2014. Integrated Remote Sensing and GIS in lineament mapping for groundwater exploration – A
case study in Ambalantota, Sri Lanka. SAITM Research Symposium on Engineering Advancements.
WWDSE & TAHAL GROUP, 2008. Megech Dam Final Fisibility Report Volume 2. Addis Ababa, Ethiopia.. .
Yazi, A., Pirasteh, S., Arvin, A., Pradhan, B., Nikouravan, B., Mansor, S., 2010. Disasters and risk reduction in groundwater: zagros Mountain Southwest Iran using geo-
informatics techniques. Interdiscip. Neurosurg. 3 (1), 51–57.
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 (1), 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. Sust. Environ. Res. 26
(1), 33–43.
Zomlot, Z., Verbeiren, B., Huysmans, M., Batelaan, O., 2015. Spatial distribution of groundwater recharge and base flow: assessment of controlling factors. J. Hydrol.
Reg. Stud. 4, 349–368 2015.

20