Journal of Hydrology: Regional Studies 37 (2021) 100917

Contents lists available at ScienceDirect

Journal of Hydrology: Regional Studies

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

Groundwater recharge estimation using empirical methods from

rainfall and streamflow records
Tesfa Gebrie Andualem a, *, Girum Getachew Demeke b, Imran Ahmed a,
Mithas Ahmad Dar c, Mesenbet Yibeltal d
a
Department of Hydraulic and Water Resources Engineering, Debre Tabor University, Debre Tabor, Ethiopia
b
Department of Natural Resources Management, Debre Tabor University, Debre Tabor, Ethiopia
c
Department for Management of Science and Technology Development, Ton Duc Thang University, Ho Chi Minh City, Viet Nam
d
Faculty of Civil and Water Resources Engineering, Bahir Dar Institute of Technology, Bahir Dar, Ethiopia

A R T I C L E I N F O A B S T R A C T

Keywords: Study region: Gumara and Ribb watersheds, Abbay Basin, Ethiopia.
Baseflow separation Study focus: In this paper, an attempt was made to estimate groundwater recharge from
Ethiopia streamflow and rainfall records. The groundwater recharge was estimated by different empirical
Groundwater recharge
methods, including the water balance method. The Ribb and Gumara rivers streamflow data were
Rib and gumara rivers
collected from the Ministry of Water, Irrigation, and Electricity of Ethiopia for estimating
groundwater recharge by using baseflow separation techniques. The climate data was obtained
from the Amhara National Meteorology Agency for estimating rainfall recharge and water bal­
ance components.
New hydrological insights for the region: Groundwater resource development needs the determi­
nation of groundwater recharge volume. The annual areal rainfall, reference evapotranspiration,
and recharge from rainfall of the study area were estimated as 1399.8 mm, 1300.21 mm, and
253.70 mm, respectively. The evapotranspiration was determined using the Penman-Monteith
method. The baseflow separation techniques of RORA, PART, WHAT, and RECESS were
employed during this study. The recharge coefficient estimated from rainfall and streamflow data
were found at 0.18 and 0.20 respectively, which indicated that the study area has high
groundwater recharge volume which could be used for different groundwater development
projects. This study provides evidence for the groundwater potential areas identified by using
geospatial methods and groundwater development projects that can be practiced in the area.

1. Introduction

Surface water resources are becoming more scarce, influencing the socio-economic development of a country. The use of
groundwater as an alternative means of water supply is essential. The estimation of groundwater quantity and quality is crucial for
most hydrological studies by supplementing the water demand for a variety of purposes. Groundwater recharge is considered an
important component of groundwater development projects (Sanford, 2002). Sustainable groundwater development needs the
quantification of recharge amounts that could be stored in groundwater aquifers (Rushton and Ward, 1979).

* Corresponding author.
E-mail address: tesfgeb@dtu.edu.et (T.G. Andualem).

https://doi.org/10.1016/j.ejrh.2021.100917
Received 28 December 2020; Received in revised form 7 September 2021; Accepted 8 September 2021
Available online 17 September 2021
2214-5818/© 2021 The Author(s). 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/).
T.G. Andualem et al. Journal of Hydrology: Regional Studies 37 (2021) 100917

Recharge might occur naturally (direct infiltration from rainfall or percolation from adjacent water bodies) or artificially (due to
induced human activities like irrigation, urbanization, development of boreholes, and diversion of rivers). Rushton (1997) discussed
the actual and potential groundwater recharge to differentiate between the infiltrated water which reaches the water table and the
unsaturated zone which may or may not reach the water table.
The development of groundwater requires the determination of groundwater recharge magnitude from rainfall, streamflow, and
aquifer characteristics. Assessment of the recharge rate can be done from discharge data records using hydrograph separation tech­
niques by distinguishing streamflow as baseflow and surface runoff (Chen and Lee, 2003). Recharge estimation could be performed
using a variety of manual or automated methods of baseflow separation (Eckhardt, 2008). Most recharge methods estimate
groundwater recharge as the product of the water level fluctuation in wells and the specific yield of an aquifer material using water

Fig. 1. Study area showing the location of Ribb, Gumara watershed, Lake Tana, Tana and Abbay basin, as well as the location of streamflow
gauging stations.

2
T.G. Andualem et al. Journal of Hydrology: Regional Studies 37 (2021) 100917

balance/water budget models (Moon et al., 2004). The groundwater balance approach is an appropriate way of assessing groundwater
recharge and evaluating the precision of widespread methods for the assessment of groundwater losses and recharge from other
sources (Cheng-Haw et al., 2006).
Even though several methods have been used to estimate groundwater discharge and recharge from streamflow records, the most
commonly used technique involves baseflow separation. Baseflow separation methods are aimed at the estimation of a continuous
record of baseflow from the streamflow hydrograph. Baseflow separation techniques require long-term streamflow records and the
exercise of various manual methods (Olmsted and Hely, 1962; Zektser, 1977; Niazi et al., 2017; Misstear et al., 2009; Dzhamalov,
1973) or analysis of baseflow records (Rutledge, 1992; Mau and Winter, 1997). The recharge from streamflow records can be estimated
using baseflow separation techniques like Digital Filter Method, PART, RORA, and RECESS (Eckhardt, 2005, 2008; Rorabaugh, 1964;
Knisel and Sheridan, 1983; Meyboom, 1961). The groundwater recharge from rainfall records can also be estimated by employing
different empirical equations which were developed for different catchments (Chaturvedi, 1973; Kumar and Seethapathi, 2002;
Krishna Rao, 1970; Maxey and Eakin, 1949; Bredenkamp et al., 1995).
Currently, surface water sources are not able to meet the growing water demands in the Guna Tana Landscape (Ribb and Gumara
watersheds). In the lower parts of the landscape, especially in the Fogera flood plain, there is a conflict regarding the amount of water
needed by agriculture. The upper parts of the study area are also experiencing water shortages for domestic use (Debre Tabor town
water users get water once a week). Despite the fact that many wells have been developed to supply domestic water needs, the shortage
of surface water requires attention. The availability of data limits the application of detailed and in-depth methods of determining
groundwater recharge. While the necessity to estimate and constrain groundwater recharge for Guna Tana is important, the avail­
ability of data hinders detailed and in depth approaches. As a result, this study investigates the potential amount of groundwater
recharge using rainfall and streamflow measurements within Guna Tana.

Fig. 2. Lithology map of the study area.

3
T.G. Andualem et al. Journal of Hydrology: Regional Studies 37 (2021) 100917

2. Methodology

2.1. Study area

The Guna Tana Landscape is situated in the northwest part of Ethiopia between 11.583 ◦ and 12.317◦ N latitude and 37.45 ◦ and
38.23 ◦ E longitudes (Fig. 1). This landscape is comprised of the Gumara and Ribb adjacent watersheds. The study watershed has an
area of 3499.62 km2 and contributes to 30.5 % of the basin inflow to Lake Tana.
The topography of the study area ranges from 1774 m a.s.l at Lake Tana to 4090 m a.s.l at Mount Guna. The Gumara and Ribb rivers
originate from small springs located near Guna Mountain at an elevation of 4090 m a.s.l and drain into the eastern part of Lake Tana.
The geology map of the study watersheds was attained from the Geological Survey of Ethiopia with a resolution of 1:250,000. The
geology is dominated by Debre Tabor basalts and trachytes, quaternary lacustrine sediment, middle basalt flows, and upper basalts and
trachytes covering 32.56 %, 24.31 %, 16.12 %, and 14.9 % of the study area respectively (Fig. 2). The quaternary lacustrine sediment
found in the downstream parts of the study area near the lake offshore is characterized by alluvial deposits which are more favorable
for groundwater recharge (Bhutta et al., 2005).
In the adjacent watersheds, water is used mainly for domestic purposes; sometimes it is used to supplement the deficit irrigation in
downstream parts of the landscape. In the Ribb watershed, the water is required for recreation and fishing activities at the Ribb
reservoir. The need for supplementary irrigation (in lower parts of the landscape) and domestic use (in upper parts of the landscape) is
very high. Most of the irrigated areas are below the gauging stations in both the watersheds.
The land use of the study area is mainly dominated by cultivated/agriculture with 64.67 % coverage, followed by grassland with
14.17 % and the remaining 21.16 % covered by shrubland, bare land and water (Andualem and Demeke, 2019).

2.2. Hydrometeorological data analysis

The long-term time series records of meteorological parameters like rainfall, temperature, wind speed, relative humidity, and
sunshine duration were considered for evaluating the hydrologic components (Shaw, 1998). To evaluate the hydrologic cycle com­
ponents of the study area, time series meteorological data was collected from the Amahara meteorology agency. The streamflow data
for the Gumara and Ribb rivers was collected from the Ethiopian Ministry of Water, Irrigation, and Electricity.
For the period from 1996–2015, ten meteorology stations data which are found in and around Guna Tana Landscape (Debre Tabor,
Woreta, Yifag, Adis Zemen, Ebinat, Dera Hamusit, Wanzaye, Mekane Eyesus, Amed Ber, and Gassay stations) were collated during this
study (Table 1).
The meteorological data was used for the determination of the areal depth of rainfall, evapotranspiration, and recharge from
rainfall.
Relative Humidity: relative humidity has an inverse relationship with evaporation (Shaw, 1988). The minimum and maximum
relative humidity of the Debre Tabor station are 38.7 and 90.3 respectively. The sunshine hour duration of the Debre Tabor station was
collected from Amhara national meteorology agency and the values ranged between 4.47 and 8.80 (Table 2).

2.2.1. Spatiotemporal distribution of rainfall

The precipitation that reaches a particular area rarely produces a uniform depth of rainfall over the entire area. The rainfall ob­
tained from a single rain gauge station is point rainfall that might not represent the areal average depth of rainfall. Therefore, an
average areal rainfall is necessary for most hydrological problems. The areal rainfall determination methods of arithmetic mean and
Theissen polygon methods were employed in this study. The spatial and temporal rainfall variability of the study area is exhibited in
Figs. 3 and 4 respectively.
The rainfall in the area is unimodal, with annual rainfall varying between 1023 mm and 1588 mm (Fig. 3). The area gets the
maximum rainfall during July and August (Fig. 4).
Arithmetic Mean Method: it is one of the simplest methods of obtaining average areal rainfall by dividing the sum of rainfall
recorded at all rain gauge stations by the number of stations (Shaw, 1998).

Table 1
Location of meteorological stations and data available for the analysis.
No. Station Name Longitude (m) Latitude (m) Altitude (m) Data Used Period Annual Rainfall (mm)

1. Debre Tabor 391983.6 1,309,904 2612 1996–2015 1535.7

2. Woreta 357898.1 1,319,046 1798 1996–2015 1381.6
3. Yifag 360220.9 1,336,276 1901 1996–2015 1088.5
4. Adis Zemen 367315.6 1,340,829 1936 1996–2015 1341.5
5. Ibnat 396735.46 1,340,334 2212 1996–2015 1009.2
6. Dera Hamusit 343324.19 1,303,513 1934 1996–2015 1592.8
7. Wanzaye 355637.06 1,303,231 1830 1996–2015 1455.2
8. Mekaneyesus 396868.56 1,283,379 2380 1996–2015 1451.9
9. Amed Ber 378685.63 1,317,287 2051 1996–2015 1454.9
10. Gassay 406853.23 1,304,248 2795 1996–2015 1421.8

4
T.G. Andualem et al. Journal of Hydrology: Regional Studies 37 (2021) 100917

Table 2
Monthly average relative humidity (%) and sunshine hour values of Debre Tabor station.
Parameter Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Relative humidity (%) 42.3 38.7 42.2 44.9 55.1 70.8 82.5 90.3 75.2 63.9 54.3 48.1
Sunshine hour 8.62 8.80 7.54 7.20 6.64 6.25 4.47 4.87 6.55 7.67 7.75 8.74

Fig. 3. Mean Annual Precipitation (MAP) spatial distribution over the study area.

Fig. 4. Long-term monthly average precipitation in Guna Tana Landscape.

P1 + P2 + … + Pn
Pav = (1)
N

Where: Pav is the areal depth of precipitation, P1, P2…, Pn are rainfall records at stations 1, 2, etc., and N is the total number of stations
in and around the landscape/basin.
Theissen Polygon Method: This method considers the areal representation of each station and it is more accurate than the simple
arithmetic mean method.
P1 A1 + P2 A2 + … + Pn An
Pav = ∑n (2)
A
i=1

5
T.G. Andualem et al. Journal of Hydrology: Regional Studies 37 (2021) 100917

Where: P1, P2…, Pn are the mean annual rainfall recorded at each rainfall station and A1, A2 …, An are the polygonal areas around each
gauging station enclosed in the catchment and A is the total area of the study area.

2.2.2. Determination of evapotranspiration

Evapotranspiration was used to quantify the loss of water from the land surface to the atmosphere through evaporation and
transpiration. There are different methods (e.g. Turc, Thornthwaite, and FAO Penman-Monteith) of estimating evapotranspiration.
Turc utilized a formula to estimate annual actual evapotranspiration (AET) for catchment regions in humid or arid climates (Shaw,
1998), whereas the Thornthwaite approach relies on mean temperature data. In this study, the most widely used and appropriate FAO
Penman-Monteith method was employed to calculate the amount of potential evapotranspiration (Bautista et al., 2009; Lang et al.,
2017).
Penman developed an equation for the estimation of evapotranspiration by combining the mass transfer method with energy
balance from temperature, sunshine, relative humidity, and wind speed records. Penman also considered aerodynamic and surface
resistance factors for the estimation of reference evapotranspiration. The surface resistance, rsi, is the resistance of vapor flow through
stomata openings, total leaf area, and soil surface; while aerodynamic resistance, rai, is the resistance of vegetation upward involving
friction from the air flowing over the vegetative surface (FAO 56). Penman-Monteith developed the equation as:

Δ(Rni − Gi) + ρa cp (esir−aieai )

ET = (3)
Δ + γ(1 + rraisi )

Where Rni is the net radiation from the surface, Gi is soil heat flux, (esi-eai) is the vapor pressure deficit of the air, rai and rsi are the
aerodynamic and surface resistances, ρa is the mean air density at constant pressure, cp is the specific heat of air, Δ represents the slope
of the saturation vapor pressure-temperature relationship, and γ is the psychrometric constant.
[ ] [ ]
ln zmzom− d *ln zhzoh− d
rai = (4)
k2 uz

where rai is the aerodynamic resistance [s/m], zm is the height of wind measurements [m], zh is the height of humidity measurements
[m], d is the zero plane displacement height [m], zom is the roughness length governing momentum transfer [m], zoh is the roughness
length governing the transfer of heat and vapor [m], k is von Karman’s constant, 0.41 [-], uz is the wind speed at height z [m/s].
r1
rsi = (5)
LAIactive

where rsi is the surface resistance [s/m], rl is the bulk stomatal resistance of the well-illuminated leaf [s/m], LAIactive is the active leaf
area index [m2 (leaf area) per m2 (soil surface)].
The FAO Penman-Monteith equation derived to estimate ETo from Eqs. 3 and 4 as:
900
0.408Δ(Rn − G) + γ T+273u2 (es − ea )
ETo = (6)
Δ + γ(1 + 0.34u2 )

Where; ETo is the reference evapotranspiration (mm/day), Rn is net radiation at the crop surface (MJ/m2day, G is soil heat flux density
(MJ/ m2day), T is the mean daily air temperature at 2 m height (0C), u2 is the wind speed at 2 m height (m/s), es is saturation vapor
pressure (kPa), ea is actual vapor pressure (kPa), es-ea is saturation vapor pressure deficit (kPa), Δ is slope vapor pressure curve
(kPa/0C), γ is psychrometric constant (kPa/0C).

2.3. Methods of recharge computation

2.3.1. Computation of recharge from rainfall

The rainwater that reaches the ground might not fully infiltrate or contribute to the groundwater since some portions will evap­
orate, transpire, and runoff from the watershed. The rainwater which has infiltrated fills the soil moisture deficiency and/or percolate
down to recharge the aquifer. The evaluation of rainfall recharge was done by using different empirical relationships between recharge
and rainfall developed for different regions having similar climates (rainfall and temperature data) (Table 3).

Table 3
Recharge computation from recorded rainfall.
No. Formula Name Equation (s) Definition of parameter and coefficient value range

1 Chaturvedi Formula (in inch) (Chaturvedi, 1973) R = 2.0(P – 15) 0.4 P = Yearly rainfall (inch)
2 Modified Chaturvedi Formula (in inch) (Kumar and Seethapathi, 2002) R = 1.35(P – 14) 0.5 P = Yearly rainfall (inch)
3 Relationship of Krishna Rao (mm) (Krishna Rao, 1970) R = 0.30 (P – 500) P = Yearly rainfall (mm)
4 The Maxey-Eakin (1949) method (in mm) R = P*a P = Yearly rainfall, a = 20 %
5 Bredenkamp et al. (1995) formula (in mm) R = 0.32 (MAP – 360) MAP = Mean Annual rainfall (mm)

6
T.G. Andualem et al. Journal of Hydrology: Regional Studies 37 (2021) 100917

In 1970; Krishna Rao developed an empirical relationship to determine the groundwater recharge in a limited climatological
homogeneous area as Eq. 7:
R = K (P – X) (7)
Krishna Rao has developed the following relationships for different levels of annual rainfall:
R = 0.20 (P – 400) for areas with annual normal rainfall (P) between 400 and 600 mm
R = 0.25 (P – 400) for areas with P between 600 and 1000 mm
R = 0.35 (P – 600) for areas with P above 2000 mm
where R and P are expressed in millimeters.
For this study area having an annual rainfall of 1399.8 mm, the recharge equation was interpolated and used to be R = 0.30 (P –
500).
Maxey and Eakin (1949) also developed the empirical relationship between mean annual precipitation of the subbasin and
recharge from precipitation by considering evapotranspiration, surface water runoff, and recharge coefficient values between 0–25%.
Groundwater Estimation Committee Norms
Groundwater recharge from rainfall was also estimated using the Groundwater Estimation Committee norms. In 1987, the
Groundwater Estimation Committee developed ad-hoc norms of rainfall infiltration for determining the recharge from rainfall as
described below.

(i) Alluvial areas: recharge could be taken as 20–25% of rainfall for sandy areas, and 10–20% for areas with high clay content
(more than 40 % clay)
(ii) Semi-consolidated sandstones: 10–15% of rainfall is considered as recharge
(iii) Hard rock areas

Granitic terrain: for weathered and fractured rocks, 10–15% of rainfall

Unweathered – 5–10% of rainfall
Basaltic terrain: 10–15% of rainfall
Weathered basalt: 4–10% of rainfall
The Recharge Coefficient is defined as the ratio of recharge to effective rainfall, which indicates the effective rainfall that con­
tributes to groundwater aquifers and is expressed in percentage (Allocca et al., 2014; Muteraja, 1986) as:
R
RCoeff = (8)
Pe

Where RCoeff is the recharge coefficient, R is recharge and Pe is effective rainfall.

2.3.2. Computation of recharge from streamflow

The study area includes gauged and ungauged parts of the watershed. Around 63.95 % of the landscape is gauged and 36.05 % is
ungauged (Table 4).
The recharge which could be estimated from streamflow records was determined using different baseflow separation techniques.
Hydrograph analysis was used to separate the streamflow records into baseflow and direct runoff, which is essential for the estimation
of recharge rates (Sloto et al., 1991; Gunduz, 2006).

A Digital Filter Method using Web-Based Hydrograph Analysis Tool (WHAT): The digital filter method is used in signal analysis and
processing to separate high-frequency signals from low-frequency signals (Eckhardt, 2005). Filtering of the direct runoff from the
baseflow is similar to signal analysis and processing by using Eq. 9 (Arnold et al., 2000).
1+α
qt = α*qt− 1 + *(Qt − Qt− 1 ) (9)
2

Where, qt is the filtered direct runoff at the t time step (m3/s); qt-1 is the filtered direct runoff at the t-1 time step (m3/s); α is the filter
parameter; Qt is the total streamflow at the t time step (m3/s); and Qt-1 is the total streamflow at the t-1 time step (m3/s).

B PART: This program used time-series records of streamflow to determine the daily record of groundwater discharge by separating
the streamflow record. Knisel and Sheridan (1983) and Shirmohammadi et al. (1984) described how the daily records of streamflow

Table 4
Gauged and ungauged percentage of the study area.
No. Watershed Name Area Coverage (km2) Gauged Area (km2) Gauged area % coverage

1 Ribb 2094.177 844 40.30

2 Gumara 1405.453 1394 99.18
Total 3499.62 2238 63.95

7
T.G. Andualem et al. Journal of Hydrology: Regional Studies 37 (2021) 100917

and linear interpolation are used to evaluate groundwater discharge. In PART, a 2N-1 days interval block was created to examine
each day in the interval to determine the hydrograph’s lowest discharge. The blocks were then rearranged the next day, and the
procedure was repeated. If the requirements were met for the day, a straight line was drawn connecting the neighboring local
minima, defining the baseflow hydrograph.
C RECESS: this program determines the master recession curve of streamflow recession during times when all flow can be considered
to be groundwater discharge and when the profile of the groundwater head distribution is nearly stable. Meyboom (1961)
expressed the potential groundwater discharge based on the relationship between the logarithm of groundwater discharge and
time:
Qd *K
Vd = (10)
2.3026

Where, Vd is total potential ground-water discharge (L3), and Qd is ground-water discharge at an initial time (L3/T).

D RORA: This program used the recession-curve-displacement method to estimate the groundwater recharge for each peak instream
flow in units of specific discharge (inches per year) (Rorabaugh, 1964; Daniel, 1976). Before a critical time, this equation is used:
( )
C
dQ = √̅̅̅̅̅̅ 11
dT

Where, dQ is the difference between the ground-water discharge and the groundwater discharge that would have occurred at the same
time in the absence of the recharge event (L3/T), C is a constant, the value of which is dependent on the magnitude of the recharge
event (L3/T0.5), and dT is the time since the recharge event (T).
After the critical time, the linear extrapolation could be:
( )⎛ ⎞
− dT
K
⎜ ⎟
Qd = Qe *10 ⎝12⎠

Where, Qd is the groundwater discharge extrapolated to a time after critical time (L3/T), Qe is the groundwater discharge extrapolated
to a critical time after the peak (from Eq. 12 and superposition (L3/T)), and dT is the period from critical time to the day of interest (T).
Glover (1964) and Rorabaugh (1964) also showed that the total potential groundwater discharge to the stream at a critical time
after a peak instream flow is equal to about half of the total volume of water that recharged the groundwater system during the peak.
( )
2*(Q2 − Q1)*K
R= 13
2.3026

Where, R is the total volume of recharge due to the event (L3), Q1 is the groundwater discharge at a critical time as extrapolated from
the pre-event streamflow recession (L3/T), and Q2 is groundwater discharge at a critical time as extrapolated from the post-event
streamflow recession (L3/T).

3. Results and discussion

3.1. Areal depth of precipitation

The rainfall stations used in this study are non-uniformly distributed with different topographic variations. The precipitation
amounts among the gauging stations have significant variability (Table 5). The arithmetic average annual precipitation of the stations
within the landscape was 1373.3 mm for the period 1996–2015. Accordingly, the average annual precipitation of the study area using

Table 5
Theisen Polygon Method of Calculating Annual Rainfall.
Stations Area Coverage (km2) Weighted Area (%) Annual RF (mm) Weighted RF (mm)

Adis Zemen 454.6 12.8 1341.6 171.9

Amed Ber 493.8 13.9 1455.0 202.5
Debre Tabor 600.4 16.9 1535.7 259.9
Ebnat 354.3 10.0 1009.2 100.8
Gasay 343.2 9.7 1421.8 137.5
Derahamusit 126.1 3.6 1592.8 56.6
Mekaneyesus 376.9 10.6 1451.9 154.2
Wanzaye 389.8 11.0 1455.3 159.9
Woreta 375.8 10.6 1381.6 146.3
Yifag 32.9 0.9 1088.5 10.1
Sum 3547.84 100 1373.3 1399.8

8
T.G. Andualem et al. Journal of Hydrology: Regional Studies 37 (2021) 100917

the Theisen polygon method was estimated to be 1399.8 mm (Table 5). The average annual precipitation estimated using arithmetic
and Theisen polygon methods gave relatively similar results.
Even though the average annual rainfall for the Guna Tana landscape was computed at 1399.8 mm, it varies between the Ribb and
Gumara watersheds (Fig. 3). Hence, the computed annual areal rainfall for the two watersheds, Gumara and Ribb, was 1479.44 mm
and 1346.06 mm respectively. Thus, a variation in rainfall leads to a recharge difference between two watersheds.

3.2. Evapotranspiration

The reference evapotranspiration computed using the most typically used FAO Penman-Monteith method was found to be 1300.21
mm (about 93 % of the precipitation) for the period 1996–2015 (Table 6).
Evapotranspiration findings showed that there can be a large potential quantity of rainfall that has been lost in the atmosphere due
to evaporation and transpiration and does not contribute to groundwater recharge (Table 6). The recharge potential is therefore low
due to the high potential evapotranspiration (Portela et al., 2019). Since FAO Penman-Monteith computes only the reference
evapotranspiration, the actual evapotranspiration might be less than this value. Asmerom (2008) also used the Penman-Monteith
method and found similar results for evapotranspiration from the Lake Tana Basin with a value of 1344 mm yr–1 of water.

3.3. Groundwater recharge

3.3.1. Recharge from rainfall

The amount of recharge for the study area using the groundwater estimation committee method revealed that 197.23 mm of
rainfall has the potential to contribute to the aquifer storage (Table 7). The groundwater committee method gave slightly smaller
results of recharge as compared to the other empirical methods of recharge estimation from rainfall (Table 9). According to the
groundwater estimation committee, the recharge from rainfall depends upon the characteristics and types of rock in the landscape
(Groundwater Estimation Committee, 1997).
The bulk of the catchment, which is characterized by Quaternary Lacustrine sediments and Debre Tabor basalts, has the largest
potential for groundwater recharge due to it’s area and percentage recharge potential (Table 7 and Fig. 2). The Quaternary lacustrine
sediment type of lithology has good postulated groundwater holding and recharge capacity.
The two watersheds’ recharge from rainfall was also computed using the groundwater estimation committee method separately. It
has been estimated that the Ribb and Gumara watersheds have a groundwater recharge potential of 204.76 mm and 184.83 mm
respectively (Table 8). The result indicated that there is a slight variation due to the presence of more quaternary lacustrine sediments
in the downstream parts of the Ribb watershed than in Gumara.
The different empirical equations used to estimate groundwater recharge resulted in almost similar and comparable values
(Table 9). In the Guna Tana Landscape, the average annual recharge from rainfall was estimated to be about 253.70 mm, which has the
potential to contribute to the aquifer. Hence, for the Guna Tana landscape, the recharge coefficient from rainfall recharge was found to
be 0.18, which revealed that about 18 % of the rainfall has the potential to recharge the groundwater aquifers.

3.3.2. Recharge from streamflow

The relationship between surface and groundwater was determined by separating the baseflow from surface runoff using PART,
RECESS, RORA, and WHAT methods.

Table 6
Monthly Evapotranspiration using FAO Penman-Monteith method.
Month Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Annual

PET (mm) 107.99 115.93 132.13 128.81 124.05 106.14 91.15 91.81 100.52 105.02 96.24 100.41 1300.21

Table 7
Recharge from rainfall using groundwater estimation committee methods.
Area
No. Lithology Percent of rainfall recharged Rainfall (mm) Recharge (mm)
km2 Percentage

1 The plateau basalts and pyroclasts 16.66 0.47 10 1246.33 0.59

2 Quaternary lacustrine sediment 859.56 24.31 20 1358.05 66.03
4 Upper basalts and trachytes 526.76 14.90 12 1214.93 21.72
5 trachyte plug 7.23 0.20 10 1473.46 0.29
6 Upper basalts and pyroclasts 17.63 0.50 10 1430.13 0.72
7 Guna trachyte 29.05 0.82 8 1430.13 0.94
8 Guna tuff 224.75 6.36 10 1430.13 9.10
9 Quaternary volcanics 132.44 3.75 10 1460.68 5.48
10 Middle basalt flows 570.09 16.12 10 1438.29 23.19
11 Debre Tabor basalts and trachyte 1151.26 32.56 15 1417.00 69.21
Total 197.25

9
T.G. Andualem et al. Journal of Hydrology: Regional Studies 37 (2021) 100917

Table 8
Recharge from rainfall using groundwater estimation committee method of the two watersheds.
Ribb Gumara
% Rainfall
No. Lithology % Area Rainfall Recharge % Area Rainfall Recharge
Recharged
(mm) (mm) (mm) (mm)

1 The plateau basalts and 10 0.64 1248.05 0.80 – – –

pyroclasts
2 Quaternary lacustrine sediment 20 29.44 1323.30 77.92 15.48 1452.50 44.97
4 Upper basalts and trachytes 12 25.23 1214.58 36.77 – – –
5 trachyte plug 10 0.20 1438.15 0.29 0.22 1527.10 0.34
6 Upper basalts and pyroclasts 10 0.64 1430.10 0.92 0.24 1430.10 0.34
7 Guna trachyte 8 1.10 1430.10 1.26 0.32 1430.10 0.37
8 Guna tuff 10 4.95 1430.10 7.08 8.64 1430.10 12.36
9 Quaternary volcanics 10 – – – 9.32 1460.32 13.61
10 Middle basalt flows 10 – – – 40.53 1438.35 58.30
11 Debre Tabor basalts and trachyte 15 37.81 1405.86 79.73 25.24 1440.92 54.55
Total 204.76 184.83

Table 9
Recharge from rainfall.
No. Formula Name Recharge from rainfall (in) Recharge from rainfall (mm) Recharge coefficient

1 Chaturvedi Formula 8.76 222.42 0.16

2 Modified Chaturvedi Formula 8.66 219.86 0.16
3 Relationship of Krishna Rao 10.63 269.94 0.20
4 The Maxey-Eakin (1949) method 11.02 279.96 0.20
5 Bredenkamp et al. (1995) formula 13.1 332.74 0.24
6 Groundwater estimation committee 7.76 197.25 0.14
Average 253.7 0.18

Table 10
Baseflow separated using the PART program.
River name Drainage area (km2) Mean baseflow period Mean streamflow (mm yr–1) Baseflow (mm yr–1) Baseflow index

Upper Ribb 844 2001–2015 335.03 119.13 0.35

Gumara 1394 2001–2015 965.02 561.59 0.59

The mean streamflow and baseflow for the two rivers separated using the PART program (Table 10) indicated that the baseflow was
561.59 mm yr–1 and 119.13 mm yr–1 respectively for the Gumara and Ribb rivers. Even though the Ribb dam was upstream (500 m) of
the gauge, the data used for recharge computation is considered prior to the construction of the dam. Asmerom (2008) compared the
baseflow using Eckhardt and found that the Ribb watershed baseflow is significantly lower than Gumara watershed with values of 52.5
MCM yr–1 and 142.59 MCM yr–1 respectively. Dessie et al. (2014) found that the runoff contribution from the Ribb river to Lake Tana is
minimal.
The baseflow determined using the WHAT; recursive digital filter method also showed that there was an annual baseflow of 84.3
m3/s and 15.75 m3/s for the Gumara and Ribb watersheds respectively (Table 11 and Fig. 5).
Where, Kmin, Kmed, and Kmax are the Minimum, Median, and Maximum recession indexes respectively among segments selected,
LogQmn and LogQmx are the minima and maximum values of the log of the streamflow, and A, B, and C are the coefficients of the
master recession index.
The recession index (K) was determined for each watershed before recharge could be estimated by the recession-curve displace­
ment method with the RORA program. K is the time in days required for groundwater discharge to decline by one log cycle after the
recession curve becomes nearly linear on a semi-log hydrograph. K is determined by the construction of a master recession curve
representing streamflow during periods when most or all flow is contributed by groundwater discharge (Rutledge, 1998) (Table 12).
The recharge determined from streamflow records using the RORA method indicated that there was 142.14 mm yr–1 and 664.08
mm yr–1 for the Upper Ribb and Gumara watersheds respectively (Table 13). The recharge estimated using RORA gave a comparable
result to PART with 119.13 mm yr–1 and 561.59 mm yr–1 respectively for the Upper Ribb and Gumara watersheds. This indicates that
there is a high amount of groundwater that could be used for different groundwater development practices. The base flow for the Ribb
watershed showed a minimal result as compared to the Gumara watershed. This can be attributed to the differential measurements of
the recorded flow. The measurements of flow for the Ribb watershed were only for 40 % of the upper watershed, with the remaining
downstream parts of the watershed not being represented by the recorded flow. The recharge potential estimated using rainfall and
streamflow records contradicts since the Ribb streamflow measurement was recorded prior to the postulated potential lithology (the
streamflow gauge is upstream and therefore does not include quaternary lacustrine sediment formation/lithology). Besides, there was

10
T.G. Andualem et al. Journal of Hydrology: Regional Studies 37 (2021) 100917

Table 11
Baseflow separated using WHAT.
Gumara River Upper Ribb

Months Stream flow (m3/s) Direct Runoff (m3/s) Baseflow (m3/s) Stream flow (m3/s) Direct Runoff (m3/s) Baseflow (m3/s)

Jan 3.76 1.82 1.94 0.44 0.21 0.23

Feb 2.96 1.63 1.33 0.37 0.21 0.15
Mar 2.74 1.75 1.00 0.38 0.27 0.12
Apr 2.35 1.61 0.75 0.39 0.29 0.10
May 4.94 4.13 0.85 1.62 1.48 0.17
Jun 15.94 14.37 1.62 3.17 2.89 0.32
Jul 84.33 76.51 7.91 20.06 18.41 1.73
Aug 131.58 111.06 20.62 32.35 27.67 4.77
Sep 90.03 65.51 24.58 16.18 11.05 5.19
Oct 27.21 13.47 13.75 3.26 1.52 1.76
Nov 9.12 2.76 6.37 1.52 0.70 0.83
Dec 6.06 2.47 3.58 0.50 0.13 0.37
Total 381.04 297.09 84.30 80.25 64.85 15.75

Fig. 5. Gumara and Upper Ribb rivers baseflow separation using WHAT.

Table 12
RECESS results of baseflow separation.
File Period # Kmin Kmed Kmax logQmn logQmx A B C

Upper Ribb 2001− 2015 8 31.7 70.9 113.8 0.915 1.802 11.1 − 101.3 146.4
Gumara 2001− 2016 4 221.4 270.4 283.6 1.832 2.297 65.1 − 531.7 877.8

Table 13
Recharge estimated from streamflow using RORA program.
River Drainage area Mean baseflow Requirement of the antecedent Number of peaks Recession index Recharge (mm
Name (km2) period recession (days) detected (days/Log Q) yr–1)

Upper 844 2001–2015 4 288 65 142.14

Ribb
Gumara 1394 2001–2015 4 247 65 664.08

a considerable precipitation and evapotranspiration difference between the two watersheds. Dessie et al. (2014) stated this might be
due to high evapotranspiration and/or the discharge of groundwater that is not captured by the gauge measurements (as it discharges
downstream of the gauges).

4. Conclusions

Groundwater recharge can be computed using different methods and models depending on the availability of data. The magnitude
of groundwater recharge potential could be estimated using different techniques. Different empirical formulas were used to estimate
the areal precipitation, evapotranspiration, and groundwater recharge quantities. The areal precipitation of the study area was used for

11
T.G. Andualem et al. Journal of Hydrology: Regional Studies 37 (2021) 100917

the estimation of the groundwater recharge using the water balance method and the groundwater estimation committee method. The
recharge estimated from streamflow was determined by separating the baseflow using different methods. In the Guna Tana landscape
(Ribb and Gumara watersheds), the annual rainfall was 1399.82 mm, with the direct runoff and groundwater recharge estimated
values of 1159.35 mm and 240.47 mm respectively. Even though there was a high amount of rain, a large amount evaporated and
transpired from the land surface and vegetation. The recharge estimated from streamflow records indicated that there was a total of
806.22 mm yr–1 and 680.72 mm yr–1 using RORA and PART methods respectively. The recharge coefficient derived from rainfall and
streamflow data was found to be 0.18 and 0.20 respectively. This implied that around 20 % of the rainfall could be considered as a
potential groundwater recharge amount. The study results indicated that the groundwater recharge estimation either from rainfall or
streamflow data is almost the same. This showed that in data-scarce areas, streamflow and rainfall data could be used for groundwater
recharge estimation. In data-scarce areas, empirical equations might be good to estimate the recharge magnitude from rainfall and
streamflow records for further groundwater development projects initiation and implementation. Therefore, groundwater develop­
ment projects could be developed to emphasize the development of the study area and the country as a whole. This would increase the
positive socio-economic value of water and reduce water security problems. People in this research region would profit from
groundwater by exploring it through shallow and groundwater wells that may be utilized for residential and irrigation purposes. The
researchers also propose that before implementing groundwater projects, the real groundwater recharge be estimated using water
balance and other direct measurement methods.

Declaration of Competing Interest

There is no conflict of interest between the authors.

Acknowledgment

The authors kindly acknowledge the South Gondar zone water, irrigation, and energy department, Geological Survey of Ethiopia,
and Amhara Meteorology agency for supplying free of charge data. The authors also give their great thanks to Debre Tabor University
for supporting the idea of this research. The authors are highly thankful to the handling editor (Prof Denis Hughes) and reviewers for
improving the quality of the paper.

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.2021.
100917.

References

Allocca, V., Manna, F., De Vita, P., 2014. Estimating annual groundwater recharge coefficient for karst aquifers of the southern Apennines (Italy). Hydrol. Earth Syst.
Sci. 18 (2), 803–817.
Andualem, T.G., Demeke, G.G., 2019. Groundwater potential mapping using GIS and remote sensing; a case study of Guna Tana landscape upper Blue Nile Basin,
Ethiopia. J. Hydrol. Reg. Stud. 24.
Arnold, J.G., Muttiah, R.S., Srinivasan, R., Allen, P.M., 2000. Regional estimation of baseflow and groundwater recharge in the Upper Mississippi river basin.
J. Hydrol. (Amst) 227 (1–4), 21–40.
Asmerom, G.H., 2008. Groundwater Contribution and Recharge Estimation in the Upper Blue Nile Flows. ITC, Ethiopia.
Bautista, F., Bautista, D., Delgado-Carranza, C., 2009. Calibration of the equations of Hargreaves and Thornthwaite to estimate the potential evapotranspiration in
semi-arid and subhumid tropical climates for regional applications. Atmsfera 22 (4), 331–348.
Bhutta, M.N., Chaudhry, M.R., Chaudhry, A.H., 2005. Groundwater quality and availability in Pakistan. Proceedings of the Seminar on Strategies to Address the
Present and Future Water Quality Issues.
Bredenkamp, D., Botha, L.J., Van Tonder, G.J., Janse van Rensburg, H., 1995. Manual on qualitative estimation of groundwater recharge and aquifer storativity, based
on practical hydrological methods. Water Res. Commission. TT 73/95. ISBN: 1 86845 1763.
Chaturvedi, R.S., 1973. A Note on the Investigation of Groundwater Resources in Western Districts of Uttar Pradesh. Annual Report. UP Irrigation Research Institute,
pp. 86–122, 1973.
Chen, W., Lee, C., 2003. Estimating groundwater recharge from streamflow records. Environ. Geol. 44 (n. 3), 257–265. https://doi.org/10.1007/s00254-002-0753-2.
Cheng-Haw, Lee, Chen, Wei-Ping, Lee, Ru-Huang, 2006. Estimation of groundwater recharge using water balance coupled with base-flow-record estimation and
stable-base-flow analysis. Environ. Geol. 51 (2006), 73–82. https://doi.org/10.1007/s00254-006-0305-2.
Daniel, J.F., 1976. Estimating groundwater evapotranspiration from streamflow records. Water Resour. Res. 12 (3), 360–364.
Dessie, M., Verhoest, N.E.C., Admasu, T., Pauwels, V.R.N., Poesen, J., Adgo, E., Deckers, J., Nyssen, J., 2014. Effects of the floodplain on river discharge into Lake
Tana (ethiopia). J. Hydrol. 519, 699–710. https://doi.org/10.1016/j.jhydrol.2014.08.007.
Dzhamalov, R.G., 1973. Ground-water Flow of the Terek-kuma Artesian Basin.
Eckhardt, K., 2005. How to construct recursive digital filters for baseflow separation. Hydrol. Proces. Int. J. 19 (2), 507–515.
Eckhardt, K., 2008. A comparison of baseflow indices, which were calculated with seven different baseflow separation methods. J. Hydrol. (Amst) 352 (1–2),
168–173.
Groundwater Estimation Committee (GEC), 1997. Detailed Guidelines for Implementing the Groundwater Estimation Methodology. Central Groundwater Board,
Ministry of Water Resources, Government of India.
Gunduz, O., 2006. Surface/subsurface interactions: coupling mechanisms and numerical solution procedures. Groundwater and Ecosystems. Springer, Dordrecht,
pp. 121–130.
Knisel, W.G., Sheridan, J.M., 1983. Procedure for characterizing hydrologic processes in the coastal plain of the southeastern United States. Proceedings Hydrology on
Large Flat Lands, UNESCO International Programme on Hydrology.

12
T.G. Andualem et al. Journal of Hydrology: Regional Studies 37 (2021) 100917

Krishna Rao, P.R., 1970. Hydrometeorological aspects of estimating groundwater potential. In: Seminar Volume. Groundwater Potential of Hard Rock Areas of India.
Bangalore, p. 1.
Kumar, C.P., Seethapathi, P.V., 2002. Assessment of natural groundwater recharge in Upper Ganga Canal command area. J. Appl. Hydrol. 15 (4), 13–20.
Lang, D., Zheng, J., Shi, J., Liao, F., Ma, X., Wang, W., et al., 2017. A comparative study of potential evapotranspiration estimation by eight methods with FAO
Penman–Monteith method in southwestern China. Water 9 (10), 734.
Mau, D.P., Winter, T.C., 1997. Estimating ground-water recharge from streamflow hydrographs for a small mountain watershed in a temperate humid climate, New
Hampshire, USA. Groundwater 35 (2), 291–304.
Maxey, G.B., Eakin, T.E., 1949. Ground Water in White River Valley, White Pine, Nye, and Lincoln Counties, Nevada.
Meyboom, P., 1961. Estimating ground-water recharge from stream hydrographs. J. Geophys. Res. 66 (4), 1203–1214.
Misstear, B.D.R., Brown, L., Johnston, P.M., 2009. Estimation of groundwater recharge in a major sand and gravel aquifer in Ireland using multiple approaches.
Hydrogeol. J. 17 (3), 693–706.
Moon, S.K., Woo, N.C., Lee, K.S., 2004. Statistical analysis of hydrographs and water-table fluctuation to estimate groundwater recharge. J. Hydrol. 292 (1–4),
198–209.
Muteraja, K.N., 1986. Applied Hydrology. Tata McGraw-Hill, New Delhi, India.
Niazi, A., Bentley, L.R., Hayashi, M., 2017. Estimation of spatial distribution of groundwater recharge from stream baseflow and groundwater chloride. J. Hydrol. 546,
380–392.
Olmsted, F.H., Hely, A.G., 1962. Relation Between Groundwater and Surface Water in Brandywine Creek Basin. US Government Printing Office, Pennsylvania.
Portela, M.M., Santos, J., de Carvalho Studart, T.M., 2019. Effect of the evapotranspiration of Thornthwaite and of penman-monteith in the estimation of monthly
streamflows based on a monthly Water balance model. Current Practice in Fluvial Geomorphology-Dynamics and Diversity. IntechOpen.
Rorabaugh, M.I., 1964. Estimating changes in bank storage and groundwater contribution to streamflow. Int. Assoc. Sci. Hydro. Publ. 63, 432–441.
Rushton, K., 1997. Recharge from permanent water bodies. In: Simmers, I. (Ed.), Recharge of Phreatic Aquifers in (Semi)Arid Areas. AA Balkema, Rotterdam,
pp. 215–255.
Rushton, K.R., Ward, C., 1979. The estimation of groundwater recharge. J. Hydrol. 41 (3–4), 345–361.
Rutledge, A.T., 1992. Methods of using streamflow records for estimating total and effective recharge in the appalachian Valley and Ridge, piedmont, and Blue Ridge
physiographic provinces. In: Regional Aquifer Systems of the United States, Aquifers of the Southern and Eastern States. American Water Resources Association
Monograph Series, 17, pp. 59–73.
Rutledge, A.T., 1998. Computer Programs for Describing the Recession of Ground-water Discharge and for Estimating Mean Ground-water Recharge and Discharge
From Streamflow Records: Update.
Sanford, W., 2002. Recharge and groundwater models: an overview. Hydrogeol. J. 10 (1), 110–120.
Shaw, E.M., 1998. Hydrology in Practice, 3rd ed. Stanley Thornes Pub, UK, p. 569.
Shirmohammadi, A., Knisel, W.G., Sheridan, J.M., 1984. An approximate method for partitioning daily streamflow data. J. Hydrol. (Amst) 74 (3–4), 335–354.
Sloto, R.A., Cecil, L.D., Senior, L.A., 1991. Hydrogeology and Ground-water Flow in the Carbonate Rocks of the Little Lehigh Creek Basin, Lehigh County,
Pennsylvania. US Department of the Interior, US Geological Survey.
Zektser, I.S., 1977. Laws of Formation of Groundwater Runoff Study. Nedra, Moscow, p. 173.

13