IMPACT OF LAND USE LAND COVER AND CLIMATE CHANGE ON

WATER BALANCE COMPONENTS HARGEISA WATERSHED IN

SOMALILAND

MSc THESIS

MOHAMED ABDISAMED KHALIF

NOVEMBER 2023

HARAMAYA UNIVERSITY, ETHIOPIA

IMPACT OF LAND USE LAND COVER AND CLIMATE CHANGE ON
WATER BALANCE COMPONENTS HARGEISA WATERSHED IN
SOMALILAND

The thesis submitted to the School of Water Resources and Environmental

Engineering,

Postgraduate Program Directorate

HARAMAYA UNIVERSITY

In Partial Fulfilment of the Requirement of Degree of MASTER OF SCI-

ENCE IN ENGINEERING HYDROLOGY

By
MOHAMED ABDISAMED KHALIF

NOVEMBER 2023
HARAMAYA UNIVERSITY, ETHIOPIA

HARAMAYA UNIVERSITY
POSTGRADUATE PROGRAM DIRECTORATE

I hereby certify that I have read and evaluated this thesis entitled: “Impact of land use land
cover and climate change on water balance components Hargeisa Watershed in Somalil-
 iv

and”, prepared under my guidance by Mohamed Abdisamed Khalif. I recommend that it be

submitted as fulfilling the Thesis requirement.

Meseret Dawit (PhD) ______________ _______________

Major Advisor Signature Date

Asfaw Kebede (PhD) _______________

_______________
Co-advisor signature Date

As a member of the Board of Examiners of the MSc Thesis Open Defences Examination, I
certify that I have read and evaluated this thesis prepared by Mohamed Abdisamed

Khalif and examined the candidate. I recommend that this thesis be accepted as fulfilling the
Thesis requirements for the degree of Master of Science in Engineering Hydrology.

__________________ ______________ _______________

Chairperson Signature Date

__________________ _______________ _______________

Internal examiner signature Date

__________________ _______________ _______________

External examiner signature Date

Final approval and acceptance of this Thesis is contingent upon the submission of its final
copy to the council of Graduate studies (CGS) through the candidate’s department or School
Graduate committee (DGC or SGC).
 I

DEDICATION

I dedicated this thesis manuscript to my beloved sister Maryam Mohamed Carte who suppor-
ted, encouraged, and being at my side during my study period and preparations for this thesis
paper and at every time.
 II

STATEMENT OF THE AUTHOR

My signature below, I declare and affirm that this Thesis is my work. I have followed all eth-
ical and technical principles of scholarship in the preparation, data collection, data analysis,
and compilation of this Thesis. Any scholarly matter included in the thesis has been recog-
nized through citation.

This Thesis is submitted in partial fulfillment of the requirements for the Master's degree of
impact of land use land cover and climate change on water balance components Hargeisa
watershed in Somaliland at Haramaya University. This thesis is deposited in Haramaya Uni-
versity Library and made available to the borrowers by following the rules and regulations of
the library. I declare that this thesis has not been submitted to any other university or college
or any institution for academic purposes.

Brief quotations from this Thesis may be made without special permission provided that ac-
curate and complete acknowledgment of the source is made. Requests for permission for ex-
tended quotations from or reproduction of this Thesis in whole or in part may be granted by
the Head of the School or Department when his or her judgment the proposed use of the ma-
terial is for a scholarly interest. In all other instances, however, permission must be obtained
from the author of the Thesis.

Name: Mohamed Abdisamed Khalif

Date: ______________________

Signature: __________________

School/Department: ___________________
 III

ACKNOWLEDGEMENTS

This hard work would be difficult and not be completed without support, encouragement, and
honest work of different parts. In the following sections, I will mention some individuals who
take part in this unforgettable work.

First and foremost, I would like to thank the almighty Allah for his continuous support at
anytime and anywhere and for making me strong, healthy, and capable of doing this hard
work. Secondly, I want to acknowledge is for my beloved friend Mr. Abebe for his honest
support and for giving constructive and continuous support at any time, by calling, by E-mail,
and face to reach my goal. In addition, I want to thank deeply to my major adviser Meseret
Dawit (PhD), and my Co-adviser Asfaw Kebede (PhD) for her/his endless support and guid-
ance.

Thirdly, I would like to thank Haramaya University, especially the postgraduate program dir-
ectorate, especially the School of Water Resource and Environmental Engineering studies,
my beloved professors, and the department head Mr Nura Boru for his/their facilitation and
for giving me this chance.

fourthly, I would like to thank my friends who helped me with the data collection, including
Eng. Ridwan Ahmed, Mss Hawo Hamid, and other best friends. Also, my classmates helped
me during my study period. In addition. And finally, my last thanks are to my beloved family,
my father Abdisamed Khalif, my brothers Mohamud/omar Abdirahman, my brothers and sis-
ters, and all other parts of my beautiful family for their support and encouragement, whenever
and wherever.
 IV

LIST OF ACRONYMS And ABBREVIATION

ADBG African Development Bank Group

CMHYD Climate Model Hydrological

CORDEX Coordinated Regional Climate Downscaling Experiments

DEM Digital Elevation Map

FAO Food and Agricultural Organization

GCMs Global Climate Models

GD Geed Deeble

GHG Green House Gases

GIS Geographic Information System

HUWSUP Hargeisa Urban Water Supply Upgrading Supply

HWA Hargeisa Water Agency

IPCC Intergovernmental Panel Climate Change

LULCC Land Use Land Cover Change

MNPD Ministry of National Planning and Development

MOA Ministry of Agriculture

MOEAWR Ministry of Energy and Water Resources

MWMR Ministry of Water and Mineral Resources

NASA POWE National Aeronautics and Space Administration of Worldwide En-

ergy Resources

NGDPS New Geed Deeble Pumping Station

NGOs Non-Governmental Organizations

PET Potential Evapotranspiration

PPPS Public Private Partnership

RCMs Representative Concentration Pathways

RCPs Regional Climate Pathways

 V

RMSE Root Mean Square Error

SWALIM Somali Water and Land Information Management

US United State

USDA United States Department of Agriculture

 VI

Table of Contents
DEDICATION I
STATEMENT OF THE AUTHOR II
ACKNOWLEDGEMENTS III
LIST OF ACRONYMS And ABBREVIATION IV
LIST OF TABLES IX
LIST OF FIGURES X
ABSTRACT XII
1. INTRODUCTION 1
1.1 Background 1
1.2. Problem Statement 3
1.3. Research Objectives 4
1.3.1. General Objective 4
1.3.2. Specific Objective 4
1.4. Research Questions 4
1.5. Significance of the Study 4
1.6. Scope of the Study 5
1.7. Limitation 5
2. LITERATURE REVIEW 6
2.1. Theoretical Literature Review 6
2.2. Land Use/Cover Change Impact on Water Resource 7
2.3. Climate change impact on water resource 9
2.4 Impact of land use/cover and climate change on water resource 11
2.5 Rainfall-Runoff-Recharge Relationship 12
2.7 Remote Sensing and GIS for groundwater recharge estimation 13
2.8 Estimation of recharge, runoff and actual evapotranspiration 13
2.8.1. Estimation of groundwater recharge 13
2.8.2 Estimation of actual evapotranspiration and runoff 14
2.9 Hydrological models 14
2.10 The Wetspass Model 15
3. MATERIALS AND METHODS 17
3.1. Description of the Study area 17
3.1.1. Socio-economic of the Study Area 17
3.1.2. Water supply situation of the area 17
3.1.3. Agriculture activities 18
 VII

3.1.4. Land use land cover 19

3.1.5. Climate 19
3.2. Data collection 21
3.2.1. Meteorological Data 21
3.2.2. Physical Data 21
3.2.3. Satellite Imaginary Data 22
3.3. Methodology 22
3.3.1. Groundwater Recharge Estimation Model 22
3.3.2. Groundwater Recharge Zone 23
3.3.3. Climate Data 23
3.3.4. Land Use and Soil Data 24
3.3.5. Methodology Framework 25
3.4. Wetspass Model Description 26
3.5. Model Calibration 27
3.6. Wetspass Model input data 28
3.7. grid maps 30
3.8. Rainfall 31
3.9. Temperature 33
3.10. actual evapotranspiration (AET) 35
3.11. Wind speed 37
3.12. Topography 39
3.13. Slope 40
3.15. Soil Data 41
3.16. groundwater depth 42
4. RESULTS AND DISCUSSION 43
4.1. Estimation of groundwater recharge 43
4.1.1. Calibration of groundwater recharge results 43
4.1.2. Model Calibration of The Recharge 44
4.2. Wetspass output 45
4.2.1. Actual Evapotranspiration (AET) 45
4.2.2. Surface Runoff 48
4.2.3. Groundwater Recharge Zone 50
4.3 Land Use Land Cover 53
4.4. Future Climate Change 55
4.4.1. Maximum Temperature 55
4.4.2. Seasonal Maximum Temperature 57
 VIII

4.5.1 Minimum Temperature 59

4.5.2. Seasonal Minimum Temperature 60
4.6.1. Rainfall 62
4.6.2. Seasonal Rainfall 63
4.7. Impact of Land Use Land Cover and Climate Change on water balance components 65
4.7.1. Land use land cover change impact 65
4.7.2. Future climate change impact 67
5. CONCLUSION AND RECOMMENDATION 71
5.1. Conclusion 71
5.2. Recommendations 72
6. REFERENCES 74
 IX

LIST OF TABLES

Table 1: Input variables for the Wetspass model 38

Table 2: Statistical parameters of Wetspass recharge compared with an empirical equation 54
Table 3: Land use types in Hargeisa Watershed 63
Table 4: land-use land cover changes in the Hargeisa Watershed 64
Table 5: Seasonal rainfall projected (%) 79
 X

LIST OF FIGURES

Figure 1: SWALIM (Somali Water and Land Information Management). 20

Figure 2. schematic water balance of hypothetical raster cell (Batelaan & DeSmedt, 2001) 25
Figure 3: Study area map 30
Figure 4: Flow Chart Diagram 35
Figure 5: monthly average rainfall for stations in the study area 41
Figure 6: monthly average rainfall for stations in the study area 42
Figure 7: Monthly average temperature stations in the study area 43
Figure 8: Seasonal average temperature map 44
Figure 9: actual evapotranspiration values stations of the study area 45
Figure 10: seasonal potential evapotranspiration map of the watershed 46
Figure 11: Monthly average wind speed values stations in the Hargeisa watershed 47
Figure 12: Seasonal average wind speed map of the Hargeisa watershed (worldclim. Org) 48
Figure 13: The topographic elevation map of the Hargeisa watershed coverage area 49
Figure 14: Slope map of the study area 50
Figure 15: Soil map of the study area (Hargeisa 51
Figure 16: Groundwater map in the Hargeisa watershed. 52
Figure 17. calibration graph 55
Figure 18: Seasonal actual evapotranspiration map of Hargeisa watershed a) winter AET b)
spring AET c) summer AET d) autumn AET 58
Figure 19: Runoff maps in the Hargeisa watershed a) Diraac/winter runoff b) Gu’/spring run-
off c) Xagaa/summer runoff d) Dayr/autumn runoff 60
Figure 20: the relationship between monthly recharge and rainfall 61
Figure 21: Seasonal variation of water balance component in the watershed 62
Figure 22 Groundwater recharge map zone in the watershed a) winter recharge b) spring re-
charge c) summer recharge d) autumn recharge 63
Figure 23: Land cover type and percentage Area change 65
Figure 24: Land use type map in the Hargeisa watershed for 2001 and 2020 66
Figure 25: mean maximum temperature and change of base period (1986-2005) for both
scenarios 67
Figure 26: mean maximum temperature and change of base period (1986-2005) for both
scenarios. 67
 XI

Figure 27: Seasonal max temperature change for RCP4.5 and 8.5 scenarios during the projec-
tion period in the 2020s and 2050s. 69
Figure 28: Seasonal max temperature change for RCP4.5 and 8.5 scenarios during the projec-
tion period in the 2070s and 2080s. 69
Figure 29: mean minimum temperature and change of base period (1986-2005) for both scen-
arios 71
Figure 30: mean minimum temperature and change of base period (1986-2005) for both scen-
arios 71
Figure 31: seasonal min temperature change for RCP4.5, 8.5 scenarios during the projection
period in the 2020s and 2050s 72
Figure 32: seasonal min temperature change for RCP4.5, 8.5 scenarios during the projection
period in the 2070s and 2080s 73
Figure 33: mean annual rainfall and change of base period (1986-2005) for both scenarios 74
Figure 34: mean annual rainfall and change of base period (1986-2005) for both scenarios 74
Figure 35: seasonal rainfall changes for RCP4.5, 8.5 scenarios during the projection period
2020s and 2050s 76
Figure 36: seasonal rainfall changes for RCP4.5, 8.5 scenarios during the projection period
2070s and 2080s 76
Figure 37: Annual recharge distribution map 2001 and 2020 77
Figure 38: annual surface runoff distribution map in 2001 and 2020 78
Figure 39: seasonal rainfall projected RCP4.5, 8.5 81
 XII

ABSTRACT

Understanding the hydrological environment and groundwater resource availability leads to

effectively planning, developing, and managing the available water resources. There are dif-
ferent physical and empirical models to understand the water balance components of a given
watershed. One of the spatially distributed physical models is the Wetspass model that’s re-
spond the impact of lulc on water balance components and climate Hargeisa watershed. The
Wetspass-M model has simulated the annual, monthly water balance components of the
Hargeisa watershed successfully, based on the model, the monthly water balance compon-
ents in the Hargeisa watershed is summed up by seasonality; the highest season recharge oc-
curs in summer were 0 to 74.9mm, sprin were 0 to 68mm, and Autumn moderate, while
winter is dry. The recharge rates performed by Wetspass model were consequently compared
with those achieved by empirical relations namely; Chaturvedi Formula (CF), Sehgal For-
mula (SF), Krishna Rao Formula (KRF), and Bhattacharya Formula (BF). It was exhibited
that average rate of yearly groundwater recharge for calibration periods during 2014–2017
was 193.02mm/yr with RMSE and R², of 46.76mm, 0.7, respectively . The model also resulted
in monthly annual runoff in the watershed, which were seasonal 69.37mm, 59.38mm, and
40.4mm in the spring, summer, and autumn, respectively. Likewise, the AET seasonal water-
shed is 138mm, 159mm, 207mm, and autumn 158mm in the winter, spring, summer, and au-
tumn, respectively. The sensitivity analysis of the different input variables was conducted and
most of the variables are highly sensitive in the Hargeisa watershed. The analysis results
show that rainfall, soil, and slope are the most important hydrologic processes in the study
area in terms of affecting the amount and rate of the different water balance components.
The parametric coefficient of alfa coefficients, interception coefficients, and Lp coefficient
are also relatively sensitive. The calibration was made between the simulated and observed
empirical recharge through the coefficient determination R^2 0.87, which indicates a good
correlation between both. the future climate projected using Regional climate model based
CMIP5 namely single model ROC-ESM-CHEM under RCP4.5, and 8.5 scenarios, in the pro-
jected temperatures to rise incessantly, although the monthly annual rainfall increased, the
seasonal rainfall increased in winter, and autumn for both scenarios, but summer decreased
RCP4.5 13.47 and 9% 2020s, 2050s, and RCP8.5 2.47% 2050s. Spring decreased for both
scenarios in all periods. The Wetspass-M model has successfully simulated the annual
monthly water balance components.
The high variable distribution of the climatic inputs (parameters) with the variation in lulc,
soil texture, topography, and slope are responsible for variations of water balance elements
within the watershed. Likewise, the future climate projected result shows the increase in tem-
perature max/min annually and season, while the rainfall increases annually, but decreases
seasonally specifically rain season in the study.

Keywords: Hargeisa watershed, WBC, land use land cover, Wetspass model, climate.
 XIII

1. INTRODUCTION

1.1 Background

The climate and land use/land cover (LULC) change are key factors that can regulate flow re-
gimes and water availability. Since the 20th century, climate variability is believed to have
caused changes in global precipitation patterns, and the effects of climate and LULC changes
on hydrological processes and water resources will probably continue to increase, particularly
in arid and semi-arid regions characterized as vulnerable(Yin et al. 2017). According to YIN
et al ( 217), the effects of climate and LULC changes on runoff can generally be recognized
by using hydrological models, these models provide precious frameworks for investigating
the modifications among numerous hydrological pathways that are due to climate and human
activities, distributed hydrological models, which use input parameters that immediately rep-
resent land surface characteristics, were applied to evaluate the effects of climate LULC
changes in runoff in water resource management areas.

During the next few decades, climate and land use change are major global issues, and proper
studies highly demand finding their relationships and their impacts for the future(Isabirye et
al., 2012). In 2014, the National Climate Assessment recognized water availability as one of
the key impacts of climate change in the southeast US, and forecasts indicate that much of the
region will experience an increasing deficit in human demand and supply. Consistently with
global trends, the US Southeast is expected to experience a doubling in urban areas in 2060,,
leading to an increase in the average annual temperature and expected greater frequency and
severity of droughts and floods (Martin et al. 2017).

Developing countries with political uncertainty and poor economic status do not have clear
and effective rural water supply, hygiene, and sanitation rules and strategies. Such policies
are written; however, the lack of structure, confined resources, and coherence makes them in-
effective. The water regulation framework authorized by the parliament of Somaliland for the
water sector to perform effectively has no longer been enforced effectively(City & Ahmed,
2019).

Water availability in Somaliland depends more on underground reservoirs than on surface

water bodies, as there aren’t any major rivers or other permanent surface water in the country.
However, berkads and earth dams contributed significantly to water availability in rural
Somaliland before 1991. In the 1980s, the Somali government, through the Rural Water De-
 XIV

velopment Project funded by The World Bank, developed six earth dams with plastic pave-
ment that could last 3–4 months after the rains. At present, the best dams retain water for long
periods. The other three collapsed because of poor maintenance, attributed mainly to the loss
of technical personnel. Many of the berkads could not be sustained due to the reduced income
of pastoralists due to war and prolonged droughts(Muthusi et al., 2007).

The water resources used in Somaliland include small dams, Berkads on the surface of the
water catchment. In contrast, the shallow well boreholes and springs are used for groundwa-
ter resources. The water resource availability and aquifer characteristics, groundwater yield
very widely, the groundwater yield depend on aquifer behaviour. Groundwater aquifers in
Somaliland are characterized from place to place regards conditions geology and hydrology.
although a seasonal behaviour of the aquifers has been reported, the recharge would rather
benefit decentralized smaller-scale water resources that use the upper aquifers in place of re-
charging the main deeper aquifers that are presently used or are planned to be developed for
bulk water supply in Hargeisa(Petersen, G., Gadain, H.M., 2012).

According to Yonis (2015), Somaliland’s water supply mainly depends on rainwater from
April to June (the spring season), and September to November (the autumn season), so if the
rainy time changes and prolonged period of rain then the population frequently experienced
drought Climate change and global warming are reasons for serious water shortages globally,
especially in arid and semi-arid regions such Somaliland, while. Hargeisa is a primary and
most populated town of Somaliland in the present-day time, due to being central for commer-
cial and seats of government. The water supply system does not reach the entire city because
of poorly managed institutions involved with water distribution and lack of capacity resulting
from poor performance and unskilled Hargeisa Water Supply Agent (HWA) staff. In addi-
tion, the biggest challenges in Hargeisa water supply are technical staff, financial and hard
situation of the environment.

The water resources situation in Hargeisa is characterized by an undersupply and reduced per
capita water availability because of climate change and populace growth. The water supply is
controlled by the Hargeisa Water Agency, a non-public water utility that manages the water
grid. The water supply is augmented by private water trucking businesses that supply water to
now no longer connected populations from loads of boreholes and shallow wells in the city
area. A master plan for Hargeisa has been developed, and goals for the supplementation of
Geed Deeble through extra good fields and respective water conveyance infrastructure to ful-
fil the demand which have been based on the land use and land cover and climate. The pos-
 XV

sibilities to enhance water balance component around Hargeisa watershed are limited, partic-
ularly considering the identified deficit. While Hargeisa features a developed stream network
with sand deposits that could be used for water balance streams do not carry significant run-
off(Abdishakur et al., 2022).

1.2. Problem Statement

In most parts of the country depended on groundwater, which caused a huge water shortage
because there’s no other source of water sources, therefore more consumption of groundwater
leads decline of groundwater level and scarcity(Jama & Mourad, 2019).

The Hargeisa Water Agency (HWA) sector does not document information regarding land
use, land cover and climate change due to the absence of a research sector or scholars who
have investigated the impact of land cover and climate change on water resources. However,
changes in land use and land cover greatly affect the climate, which in turn directly impacts
water resources. While the Somali people are well-informed about the effects of climate
change, the problem lies in the lack of land information. Somali scholars such as engineers,
geologists, and hydrologists have not sufficiently investigated the impact of land use and land
cover on water resources.

Despite several challenges in evaluating the impact of climate change on groundwater re-
sources, no studies have been conducted to explore the effect of land use, land cover and cli-
mate change on water resources using the Wetspass model in the water balance components
Hargeisa watershed in Somaliland. The Hargeisa watershed has great potential for investigat-
ing annual or seasonal water balances to assess the availability of water resources. Ground-
water is the primary source of water for domestic and agricultural use in the Hargeisa water-
shed. However, the abstraction and consumption of groundwater in the area is not based on
an understanding of the Hargeisa water balance component, or the impact of land use and cli -
mate change on the hydrological process. Nevertheless, the water balance of the Hargeisa wa-
tershed has not yet been estimated using Wetspass modelling or other methods.

The purpose of this research work is to fill the aforementioned gaps and resolve the problem
of future proper utilization of the water balance in the Hargeisa watershed by estimating the
seasonal and annual values based on the biophysical and meteorological characteristics of the
watershed.
 XVI

1.3. Research Objectives

1.3.1. General Objective

The general objective is to evaluate the impact of land use land cover and climate change on
water balance component Hargeisa watershed in Somaliland

1.3.2. Specific Objective

1. To assess the effect of land use land cover on the water balance component of
Hargeisa watershed
2. To assess and mapping water balance components of Hargeisa watershed
3. To evaluate climate change's impact on groundwater resource of the Hargeisa water-
shed

1.4. Research Questions

The study attempts to the impact of land use, land cover, and climate change on water
balance components Hargeisa watershed in Somaliland.
1. What are the effects of land use land cover on the water balance components
of Hargeisa watershed
2. Where is the groundwater recharge zone in Hargeisa watershed?
3. What is the impact of climate change on the groundwater resource of the
Hargeisa watershed?

1.5. Significance of the Study

land use and climate changes have a significant impact on hydrological processes in water-
sheds, which can alter surface hydrology and water balance. This ultimately affects the avail-
ability and security of water resources globally, as noted by Warku et al. (2022). In particular,
Somalia, including Somaliland, is highly vulnerable to the effects of climate change due to
frequent droughts. Therefore, this study aims to investigate the impact of land use, land
cover, and climate change on the water balance components of the Hargeisa watershed. Un-
derstanding the biophysical and meteorological parameters that affect the water balance is
crucial for the sustainable development, management, and proper utilization of water re-
sources in the watershed. This study will serve as a baseline for future research on water re-
sources in the area
 XVII

1.6. Scope of the Study

The study was aimed to understand the source, process and availability of the main water bal-
ance components of a given watershed (Hargeisa watershed). Which includes the recharge,
runoff, interception and the actual evapotranspiration lost from the watershed. this was con-
ducted in Hargeisa watershed of Ogo Mountains, at an elevation of 1,334 meters above sea
level. Waqooyi Galbeed is bordered by the Gulf Adan to the north and Ethiopia to the south.
Hargeisa city is the most populace city in Somaliland, and it’s the capital city of Somaliland,
and it is located between scope 9°33′47″ N and 44°4′3″E.

1.7. Limitation

The observed climate data were highly challenged by the lack of national meteorological ser-
vices, this was overcome using open source data such as USGS, World Climate data, FAO,
and SWALIM, which was blended satellite data. The availability of data boreholes in the wa-
tershed is also very limited. In addition to this, the presence of monitored wells was a big
challenge to conduct model validation tests, which led to the destruction of the central gov-
ernment in 199. The observed data was also a big problem in getting full, reliable, representa-
tive, and consistent data.
 XVIII

2. LITERATURE REVIEW

2.1. Theoretical Literature Review

Ecosystem services are benefits humans derive from ecosystems, including provisioning, reg-
ulatory, supporting, and cultural offerings(Geng et al., 2015). Environmental services are the
basis of human survival and are closely related to human well-being(Deng et al., 2013). Ac-
cording to Geng et al. (2015), the environment is one of the important needs to manage,
which might be vital for maintaining the provision of natural services. It is necessary to con-
sider climate change as well as land use/land cover change (LUCC) in the relevant ecosystem
carrier assessments, particularly when analysing water-related environmental services.
Moreover, water supply is one of the most essential environments since adequate freshwater
supply is fundamental for ensuring the sustainability of agriculture, industry, and the natural
environment(Belete et al., 2020).

The water resource and its driving mechanism have received more and more attention, and
analysis of the impact of LUCC and climate change on the water resources are essential for
formulating adaptive management strategies(Boithias et al., 2014). Drinking water and de-
mand can vary from place to place with availability and accessibility sources of water supply,
including groundwater, rivers, lakes, springs, and small streams. At present, the world’s
people are withdrawing 30% of the runoff that is accessible; however, about 20% of the total
runoff is remote and not readily available to meet water consumption. Water demand is the
outcome of unplanned urbanization, industrialization, and rapid population growth. Water de-
mand also can exist if groundwater is tapped out without replenishment when a rate of ex-
ploitation exceeds its restoring capacity(Ali, Mushir, Terfa, 2012).

The changes in land use/cover and climate affect the groundwater supply and significantly
distress water demand in crop production because of a loss of freshwater availability(Kirby et
al., 2016). This requires hydrological modelling to understand climate changes and LULC's
influences on water balance components. Therefore, the effect of climate and LULC changes
on water resources has become a topical hydrological issue that requires investigation. Again,
the future climate variability is essential information in impact assessment to useful resource
decision-making (Nyatuame et al., 2020). Land use/cover change is motivated by several
reasons from environmental and social dimensions in a land system(Gong et al., 2015). Thus,
the land cover dynamic becomes an issue of the 21st century with dramatic implications for
human life(Elias et al., 2019).
 XIX

Different land-use type represents diverse degrees of risk to water resources, with urban and
agricultural areas being the land-use types most liable for water quality degradation globally.
Thus, managing the instant human necessities whilst retaining a long-time period of water
supply capacity is an urgent need and a great challenge to the country. LULC change is re-
lated to increased water demand; population growth and climate change are probably going to
greatly affect water resources(Mello et al., 2020).

Worldwide climate change has wide-ranging impacts on natural resources, including fresh-
water ecosystems(Twisa & Buchroithner, 2019). Around 75% of Earth’s land surface has
now been impacted through human development, which exerts an expanding footprint on wa-
ter resources. There are distinct patterns in the evolution of land-use change, and land use is a
template on which climate interacts to influence the quantity and quality of Earth’s wa-
ter(Kaushal et al., 2017). According to Kaushal et al. (2017), the global main impacts are be-
ing caused by the increasing irrigation and accompanying dam construction and groundwater,
the land use and the climate change can lead water shortage. Also, the increase of urbaniza -
tion surface cover and artificial drainage can reduce infiltration and groundwater shortage.
The tendencies in sub-Saharan Africa (SSA) show an increasing trend of converting the nat-
ural land cover into arable land. Drivers of change are manifold and may be directly connec-
ted to human activities inclusive of populace growth, economic development, and globaliza-
tion. Natural processes like floods, landslides, droughts, and climate change affect land use
and land cover change (LULCC). These conversions into arable land hurt several ecosystem
services as a trade-off for accelerated agricultural outputs such as food and timber production.
Several researchers investigated the impact of LULCC and climate change on water sources
separately or simultaneously(Näschen, Diekkrüger, Evers, et al., 2019)

2.2. Land Use/Cover Change Impact on Water Resource

In current years, substantial changes in land-use/land-cover (LU/LC) have taken place due to
human activities. LU/LC change is linked to human factors, such as agricultural loss, overex-
ploitation of forests, and urbanization has caused natural resource shortages, consisting of
widespread and permanent losses of biodiversity across the sphere, and the population growth
continues to adjust the landscape and natural lands through socio-ecological and socio-eco-
nomic phenomena at extremely high rates, causing effects of LU/LC on the environ-
ment(Twisa et al., 2020). This change could motivate several variations in services and roles,
 XX

consequently causing degradation in the provisioning of environmental services from the nat-
ural resources on the earth(Nobert & Jeremiah, 2012). As the population increases and con-
sumption styles vary, additional land will be needed for agricultural production and residing
space. The challenge facing society as an entire is figuring out how to meet individuals’
growing demands for food, living space, fuel, and other supplies while sustaining environ-
ment services in LU/LC variations(Birkhofer et al., 2015).

Since the mid-20th century, human activities have substantially altered ecosystems, and water
stress has increased because of water pollution, withdrawal, and contamination(Haddeland et
al., 2014). LU/LC changes are regarded as the dominant form of anthropogenic stress on the
environment, causing changes in environment carrier styles and affecting groundwater re-
charge(Ouchi et al., 1982). Sustainable land management plans could ensure the constant
provision of the environmental system. Studies have shown that the outcomes of LU/LC on
environment services differ temporally and spatially (Egarter Vigl et al., 2017). According to
Egatrer et al. (2017), water supply systems are susceptible because they may be exposed to
severe natural stresses related to interactions among biophysical factors, which considerably
increase their heterogeneity from a temporal and spatial perspective. Several relatively static
influences (soil, topography, and geology) and dynamic influences (land use, land manage-
ment, and climate) interact to control water access and how it will be distributed to competing
users.

Water ecosystem services for drinking are intensely affected by the quantity and quality of
water supplies to the basin and how divided between the processes of surface water runoff,
evaporation, groundwater recharge, and transpiration. Therefore, understanding the effects of
LU/LC on the water resources services for drinking is important for understanding the signi-
ficance of decisions and policies and might support the development of appropriate plans(He
& Hogue, 2012). After that, land use management and plan assessment require in-depth
knowledge of the different effects on water resources (Arunyawat & Shrestha, 2016). In addi-
tion, the nature of interactions between ecological, physical, and hydrological traits that de-
cide the effects of land cover change on surface and sub-surface hydrology is not properly
understood in natural and human-dominated environments. Changes in agricultural land uses
are central to environmental change studies because situated at the interface between water
ecosystems and society(Chemura et al., 2020).

The changes in land use and land cover (LULC) cause a big impact on water resources and
significant pressure on biodiversity and environmental services. The LULC changes can also
 XXI

result in water yield variations, such as an increase in discharge, a decrease in infiltration rate,
and an increase in rainfall intensity or variability. Moreover, it may increase water resource
shortage and food lack of confidence in Africa. (Measho et al., 2020).

According to Maitima, Mulligan (2009, 2015). The impacts can be more critical in the Horn
of Africa, which has been exposed to recurrent drought and has high climate variability. In
many parts of the Horn of Africa, LULC changes had been occurring quickly, and the change
was an increase in cropland compromised with a decrease in forest, bushlands, and grass-
lands.

In a recent review for Eastern Africa, LULC change due to deforestation, for example, turned
into evident with high effects on hydrological fluxes, mainly in discharge and surface runoff.
However, different research has stated contradicting results(Guzha et al., 2018). In addition,
groundwater, the primary drinking water supply, is deteriorating due to the increasing popula-
tion growth, urbanization, land use/land cover changes, water demand, and climate change.
The combined outcomes of those changes and natural activities, such as droughts and water
resources, particularly freshwater, are becoming inadequate(Ahmad et al., 2021).

2.3. Climate change impact on water resource

Water resources of the world in general and in Africa are under heavy stress because of the
increased impact of climate change groundwater recharge. However, the severity of the im-
pact varies from one region to another(Luis & Moncayo, n.d.2018). So, water resources are
sources of water that can be beneficial or potentially useful to humans, and It is important be-
cause it is needed for life to exist. Many uses of water include agricultural, industrial, do-
mestic, recreational, and environmental activities. Virtually all human uses need fresh water.

Climate change may be due to internal processes or external forces. Some external influ-
ences, such as changes in solar radiation and volcanism, occur naturally and contribute to the
climate system's natural variability. Other external changes, including the change in the at-
mosphere's composition that commenced with the Industrial Revolution, are the result of hu-
man activity(Gebre, 2015). Therefore, according to Gebre, (2015), climate change (changes
in frequency and intensity of extreme climate events) is likely to have main impacts on nat-
ural and human systems concerning hydrology; climate change can motivate significant influ-
ences on groundwater resources by resulting in changes in the hydrological cycle.
 XXII

Water resources evaluation is, therefore, a countries-wide responsibility that requires special
arrangement and capability toward achieving an appropriate water resources assessment.
Such evaluation is necessary since the unabated increases in water demands and their fluctu-
ation constitute issues of concern in the planning and development of water sources, most es-
pecially in the area of water production and water utilization, particularly in developing coun-
tries(Adeaga et al., 2019). In addition, the effect of climate change and variability on water
resources development is well-identified globally. It has been identified as a major issue fa-
cing the availability of water resources such as water surface, groundwater, and supplies(Ab-
bas et al., 2017). The importance of wetlands in East Africa for the provision of numerous
environmental services, ranging from the improvement of mental well-being to water and cli-
mate regulation, is well proven(Näschen, Diekkrüger, Leemhuis, et al., 2019). Some re-
searchers speculated the prospect of increases in hydrological extremes about climate change.
The variability of runoff and groundwater resources is specifically higher for drier cli-
mates(Xu et al., 2004). Water resource scholars need to deal with the effects of climate
change on hydrology regimes and water resources. Therefore, a good understanding of the re-
lationship among climate change, human activities, and water resources, as well as their with-
drawal and use, will enable water resource managers to make rational decisions and water al-
locations.

There is little doubt that the higher annual precipitation variability has already led to con-
sequences in Somaliland. The former more or less regular rainfalls during the rainy season
have become less predictable, which results in more frequent drought years and a severe ef-
fect on the population, especially those relying on surface water sources (e.g. nomads), even
if the average precipitation has not changed. In many cases as soon as reliable surface water
resources are no longer dependable(African Development Bank Group, 2016). Therefore, the
Somaliland water resource has different situations on climate impact, for groundwater is less
directly and slowly affected by climate change when compared to surface water. But only
when the droughts or decreased rain will the groundwater decline. In general, the Somaliland
water resource has no river flows and lakes, but the water demand in the country (Somalia)
depends on groundwater, from which the majority of the population gets their water supplies,
as shown in Figure 1.
 XXIII

Figure 1: SWALIM (Somali Water and Land Information Management).

According to the African Development Bank, the climate in Somaliland is arid to semi-arid,
and the predicted increase of average precipitation by 10% might, in the long term, have pos-
itive effects on groundwater levels and availability. The observed declining groundwater
level in some wells in Somaliland is most likely related to groundwater pumping exceeding
the groundwater recharge rate and not to a climate-related decrease in groundwater recharge.

2.4 Impact of land use/cover and climate change on water resource

Persistent global climate and land use changes have severely impacted environmental sys-
tems and their abilities, leading to the degradation of environmental services(Pan et al.,
2015). Climate change affects groundwater recharge, habitats for biodiversity, and other eco-
system services(Yang et al., 2019). Land-use changes modify underlying ground surface situ-
ations, residences, and environment types, further altering the environment's structure and
functioning(Mamat et al., 2018). Understanding the relative impacts of land use and climate
change on groundwater recharge and water resources management is essential to develop ef-
fective climate change adaptation policies and optimal land use management(Brauman,
2015). Land use and climate changes can impact hydrological processes and attributes, pos-
ing considerable challenges for water resource management(Collet et al., 2015).
 XXIV

Climate change is causing more frequent droughts and floods, and increasing demands for
water due to urbanization, agriculture, and tourism(Sisto et al., 2016). Assessing water re-
source vulnerability and risk, estimating water shortages, and analysing droughts are crucial
for avoiding water crises(Hishe et al., 2020). Climate change research, including its impact
on water resources and land use changes, is ongoing. Land use changes affect the hydrolo-
gical system of watersheds, especially in arid and semi-arid regions, and can impair water
quality(Choukri et al., 2020).

2.5 Rainfall-Runoff-Recharge Relationship

Understanding rainfall-runoff relationship is critical in order to quantify ground water re-

charge. Because the source of most natural ground water recharge is precipitation (Neven,
2009). the principal factors to the rainfall-runoff relationship is the land cover, the soil type
and the antecedent moisture condition. Soils that are well drained generally have high effect-
ive porosities and high hydraulic conductivities, whereas soils that are poorly drained have
higher total porosities and lower hydraulic conductivities (Chemura et al., 2020), Wet and
poorly drained soils will readily produce runoff, while dry, well-drained soils will readily
contribute to groundwater recharge. Land use landcover also affect the rainfall runoff rela-
tionship. after subtracting what is lost to overland flow (runoff) and evapotranspiration (ET)
from rainfall the amount of ground water recharge will be known.

2.6 Groundwater Recharge

Groundwater recharge as water that percolates to aquifer from any direction, i.e., down from
surface through zone of aeration, up from confined aquifers, or laterally (Dereje & Nedaw,
2019). Recharge is the actual entering of water into an aquifer. Usually, the recharge source is
precipitation and to a lesser degree, irrigation or artificially constructed recharge pit or losing
streams. recharge from precipitation is the most important category of recharge (Lerner D.,
1990). focus will be given to it in the study area. The actual recharge rate is controlled by
several factors: precipitation intensity, soil moisture content, elevation of the area, the hori-
zontal hydraulic conductivity of the aquifer and its hydraulic gradient and the vertical hy-
draulic conductivity of the soil (Halder et al., 2016). The identification of recharge areas is
becoming increasingly important because of the expanding use of the land surface for waste
disposal. Recharge rates vary from year to year, depending on land cover types of the area,
 XXV

the intensity of precipitation and its seasonal distribution, air temperature, and other factors.
Among the other factors are land use. For example, recharge rates are much higher in highly
vegetated area than in impermeable cities and asphalt roads(Armanuos et al., 2016).

2.7 Remote Sensing and GIS for groundwater recharge estimation

GIS and remote sensing are the latest, time and cost-effective technology for groundwater ex-
ploration by acquiring full information and access all parameters of factors which controls
groundwater potentials and recharge zone areas by using different software’s easily(Wang et
al., 2015). GIS can be utilized to evaluate and delineate patterns of recharge within a region
by incorporating available information about observed temporal fluctuations of recharge at
specified locations. Continuous remote acquisition of water table depths combined with GIS
capabilities for distributing the flux spatially across the study area can result in near-real-time
monitoring of spatial variability in recharge flux. Recharge can also be estimated using in -
formation including current climate, soil, and vegetation/land use patterns and estimates of
recharge for various soil-vegetation combinations(Singh et al., 2010).

2.8 Estimation of recharge, runoff and actual evapotranspiration

2.8.1. Estimation of groundwater recharge

Groundwater recharge is difficult to estimate directly, because recharge is a result of various

parameters and highly variable depending on different factors of the area (Neven, 2009).
Therefore, it should be applying another indirect method. Recharge can be expressed as a per-
centage of the annual rainfall or as an average rate of water in millimetres per year. The
volume of recharge, expressed as cubic meters per year, can be obtained by multiplying the
recharge rate by the land area under consideration (Kahsay et al., 2018).
There are different methods to estimate ground water recharge depending on actual physical
processes of the recharge. The three hydrological zones of study are surface water, unsatur-
ated zone and saturated zones, Physical, numerical modelling, and tracer techniques are
among the methods. Physical methods, e.g., channel-water budget, seepage meters and base
flow discharge; tracer methods, e.g., stable isotopes of oxygen and hydrogen; numerical mod-
elling methods, e.g., Groundwater flow numerical models are included in methods based on
groundwater studies(Teklebirhan et al., 2012). To select an appropriate method, the possible
 XXVI

flow mechanisms and the important features influencing the recharge in an area should be
identified. This is because the recharge processes vary from one place to another, this is true
for methods working for one area may be invalid for other.

2.8.2 Estimation of actual evapotranspiration and runoff

Evapotranspiration (ET)is a sum of evaporation from open water bodies and bare soil, inter-
ception loss by vegetation and transpiration through plants (Dereje & Nedaw, 2019). atmo-
spheric humidity, net radiation, wind speed and temperature of the area are the governing
factors (Gautam et al., 2018). It is important for hydrological basin in order to control
groundwater recharge, overland flow and soil moisture. It is expressed by potential evapo-
transpiration (PET) which is the possible water to evaporate from the watershed through plant
and directly from open water bodies as well as bare soil from land surface, and actual evapo-
transpiration (AET) which is the actual water that is evaporate to the atmosphere. The runoff
coefficient which is a function of soil texture, slope and vegetation type is inputs for compet-
ing actual surface runoff (Mustafa & Ali, 2013). It is an idea of how much water from precip-
itation will flow on land surface through rivers and streams.

2.9 Hydrological models

Hydrological models are simplified systems to compute the processes of the hydrological
cycle (Precipitation, Infiltration, Interception, Evaporation, Runoff, etc.) in an entire river
basin based on a set of interrelated equations that try to convert the physical laws, which gov-
erns the complex natural phenomena (Barbara, 2008).
Based on the description of physical process, the hydrological models can be classified into
three groups. conceptual models, physically based distributed models and empirical models
(Beven, 2001). physically based distributed models (sometimes called white-box models)
presumably are the consequences of the most important laws governing the phenomena. The
model has a logical structure similar to the real-world system and may be helpful under
changed circumstances. It is designed by the principles of physical processes based on con-
tinuity and the conservation of mass, energy and momentum. This model uses a number of
model parameters that represents the different heterogeneities in the catchment. e.g. infiltra-
tion model, evaporation model, SHE models. In the recent time, advances have been made in
 XXVII

hydrological process understanding and modeling using topography driven, conceptual, flex-
ible, semi-distributed model structures (Gharari et al., 2013).

Regional groundwater models used for identifying and analysing groundwater system (infilt-
ration-percolation-discharge relations) are often quasi-steady state and therefore it needs a
long-term average input data. Thus, Wetspass, which yields spatially varying groundwater re-
charge using spatially varying soil, land-use, and meteorological inputs, can be used for the
purpose of understanding the characteristics of groundwater recharge (Batelaan and
Woldeamlak, 2003). Empirical models (Black box) which mathematical correlations are ob-
tained based on the observed data analysis rather than based on the physical processes in the
catchment e.g. Unit hydrograph-based model, Linear regression model. Conceptual models
(sometimes called grey-box models) are intermediate between theoretical and empirical mod-
els which is consider physical laws in highly simplified form (Dandy & Maier, 2005).

2.10 The Wetspass Model

Wetspass was built as a physically distributed based methodology for estimation of the long-
term average, spatiotemporally varying, water balance components: groundwater recharge,
surface runoff, and actual evapotranspiration. It is an acronym for water and energy transfer
between soil, plants and atmosphere under quasi-steady state that was built upon the founda-
tions of the time dependent spatially distributed water balance model (Batelaan & DeSmedt,
2001). The Wetspass model estimates seasonal and annual long-term spatial distribution
amounts of groundwater recharge by subtracting the seasonal and annual surface runoff and
evapotranspiration from the seasonal and annual precipitation respectively (Arefaine et al.,
2012).
 XXVIII

The model computes different water balance components, surface runoff, actual evapotran-
spiration and groundwater recharge based on distributed data. WetSpass was successfully ap-
plied in Belgium (Batelaan and De Smedt, 2001), Geba watershed, Ethiopia (Tesfamichael, et
al., 2010) and in Werii watershed in Ethiopia. Based on those authors groundwater recharge
was successfully simulated which is the main interest of this research. In Ethiopia, several au-
thors were used WetSpass model for their groundwater recharge estimation studies.

Figure 2. schematic water balance of hypothetical raster cell (Batelaan & DeSmedt, 2001)
 XXIX

3. MATERIALS AND METHODS

3.1. Description of the Study area

The study area (Hargeisa City), as indicated in figure 2 is located in a steep zone according to
Somaliland. It’s the semi-arid region and its above sea level 1,334 meters, Geographically
(ogo), it is located latitude 9°33′47″ N and 44°4′3″E, and longitude 9°33′0″ N and
44°4′0.0012″E. Also, its largest capital is the city of Somaliland, which is 128km² in the re-
gion horn of Africa. other hand is the second capital of my country after the capital
Mogadishu. Hargeisa is situated in a valley in the Galgodon (Ogo) highland. The population
of Hargeisa city is estimated at 1.2 million (MNPD, 2011).

3.1.1. Socio-economic of the Study Area

The population of Somaliland in 2009 was estimated at 3.85 million with 1.2 million in
Hargeisa City, according to the Ministry of National Planning and Development Somaliland.
About 55% of the people are nomadic pastoralists, and 45% are urban and rural dwell-
ers. Somaliland is predominantly a nomadic pastoral community, and historically, the ma-
jor livelihood is livestock husbandry, which is the main source of food and income for 70%
of the population, including rural pastoral and urban communities.

Crop production ranks second to livestock. There are three main sources of revenue for
Hargeisa which are livestock, remittance from the Diasporas, and agriculture. The livestock is
the backbone of the Somaliland economy, and the diaspora investment plays the city's main
role in the development of the City. According to ADBG (2016), agricultural production is
less than 10% of the whole of Somaliland which is located northwest of the region. We can
realize the influence of agriculture in the area. Lack of access due to poor infrastructure, es-
pecially roads, is inhibiting socio-economic development in the region. Access
to health and education in Somaliland is generally limited. Somaliland is facing rapid urban-
ization, creating another vulnerable group, the urban poor. This group includes destitute pas-
toralists, economic migrants, and people who cannot make ends meet through loss of liveli-
hood opportunities. Private sector business remains relatively limited, so there are few em-
ployment opportunities.
 XXX

3.1.2. Water supply situation of the area

HWA is a state-owned enterprise that currently provides water services to over

200,000 people in Hargeisa, the capital of Somaliland. The utility operates a water network of
21,000 household connections, 400 water kiosks, and three water tanker standpipes. Des-
pite limited infrastructure and storage capacity, the utility presently extends services to 30%
of the city’s population. The majority of this connection area is concentrated in the northern
half of the city, with households in the south depending on alternative water sources intro-
duced by tankers, which are regularly 5-6 times as costly. HWA draws all of its water from
12 boreholes in Geed Deeble, located 20km from the city(FACILITY, 2019). The aquifers in
this area have been a reliable water source for the city, with yields remaining stable even dur-
ing the current drought period in 2016/17. Water from the boreholes is pumped to a reservoir
at the Geed Deeble Pumping Station, where four booster pumps push the water up a 300m
gradient to the “Chinese Reservoir” Hargeisa main reservoir through a booster station at
Biyo-Khadir. The borehole pumps are operated for 23 hours daily and extract an average of
50m3 of water per hour.

Total water production is not currently metered but is estimated at 13,500m3/day (HWA).
This is higher than HWA figures, based on an estimated volume of water distributed from
the Chinese Reservoir(FACILITY, 2018). According to HWA (2019), the current transmis-
sion pipeline was built under a project financed by China in the 1970s and consists of parallel
300mm pipes that supply up to 9,000m3 of water to the Chinese Reservoir daily. The ongo-
ing donor-financed Hargeisa Urban Water Supply Upgrade Project (HUWSUP) is building a
new 600mm pipeline to increase capacity to 25,000m3/day. The infrastructure works involve
rehabilitating the wellfield collector network, building the new Geed Deeble pumping station
(NGDPS), and increasing water production by rehabilitating existing boreholes and drilling
new boreholes in Geed Deeble, Las Dur, and Haro-Haadley.

3.1.3. Agriculture activities

Somaliland’s agriculture area contributes to 10% of the country’s livelihoods and is

dominated by subsistence farmers, who practice rain-fed farming and de-
velop mainly sorghum and maize. Dryland farming accounts for almost 90% of all agricul-
tural activities. Irrigated agriculture debts for simplest 10% (Master plan for Somaliland agri-
culture sector(MNPD, 2011). The most common plants grown in Somaliland are maize,
 XXXI

sorghum, millet, and cowpea. Typically, local varieties, including Elmijama, the late-matur-
ing sorghum variety, are grown. Intercropping of different plant life in
an identical field is common. Watermelons and vegetables are also cultivated in some de-
veloped watersheds with the use of rainfall-runoff harvested in reservoirs along seasonal
stream flow banks. The particularly high rainfall areas of the plateau (Awdal and Waqooyi
Galbeed (Gabiley) regions as well as Odweine), which are suitable for sorghum and
maize production, constitute the food basket of Somaliland. Rain-fed agriculture is prac-
ticed around Gabiley, Borama, and Hargeisa. Low inputs characterize it, that’s a tradi-
tional low-risk practice for staple food production under erratic rainfall conditions. The
USGS land use classes did not register the agricultural Area in Hargeisa land cover, but FAO
land uses there are certain rainfed areas in Odwayne and Arabsia.

3.1.4. Land use land cover

The main land use of the study area is extensive agro-pastoralism and urbanization. The
land cover of the areas consisted of savannah and general open shrubs; there also are appro-
priate land cover types: Irrigated Agriculture (timber and cash crops), General Open Trees,
and Shrubs. The land is mainly used for livestock production and rearing or mix farming
(crop and livestock production). Most of the regions in Somaliland are dry and is not sup-
port rain-fed agriculture except for small pockets of land in the areas around Hargeisa, Gebi-
ley, and Borama, which receive amounts of rainfall that can support rainfall-dependent agri-
culture (MOA).

3.1.5. Climate

Based onthe climate data collected from representative weather stations Hargeisa, Malowle,
and Aburin, the study area is placed in a semi-arid zone. The diverge of most temperat-
ures across the country ranges from 30°C to 40°C, except at the highest elevations
and along the sea coast. The average day-by-day minimum temperatures vary from 20 ° C to
over 30 ° C. The temperatures, however, increase from the highlands towards the sea.

In Hargeisa, the maximum and the minimum temperatures are 32° C and 19° C, while the an-
nual temperature varies between 17 to 24° C. Most of the country holds under 500 mm of
rain annually, and a large area that includes the northeast and distant Somaliland receives
only 50 to 150 mm. However, some higher elevation areas in Somaliland, such as coastal
areas, record more than 500 mm per year. In general, rain occurs in showers or localized tor-
 XXXII

rential rains and is very changeable. The annual rainfall is higher in the Maroodijex catch-
ments, with Aburin and Hargeisa stations recording over 260 mm and 400 mm of annual rain-
fall, respectively (SWALIM).

Figure 3: Study area map

 XXXIII

3.2. Data collection

To conduct this study the following data were necessary; climatological data which
includes monthly precipitation, potential evapotranspiration, minimum and maximum tem-
peratures, sunshine hours, groundwater depth, and wind speed data.

Physical data of the area that including land use land cover, soil textural type, and slope of
the study area were collected and used as input for the model to compute groundwater re-
source of the study area.

3.2.1. Meteorological Data

The hydro-meteorological parameters were prepared based on the seasonal climatic condi-
tions (Dirac, Gu’, Xagaa, and Dayr seasons) identified in the region. The parameters that
were used in this study include rainfall, temperature (annual, maximum, and minimum), po-
tential evapotranspiration (PET), wind speed, and groundwater level were obtained from the
Hargeisa Water Agency (HWA).
The rainfall data was obtained from the Ministry of Agriculture Somaliland Department Me-
teorological from 2005 to 2021; there are no other data available in the meteorology, were
used open-source data, which is Global climate and weather data
(https://worldclim.org/data/index.html), which are in a grid format, and the future climate
project was use the World Bank climate change knowledge portal
(https://climateknowledgeportal.worldbank.org/). Furthermore, potential evapotranspiration
(PET) data is used (FAO, 2011). Then, the seasonal potential evapotranspiration was inter-
polated to obtain distributed maps of seasonal PET.

3.2.2. Physical Data

Biophysical data, which are soil texture and groundwater data, are essential to understand-
ing the groundwater recharge aspects. The soil map was obtained from FAO, and the ground-
water data were collected directly accessible existing Geed Deeble boreholes in Hargeisa,
which are supervised Hargeisa Water Agency (HWA). The study area's land use/
cover maps were produced from cloud-free Landsat 8TM Multispectral images and Sentinel
2 multispectral images from 2001 and 2020 https://www.usgs.gov/. Land use analysis for
spatial analysis tool in GIS, considers only USGS-NASA classes in defining potential runoff
coefficient and depression storage capacity, i.e., irrigation area, short grass, forest, bare soil,
and urban areas. These datasets were prepared, stored, manipulated, analysed, and conver-
 XXXIV

ted from raster format to ASCII grid cell format using conversion ArcGIS tools. Thirdly, the
groundwater recharge of the study area was quantified using the Wetspass model.

3.2.3. Satellite Imaginary Data

The study area was highly dependent on remote sensing data, which include land use, topo-
graphic elevation, soil, and slope. The soil data was obtained FAO, and land use, slope, and
topographic elevation were obtained from https://www.usgs.gov/ DEM and processes Arc-
GIS.

3.3. Methodology

3.3.1. Groundwater Recharge Estimation Model

The Wetspass model was used to estimate the monthly annual average temporal and spatial
difference of surface runoff, actual evapotranspiration, and groundwater recharge on a sea-
sonal and annual basis for the Hargeisa watershed. The major step to conducting this study
area was the availability of the data, which acquired possible information about the study
area. To accumulate the facts that encompassed the data mentioned above, input data for the
Wetspass model were generated. The ArcGIS software is a promising tool for preparing, ma-
nipulating, and analysing the digital.
Generally, running the Wetspass model requires an input file format. Grid map (AS-
CII file format) and parameter tables (TBL file format). The land use, soil, and number of
rain days parameters tables are required for the model. The arc GIS version 10.4 spatial ana-
lysis extension was used to interpolate to prepare these input parameters, and parameter
tables were added as a map in the attribute. Land use land cover, topography, slope, soil, po-
tential evapotranspiration, and groundwater level (GW depth) grid maps have been prepared
(Yirgalem Enedeg. A, 2020).
Hence, the topographic/elevation grid and slope maps were derived from the SRTM DEM.
Next, all the raster data were converted into ASCII format using the Wetspass model to cal-
culate the annual monthly groundwater recharge, annual runoff, and actual evapotranspiration
map of the study area. The sensitivity analysis for all input parameters as default was per-
formed to identify which input parameters are sensitive to water balance components of the
watershed. This was done through ArcGIS 10.4 and the Wetspass model. The annual average
raster map for each parameter were prepared by increasing the value of a single parameter
 XXXV

with different percentage. The model was processed to determine the sensitivity of a single
input variable by keeping constant the value of other parameters. In the same way, the pro-
cess was frequently done for other parameters. The detailed method of computation used to
obtain the value of the water balance components of the watershed is discussed below.

3.3.2. Groundwater Recharge Zone

The recharge zones were identified from the recharge output maps obtained from the Wets-
pass model. From the recharge output maps, the recharge estimation showed the places with
categories with high, medium, and low recharge zones during the computation analysis of the
recharge zone.

3.3.3. Climate Data

The climate data of this study area were used for the monthly seasonal rainfall and temperat-
ure data from RCMS-based CMIP5. This climate data was historical from (1986 to 2005) and
the future climate projected from (2020 to 2100) statistically downscaled regional climate
model (RCMs) bias-corrected coordinated regional climate downscaling experiments
(CMIP5) with a model of MIROC-ESM-CHEM which obtained (
https://climateknowledgeportal.worldbank.org/country/ethiopia/cmip5 ), at the resolution 1.0º
x 1.0º, Precipitation and max/min temperature for Somalia under RCP 4.5 and RCP 8.5 emis-
sion scenarios were used with historical data. The bias correction Has been approved for the
simulation of precipitation and temperature in RCMs output for any climate impact modeling
(Access et al., 2016; Endris et al., 2013; Onyutha, 2020).

The Standard season were adopted to determine the climate model used for the projected fu-
ture climate change scenarios for W/Galbeed region Somalia for the standard seasons. Still,
there are two standard rainfall seasons Gu (April to June) and Dayr (September to November)
whole country. Still, I was looked the four standard seasons because Hargeisa receives rain-
fall 75% hole the year which between 10mm up to 300mm annually. The temperature simula-
tions covered all of Somalia's typical seasons, namely from December to February (DJF),
March through May (MAM), July through August (JJA), and September through November
(SON) all the year. Regression analysis, correlation, and different time series methods are
some of the techniques that were used. The correlation method was investigating the relation-
ship between observed and simulated data. The variance of the observed data that the anticip-
ated variable was account for is represented by the coefficient of determination (r^2), which
 XXXVI

is the square of the correlation coefficient. The statistical significance of suitable linear re-
gression trends was be examined using graphical approaches and the common statistical test,
which includes the use of Mann-Kendall Rank statistics, in the regression method, in an at-
tempt to fit the best linear mathematical equation for the observed relationship. Similar re -
search has been applied or implemented in Africa and other regions that focused on the con-
nection between climate change and water supply; I'll concentrate on two variables (rainfall
and temperature).

3.3.4. Land Use and Soil Data

The land use and soil type information are important for Wetspass analysis of the study area
using remote sensing and ArcGIS techniques with the option available in GIS software. Land
use analysis considers only USGS-NASA classes in defining potential runoff coefficient and
depression storage capacity, i.e., irrigation area, short grass, forest, bare soil, and urban areas.
For simulation purposes, the percentage of bare soil and impervious areas are estimated for
each grid cell based on the land use map and classified into dry and wet season land use. Soil
types of the catchment are obtained from the soil information furnished by soil maps. The soil
code system used in Wetspass Extension is based on the soil texture triangle developed by the
United States Department of Agriculture (USDA), which is characterized by its percentage of
clay, silt, and sand, ranging from the fine textures (clay), through the intermediate textures
(loam); and the coarser textures (sand).
Therefore, the original soil coverage map was converted to a raster map with Wetspass soil
codes in the data preparation phase. The grid must be adjusted to the same grid structure as
the DEM and limited to the same area by using the mask grid of the catchment. The reclassi-
fication can be done within the GIS framework, which uses a reclassification table prepared
in Arc Info GIS Spatial Analyst.
 XXXVII

3.3.5. Methodology Framework

Figure 4: Flow Chart diagram

The above diagram is a methodological framework, which indicates the whole processes,
models, and data that were used and designed to be understood easily by readers.
 XXXVIII

3.4. Wetspass Model Description

For this study, the Wetspass model written in Python (Wetspass-M) is applied. The Wetspass
(an acronym for Water and Energy Transfer in Soil, Plants, and Atmosphere under quasi
Steady State), first has been developed by (Batelaan, 2001), and later modified by (Batelaan
& De Smedt, 2007). Its numerical model simulates the average spatial distribution of hydro-
logical parameters and processes at the basin scale in a quasi-steady state, which means the
model is restricted to temporal variation for only the seasonal or monthly time scale. The time
series data are averaged into single seasons/months.

The Wetspass model was preferable for my study, where the hydrological data were limited.
In addition, the model has been applied in different countries for annual and seasonal ground-
water resources for spatial variation estimation, and the result has a strong relationship with
the traditional hydrological model result simulating the same area(Warku et al., 2022).

Wetspass has several uncertainties, even though its results were valid for groundwater re-
sources management plans and policy formulations(Meresa & Taye, 2019). The spatial pat-
terns of the water balance components are obtained by summing up the individual water bal-
ance raster cells equation, which, according to Batelaan & de Smetd.

ET c =av ET v + as Es + as Es + ai E i……………………………………….1

Sc =av S v +a s S s +ao So + ai S i………………………………………….2

Rc =av Rv +a s R s +a o R o+ ai R i…………………………………………3

Where ETc, Sc and Rc are the total evapotranspiration, surface run-off, and groundwater re-
charge of a raster cell, respectively. The indices s, o and i stand for the bare area, open water
area and impervious surfaces, respectively. While Ss, So and Si; Es, Eo and Ei; and Rv, Rs,
Ro and Ri are the surface run-offs, evaporations and groundwater recharge in the bare area,
open water area and impervious surface.

The groundwater recharge Rv was calculated.

R v =p−S v + ET v ……………………………………………………………….4

Where P is the average seasonal precipitation.

ETv is the actual evapotranspiration [LT−1] (per litre).

Sv is the surface runoff over the land surface beneath the vegetation, and I is the interception
by vegetation. All variables have the unit of LT−1.
 XXXIX

3.5. Model Calibration

Due to insufficient data on observed groundwater data in Hargeisa watershed, the empirical
equations were selected to be a representative of observed groundwater recharge in
the study area, such as Chaturvedi Formula (CF), Sehgal Formula (SF), Krishna Rao Formula
(KRF), and Bhattacharya Formula (BF). These empirical relationships between groundwater
recharge and rainfall data have been developed (Teerapunyapong et al., 2020). There-
fore, these empirical equations were used for model calibration and validation compared with
the model results of Wetspass (Teerapunyapong et al., 2020). The empirical data has been ap-
plied to estimate groundwater recharge worldwide. The detailed equation is below.

1= Chaturvedi formula (CF).

R = 2(P-15) ^0.4…………………………………………………….…………5

where R is groundwater recharge (cm/yr) and P is annual rainfall (in)

2= Sehgal formula (SF).

R = 12.6(P -406.4) ^0.5……………………………………………………….…6

Where, R is groundwater recharge (mm/yr) and P is annual rainfall (mm)

3= Krishna Rao formula.

R = K(P-X) ……………………………………………………………………….7

Where, R is groundwater recharge (mm/yr), P is annual rainfall (mm), and values of K and X
depend on values of P as shown below;

If P between 400–600 mm; 𝑅 = 0.20(𝑃 − 400) ……………………………….8

If P between 600–1,000 mm; = 0.25(𝑃 − 400) ……………………………….9

If P between 1,000–2,000 mm; = 0.30(𝑃 − 500) ……………………………….10

If P more than 2,000 mm; 𝑅 = 0.35(𝑃 − 600) ………………………………….11

4= Bhattacharya Formula (BF)

𝑅 = 3.47 (𝑃 − 38) ^0.4……………………………………………………………12

Where, R is groundwater recharge (cm/yr) and P is annual rainfall (cm).

The observed groundwater level can be calibrated parameter instead of found groundwater re-
charge. The incomplete observation, the groundwater level was obtained insufficiently.
 XL

Therefore, observed groundwater levels obtained from HWA were completely unreliable to
the calibrated parameters for this study. The statistics parameters can be applied to evaluate
the model performance by comparing the observed and simulated values (Singh et al., 2010).
The Empirical recharges are used as observed values, while the Wetspass recharges are signi-
fied as simulated recharges. Wetspass model performance is evaluated by statistical paramet-
ers namely; Root Mean Square Error (RMSE), and coefficient of determination (r2). De-
tails of the parameter are defined below.

1= Root Mean Square Error (RMSE).

n
RMSE = √ ∑ (¿ Pi−Oi)/2 ¿
i=0

……………………………………………………………………….13

Where O is the observed value and P is the simulated value.

Root mean square error (RMSE) is used in the comparison and evaluation of simulation mod-
els. RMSE measures the differences between observed and simulated values (Willmott). The
range of RMSE is not determined directly by specific criteria.

2= Coefficient of determination (r^2).

……………………………………………….14

Where, O is the observed value and P is the simulated value The range of coefficient of de -
termination (r^2) lies from 0 to 1, which explains the dispersion between observed and simu-
lated values. The zero value of r^2 describes no correlation and the 1 value of r^2 describes
dispersion of observed values and simulated values which are definitely equal (Krause).

3.6. Wetspass Model input data

The Wetspass model needs different types of long-term average input data. Which are hydro-
meteorological data (precipitation, wind speed, temperature, potential evapotranspiration,
groundwater elevation) and spatial patterns of watershed-based biophysical layer data (slope,
land use/land cover, elevation, and soil). The magnitude and even the change in direction re-
charge depend on the local soil, vegetation, and climatic region (Craig et al., 2010). Wetspass
 XLI

needs the input variables every month, as a result, four seasons (Xagaa, Dayr, Diraac, and the
Gu) which are based on winter, spring, summer, and autumn

Grid maps and parameter tables are required as inputs for the model simulation and were pre-
pared with the help of ArcGIS 10.4 tools. The grid maps were a land-use land cover, soil tex-
ture, topographic elevation, slope, groundwater levels, precipitation, potential evapotranspira-
tion and wind speed, and temperature. The input files prepared as parameter tables were also
prepared in a database file format (tbl): land use, soil texture, and number of rainy days.

Table 1: Input variables for the Wetspass model

No Input variables source resolu- Size of the data
maps/table tion (cx*cy)

1 Topography https://www.usgs.gov/. 30*30 4740*5599

(DEM)

2 Slope https://www.usgs.gov/. 30*30 4740*5599

3 Land use https://www.usgs.gov/. 30*30 4740*5599

(2001-2020)

4 Precipitation https://worldclim.org/data/index.htm 30*30 4740*5599

5 Temperature https://worldclim.org/data/index.htm 30*30 4740*5599

6 Wind speed https://worldclim.org/data/index.htm 30*30 4740*5599

7 Groundwater Hargeisa water agency 30*30 4740*5599

depth

8 Soil data FAO DATA CATALOG 30*30 4740*5599

9 Potential evap- https://worldclim.org/data/index.htm 30*30 4740*5599

oration

10 No of rain days WETSPASS USER GUIDE

look-up table

11 Soil parameter WETSPASS USER GUIDE

look-up table

12 Land use pa- WETSPASS USER GUIDE

rameter look-
up table
 XLII

The relationship between rainfall and groundwater recharge depends on several meteorolo-
gical and biophysical factors, including wind speed, temperature, slope, land use, and soil
types of the area.

Rainfall is taken as the starting point for the computation of the water balance of each of the
components of a raster cell. There are four distinct hydrological seasons in the watershed.
Those are the rainy seasons Spring (MAM), summer (JJA), autumn (SON), and winter (DJF)
which is the arid season. The monthly rainfall distribution in and around the watershed is
shown in the Figure below.

Figure 5: monthly average rainfall for stations in the study area

The Hargeisa watershed received an average rainfall in the seasonally are Diraac (winter) of
16mm and 10mmmm, Gu (spring) of 186mm and 81mm, Xagaa (summer) of 220mm, 79mm,
Dayr (autumn) of 120mm, 71mm, with maximum and minimum respectively.
 XLIII

Figure 6: monthly average rainfall for stations in the study area

 XLIV

3.9. Temperature

Temperature is the main controlling factor of evapotranspiration in a certain watershed. In the

study area, the temperature data average value ranges in the four seasonal temperatures are
Diraac/ winter (DJF) 27.5C˚, 24.1C˚, Gu/spring (MAM) 31C˚, 28.5C˚, Xagaa/summer (JJA)
33.1C˚, 28.7C˚, and Dayr/autumn (SON) 30.3C˚, 28.7C˚, with maximum and minimum re-
spectively.

Monthly average temperature

40
35
30
Temperature C֯

25
20
15
10
5
0
feb mar apr may jun jul aug sep oct nov dec DJF
Monthly

bulohar geed deeble cadadley hargeisa

cada gumburaha gacanlibah salahley

Figure 7: Monthly average temperature stations in the study area

 XLV

Figure 8: Seasonal average temperature map

 XLVI

3.10. actual evapotranspiration (AET)

Actual evapotranspiration describes for the water losses that will occur under a given climatic
condition with no deficiency of water for vegetation. Since the actual evapotranspiration ac-
counts for the field condition, it depends on water availability.

Actual evapotranspiration (AET) in the Hargeisa watershed was obtained from world climate
data in 30 years. The monthly AET results were divided into four hydrological seasons in the
study area which are Diraac, Gu, Xagaa, and Dayr based on winter, spring, summer, and au-
tumn to get the distribution of AET of the watershed using inverse distance weighted (IDW)
spatial interpolation technique were used for both hydrological seasons to get the distribution
value over an area. After interpolation, the raster map was converted into ASCII format and
the resulting grid map was incorporated with other input parameters of the Wetspass model to
calculate recharge and actual evapotranspiration.

Monthly Actual evapotranspiration

80
70
60
50
PET (mm)

40
30
20
10
0
jan feb mar apr may jun jul aug sep oct nov dec
month
Figure 9: ac-
tual evapo-
bulohar geed deeble cadadley hargeisa transpiration
cada gumburaha gacanlibah salahley
values sta-
tions of the study area

The seasonal actual evapotranspiration of the watershed is winter (DJF) 138mm, 123mm,
spring (MAM) 159mm, 139mm, summer (JJA) 207mm, 158mm, and the autumn (SON)
159mm, 138mm with maximum and minimum respectively.
 XLVII

Figure 10: seasonal potential

evapotranspiration map of the watershed

3.11. Wind speed

Wind speed governs the rate and amount of

evapotranspiration and water balance compon-
ents of a given watershed. With the same pro-
cess as other Wetspass input parameters, the
wind speed distribution over the area was ob-
tained using the IDW spatial interpolation
method. The average speed of wind in the
Hargeisa watershed during winter (DJF) is 3.34 m/s,2.8m/s, spring (MAM) 3.8m/s, 2.77m/s,
 XLVIII

summer (JJA) 6.03m/s, 2.97m/s, and autumn (SON) 4.23m/s, 2.03m/s with maximum and
minimum respectively.

Figure 11: Monthly average wind speed values stations in the Hargeisa watershed
 XLIX

Figure 12: Seasonal average wind speed map of the Hargeisa watershed (worldclim. Org)

3.12. Topography

The topographic map of the study area is processed from the Shuttle Radar Topographic Mis-
sion (SRTM) data set. The SRTM provides a 30*30m resolution Digital Elevation Model
(DEM). The highest point in the watershed is located on the top of Cada/upstream with an el -
evation of 1725m and the lowest point 187m is located in the Bulahar/downstream part of the
area.

Figure 13: The topographic elevation map of the Hargeisa watershed coverage area
 L

3.13. Slope

The slope is an important parameter to identify the hydrological characteristics of the water-
shed. The steep slope part of the watershed is a recharge area where a gentle slope is con-
sidered to discharge areas(Chemura et al., 2017). Most of the time it has a direct relation with
topography. The slope map of the study area is also obtained from USGS 30m*30m DEM us-
ing the ArcGIS 10.4 slope spatial analyst tool. It is classified by degree of steepness ranging
from 0 to 70. The value 0 indicates gentle/lowland and 70 represents steep/escarpment. The
Hargeisa watershed contains both a wide area of high slope/steep, while not suitable for agri-
cultural activities, and a flat/ gentle slope of lowland area/plane which is suitable for agricul-
tural activities—the areas below 10 degrees of slope which is highly suitable for recharge/dis-
charge hydrological process(Kahsay et al., 2018).
 LI

Figure 14: Slope map of the study area

3.15. Soil Data

The second most important and sensitive parameter for the recharge estimation process using
Wetspass modeling is soil texture. The soil type map of the Hargeisa watershed was found
from FAO. The soil texture of the study area is classified into three classes using the United
States Department of Agriculture's (USDA) textural classification methods: sandy loam,
loam, and sandy clay (Figure 16). Most of the area is covered by sandy loam and loam. The
Wetspass model has been prepared as the attribute table for recognizing soil and other bio-
physical maps. The texture class obtained from USDA was prepared concerning the model
code. Finally, the raster grid map was converted to ASCII format for suitability of model pro-
cessing with 30m*30m resolution.
 LII

Figure 15: Soil map of the study area (Hargeisa

3.16. groundwater depth

The groundwater level map is a critical parameter for estimating groundwater recharge using
the Wetspass model. The groundwater depth in the Hargeisa watershed is variable. 14 bore-
hole data were collected from Hargeisa Water Agency (HWA) which is local. The groundwa-
ter depth was calculated by deducting static water level from topographic elevation and the
map was produced. These data were interpolated using the ArcGIS IDW method spatial ana-
lyst tool to represent the total area of the watershed. Most of the borehole data shows that the
static water level is shallow and artesian wells have been recorded. Topographically, low el-
evated areas have shallow groundwater levels because most water tables in unconfined
aquifers follow topography. In the Hargeisa watershed, the static water level varies from
6.9m to 56.76m below the surface, representing the watershed's low elevated area.
 LIII

Figure 16: Groundwater map in the Hargeisa watershed.

4. RESULTS AND DISCUSSION

4.1. Estimation of groundwater recharge

The study results of groundwater recharge were obtained from the Wetspass model. Results
of Wetspass recharge were compared with empirical recharge as representative of observed
groundwater recharge. As indicated in section 3.5, the Empirical recharge is determined by
empirical equations, including Chaturvedi Formula (CF), Sehgal Formula (SF), Krishna Rao
Formula (KRF), and Bhattacharya Formula (BF).
 LIV

The estimation of groundwater recharge using the Wetspass model shows that the average
simulated recharge from 2001 and 2020 is 193 mm/yr, quantified as 37.12% of the average
annual rainfall (520mm). The results of empirical recharge performed by CF, SF, KRF, and
BF equations during 2001–2020 are 88.95, 134.29, 60.96, and 106 mm/yr, which are about
9.33%, 21.5%, 34.2%, and 31.87% of average annual rainfall, respectively. According to the
empirical equations, the recharge performance varies with annual rainfall data only. Compar-
ison of the empirical recharge rates performed by CF, SF, KRF, and BF equations with those
obtained by Wetspass are explicit.

The recharge obtained from manual equations is relatively different from each other. SF re-
charge is higher than CF, BF, and KF, but less than Wetspass recharge. On the other hand,
KF recharge is lower than all recharge empirical and Wetspass.

4.1.1. Calibration of groundwater recharge

Model calibration and validation were conducted using statistical parameters including Root

Mean Square Error (RMSE) and coefficient determination (R ²) to evaluate the Wetspass
model's performance for the groundwater recharge estimation. Annual Wetspass recharge
(simulated data) is compared with empirical recharges (as observed recharge) to calculate the
value of RMSE.

4.1.2. Model Calibration of The Recharge

Model calibration was conducted. The values of statistical parameters are shown in Table 4
RMSE of Wetspass recharge compared with empirical recharge is about 46.7mm, which is
about 11.97% of the average empirical recharge. The highest RMSE is 66.02 mm (16.92% of
average empirical recharge) and 61.52mm (15.77% of average empirical recharge), respect-
ively, by comparing Wetspass recharges with KF and BF recharges. On the other hand, the
RMSE of Wetspass recharges compared with CF and SF recharges are lower than those ob-
 LV

tained with KF and BF, which are about 18.6 mm (4.61% of average empirical recharge) and
41.5 mm (10.64% of average empirical recharge), respectively.
The values of R^2 between Wetspass recharge and the empirical recharges are almost the not

big range between 0.8–0.92 as shown in Table 4. Even the values of R² show a good correla-
tion between Wetspass seasonal recharge and observed empirical recharges, which is 0.87.
However, the difference between Wetspass recharge and SF and BF recharges is relatively
high. In the modal validation, it is not possible to incorporate because of very limited data.
So, I prefer the model output in the sensitivity analysis.

Table 2: Statistical parameters of Wetspass recharge compared with an empirical equation

statistics CF KF SF BF average
RMSE (mm) 18 41.5 66.02 61.52 46.76
%RMSE 9.33 21.5 34.2 31.87 24.225
R^2 0.92 0.8 0.81 0.27 0.7

Figure 17. calibration graph

4.2. Wetspass output

The
calibration grph
70 Wets-
66.02
60 61.52
50 pass
46.76
40 41.5 monthly
34.2 31.87
30
21.5 24.225 model
20 18
10 9.33 has dif-
0 0.92 0.8 0.81 0.27 0.7
CF SF KF BF Average ferent
RMSE(mm) RMSE % R²
 LVI

parameters and needs input variables to give a reasonable output for understanding the water
balance component of a given watershed. To use the parameters of the Wetspass monthly
model, the sensitivity of all parameters and input variables was used. This was used to under-
stand the model response in terms of the variation of parameters. A sensitivity analysis was
carried out for the monthly model. To conduct the sensitivity analysis of the parameters, the
average value of all meteorological data in the Wetspass -M model input data set should be
the default. The other biophysical data is left as it is. A reasonable range has been assigned to
assess the parameter sensitivity (“a” interception, alfa coefficient, Lp coefficient, etc.).

Calibration of the model result was performed by manually adjusting the different parameters
placed on the model with range values. The parameters are namely: alfa coefficient, “a” inter-
ception, Lp coefficient, and runoff delay factor “x”.

The monthly simulated raster map output was summed up to get seasonal values, the
Hargeisa watershed is described well by the Wetspass-M model, because the model is spati-
otemporally distributed and represents all characteristics of the watershed (hydrological,
physical, and meteorological) and detects both spatially and temporal changes of the water
balance components of the watershed. The monthly water balance components of the water-
shed are the main output of the model which is based on the specific objectives of this re -
search. The seasonal values were obtained by summing up the monthly results, which are sea-
sonal actual evapotranspiration (AET), surface runoff, and recharge. The results based on cli-
matic seasons in the country, are Diraac/winter (December to February), which is dry, Gu’/
spring (March to May), which starts the rainfall all the country, then Xagaa/summer (June to
August), and the last one Dayr/autumn (SON) which is same as Gu’, but the Hargeisa water-
shed climatically semi-arid region. The outputs are a raster grid map in which each pixel rep-
resents the magnitude of different water balance components of the Hargeisa watershed.

4.2.1. Actual Evapotranspiration (AET)

The Wetspass-M model calculates the actual evapotranspiration as a sum of evaporations

from transpiration through vegetated cover, interception loss by vegetation, and evaporation
from bare soil. Actual evapotranspiration is a major component of water balance to determine
groundwater recharge of the Hargeisa watershed. The result showed that a minimum and
maximum value of each season in winter 123mm and 138mm, spring 139mm and 159mm,
summer 158mm, and 207mm, and the last autumn 138mm and 158mm as indicated Fig 18,
which water lost through evapotranspiration in each season from the watershed. This
 LVII

happened due to solar radiation, dry wind availability, and the presence of different land use
land cover of the study watershed.

The Hargeisa watershed is mostly covered by different vegetation, such as open shrubland,
which is 50.9% and 76.54% and next is non-vegetation land, 46.52% and 19.1% in 2001 and
20 20, as indicated Fig 18. This facilitates evapotranspiration by reducing surface runoff in
the watershed.

The Wetspass model simulates the actual evapotranspiration monthly. To get the seasonal
values for the water balance components of the watershed, the monthly results have to be ad-
ded into four seasons in the Somalia winter (December to February), spring (march to May),
summer (June to July) and autumn (September to November) seasons. Similarly, the season
values for the different water balance components have been obtained by summing up the
monthly results. The soil moisture deficit appears to be scarce in rainfall and is responsible
for the low evapotranspiration in the season. During the dry season, water demand increases
because the temperature is high, in the study area there is no cropland which affects water de-
mand. As in the summer season, there is high duration, intensity, and amount of rainfall dis-
tribution, high soil moisture, healthy vegetation, and other land covers which accelerates
evapotranspiration.

According to the Wetspass model result, land use/land cover and rainfall are found to be the
main input variables to control evapotranspiration in the watershed and evapotranspiration is
the main hydrological process of water loss in the watershed which is a major water balance
component that affects the recharge.
 LVIII

Figure 18: Seasonal actual evapotranspiration map of Hargeisa watershed a) winter AET b)
spring AET c) summer AET d) autumn AET
In addition, the intercept value of the area is computed using the Wetspass model. Depending
on the type of vegetation, the interception fraction represents a constant percentage of the
seasonal rainfall value. Thus, the fraction decreases with an increase in seasonal rainfall
(since the vegetation cover is assumed to be constant throughout the simulation period)
(Batelaan & Woldeamlak, 2007). The interception value in the Hargeisa watershed is con-
sidered the model output which is the decrease and increase depending on rainfall variation,
 LIX

10mm, 67.82mm, 55.83mm, and 49.56mm (winter, spring, summer, and autumn) respect-
ively.

4.2.2. Surface Runoff

The Wetspass-M model uses different input variables and parameters to calculate monthly
surface runoff (SV). It is a function of the watershed's vegetation type, soil texture, slope, and
other physical characteristics. The amount of runoff is highly variable depending on the
above input variables in the Hargeisa watershed.

The surface runoff SV value in the Hargeisa watershed is highly variable spatially from place
to place as well as seasonally. The model result shows the monthly value of runoff in the
Hargeisa watershed, summed up seasonally. The seasonal variation of the runoff values in the
watershed is winter (DJF) 1.86mm, spring (MAM) 69.37mm, summer (59.38), and autumn
(SON) 40.4mm. The runoff value is highly dependent on the runoff coefficients, which are
also highly influenced by changes in soil moisture, infiltration capacity, land use/land cover,
and the amount and intensity of rainfall. The seasonal variation in the Hargeisa watershed is
mostly due to the amount of rainfall. During the wet season, the recharge exceeds rainfall be-
cause of low intensity and soil infiltration capacity, leading to high recharge. The simulated
maps of the surface runoff in the Hargeisa watershed vary from place to place, as indicated in
the maps.

The model calibration was done by simulating the monthly runoff values. The model result
shows that the summed-up seasonal value of the runoff in the Hargeisa watershed varies in
range from Diraac (winter) 0 to 1.86mm, Gu’ (spring) 0 to 69.37mm, Xagaa (summer) 0 to
59.38mm and the Dayr (autumn) 0 to 40mm as a minimum and maximum value respectively.
The seasonal variation of runoff value in the watershed rainfall is 53 to 197mm, 31 to
201mm, and 25 to 147mm in spring (MAM), summer (JJA), and autumn (SON), with seasons
respectively, so the surface runoff has simulated the Wetspass.
 LX

Figure 19: Runoff maps in the Hargeisa watershed a) Diraac/winter runoff b) Gu’/spring run-
off c) Xagaa/summer runoff d) Dayr/autumn runoff

The spatial variation of the surface runoff values in the watershed highly depends on the dif-
ferent biophysical factors, such as slope, which significantly affects surface runoff in the wa-
tershed.
 LXI

4.2.3. Groundwater Recharge Zone

The amount of infiltration-percolation into groundwater recharge depends on different factors

such as slope, land use soil texture, and groundwater level (Al Kuisi & El-Naqa, 2013). Dif-
ferent models estimate recharge in a given area depending on actual areal conditions. In this
case, the Wetspass model estimates the monthly long-term spatiotemporal distribution of
groundwater recharge of the Hargeisa watershed by subtracting surface runoff and evapotran-
spiration from precipitation. The model determines the monthly groundwater recharge de-
pending on different input variables. Recharge is estimated as the residual of the water bal-
ance components from rainfall. Due to this, it responds to variations and is sensitive to vary-
ing in different input variables and model parameters.

According to the Wetspass result, the seasonal recharge of the Hargeisa watershed is between
0 and 74.9mm as minimum and maximum values, respectively. The highest groundwater re-
charge occurs in the three seasons, and the highest one in summer (Xagaa) 74.9mm, spring
(Gu’) 68mm, and autumn (Dayr) 47.7mm, the remaining one is in the dry season according to
the country but, the simulated map in the model computed 11.9mm. This temporal variation
is made due to the variation of different climatological and biophysical input parameters,
mainly rainfall. There is low duration, intensity, and amount of rain distribution and high soil
moisture in the summer season, which accelerates high groundwater recharge in the area.

Recharge vs Rainfall
40
monthly Recharge (mm)

35
30 R² = 0.917523904266426
25
20
15
10
5
0
0 10 20 30 40 50 60 70 80

monthly rainfall (mm)

Figure 20: the relationship between monthly recharge and rainfall

 LXII

Figure 21: Seasonal variation of water balance component in the watershed

As mentioned in the above figure, groundwater recharge is highly sensitive to variations in
rainfall, actual evapotranspiration, and other biophysical elements. Groundwater recharge can
have negative zero values in a given watershed. The negative value indicates that evapotran-
spiration exceeds precipitation. In the case of the Hargeisa watershed, the minimum value is
zero.
 LXIII

Figure 22 Groundwater recharge map zone in the watershed a) winter recharge b) spring re-
charge c) summer recharge d) autumn recharge

The groundwater recharge distribution in the Hargeisa watershed also varies spatially from
place to place. The marking part of the maps is receiving more recharge. This is due to the
good permeability of the soils, the fact that the land use cover in this marking area is covered
by grassland as indicated in the land use map, and the gentle slope topography. The other part
of the area receives relatively low/moderate groundwater recharge.

4.3 Land Use Land Cover

Land use is an important characteristic of governing the surface runoff process that affects in-
filtration(recharge), evapotranspiration, and erosion. Different land use types have different
influencing capacities on recharge, runoff process, and evapotranspiration rates, due to their
different leaf area indices, root depth, and size (Batelaan and DeSmedt, 2001). It is a highly
sensitive parameter for recharge estimation using Wetspass modeling.

USGS-NASA prepared the land use/land cover map of Somalia. The Hargeisa watershed land
use land cover was derived and modified from this map. Different land uses cover the study
watershed in 2001 and 2020. In terms of area coverage, the important land cover units are
open shrubland, grassland, urban and built-up land, and non-vegetation land. the land cover
percentage areas open shrubland 50.9%, grassland 2.36%, urban & built-up land 0.23%, and
non-vegetation land 46.52%, while the land cover 2020 is open shrubland 76.54%, grassland
4.12%, urban & built-up land 0.24% and non-vegetation 19.10% for each in the total area. In
the land use, there is no significance with seasonal, and it does not account for the Wetspass
model with respect to the parameter table in the seasonal. Land-use/land-cover fractions are
used as weighting factors for calculating the water balance at each grid cell level. The total
coverage of the Hargeisa watershed is shown below in a table and map with all areas' cover -
age.

Table 3: Land use types in Hargeisa watershed

No Land use type 2001 2020 Area percentage

code land coverage Area sqkm Area sqkm 2001 % 2020 %

7 Open shrub land 6741.66 10140.33 50.9 76.54

 LXIV

10 Grassland 312.58 545.63 2.36 4.12

13 Urban & and built-up 30.78 31.21 0.23 0.24

land

15 Bare land 6162.73 2530.65 46.52 19.10

Total 13247.75 100 100

Table 4: land-use land cover changes in the Hargeisa watershed

change (2001-2020)
Open shrubland - Open s
Open shrubland - Grassl
Open shrubland - Non-v
Grassland - Open shrubl
Grassland - Grassland
Grassland - Non-vegetat
Urban & Built-up land -
land
Bare land
bare land - Grassland
Bare land - Urban & Bui
Bare land - Bare land
total
 LXV

Figure 23: Land cover type and percentage Area change

Figure 24: Land use type map in the Hargeisa watershed for 2001 and 2020

4.4. Future Climate Change

In this study, the analysis was performed on two variables, precipitation (rainfall) and tem-
perature (T max and T min) data.
 LXVI

4.4.1. Maximum Temperature

The mean monthly, annual maximum temperature and the change for the base period and the
future climate scenarios for (RCP4.5, 2020s, 2050s, 2070s, and 2099s), and (RCP8.5, 2020s,
2050s, 2070s, and 2099s) are shown in the figures 23 and 24.

The annual mean temperature (RCP4.5, 2020s, 2050s, 2070s, and 2099s), and (RCP8.5,
2020s, 2050s, 2070s, and 2099s) are 30C˚, 30.96C˚, 31.35C˚ 31.21C˚, and 30.41C˚, 31.15C˚,
31.36C˚ and 32.25C˚ each value is respectively. The average annual increment of maximum
temperature ranges increases by +0.25C˚, 1.20 C˚, 1.59C˚, and 1.45 C˚ from the base period
to RCP 4.5 2020s, 2050s, 2070s, and 2099s, respectively, which is intermediate emission sce-
narios, and the highest baseline scenarios emissions which are RCP 8.5 2020s, 2050s, 2070s,
and 2099s increment increase by +0.64C˚, 1.38 C˚, 1.60C˚, and 2.50 C˚, respectively.

The monthly maximum temperature in RCP4.5 scenarios will decrease in January, October,
and November -0.23C˚, -0.78C˚, and -0.35C˚ in the first decay, while annually projects will
increase in all decades, as indicated in the figure.

Figure 25: mean maximum temperature and change of base period (1986-2005) for both
scenarios
 LXVII

Figure 26: mean maximum temperature and change of base period (1986-2005) for both
scenarios.

4.4.2. Seasonal Maximum Temperature

The study revealed that seasonal maximum temperature increased for all seasons in RCP4.5
and RCP8.5 scenarios during the projected periods except for the first autumn decay, as indi-
cated in the figure below. Somalia has four seasons: Gu’, Xagaa, Dayr, and Dirac, based on
the spring, summer, autumn, and winter, as shown in Figures 29 and 30.

The temperature would be increased in all seasons for RCP4.5 winter (DJF) 0.29cͦ, spring
(MAM) 0.62cͦ, and summer (JJA) 0.38cͦ, but the autumn (SON) would be decreased in 0.28cͦ,
while the RCP8.5 winter (DJF) 0.29cͦ, spring (MAM) 1.1cͦ, and summer (JJA) 0.61cͦ, and the
autumn (SON) 0.59cͦ for the 2020s. in the 2050s RCP4.5 and RCP8.5, winter (DJF) 1.04cͦ,
spring (MAM) 0.621.44cͦ, summer (JJA) 1.59cͦ, autumn (SON) 0.72cͦ, and winter (DJF) 1.1cͦ,
spring (MAM) 1.91cͦ, summer (JJA) 1.81cͦ, and autumn (SON) 0.72cͦ respectively. In the
other two decay for 2070s and 2080s in the RCP4.5 would be increased by winter (DJF)
1.51cͦ, 1.23cͦ, spring (MAM) 2.41cͦ, 2.26cͦ, and summer (JJA) 1.72cͦ, 1.60cͦ, and autumn (SON)
0.72cͦ, 0.69cͦ respectively.

In the 2080s, the temperature would decrease according to the 2070s. the RCP8.5 in 2070s
and 2080s is winter (DJF) 1.56cͦ,1.98cͦ, spring (MAM) 2.68cͦ, 3.76cͦ, summer (JJA) 1.45cͦ,
2.72cͦ and autumn (SON) 0.68cͦ, 1.51cͦ, with orderly. As indicated in the table for the projec-
tion periods in both scenarios the spring season shows an increase for all periods except sum-
 LXVIII

mer in RCP4.5 2040-2059 which the temperature increases by 1.59cͦ, while the spring is
1.44cͦ. in general, this season for Somalia is known as rain season, in all projection periods
the temperature maximum and minimum increase in this season, in as the rise up the tempera-
ture max/min, the rainfall decreases between -4.71% to -14.91% for 2040-2099, but in the
first decay is 11.47%.

Figure 27: Seasonal max temperature change for RCP4.5 and 8.5 scenarios during the pro-
jection period in the 2020s and 2050s.

Figure 28: Seasonal max temperature change for RCP4.5 and 8.5 scenarios during the pro-
jection period in the 2070s and 2080s.
 LXIX

4.5.1 Minimum Temperature

As can be seen in figures 25 and 26, the mean monthly, annual temperature and change in
minimum temperature from the base period (1986-2005) to future climate scenarios for
(RCP4.5, the 2020s, 2050s, 2070s, & 2099s), and for (RCP8.5, 2020s, 2050s, and 2070s, &
2099s) are shown the figures below. The minimum mean temperatures (RCP4.5, 2020s,
2050s, 2070s, & 2099s), and (RCP8.5, 2020s, 2050s, 2070s, & 2099s)) are 17.74C˚, 18.98C˚,
18.98C˚, and 19.03C˚, which are intermediate emission scenarios. While the highest baseline
emission scenarios are 17.97C˚, 18.49C˚, 19.63C˚ and 20.91C˚, each value is respectively.
The average annual minimum temperature would be increased by the range +0.78 C˚, 1.53C˚,
2.02C˚ and 2.06C˚, with (RCP4.5, 2020s, 2050s, 2070s, and 2099s) respectively. Also, high-
est emission scenarios would be increased for (RCP8.5 2020s, 2050s, 2070s and 2099s) 1.0
C˚, 1.95C˚, 2.67C˚ and 3.95C˚, with orderly.
In the minimum monthly temperature increase in both scenarios in all project periods. Re -
garding the annual RCMs, the maximum temperature will increase by an average of 0.25C˚
and 1.45C˚ under medium emission scenarios (RCP4.5), and the minimum temperature will
increase by 0.78C˚ and 2.1C˚ in the mid-term of the future, respectively. Likewise, the max-
imum temperature will increase by 0.64C˚ and 2.50C˚ and the minimum temperature will in-
crease by 1.02C˚ and 3.95C˚ for the near future and mid-future respectively under high emis-
sion scenarios RCP8.5. in all projection years in the RCP8.5 is warmer than the RCP4.5 pro-
jection scenarios.
The changes in projected temperature were not only expected to vary annually but also
vary in all seasons. However, the seasonal variations are higher for the minimum than the
maximum temperature, as discussed in the section of the seasonal. The seasonal changes of
maximum temperature are higher in the Gu’ seasons (March, April, and May), while the rise
in minimum temperature is higher in the Gu’ season in the maximum temperature in season.
Moreover, the highest rise of temperature in both scenarios (RCP4.5, 8.5) 2.38C˚, and
3.03C˚, IN 2060-79, likewise 2080-99 the temperature rises 3.92C˚ and 4.2C˚. furthermore,
the maximum temperature rise in projection scenario is 2.41C˚, and 2.68C˚, in 2060-2079,
and 2080-2099 2.26C˚, and 3.76C˚, for both scenario with respectively.
 LXX

Figure 29: mean minimum temperature and change of base period (1986-2005) for both
scenarios

Figure 30: mean minimum temperature and change of base period (1986-2005) for both
scenarios

4.5.2. Seasonal Minimum Temperature

The seasonal minimum temperature of the study revealed the increase for all seasons in the
RCP4.5 and 8.5 scenarios during the projected period as indicated in the figure below with ta-
ble. Somalia has four seasons as I mentioned above which are based on climatic seasons in
the world. The minimum temperature would be increased for the 2020s and 2050s winter
(DJF) 0.74cͦ, 1.38cͦ, spring (MAM) 1.01cͦ, 1.72cͦ, summer (JJA) 0.81cͦ, 1.7cͦ, and autumn
(SON) 0.73cͦ, 1.35cͦ, based on RCP4.5 scenario respectively. while the RCP8.5 2020s and
 LXXI

2050s increased the minimum temperature are winter (DJF) 0.89cͦ, 1.96cͦ, spring (MAM)
1.19cͦ, 2.1cͦ, summer (JJA) 1cͦ, 2cͦ, and autumn (SON) 0.89cͦ, 1.72cͦ, respectively.

As same time the 2070s and 2080s based on RCP4.5 the minimum temperature increased by
winter (DJF) 2cͦ, 1.98cͦ, spring (MAM) 2.38cͦ, 3.92cͦ, summer (JJA) 2.1cͦ, 2.1cͦ, and autumn
(SON) 1.6cͦ, 1.84cͦ, respectively. while the RCP8.5 based on the 2070s and 2080s in the pro-
jection period the temperature increased by winter (DJF) 2.62cͦ, 3.98cͦ, spring (MAM) 3.03cͦ,
4.2cͦ, summer (JJA) 2.57cͦ, 3.92cͦ, and autumn (SON) 2.46cͦ, 3.71cͦ, orderly. In eastern semi-
arid Africa, the minimum and the maximum temperatures for both future time horizons are
within the projected range by IPCC and agree with a range produced by other regional re-
searchers. According to Ethiopian researchers, climate change in different regions has shown
that the temperature is expected to increase, which depends on the strength of change, vary-
ing within the techniques of downscaling and the climate model type(Chemura et al., 2017).

Figure 31: seasonal min temperature change for RCP4.5, 8.5 scenarios during the projection
period in the 2020s and 2050s
 LXXII

Figure 32: seasonal min temperature change for RCP4.5, 8.5 scenarios during the projection
period in the 2070s and 2080s

4.6.1. Rainfall

As indicated from the figure below the mean monthly and annual rainfall and the change in
rainfall of the Waqooyi Galbeed region (northeast region) for the base period and the future
climate scenarios (RCP4.5, the 2020s, 2050s, 2070s, and 2099s) are 210.86mm, 226.9mm,
222.65mm, and 261.16mm, respectively. At the same time, the (RCP8.5, the 2020s, 2050s,
2070s, and 2099s) are 215.61mm, 242.47mm, 270.35mm, and 277.89mm, respectively.

The mean annual rainfall has increased in each scenario for the selected time scale. The mean
annual rainfall would be increased in percentage when comparing the base period by 1.51%,
5.45%, 8.19%, and 19.44%, in the RCP4.5, also in RCP8.5, 4.68%, 11.80%, 22.90%, and
31.34%, respectively the above time scale. As indicated in the figures the annual rainfall will
increase in the future. Still, Somalia has four seasonal climatic that are based globally, gener-
ally, the country receives rainfall in two seasons Gu’ (spring) and Dayr (autumn). Still, the
study area is exceptional because it receives rainfall 65% of the year which is low intensity,
and will be discussed in detail in the section 4.4
 LXXIII

Figure 33: mean annual rainfall and change of base period (1986-2005) for both scenarios

Figure 34: mean annual rainfall and change of base period (1986-2005) for both scenarios

4.6.2. Seasonal Rainfall

The seasonal rainfall changes in the projection periods 2020s and 2050s based on RCP4.5
scenarios would be winter (DJF) -0.6mm, 6.02mm, spring (MAM) 11.47mm, -5.74mm, sum-
mer (JJA) -37.33, -25.02mm, and autumn (SON) 44.62mm, 90.17mm, respectively. In the
same way, the 2020s and 2050s in RCP8.5 scenarios would be winter (DJF) 2.77mm,
 LXXIV

19.65mm, spring (MAM) -14.35mm, -9.8mm, summer (JJA) 5.4mm, -6.85mm, and autumn
(SON) 62.28mm, 138.57mm, respectively. At the same time, the rainfall in the 2070s and
2080s in RCP4.5 would be winter (DJF) 13.55mm, 31.34mm, spring (MAM) -10.57mm, -
4.71mm, summer (JJA) 5.69mm, 43.21mm, and autumn (SON) 89.55mm, 163.38mm, with
orderly. Also, the RCP8.5 scenarios in the projected period of the 2070s and 2080s would be
winter (DJF) 19.24mm, 51.15mm, spring (MAM) -14.44mm, -14.91mm, summer (JJA)
70.24mm, 101.5mm, and autumn (SON) 199.81mm, 238.28mm, respectively.

The percentage change in rainfall for all seasons during the projection periods is based on the
scenarios (RCP4.5 and 8.5), which are increase and decrease. The increase for the RCP4.5 in
the 2020s is spring (MAM) 11.95% and the autumn 10.1%, in the 2050s winter (DJF) 8.95%,
summer (JJA) 9%, and autumn (SON) 20.42%, in the same way, the RCP8.5 scenario for
projection period 2020s and 2050s winter (DJF) 4.12%, 29.2%, summer (JJA) 1.95%, and au-
tumn (SON) 14.1%, 31.37%. the decrease for seasonal rainfall changes for RCP4.5 based
2050s the spring is 5.98%, but the summer for both 2020s and 2050s would decrease by
13.47%, and 11%, while the RCP8.5 for spring would decrease 14.95%, 10.21% for both pro-
jection period, but in 2050s the summer rainfall decreases 2.47%. also, the 2070s and 2080s
based scenario RCP4.5, winter (DJF) 20.14%, 46.59%, summer (JJA) 2.1%, 15.59% and au-
tumn (SON) 20.27%, 36.98%, at same time the RCP8.5, winter (DJF) 28.6%, 76%, summer
(JJA) 25.35%, 36.63%, and autumn (SON) 45.25%, 53.94%, each one respectively. the
spring decrease for the 2070s and 2080s 11%, 4.91% and 15%, 15.53% for both scenarios
RCP4.5, 8.5 respectively.

By projection periods in both scenarios, the annual rainfall will increase as indicated above,
but when we look for the seasonality change, the summer in 2020-39 will increase by 13.47%
and 2040-59 -9%, while other periods will rise. In the Gu’ season, the projected rainfall is
less than normal, and the spring season decrease with different fluctuation in the Dayr/au-
tumn increase as indicated in Figure 33 and 34
 LXXV

Figure 35: seasonal rainfall changes for RCP4.5, 8.5 scenarios during the projection period
2020s and 2050s

Figure 36: seasonal rainfall changes for RCP4.5, 8.5 scenarios during the projection period
2070s and 2080s

4.7. Impact of Land Use Land Cover and Climate Change on water balance components

4.7.1. Land use land cover change impact

The seasonal groundwater recharge map was discussed in the figure recharge zone which is
mm/season, in this section as indicate the impact of LULC in 2001 and 2020 change in the
groundwater recharge and the surface runoff in the model simulated table of the study area
with annual yearly.
 LXXVI

The LULC change produced an average groundwater recharge monthly in a map form and
simulated table in the output of 2001 and 2020, these outputs were summed up by yearly, the
simulated table annual recharge 175.84mm/year, and 190.42mm/year in 2001 and 2020 re-
spectively, which is the percentage of annual rainfall 36.62% and 33.84% (rainfall 520mm/
year), the high recharge occur in the summer and spring 75mm and 68mm, the recharge al-
most zero in the winter season, while the autumn is moderate when comparing other two sea-
sons. In general, the recharge in the watershed increased by 0.8mm/year with the change of
land use land cover because as indicated by the LULC there’s no agricultural area and urban
expansion which highly affects the land use change, but in 2001 the watershed cover open
shrubland on non-vegetation land 50.9% and 46.52% which is total cover 97.42% of the
whole watershed, and in 2020 open shrubland cover 76.54% and non-vegetation land 19.1%
by a total of 95.64%, the open shrubland increased 25.6% and non-vegetation land decreased
27.42%, the increase of open shrub was reduced the soil erosion and surface runoff. Like-
wise, the soil classes in the watershed mainly two types, sandy loam and loam, cover the area
that has good infiltration capacity because it contains a mixture of soil types. Furthermore, in
the map of annual recharge in 2001 and 2020 there is no change in value but the distributions
differ from one another as indicated figure 35.

Figure 37: Annual recharge distribution map 2001 and 2020

The runoff was discussed seasonally in the surface runoff section in the model output, as il-
lustrated in figures, how the runoff was unevenly distributed throughout the seasons. The an-
 LXXVII

nual surface runoff decreased from 178.8mm/yr to 152.1mm/yr, and the Wetspass simulated
table 114.27mm/yr and 71.82mm/yr due to land use land cover change from 2001 to 2020
over the Hargeisa watershed, so the decrease in surface runoff affected the increase of open
shrubland and the drawdown of non-vegetation land as mention land use figures. Therefore,
the maps show the variation runoff in 2001 and 2020. In general, the increase/decrease of sur-
face runoff influences the availability of vegetation, soil type, and the slope of the watershed.
these maps display the decrease in surface runoff in the whole watershed.

Figure 38: annual surface runoff distribution map in 2001 and 2020

4.7.2. Future climate change impact

Somalia is one of the countries most affected by the impacts of climate change due to its geo-
graphical location in the Horn of Africa and its reliance on climate-sensitive sectors such as
agriculture and pastoralism as its main source of food and income. Climate change is impact-
ing the availability and quality of water resources in Somalia, bringing several challenges to
the country's water supply(Ali et al., 2023). Climate change is a major issue in Somalia, and
the effects on water resources have become apparent annually. with a growing population,
 LXXVIII

limited access to clean, safe drinking water, and proper water resource management is critical
in the country.

Sugulle, 2011, Indicated that climate change is already affecting water resources in Somalia,
rising sea levels from the temperature increase caused saltwater intrusion in coastal areas.
This has led to a decrease in the availability of freshwater resources, making it increasingly
difficult for people to access clean drinking water. Additionally, extreme weather events and
prolonged droughts decrease surface water (Barkads) and groundwater levels. This has resul-
ted in an increase in water shortages and scarcity, which directly impact Somalis' livelihoods
and the country's sustainable development.

The groundwater system is a vital part of the water resource in Somalia and is heavily im-
pacted by climate change, while the groundwater resources are mainly in the form of aquifers
that are limited capacity. Climate change is having a direct and indirect impact on the ground-
water systems in Somalia. The direct impact is through temperature changes and rainfall lead-
ing to the decline in the availability of groundwater resources. In addition, the indirect im -
pacts of climate change on the groundwater systems include changes in land use, soil erosion,
and sedimentation.

The impacts of climate change on the groundwater systems in Somalia have led to numerous
consequences for the local population, including reduced access to drinking water, increased
water scarcity, and the increasing temperatures have put additional stress on the existing wa-
ter resources. According to Ahmed and Ali (2023), a wide range of interventions is required
to effectively address the impacts of climate change on the water resource systems in
Somalia. These interventions should include improved water management systems and
policies, increased water harvesting and storage, improved irrigation techniques, and im-
proved water quality standards.

Generally, the monthly annual rainfall and the projected temperature will increase as com-
pared to the baseline for both scenarios in the above indicated.

The projected climate change was varying yearly, monthly annually, and the seasonally.
However, the seasonal variations have higher minimum temperatures than maximum temper-
atures, as shown in the seasonal figure. The seasonal change maximum temperature is higher
spring (MAM) in all projections in both scenarios except the 2040-2059 for RCP4.5. Mean-
while, the minimum temperature rise is higher than the maximum temperature in all seasons.
 LXXIX

The highest seasonal rise is RCP8.5 in the 2060s and 2080s, and the lowest is RCP4.5 in the
2020s and 2050s for both maximum and minimum temperatures.

The changes in projected rainfall are more profound on seasonal bases compared to the sea-
sonal baseline, the seasonal projection rainfall shows the decline spring season started 2040
to 2099 for both scenarios with different fluctuations, in the summer 2020s rainfall declined -
37.33mm for RCP4.5, while, 2040s the rainfall decreased -25.02mm and -6.85mm for
RCP4.5,8.5. Somalia mainly has two rain seasons: spring (Gu’) and autumn (Dayr). As indi-
cated in the table and the figure the spring is below zero with different fluctuation, and the
autumn increase for all projected seasons as the table and the figure for both scenarios
Hargeisa City depend on groundwater resources in the Geed deeble aquifer. The recharge
aquifer depending on rainfall and one of two main rain seasons under zero. Likewise, the
summer decrease for 2020s and 2040s is both RCP4.5,8.5, so that means the groundwater
recharge decreases in these seasons. Inversely the autumn season the rainfall rises as com-
pared to the baseline. This clarifies that the rainfall and the groundwater recharge are directly
proportional because the groundwater recharge depends on only rainfall because farmers and
open water surface are part of groundwater recharge.

Table 5: Seasonal rainfall projected (%)

Scenario RCP 4.5 % RCP 8.5 %

Projection 2020- 2040- 2060- 2080- 2020- 2040- 2060- 2080-

years 2039 2059 2079 2099 2039 2059 2079 2099

Winter 0.9 8.95 20.14 46.59 4.12 29.2 28.6 76

Spring 11.95 5.98 11 4.91 14.95 10.21 15 15.53

Summer 13.47 9 2.1 15.59 1.95 2.47 25.35 36.63

autumn 10.1 20.24 20.27 36.98 14.1 31.37 45.25 53.94

 LXXX

projection rainfall RCP4.5,8.5

Winter Spring Summer autumn
300
rainfal % change

200
100
0
2020- 2040- 2060- 2080- 2020- 2040- 2060- 2080-
-100 2039 2059 2079 2099 2039 2059 2079 2099
RCP 4.5 RCP 8.5
seasonal rainfall

Figure 39: seasonal rainfall projected RCP4.5, 8.5

 LXXXI

5. CONCLUSION AND RECOMMENDATION

5.1. Conclusion

Water resources are critical elements to sustain life, and the hydrological environment should
be investigated and understood for planning, managing, and properly utilizing water re-
sources. A nowadays. Water resource use is decreasing due to contamination and pollution
and the water resource concentration leading to groundwater. So, the available resources
should be investigated. There are different physical and empirical models to understand the
water balance components of a given watershed. One of the spatially distributed physical
models is Wetspass. The model considers all meteorological, hydrological, and biophysical
factors of the area. To evaluate water balance component of the watershed, hydro-meteoro-
logy, land use, soil texture, topography and slope of the area has been investigated to manage
water resource.

The Wetspass-M model has successfully simulated the annual monthly water balance com-
ponents of the Hargeisa watershed. The highly variable distribution of the climatic inputs
(parameters) associated with variations in land use/land cover, soil texture, topography, and
slope are responsible for variations of the water balance element within the watershed. Based
on the model, the annual groundwater recharge in the Hargeisa watershed is 5mm,
192.1mm/year, and 6mm 193mm/year with a minimum and maximum as indicated in figure
24. Which represents 36.62% and 33.84% of the total annual rainfall. In the seasonality, the
main recharge occurs in summer, and the spring.

The study investigated of LULC changes on spatially distributed water balance components
in the Hargeisa watershed. All data was processed into raster form and used as input in the
water balance model Wetspass to simulate monthly water balances. From the land cover
change analysis, the recharge increased from 2001 to 2020 LULC due to the increase in open
shrubland areas and the decrease in non-vegetation land.

The sensitivity analysis in the watershed shows that the rainfall, temperature, and slope are
the most important hydrological processes in the study area in terms of affecting the rate and
the amount of water balance component. The Hargeisa watershed was not conducted a scien-
tific study quantification of the future climate change and the water balance component was
not well specified in the country. The model considers all of the area's meteorological, hydro-
logical, and biophysical aspects. The area's land use, soil texture, topography, and slope were
 LXXXII

researched to estimate water balance components of the watershed hydrometeorology. The

water balances evaluate using the Wetspass model was not correctly specified because of the
limitation of the data and other problems.

Climate change has had a profound impact on the availability of water resources in Somalia,
and its impacts are likely to become more severe shortly. The Somali people have been adapt-
ing to the impacts of climate change on water resources, but the scale of the challenge re-
quires a system-wide approach to interventions with support from the global community. In
addition, global responsibility must be taken to ensure that Somalia is supported in its efforts
to address the effects of climate change on water resources. Through these measures, Somalia
can ensure that its water resources remain abundant and resilient in the face of a changing cli -
mate.

The projection of future climate change results shows that the rainfall and the maximum and
minimum temperature increased annually for both scenarios, likewise, the seasonal projected
temperature increased when the minimum temperature was higher than the maximum temper-
ature, in the rainfall projected seasons were increased in winter, summer and the autumn in
the 2020s to 2099s for both scenarios, but the spring were decreased in 2020s to 2099s with
fluctuation 4.91% to 15.53%. as the rainfall decreases spring for all projected period the tem-
perature is higher than other seasons minimum and maximum. while the winter and the sum-
mer rainfall decreased 0.9% and 13.47% in the 2020s.

5.2. Recommendations

The results obtained from the study can be taken as hints and initial investigation into the
groundwater recharge modelling of the Hargeisa watershed. The groundwater recharge and
future climate results can be improved further and reached to more accurately monitor
groundwater resource data and management. The new and sufficient borehole data, soil, and
land use data are prepared in fine resolution, and new meteorological stations become avail-
able at the watershed because the GIS-based Wetspass model is highly sensitive to those in-
put data.

The study will be an input for further investigation and water resources planning purposes,
design, and developmental activities.
 To understand in general, the fluctuation of groundwater level, groundwater level
monitoring as suitable as can. This helps to estimate the future groundwater recharge
with different methods and the important Wetspass model recharge to validation test.
 LXXXIII

 The groundwater resources in Somalia/Somaliland, especially Hargeisa, depend on

rainfall. The water supply to Hargeisa city from Geed Deeble aquifer (boreholes)
received the recharge from the rainfall, so the Hargeisa watershed needs detailed
investigation for the present and future.
 To investigate the climate change impact on water resources, full meteorological
stations should be established at least one-third of the country.
 Furthermore, it should be paid attention to when using the hydrological models to
simulate hydrological response under climate change scenarios in which the result
shows seasonal rainfall below normal and the temperature rising.
 LXXXIV

6. REFERENCES

Abbas, N., Wasimi, S., & Al-Ansari, N. (2017). Impacts of Climate Change on Water Re-
sources of Greater Zab and Lesser Zab Basins, Iraq, Using Soil and Water Assessment
Tool Model. International Journal of Environmental, Chemical, Ecological, Geological
and Geophysical Engineering, 11(10), 823–829.

Abdishakur, A. E., Mohamed, A., Ali, S., Hassan, A. M., & Identification, H. (2022).
Journal of Civil Engineering Assessment of Drinking Water Quality in Mogadishu ,
Somalia. 2022(April), 0–7. https://doi.org/10.47363/JCERT/2022(4)125

Access, O., Kerkeni, L., Ruano, P., Delgado, L. L., Picco, S., Villegas, L., Tonelli, F., Merlo,
M., Rigau, J., Diaz, D., & Masuelli, M. (2016). We are IntechOpen , the world ’ s lead-
ing publisher of Open Access books Built by scientists , for scientists TOP 1 %. Intech,
i(tourism), 13.
https://www.intechopen.com/books/advanced-biometric-technologies/liveness-
detection-in-biometrics%0Afile:///D:/Google Drive/Organized Folder/2016/Kerkeni et
al/We are IntechOpen , the world ’ s leading publisher of Open Access books Built by
scientists ,

Adeaga, O. A., Bello, A. A., & Akinbaloye, T. (2019). Water Resources Uncertainty in Yewa
River Basin , Ogun state , South-West Nigeria. 131(May), 37–53.

African Development Bank Group. (2016). Water Infrastructure Development for Resilience
in Somaliland Program: Environmental and Social Management Framework. June.

Ahmad, W., Iqbal, J., Nasir, M. J., Ahmad, B., Khan, M. T., Khan, S. N., & Adnan, S.
(2021). Impact of land use/land cover changes on water quality and human health in dis-
trict Peshawar Pakistan. Scientific Reports, 11(1), 1–14. https://doi.org/10.1038/s41598-
021-96075-3

Ali, Mushir; Terfa, A. (2012). State of Water Supply and Consumption in Urban Areas at
Household Level: A Case Study of East Wollega Zone, Ethiopia. British Journal of Hu-
manities and Social Sciences, 5(2), 15.

Ali, A. I., Kassem, Y., & Gökçekuş, H. (2023). Examining the impact of climate change on
water resources in Somalia: The role of adaptation. Future Technology, 2(4), 45–58.
https://doi.org/10.55670/fpll.futech.2.4.5
 LXXXV

Armanuos, A. M., Negm, A., Yoshimura, C., & Valeriano, O. C. S. (2016). Application of
WetSpass model to estimate groundwater recharge variability in the Nile Delta aquifer.
Arabian Journal of Geosciences, 9(10). https://doi.org/10.1007/s12517-016-2580-x

Arunyawat, S., & Shrestha, R. P. (2016). Assessing land use change and its impact on ecosys-
tem services in northern Thailand. Sustainability (Switzerland), 8(8).
https://doi.org/10.3390/su8080768

Batelaan, O., & De Smedt, F. (2007). GIS-based recharge estimation by coupling surface-
subsurface water balances. Journal of Hydrology, 337(3–4), 337–355.
https://doi.org/10.1016/j.jhydrol.2007.02.001

Belete, M., Deng, J., Abubakar, G. A., Teshome, M., Wang, K., Woldetsadik, M., Zhu, E.,
Comber, A., & Gudo, A. (2020). Partitioning the impacts of land use/land cover change
and climate variability on water supply over the source region of the Blue Nile Basin.
Land Degradation and Development, 31(15), 2168–2184.
https://doi.org/10.1002/ldr.3589

Birkhofer, K., Diehl, E., Andersson, J., Ekroos, J., Früh-Müller, A., Machnikowski, F.,
Mader, V. L., Nilsson, L., Sasaki, K., Rundlöf, M., Wolters, V., & Smith, H. G. (2015).
Ecosystem services-current challenges and opportunities for ecological research. Fronti-
ers in Ecology and Evolution, 2(JAN), 1–12. https://doi.org/10.3389/fevo.2014.00087

Boithias, L., Acuña, V., Vergoñós, L., Ziv, G., Marcé, R., & Sabater, S. (2014). Assessment
of the water supply: DEmand ratios in a Mediterranean basin under different global
change scenarios and mitigation alternatives. Science of the Total Environment, 470–
471, 567–577. https://doi.org/10.1016/j.scitotenv.2013.10.003

Brauman, K. A. (2015). Hydrologic ecosystem services: linking ecohydrologic processes to

human well‐being in water research and watershed management. WIREs Water, 2(4),
345–358. https://doi.org/10.1002/wat2.1081

Chemura, A., Rwasoka, D., Mutanga, O., Dube, T., & Mushore, T. (2020). The impact of
land-use/land cover changes on water balance of the heterogeneous Buzi sub-catchment,
Zimbabwe. Remote Sensing Applications: Society and Environment, 18, 100292.
https://doi.org/10.1016/j.rsase.2020.100292

Chemura, A., Rwasoka, D., Mutanga, O., Dube, T., Mushore, T., Lyu, L., Wang, X., Sun, C.,
Ren, T., Zheng, D., Khoi, D. N., Dibaba, W. T., Demissie, T. A., Miegel, K., Hossain,
 LXXXVI

S., Albhaisi, M., Brendonck, L., Batelaan, O., Liu, J., … Martins, L. D. (2017). Impacts
of land-use change on streamflow and sediment load in Be River Catchment , Vietnam.
Energy Procedia, 2017(2), 5–7. https://doi.org/10.1016/j.rsase.2020.100292

Choukri, F., Raclot, D., Naimi, M., Chikhaoui, M., Nunes, J. P., Huard, F., Hérivaux, C.,
Sabir, M., & Pépin, Y. (2020). Distinct and combined impacts of climate and land use
scenarios on water availability and sediment loads for a water supply reservoir in north-
ern Morocco. International Soil and Water Conservation Research, 8(2), 141–153.
https://doi.org/10.1016/j.iswcr.2020.03.003

City, H., & Ahmed, B. B. (2019). COLLEGE OF URBAN DEVELOPMENT STUDIES

URBAN INFRASTRUCTURE PROVISION AND Appraisal of Urban Water Supply Ser-
vice : In Somaliland Advisor : Amdemariam Shiferie A thesis Submitted to the College of
Urban Development Studies.

Collet, L., Ruelland, D., Estupina, V. B., Dezetter, A., & Servat, E. (2015). Water supply sus-
tainability and adaptation strategies under anthropogenic and climatic changes of a
meso-scale Mediterranean catchment. Science of the Total Environment, 536, 589–602.
https://doi.org/10.1016/j.scitotenv.2015.07.093

Deng, X., Li, Z., Huang, J., Shi, Q., & Li, Y. (2013). A revisit to the impacts of land use
changes on the human wellbeing via altering the ecosystem provisioning services. Ad-
vances in Meteorology, 2013. https://doi.org/10.1155/2013/907367

Dereje, B., & Nedaw, D. (2019). Groundwater Recharge Estimation Using WetSpass Model-
ing in Upper Bilate Catchment, Southern Ethiopia. Momona Ethiopian Journal of Sci-
ence, 11(1), 37. https://doi.org/10.4314/mejs.v11i1.3

Egarter Vigl, L., Tasser, E., Schirpke, U., & Tappeiner, U. (2017). Using land use/land cover
trajectories to uncover ecosystem service patterns across the Alps. Regional Environ-
mental Change, 17(8), 2237–2250. https://doi.org/10.1007/s10113-017-1132-6

Elias, E., Seifu, W., Tesfaye, B., & Girmay, W. (2019). Impact of land use/cover changes on
lake ecosystem of Ethiopia central rift valley. Cogent Food and Agriculture, 5(1).
https://doi.org/10.1080/23311932.2019.1595876

Endris, H. S., Omondi, P., Jain, S., Lennard, C., Hewitson, B., Chang’a, L., Awange, J. L.,
Dosio, A., Ketiem, P., Nikulin, G., Panitz, H. J., Büchner, M., Stordal, F., & Tazalika, L.
(2013). Assessment of the performance of CORDEX regional climate models in simulat-
 LXXXVII

ing East African rainfall. Journal of Climate, 26(21), 8453–8475.

https://doi.org/10.1175/JCLI-D-12-00708.1

FACILITY, P.-P. I. A. (2018). Hargeisa Water Agency. December.

Gautam, S., Costello, C., Baffaut, C., Thompson, A., Svoma, B. M., Phung, Q. A., & Sadler,
E. J. (2018). Assessing long-term hydrological impact of climate change using an en-
semble approach and comparison with global gridded model-A case study on Goodwater
Creek Experimental Watershed. Water (Switzerland), 10(5).
https://doi.org/10.3390/w10050564

Gebre, S. L. (2015). Potential Impacts of Climate Change on the Hydrology and Water re-
sources Availability of Didessa Catchment, Blue Nile River Basin, Ethiopia. Journal of
Geology & Geosciences, 04(01), 1–7. https://doi.org/10.4172/2329-6755.1000193

Geng, X., Wang, X., Yan, H., Zhang, Q., & Jin, G. (2015). Land use/land cover change in-
duced impacts on water supply service in the upper reach of Heihe River Basin. Sustain-
ability (Switzerland), 7(1), 366–383. https://doi.org/10.3390/su7010366

Gong, J., Yang, J., & Tang, W. (2015). Spatially explicit landscape-level ecological risks in-
duced by land use and land cover change in a national ecologically representative region
in China. International Journal of Environmental Research and Public Health, 12(11),
14192–14215. https://doi.org/10.3390/ijerph121114192

Guzha, A. C., Rufino, M. C., Okoth, S., Jacobs, S., & Nóbrega, R. L. B. (2018). Impacts of
land use and land cover change on surface runoff, discharge and low flows: Evidence
from East Africa. Journal of Hydrology: Regional Studies, 15(May 2017), 49–67.
https://doi.org/10.1016/j.ejrh.2017.11.005

Haddeland, I., Heinke, J., Biemans, H., Eisner, S., Flörke, M., Hanasaki, N., Konzmann, M.,
Ludwig, F., Masaki, Y., Schewe, J., Stacke, T., Tessler, Z. D., Wada, Y., & Wisser, D.
(2014). Global water resources affected by human interventions and climate change.
Proceedings of the National Academy of Sciences of the United States of America,
111(9), 3251–3256. https://doi.org/10.1073/pnas.1222475110

Halder, S., Saha, S. K., Dirmeyer, P. A., Chase, T. N., & Goswami, B. N. (2016). Investigat-
ing the impact of land-use land-cover change on Indian summer monsoon daily rainfall
and temperature during 1951-2005 using a regional climate model. Hydrology and
Earth System Sciences, 20(5), 1765–1784. https://doi.org/10.5194/hess-20-1765-2016
 LXXXVIII

He, M., & Hogue, T. S. (2012). Integrating hydrologic modeling and land use projections for
evaluation of hydrologic response and regional water supply impacts in semi-arid envir-
onments. Environmental Earth Sciences, 65(6), 1671–1685.
https://doi.org/10.1007/s12665-011-1144-3

Hishe, S., Bewket, W., Nyssen, J., & Lyimo, J. (2020). Analysing past land use land cover
change and CA-Markov-based future modelling in the Middle Suluh Valley, Northern
Ethiopia. Geocarto International, 35(3), 225–255.
https://doi.org/10.1080/10106049.2018.1516241

Isabirye, M., Raju, D. V. ., Kitutu, M., Yemeline, V., Deckers, J., & J. Poesen Additional.
(2012). We are IntechOpen , the world ’ s leading publisher of Open Access books Built
by scientists , for scientists TOP 1 %. Intech, 13. http://dx.doi.org/10.1039/C7RA00172J
%0Ahttps://www.intechopen.com/books/advanced-biometric-technologies/liveness-
detection-in-biometrics%0Ahttp://dx.doi.org/10.1016/j.colsurfa.2011.12.014

Jama, A. A., & Mourad, K. A. (2019). Water services sustainability: Institutional arrange-
ments and shared responsibilities. Sustainability (Switzerland), 11(3), 1–15.
https://doi.org/10.3390/su11030916

Kahsay, K. D., Pingale, S. M., & Hatiye, S. D. (2018). Impact of climate change on ground-
water recharge and base flow in the sub-catchment of Tekeze basin, Ethiopia. Ground-
water for Sustainable Development, 6, 121–133.
https://doi.org/10.1016/J.GSD.2017.12.002

Kaushal, S. S., Gold, A. J., & Mayer, P. M. (2017). Land use, climate, and water resources-
global stages of interaction. Water (Switzerland), 9(10), 815.
https://doi.org/10.3390/w9100815

Kirby, J. M., Mainuddin, M., Mpelasoka, F., Ahmad, M. D., Palash, W., Quadir, M. E., Shah-
Newaz, S. M., & Hossain, M. M. (2016). The impact of climate change on regional wa-
ter balances in Bangladesh. Climatic Change, 135(3–4), 481–491.
https://doi.org/10.1007/s10584-016-1597-1

Luis, F., & Moncayo, G. (n.d.). No 主観的健康感を中心とした在宅高齢者における 健康関連指標に関する共分散構造分析

Title.

Mamat, Z., Halik, Ü., Keyimu, M., Keram, A., & Nurmamat, K. (2018). Variation of the
floodplain forest ecosystem service value in the lower reaches of Tarim River, China.
 LXXXIX

Land Degradation and Development, 29(1), 47–57. https://doi.org/10.1002/ldr.2835

Martin, K. L., Hwang, T., Vose, J. M., Coulston, J. W., Wear, D. N., Miles, B., & Band, L. E.
(2017). Watershed impacts of climate and land use changes depend on magnitude and
land use context. Ecohydrology, 10(7). https://doi.org/10.1002/eco.1870

Measho, S., Chen, B., Pellikka, P., Trisurat, Y., Guo, L., Sun, S., & Zhang, H. (2020). Land
Use/Land Cover Changes and Associated Impacts on Water Yield Availability and Vari-
ations in the Mereb-Gash River Basin in the Horn of Africa. Journal of Geophysical Re-
search: Biogeosciences, 125(7), 0–3. https://doi.org/10.1029/2020JG005632

Mello, K. de, Taniwaki, R. H., Paula, F. R. de, Valente, R. A., Randhir, T. O., Macedo, D. R.,
Leal, C. G., Rodrigues, C. B., & Hughes, R. M. (2020). Multiscale land use impacts on
water quality: Assessment, planning, and future perspectives in Brazil. Journal of Envir-
onmental Management, 270(March), 110879.
https://doi.org/10.1016/j.jenvman.2020.110879

Meresa, E., & Taye, G. (2019). Estimation of groundwater recharge using GIS-based WetS-
pass model for Birki watershed, the eastern zone of Tigray, Northern Ethiopia. Sustain-
able Water Resources Management, 5(4), 1555–1566. https://doi.org/10.1007/s40899-
018-0282-0

MNPD. (2011). Republic of Somaliland Somaliland Food & Water Security Strategy Somali-
land Vision 2030.

Muthusi, F., Mahamud, G., Abdalle, A., & Gadain, H. (2007). Rural Water Supply Assess-
ment. http://www.faoswalim.org/downloads/Pages from W-08 Rural Water Supply As-
sessment.pdf

Näschen, K., Diekkrüger, B., Evers, M., Höllermann, B., Steinbach, S., & Thonfeld, F.
(2019). The Impact of Land Use/Land Cover Change (LULCC) on Water Resources in a
Tropical Catchment in Tanzania under Different Climate Change Scenarios. Sustainab-
ility (Switzerland), 11(24). https://doi.org/10.3390/su11247083

Näschen, K., Diekkrüger, B., Leemhuis, C., Seregina, L. S., & van der Linden, R. (2019). Im-
pact of climate change on water resources in the Kilombero Catchment in Tanzania.
Water (Switzerland), 11(4). https://doi.org/10.3390/w11040859

Nobert, J., & Jeremiah, J. (2012). Send Orders of Reprints at reprints@benthamscience.org

78 The Open Hydrology Journal. 6, 78–87.
 XC

Nyatuame, M., Amekudzi, L. K., & Agodzo, S. K. (2020). Assessing the land use/land cover
and climate change impact on water balance on Tordzie watershed. Remote Sensing Ap-
plications: Society and Environment, 20(March), 100381.
https://doi.org/10.1016/j.rsase.2020.100381

Onyutha, C. (2020). Analyses of rainfall extremes in East Africa based on observations from
rain gauges and climate change simulations by CORDEX RCMs. Climate Dynamics,
54(11–12), 4841–4864. https://doi.org/10.1007/s00382-020-05264-9

Ouchi, T., Ibrahim, A. K., & Latham, G. V. (1982). Seismicity and crustal structure in the
Orozco fracture zone: Project ROSE Phase II. Journal of Geophysical Research,
87(B10), 8501–8507. https://doi.org/10.1029/JB087iB10p08501

Pan, T., Wu, S., & Liu, Y. (2015). Relative contributions of land use and climate change to
water supply variations over yellow river source area in Tibetan Plateau during the past
three decades. PLoS ONE, 10(4), 1–19. https://doi.org/10.1371/journal.pone.0123793

Petersen, G., and Gadain, H.M. (2012). Climate Change Impacts on Water Resources of
Somaliland and Puntland. Technical Report No W-21, FAO-SWALIM, Nairobi, Kenya.
December, 85.

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. Central European Journal of Geosciences, 2(2), 124–131.
https://doi.org/10.2478/v10085-010-0003-x

Sisto, N. P., Ramírez, A. I., Aguilar-Barajas, I., & Magaña-Rueda, V. (2016). Climate threats,
water supply vulnerability and the risk of a water crisis in the Monterrey Metropolitan
Area (Northeastern Mexico). Physics and Chemistry of the Earth, 91, 2–9.
https://doi.org/10.1016/j.pce.2015.08.015

Teerapunyapong, P., Phankamolsil, Y., Rittima, A., Talaluxmana, Y., & Tabucanon, A. S.
(2020). Estimation of groundwater recharge using wetspass model in the Phanom Thuan
Song Phi Nong-Bang Len Operation and maintenance projects, Thailand. 22nd Con-
gress of the International Association for Hydro-Environment Engineering and Re-
search-Asia Pacific Division, IAHR-APD 2020: “Creating Resilience to Water-Related
Challenges,” 1–8.

Teklebirhan, A., Dessie, N., & Tesfamichael, G. (2012). Groundwater Recharge, Evapotran-
 XCI

spiration and Surface Runoff Estimation Using WetSpass Modeling Method in Illala
Catchment, Northern Ethiopia. Momona Ethiopian Journal of Science, 4(2), 96.
https://doi.org/10.4314/mejs.v4i2.80119

Twisa, S., & Buchroithner, M. F. (2019). Seasonal and annual rainfall variability and their
impact on rural water supply services in the Wami River Basin, Tanzania. Water
(Switzerland), 11(10). https://doi.org/10.3390/w11102055

Twisa, S., Mwabumba, M., Kurian, M., & Buchroithner, M. F. (2020). Impact of land-use/
land-cover change on drinking water ecosystem services in Wami River Basin, Tan-
zania. Resources, 9(4), 1–18. https://doi.org/10.3390/RESOURCES9040037

Wang, K., Zhang, Q., Chen, Y. D., & Singh, V. P. (2015). Effects of land-use/cover change
on hydrological processes using a GIS/RS-based integrated hydrological model: case
study of the East River, China. Hydrological Sciences Journal, 60(10), 1724–1738.
https://doi.org/10.1080/02626667.2014.949723

Warku, F., Korme, T., Wedajo, G. K., & Nedow, D. (2022). Impacts of land use/cover
change and climate variability on groundwater recharge for upper Gibe watershed,
Ethiopia. Sustainable Water Resources Management, 8(1).
https://doi.org/10.1007/s40899-021-00588-8

Xu, Z. X., Chen, Y. N., & Li, J. Y. (2004). Impact of climate change on water resources in
the Tarim River basin. Water Resources Management, 18(5), 439–458.
https://doi.org/10.1023/B:WARM.0000049142.95583.98

Yang, C., Wu, G., Chen, J., Li, Q., Ding, K., Wang, G., & Zhang, C. (2019). Simulating and
forecasting spatio-temporal characteristic of land-use/cover change with numerical
model and remote sensing: a case study in Fuxian Lake Basin, China. European Journal
of Remote Sensing, 52(1), 374–384. https://doi.org/10.1080/22797254.2019.1611387

Yin, J., He, F., Jiu Xiong, Y., & Yu Qiu, G. (2017). Effects of land use/land cover and cli-
mate changes on surface runoff in a semi-humid and semi-arid transition zone in north-
west China. Hydrology and Earth System Sciences, 21(1), 183–196.
https://doi.org/10.5194/hess-21-183-2017

Yirgalem Enedeg. A. (2020). College of Development Studies (Cds) Water Resource Man-
agement Program Groundwater Recharge Estimation Using Gis Based Wetspass
Model. a Case of Jewha Watershed, Middle Awash River Basin, Ethiopia.
XCII