MSc_GroundwaterModellingOptimization

Groundwater Modelling and
Optimization of Irrigation Water Use

Efficiency to Sustain Irrigation in
Kobo Valley, Ethiopia
Getahun Wendmkun Adane
MSc Thesis Number:WSE-HELWD-14.09
April 2014
Groundwater Modeling and optimization of Irrigation water
use efficiency to sustain irrigation in Kobo Valley, Ethiopia
Master of Science Thesis

by
Getahun Wendmkun Adane
Supervisors
Profes. Charlotte De Fraiture (PhD, MSc)
Mentors
Abraham Mehari Haile (PhD, MSc)
Examination committee
Profes. Charlotte De Fraiture ,PhD,MSc (UNESCO-IHE)
Abraham Mehari Haile,PhD,MSc (UNESCO-IHE)
F.Van Steenbergen, PhD,MSc (Meta Meta's Hertogenbosch)
This research is done for the partial fulfilment of requirements for the Master of Science degree at the
UNESCO-IHE Institute for Water Education, Delft, the Netherlands
Delft
April 2014
©2014by Getahun Wendmkun Adane. All rights reserved. No part of this publication or the information
contained herein may be reproduced, stored in a retrieval system, or transmitted in any form or by any
means, electronic, mechanical, by photocopying, recording or otherwise, without the prior permission of
the author. Although the author and UNESCO-IHE Institute for Water Education have made every effort to
ensure that the information in this thesis was correct at press time, the author and UNESCO-IHE do not
assume and hereby disclaim any liability to any party for any loss, damage, or disruption caused by errors
or omissions, whether such errors or omissions result from negligence, accident, or any other cause.
Abstract
Agriculture, Industry and domestic activities use enormous of water which results in the over-pumping and
leading to continuous decline of groundwater level. Farmers often fail to satisfy the required soil moisture
conditions for growing crops due to erratic low rainfall distribution. Poor on-farm water management
practices resulted in excessive use and this leads to high energy cost. This study mainly focuses on
groundwater modelling of kobo valley so as to predict the current and future groundwater level under
different hydrologic and pumping scenarios. Aqua crop model was also used as a tool for assessing crop
and water productivity under irrigated agriculture - the main production system in Kobo valley, the main
study area of this research, located in Northern Ethiopia is enclosed by high mountain ranges on the edge of
Afar Rift system and measures about 1200km2 area. It has two main sub-basins, Hormat-Golina and Waja-
Golesha that are characterized by high abundant resources with respect to groundwater, fertile land and
livestock potential.
The groundwater flow system in unconsolidated deposit of Kobo valley was modelled using MODFLOW
(McDonald and Harbaugh, 1988). The model was run for steady-state conditions in unconfined and
confined aquifer. The grid cell size of the model was taken 300 x 400m and contains two layers. Model
area and the layer top elevation were delineated by the ASTER DEM processing and use of topographic
maps. The hydraulic conductivity values were determined from pumping test data analysis and literature
review for the alluvial sediment aquifer and the fractured volcanic aquifer respectively. Recharge was
estimated from water balance and Darcy's approach method and has a value of 95MCM and 83 MCM
respectively. The model was calibrated using observed hydraulic heads from 35 wells from Hormat-Golina
sub-basin using trial and error method and resulted in a Root Mean Square Error (RMSE) value of about
7m, which is considered to be good and an indicator of reliable model results. The model simulated water
budget showed that the valley receives a total recharge of 118MCM /year. The steady state model with
pumping scenarios:- current scenario (11 wells are operated simultaneously), scenario-one (35 wells are
operated simultaneously) & scenario-two (70 wells are operated simultaneously) indicated respective
groundwater abstraction of 5192 m3d-1, 27878 m3d-1 and 55825 m3d-1 .This resulted in an average
groundwater level decline (at the pumping well) of about 7m, 14m and 32m respectively. A maximum of
35 wells simultaneous operation is recommended as this will maintain a 20% stabilized drawdown, which
results in a balance between abstraction and recharge.
The current irrigation system is operating at an efficiency of 55% as the actual amount of water applied is
about 730 mm while the net requirement obtained from Aqua Crop is 404 mm. Under groundwater based
irrigation, the efficiency could be improved to 80% through the use of piped conveyance canals and good
field water distribution that reduces runoff and deep percolation losses. Under the current actual harvest by
the farmers of 1.8 ton/ha, and assuming only the fuel cost, the difference in net income between operating
at 55% and 80% efficiency is 109 Euros/ha. If, however, the yield could be improved to 4.7 ton/ha (this
yield is obtained by farmers in others regions of Ethiopia), the difference in net income is nearly 1000
Euros/ha.
Key words: Kobo valley, Groundwater modelling, Recharge, Aqua Crop model, Irrigation water use
efficiency
i
Acknowledgements
I would like to express my deepest gratitude to the Government of Netherlands through
Netherland Fellowship programme (NFP) and IFAD for grating me the chance to study
Master of Science degree at UNESCO-IHE.
I greatly thank my mentor Abraham Mehari Haile (PHD) for his excellent guidance and
encouragements. I highly appreciate his constructive advice and fruitful comments. Special
thanks go to my supervisor Prof. Charlotte De Frature for her special comments as well as
Suryadi (PHD) for his technical support in Groundwater modelling.
I gratefully acknowledge all Amhara Water Work Construction Workers especially kobo
project. There is no proper word for Mr.Alelegn Yesuf; Kobo project manager, Mr.Ermiya,
Mr. Amhayes Zewudu, Mr. Mohammed Ebirahim, Mr.Jemal, Mr.Birhanu, Mr.Tilahun
Mekonen and Mr.Alemu for their special help during field work activities. I thank you all. I
also thank Mr.Nafiyad and Mr.Wassiye Abate for their help in Groundwater modelling and
data collection in Addis Ababa.
My deepest gratitude must go to my sister, Genzebua Fentaw for her encouragement and
support in my academic life.
Finally, I offer my deepest gratitude and appreciations to my family and friends for their
support and love throughout my life.
iii
Table of Contents
Abstract i
Acknowledgements iii
List of Figures vii
List of Tables viii
Abbreviations ix
1. Introduction 1
1.1. Back ground 1
1.2. Problem statement 2
1.3. Objectives of the Research 2
1.4. Research questions 2
1.5. Possible scenarios 2
1.6. Methodology 2
1.7. Outline of the thesis. 4
2. 2.Literature Review 6
2.1. Review of previous study 6
2.2. Groundwater Modelling 7
3. General description of the study area 9

3.1. Location 9
3.2. Drainage 10
3.3. Climate 10
3.4. Land use and cover 10
3.5. Geology and Hydrogeology 10
3.5.1. Geology 10
3.5.2. Hydrogeology 12
4. Analysis and Model input data preparation 13

4.1. Meteorological data Analysis 13
4.1.1. Rainfall 13
4.1.2. Temperature 15
4.1.3. Wind speed 16
4.1.4. Solar Radiation 16
4.1.5. Evapotranspiration (ETo) 16
4.2. Groundwater recharge estimation 18
4.2.1. Water Balance Method 18
4.2.2. Darcy Approach 19
4.3. Pump test data Analysis. 19
5. Groundwater Modelling of Kobo Valley 21

5.1. Introduction 21
v
5.1.1. The modelling Process 21
5.2. Conceptual Model 22
5.2.1. Boundary Conditions 24
5.2.2. Stratigraphic Units 24
5.2.3. Sources and sinks of the Model area 24
5.2.4. The Model area 28
5.2.5. Aquifer Geometry 28
5.3. Numerical Model 29
5.3.1. Data input for the Model 29
5.3.2. Model execution and Calibration 30
5.3.3. Calibration target and Uncertainty 30
5.3.4. Trial and Error Calibration 30
5.3.5. Evaluation of calibration 31
6. Result and Discussion 33

6.1. Results 33
6.1.1. Water Budget of the model domain 33
6.1.2. Sensitivity Analysis 33
6.1.3. Pumping Scenario Analysis 34
6.1.4. Estimation of Radius of influence and well interference 38
6.1.5. Groundwater Reserve in Kobo Valley 40
6.1.6. Estimation of Allowable Exploitation of Groundwater 41
6.1.7. Irrigation Water use efficiency 42
6.1.8. Aqua crop Model to determine seasonal water requirement 42
6.1.9. Pumping Cost. 45
6.2. Discussions 47
6.2.1. Water Balance Method and Darcy's Approach 47
6.2.2. Groundwater modelling 47
6.2.3. Water budget of the model domain 48
6.2.4. Pumping Scenario 48
6.2.5. Radius of Influence and well interference 48
6.2.6. Groundwater reserve and allowable exploitation 48
6.2.7. Aqua Crop model 48
6.2.8. Model Limitation 49
6.2.9. Groundwater model and Aqua crop model 49
7. Conclusion and Recommendation 50

7.1. Conclusion 50
7.2. Recommendation 51
Appendix A Metrological Data 54
Appendix B Pump Test Data 61
Appendix C Error evaluation 65
Appendix D Groundwater Abstraction 66
Appendix E Radius of influence and drawdown 72
vi
List of Figures
Figure 1.1 Groundwater level measuring using deep meter at PK8 borehole............................................ 3
Figure 1.2 Taking bore location & elevation using GPS at K5 on left and river data around Golina
river outlet on the right. .......................................................................................................... 3
Figure 1.3 Flow chart of the methodology................................................................................................. 5
Figure 3.1 Location map of the study area................................................................................................. 9
Figure 3.2 Geology and structural map of Kobo-Girana Valley: source geological map of Ethiopia,
1996 ........................................................................................................................................ 11
Figure 4.1 Mean monthly rainfall of the stations (1996- 2010) ............................................................... 14
Figure 4.2 Mean annual rainfalls of the stations ...................................................................................... 14
Figure 4.3 Kobo station annual rainfall (1996-2010)............................................................................... 15
Figure 4.4 Mean Monthly Maximum Temperatures of Kobo & Korem Station ..................................... 15
Figure 4.5 Mean monthly minimum temperature (oc) of Kobo and Korem ............................................ 16
Figure 4.6 Mean monthly Wind speed of kobo ....................................................................................... 16
Figure 4.7 Precipitation (P), potential evapotranspiration (ETO) and average monthly Temperature
(Tavg) for Kobo station (1996 to2010) .................................................................................... 17
Figure 5.1 Steps in modelling protocol (after Anderson and Woessner, 1992). ...................................... 22
Figure 5.2 Schematic diagram illustrating the simplified conceptual model. .......................................... 23
Figure 5.3 Simplified Conceptual model as taken from Global Mapper. ................................................ 23
Figure 5.4 Golina river near the outlet to Afar. ....................................................................................... 25
Figure 5.5 Plan View of the Entire Modelled Area ................................................................................. 28
Figure 5.6 Cross-Sectional View of the Model Area along Line North to South of Fig. 5-5 .................. 29
Figure 5.7 Trial and error calibration procedures (adapted from Anderson and Woessner, 1992) .......... 30
Figure 5.8 Graphical comparison between the observed and simulated heads ........................................ 31
Figure 5.9 The scatter diagram showing the comparison of measured and simulated heads ................... 32
Figure 6.1 The model simulated groundwater heads for pumping scenario-one ..................................... 35
Figure 6.2 The model simulated groundwater heads for pumping scenario-two ..................................... 36
Figure 6.3 Comparison between the observed and simulated heads of different scenarios ..................... 37
Figure 6.4 The scatter plots of simulated versus observed heads for different pumping scenarios ......... 38
Figure 6.5 well interference in Hormat-Golina sub-basin ....................................................................... 39
Figure 6.6 Net irrigation requirement from Aqua Crop model. ............................................................... 43
Figure 6.7 Soil water balance from Aqua Crop model. ........................................................................... 44
Figure 6.8 Soil water balance in generation of irrigation schedules. ....................................................... 45
vii
List of Tables
Table 4.1 Mean monthly rainfall distribution mm (1996-2010) ............................................................ 14
Table 4.2 Computed ETO of Kobo station using Penman-Monteith equation. ...................................... 17
Table 4.3 Darcy approach recharges estimation of Hormat-Golina and Waja-Golesha sub-basins ....... 19
Table 5.1 The estimated amount of water abstracted from boreholes per annual .................................. 27
Table 5.2 Error summary for the calibrated model................................................................................. 32
Table 6.1 Water budget of the entire model domain in m3d-1 ................................................................. 33
Table 6.2 Estimated groundwater abstraction rate and the average decline in groundwater level for
different pumping scenarios ................................................................................................... 34
Table 6.3 Simulated water budget of Kobo-valley for scenario-one in m3d-1 ........................................ 37
Table 6.4 Simulated water budget of Kobo-valley for scenario-two in m3d-1 ........................................ 37
Table 6.5 Well interference computation in Hormat-Golina sub-basin .................................................. 40
Table 6.6 Groundwater reserve in Kobo valley ...................................................................................... 41
Table 6.7 Allowable abstraction for 60% drawdown of the saturated thickness in Kobo Valley .......... 41
Table 6.8 Estimation of the Available Groundwater Potential and Number of Wells in Kobo Valley .. 41
Table 6.9 Cost- Analysis of pumping in well ......................................................................................... 45
Table 6.10 Pump detail ............................................................................................................................. 46
Table 6.11 Fuel cost-benefit analysis of Maize production ..................................................................... 46
Table 6.12 Fuel cost-benefit analysis of optimized maize production ..................................................... 47
Table A.1 Monthly Rainfall (mm) at Kobo station ................................................................................. 54
Table A.2 Monthly Rainfall (mm) at Alamata Station ............................................................................ 54
Table A.3 Monthly Rainfall (mm) at Zoble Station ................................................................................ 55
Table A.4 Monthly Rainfall (mm) at Korem station ............................................................................... 55
Table A.5 Monthly Rainfall (mm) at Lalibela station ............................................................................. 56
Table A.6 Monthly Tmax (0c) at Kobo station ........................................................................................ 56
Table A.7 Monthly Tmax. (o c) at Alamata Station ................................................................................. 57
Table A.8 Monthly Tmax. (o c) at Korem station ..................................................................................... 57
Table A.9 Monthly Tmin. (0c) at Kobo station ....................................................................................... 58
Table A.10 Monthly Tmini. (o c) at Alamata station ................................................................................. 58
Table A.11 Monthly Tmini. (oc) at Korem Station .................................................................................... 59
Table A.12 Monthly average wind speed (m/s) at Kobo Station .............................................................. 59
Table A.13 Monthly average sunshine (hour) at Kobo station .................................................................. 60
Table B.1 Hydraulic parameters of the alluvial sediment aquifer .......................................................... 61
Table C.1 Comparison of the Observed and Simulated Heads and Error Calculation ............................ 65
Table D.1 The estimated amount of abstracted water used in scenario-one ............................................ 66
Table D.2 The estimated amount of water abstracted from 70 boreholes in scenario-two...................... 68
Table D.3 Estimating the decline groundwater level for scenario-one and scenario-two ....................... 71
Table E.1 Radius of influence and drawdown ........................................................................................ 72
viii
Abbreviations
ASTER Advanced Space borne Thermal Emission and Reflection Radiometer
Co-SAERAR Commission for Sustainable Agriculture and Environmental Rehabilitation in
Amhara Region
DD Drawdown
DEM Digital Elevation Model
DWL Dynamic Water Level
EIGS Ethiopian Institute of Geological Survey
ETo Potential Evapotranspiration
GES Geo-Engineering Service
GIS Geographic Information System
GPS Global Positioning System
Hm Measured head
Hs Simulated head
KGVDP Kobo-Girana Valley Development Project
m.a.s.l Meter above sea level
MAE Mean absolute error
ME Mean error
MCM Million Cubic Meters
RMS Root mean square error
RVDP Raya Valley Development Project
SCS Soil Conservation Service
SRTM Shuttle Radar Topographic Mission
SWL Static Water Level
ix
List of Symbols
d day
ETo Potential Evapotranspiration (mm)
ETa Actual Evapotranspiration (mm)
ha hectare
I Hydraulic gradient
k Hydraulic conductivity (m/d)
Kav Average Hydraulic conductivity (m/d)
l Liter
m Meter
mm Millimeter
o
c Degree centigrade
p Annual precipitation
Q Discharge (m3/d)
R Recharge to groundwater
S Soil moisture content
Sr Annual surface runoff
Sy Specific Yield
T Transmissivity (m2/d)
Tav Average Transmissivity (m2/d)
x
CHAPTER 1
Introduction
1.1. Back ground

Groundwater is one of the most valuable natural resources, which supports human health, economic
development, and ecological diversity. Because of its several inherent qualities (e.g., consistent
temperature, widespread and continuous availability, excellent natural quality, limited vulnerability, low
development cost, drought reliability), it has become an important and dependable source of water supplies
in all climatic regions including both urban and rural areas of developed and developing countries (Todd,
2005). Of the 37 Mkm3 of freshwater estimated to be present on the earth, about 22% exist as groundwater,
which constitutes about 97% of all liquid freshwater potentially available for human use (Foster, 1998).
Approaches of sustainable development and integrated groundwater resources management must be
developed and implemented to guarantee the right of use of the limited water resources for future
generation. Groundwater resource management of an aquifer system involves developing a quantitative
understanding of the flow processes that operate within the aquifer. Three main futures must be considered:
how water enters the aquifer system; how water passes through the aquifer system and how water leaves
the aquifer system. The best tool available to help groundwater hydrologists to formulate technically-sound
ground water resources management is usually a ground water model. Groundwater models have been used
as interpretation tools for investigating groundwater system dynamics, assessment tools for evaluating
recharge and quantifying sustainable yield (Anderson and Woessner, 1992).
Irrigation water increases crop yields and quality in semi-arid areas like kobo, Northern Part of Ethiopia. It
is essential especially during periods of erratic rain fall and drought. Since there is a declining groundwater
resource, the main sources for irrigation in the study area. The irrigation water efficiency has to be
increased as much as possible. The right amount of irrigation water has to be reached the right place at the
right time in order to have effective irrigation.
Water efficiency of irrigation can be improved by making the right decision regarding to crop selection,
irrigation scheduling and irrigation methods. The actual irrigation system capacity, the crop water demand
is computed by Aqua crop model. This helps to reduce the amount of irrigation water pumped and avoid
excessive energy use. Quantification of the actual irrigation water demand also provides critical
information to the farmers, local groundwater conservation, Irrigation and regional water planning groups.
The study area mainly focuses on the two main sub-basin of Kobo valley named as Hormat-Golina and
Waja-Golesha located in Northern Ethiopia. These sub-basins have a high groundwater potential and also
known by semi-arid climate. Because of low and variable annual rainfall, groundwater irrigation is used to
alleviate draught problem in the area but there is low attention given to groundwater resource management.
Groundwater Modeling and optimization of Irrigation water use efficiency to sustain irrigation in Kobo
Valley, Ethiopia 1
As observed from the field visit, history of bore holes are not recorded properly, no surface water gauging
stations over the area and along the main rivers. However, for sustainability of groundwater resource, there
should be safe abstraction and proper management.
In this thesis, numerical groundwater modelling of kobo valley using MODFLOW-version 8 for steady
state condition has given higher priority in addition to optimization of irrigation water use efficiency in the
area. Therefore clear understanding of the response of the aquifer is crucial for better management of
groundwater resources.
1.2. Problem statement

Even though the kobo valley is rich in fertile soil for agricultural production, the lowland area has low
rainfall which is insufficient for higher agricultural production. Due to erratic rainfall distribution in the
area, the farmers often fail to satisfy the required soil moisture conditions for growing crops. Consequently,
the area was affected by drought in a number of times. Now days, using ground water is growing
continuously and increasingly being used as a main source of water for irrigation. However, there is no
control or management of groundwater resources, hence developing ground water model is essential.
The kobo valley has a potential of over 10,000ha of land to irrigate however, the current irrigated land is
less than 1000ha.Due to high energy cost for pumping water from wells and problem of on-farm water
management, there is a need to optimize water use efficiency and maximize crop yield. Aqua crop program
me is useful to determine the actual crop water requirement.
1.3. Objectives of the Research

• To quantify the recharge and abstraction of groundwater.
• To analyze the impacts or drawdown of wells under different well operation scenarios.
• To recommend on-farm water management for improving irrigation water use efficiency and
minimize costs.
1.4. Research questions

• What is the recharge and abstraction of groundwater in the study area?
• What are the impacts or drawdown of wells under different well operation scenarios?
• How efficient is the current irrigation application system? How and to what extent can this be
improved? And what will be the implication on reduction of pumping costs and the net income?
1.5. Possible scenarios

Analysis of rainfall data (at least 15 year) to assess the variations in recharge and abstractions of
groundwater.
The impacts or drawdown in operating all the wells at its maximum discharge especially in
concentrated wells at a time.
The impacts or drawdown in operating wells by dividing in to two groups at different time.
1.6. Methodology
Appropriate methods and materials should be used in order to achieve the objectives stated above (section
1.3).The main activities can be classified in to pre-field, field and post-field works. All the three stages
were illustrated in Figure 1.3.
MSc thesis, G.W. Adane 2

Pre-filed work
Data have been collected from different offices (Ethiopia National Meteorological Agency, Ministry of
Water resource).Equipments like Groundwater level measuring device (deep meter) and GPS were taken
from Amhara Water Work Construction Enterprise office and Kobo town water supply office respectively.
Soil and crop data were taken from Kobo Girana Valley Development Project office. Delineation of the
study area and literature review of groundwater modelling and irrigation water use efficiency was included
in this stage.
Field work
This stage was conducted in order to get primary data from the study area and secondary data from
different sources. Measuring ground water level measuring using deep meter, taking reading of borehole
locations and elevation using GPS, river out let and longitudinal data using Total stations, determination of
physical boundaries were the main primary data conducted in the field.(see Figure 1.1 and Figure 1.2). The
secondary data such as borehole logs, pumping test data and hydrogeology feasibility report and well
completion report were collected from Kobo-Girana Valley Development Project Office.
Figure 1.1 Groundwater level measuring using deep meter at PK8 borehole
Figure 1.2 Taking bore location & elevation using GPS at K5 on left and river data around Golina river
outlet on the right.
Valley, Ethiopia 3
Post -field work
Data processing and analysis was the primary activity in this stage. Primary and secondary data was
processed and analyzed in order to prepare database and conceptual model. The conceptual model was used
for the core input for the modelling process.
Geographic Information System (GIS) was used to enter store, retrieve, and process and display spatial
information in the form of maps or images. Advanced Space borne Thermal Emission and Reflection
Radiometer (ASTER) DEM with 90m spatial resolution was processed to delineate the catchment.
1.7. Outline of the thesis.

The content of the thesis is briefly outlined as follows;
Chapter one: Describes the introduction.
Chapter two: Discusses the review of previous studies.
Chapter three: Describes the study area.
Chapter four: Data processing and analysis.
Chapter five: Depicts the numerical modelling and deals with the pumping scenario analysis.
Chapter six: Deals Optimization of irrigation water use efficiency and Aqua crop Model.
Chapter seven: Results and discussion.
Chapter eight: Conclusion and Recommendation.

Figure 1.3 Flow chart of the methodology
Valley, Ethiopia 5
CHAPTER 2
2.Literature Review
2.1. Review of previous study

The kobo valley has been known as one of the most drought prone regions in Ethiopia. According to the
historical analysis of rain fall, the major part of the valley once every decade is a drought year on the
average. The groundwater potential for irrigation at the valley has been the focus since long due to the
expected high groundwater potential. Following this, a number of studies were conducted on the
physiographic setup, geology, tectonics and groundwater potential at different levels and localities in Kobo
valley. The study area is one part of Kobo-Girana valley development project that mainly focus sustainable
development to alleviate the drought and famine from the area.
Among the major studies conducted in Kobo-Girana valley are the investigations of groundwater potential
for multipurpose by Co-SAERAR from 1996 to 1999 and review and appraisal of Hydro geological studies
by Geo-Engineering Service in 2002/3.
CO-SAERAR study
The Commission for Sustainable Agriculture and Environmental Rehabilitation in Amhara Region (Co-
SAERAR) studied the Kobo-Girana valley for its groundwater potential and then by implement irrigation
through groundwater source (Co-SAERAR 1999).The study made preliminary estimation of annual
recharge capacity of the Kobo, Alawha, Chireti and Gelana sub basins to be 119, 9, 15 &
27MCM, respectively. The deeply weathered and fractured zones of the volcanic rocks frequently exceed
100meters. Depth to groundwater varies from less than 20meters in the sediment deposit in the western part
and along the river courses to over 100 meters in the volcanic rocks. The predominant groundwater flow
direction is from west to east coinciding with topographic gradient. The hydraulic gradient of the Kobo
basin is 0.012 (Co-SAERAR 1999). The groundwater recharge of the Kobo-Robit basin was estimated
from three different methods. These are surface water balance method using SCS model, groundwater level
fluctuation rate, and Darcy’s approach; and the results were 59.3 MCM, 64.8 MCM and 49.82 MCM
respectively. The surface water balance method, despite its crude and lumped inputs was found relatively to
be comprehensive and practical to estimate the valley’s recharge (Co-SAERAR 1999).
EIGS study
Hydro geological and environmental isotope investigations have been done by Sileshi Mamo from the
Geological Survey of Ethiopia (2007), the total dynamic groundwater in the graben sediments estimated to
amount 68.9MCM in Kobo valley. Recharge estimation using Chloride (Cl) mass balance has given
recharge rates of 60.07 mm/year for the western plateau and 52.00 mm/year for the graben fill sediments

and adjoining escarpment. A total of 192.78 MCM/yr dynamic groundwater resource is estimated for the
graben sedimentary aquifer; 123.89 and 68.9 MCM/yr for Raya and Kobo valleys, respectively (Sileshi
Mamo, 2007).
2.2. Groundwater Modelling

A groundwater model may be defined as a simplified version of the real groundwater system that
approximately simulates the excitation- response relations of the groundwater system. The real system is
very complicated and difficult to use it directly for the purpose of planning and making management
decisions. The simplification is introduced in the form of a set of assumptions that express our
understanding of the nature of the system and its behaviour. These assumptions will tend to smooth out the
effect of various heterogeneities. Because the model is a simplified version of the real system, there exists
no unique model for a given groundwater system (Bear, 1979).
There are several ways to classify groundwater flow models, models can be transient or steady state and
one, two, or three spatial dimensions. Steady state flow occurs when at any point in a flow field the
magnitude and direction of the flow are constant with time (Anderson and Woessner, 1992).
Groundwater models are an attempt to represent the essential features of the actual groundwater system by
means of a mathematical counterpart (Todd & Mays, 2005). These models have a capacity to test and
quantify the consequences of various errors and related model-based forecasts. Groundwater models
according to Todd are physically based mathematical models derived from Darcy’s law and the law of
conservation of mass. Various established solution techniques based upon either finite difference or finite
element approximations, or a combination of both, are available for solving the governing equations of the
model. The accuracy of the solutions (model predictions) is dependent upon the reliability of the estimated
model parameters and the accuracy of the prescribed boundary conditions.
Computer program or code solves a set of algebraic equations generated by approximating the partial
differential equations (governing equation, boundary conditions, and initial conditions) that form the
mathematical mode (Anderson and Woessner, 1992). With the introduction of computers and their
application in the solution of numerical models, physical models and analogy have become laid off as tools
for predicting future groundwater regimes. The selection of the appropriate model to be used in any
particular case depends on the objective or objectives of the investigation and the available resources. The
later include time, budget, skilled manpower, high capacity computers and codes (Bear, 1979).
The finite difference method requires a rectangular element shaped discretization of the aquifer and the
finite element method consists of a triangular discretization. Discretization is the process of subdividing the
continuous hydro geologic units into discrete segments or cells. Finite element method is easy to define the
boundaries of irregularly shaped aquifers and to ensure that node points coincide with monitoring wells or
varies types of geographic features. The mathematical basis for finite element methods is more complex
than for the finite difference method (Todd & Mays, 2005).
Selecting the appropriate conceptual model for a given problem is one of the most important steps in
modelling process. The key data requirements in the process of conceptualization include data about hydro-
stratigraphic units, surface water bodies, physical and hydraulic boundaries, recharge and discharge zones.
The most common numerical methods to solve flow problems are finite differences and finite elements.
Finite-difference grids are easy to understand and require less input data than finite element grids
(Anderson and Woessner, 1992). The finite difference method, as applied in the computer code
MODFLOW, was used in this study. The code is based on the physical theory of groundwater movement
Darcy’s law and the continuity equation. The program supports seven additional packages, which are
integrated with the original MODFLOW (Chiang and Kinzelbach, 2001).
Once the conceptual model is translated into a numerical model in the form of governing equations, with
associated boundary and initial conditions, a solution can be obtained by transferring it into a numerical
model and writing a computer program (code) for solving it. This includes, design of grid, setting boundary
Valley, Ethiopia 7
and initial conditions and preliminary selection of values for aquifer parameters. The input parameters
include model grid size, layer elevations, boundary conditions, hydraulic conductivity, recharge, and
additional model input for steady state condition. Model calibration consists of changing values of model
input parameters in an attempt to match field conditions within some acceptable criteria (Anderson and
Woessner, 1992). Sensitivity analysis is useful in determining which parameter or parameters most
influence the model results. These parameters will be emphasized in the future data collection attempting to
improve model accuracy.

CHAPTER 3
General description of the study area
3.1. Location
The studied area is found in the North Wollo administration zone of Amhara National Regional state of
Ethiopia. It has a geographical zone of 11°56’ to 120°18’N and 39°23’ to 39°47’E.The kobo valley is
one part of the Kobo Girana Valley Development Project. The valley is surrounded by western high lands
in the West, Zoble Mountain in the East, Raya valley in the North and Volcanic ridges in the South. The
plain area is known by flat topography up to 1500m altitude and the mountain rises dramatically from
1500m to greater than 3000m.
Figure 3.1 Location map of the study area.
Valley, Ethiopia 9
3.2. Drainage
The major drainage system is associated with valley plains. The rivers in the valley originate from the
western mountains. Golina, Hormat, Kelkelit and Dikala Rivers drain in to the valley. The valley can be
classified in to three major sub-basins namely, Waja-Golesha, Hormat- Golina and Kobo-Arequaite-Gerbi
sub -basins.
The Waja-Golesha sub-basin is drained by Gobu and Waja streams which disappear in Waja plain. There
is one intermittent stream named Dikala stream which starts from the western ridge of Kobo Town and
flows towards the Garalencha Mendefera before it disappears in the Chobe-Golesha plain.
The Hormat-Golina sub-basin constitutes the drainage systems of Hormat, Golina, and Kelkelit. Most of
the flows of the rivers of this sub-basin too are lost in the plain before reaching their outlets through Golina
River. Hormat, Golina and Kelkeli are perennial rivers in general. However, during dry season, Hormat and
Kelkeli lose their discharge in the plain before joining Golina that ultimately discharge through the Golina
gorge to the Afar Depression.
The kobo-Arequaite-Gerbi sub-basin is a closed sub-basin that some intermittent streams are flowing from
Zobul ridge, Gedemyu and Mendefra hills into the Arequaite-Gerbi plain-depression. No surface drainage
out let is observable from this depression. Wet Season Lake at Gerbi disappears in the dry season by
evaporation.
There is high drainage density in the western highlands and, low both in the valley floor and eastern
highlands. All rivers and streams carry large volumes of sediments from the mountains in the rainy seasons
and deposit on the valley plain.
3.3. Climate
The main feature of rainfall in the area is seasonal, poor distribution and variable from year to year.
Rainfall distribution over the valley is bimodal, followed by the long and short rainy season that occurs in
July-October and February -April respectively. The rest of the months are generally dry. The mean monthly
temperatures in kobo valley vary from in 12oC in December to about 35oC in June. This is shown in
Appendix A,Table A.6 and Table A.9.The studied area also has a monthly maximum and minimum
sunshine of 9 hour in November and 5 hour in July respectively.
3.4. Land use and cover

Land use is essential in the hydrological and groundwater studies since it is a prominent factor influencing
the recharge. From field observation and Arial photos, the land use was identified as agricultural area,
woodland, forest, and bare land. The first two were the dominant land uses.
3.5. Geology and Hydrogeology

3.5.1. Geology
The geology of north and central Ethiopia, which also includes the current study area, is dominated by
Tertiary volcanic strata underlain by Mesozoic sedimentary rocks. The dominant outcrops on the mountains
are fissured basalts with silica varieties. The first geologist in Ethiopia, Branford, 1869 classified the
northern Ethiopia volcanic into Ashange and Magdala group. Two Volcanic successions occurred in the
period of Palaeocene to Miocene, recognized as the Ashangi and Magdala groups. (KGVDP).
The geological structure of the area is controlled by tectonic events that led to the development of the Rift
System. These events are characterized by tensional movements which gave rise to fissure volcanism
followed by block-faulting and tilting to form the escarpment zone including marginal grabens. These

marginal grabens are narrow elongated depressions bounded on both sides by normal faults facing each
other. The eastern and western ridges of the Waja Golesha and Hormat Golina bounding the plain area are
characterized by a system of opposite dipping faults oriented parallel to the plateau escarpments.
The Waja-Golesha-Hormat-Golina plain in the study area has a length of about 33 kilometres and a width
of about 10 to 17 kilometres. The widest basin in the study area reaches to 17 km at Waja-Adis Kigny and
the narrowest corridor is about 10 km at Kobo-Gerbi stretch. The eastern margin of this graben is a steep
slope fault downthrown to the west about 800 meters as measured from the foot of the hill to the top on the
road Kobo-Zobul all weather roads (Co-SAERAR 1999).
Figure 3.2 Geology and structural map of Kobo-Girana Valley: source geological map of Ethiopia, 1996
Valley, Ethiopia 11
3.5.2. Hydrogeology
The regional hydrogological set up of the study area and its surrounding can be summarized as localized
graben filling unconsolidated sediment composed of clay, silt, sand, gravel, boulders and pebbles above
the Ashangi group volcanic which are intern underlain by Mesozoic sedimentary rocks. The Trepan
volcanic and the underlying formation are strongly affected by faulting and displacements prior to the
deposition of the quaternary sediments in the grabens and troughs. Regionally the Ashangi volcanic are the
most extensive formations above the sedimentary basin.
With regards to groundwater movement and storage, the unconsolidated sediments in the grabens and the
sedimentary rock beneath the Ashangi Group volcanic have high potential. Although localized in
occurrence, the unconsolidated sediments are relatively thick with good hydraulic permeability and these
sediments get recharge from the weathered part of Ashangi volcanic surrounding the grabens. The
thickness of the sediment deposit increases as one move from west to east in the valley. Geological logs of
the boreholes and the geophysical surveying results show that the thickness of the sediments of the sub-
basins vary from about 300 m in the east to less than 50 m near the mountains to the west. The lateral and
vertical variations in grain composition of the sediment are common everywhere in the valley attributed to
mixing of the proximal and distal deposits following flood and depositional cycles. As a result, the
unconsolidated sediment has heterogeneous aquifer both vertically and horizontally.
Therefore, the unconsolidated sediment is recharged mainly as subsurface inflow from the locally
weathered and fractured zone of the volcanic rock of the mountains surrounding the plain area. Major
groundwater out flow is at the Selenwuha and Golina streams out let to Danakil Depression and Mile-
Awash, respectively, in Afar Region. The outlets have perennial flows from groundwater discharge.

CHAPTER 4
Analysis and Model input data

preparation
4.1. Meteorological data Analysis

4.1.1. Rainfall
Rainfall records of five stations were selected to describe the rainfall regime of the studied valley. Lalibela
station is located near on the western high lands part of the kobo valley and the other stations are aligned
linearly in North-South direction of the studied area. All stations are located outside the study area except
kobo station. Fifteen year rainfall records (1996 -2010) were collected for the analysis purpose of this study
in order to have adequate data (Table A.1). The average of monthly rainfall data of these fifteen years was
taken for analysis (Table 4.1). The mean annual precipitation has a bimodal distribution with most of the
rainfall occurring during the months July to September while there is a short rainy season from March to
April. The other months are generally dry (Figure 4.1).Lalibela and Korem stations have a maximum mean
monthly rainfall of 277mm in July and 290mm in August respectively. Similarly, a maximum value of
mean annual rainfall of 968mm and a minimum value of 674mm were recorded at Korem and Kobo
stations respectively.
Kobo annual rainfall distribution for 15 year showed that there is minimum rainfall of about 200mm in
2001 (see Figure 4.3).This was the turning point for the government officials to develop intensively Kobo
Girana Valley Development Project since the farmers fail to grow crops at that year. The maximum rainfall
about 1216mm was recorded in 2002 especially in the month of March and April for a value of about
499mm and 494mm respectively.
Since Kobo, Alamata and Zoble stations are closely located on the floor of Raya-Kobo valley, arithmetic
mean method was used to determine the areal depth of precipitation for the study area. The weighted mean
of the precipitation was calculated using equation 4.1 which is resulted in 759 mm of mean annual rainfall
for the kobo valley (Table 4.1).
∑
= (4.1)
=Mean annual rainfall for the kobo valley (mm)

= Measured precipitation at a given station and time (mm)
= Number of rain gauges
Valley, Ethiopia 13
Table 4.1 Mean monthly rainfall distribution mm (1996-2010)
300
mean monthly rainfall (mm)
250
200 Alamata
kobo
150
Korem
100 Zoble
Lalibela
50
0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Figure 4.1 Mean monthly rainfall of the stations (1996- 2010)
1200
968
Mean annual rainfall (mm)
1000
842
760 798
800
674
600
400
200
0
Alamata kobo Korem Zoble Lalibela
Rain gauge stations
Figure 4.2 Mean annual rainfalls of the stations

1400
kobo annual rainfall (1996-2010)
1200
Annual precipitation (mm)
1000
kobo
800
600
400
200
Time in Years
Figure 4.3 Kobo station annual rainfall (1996-2010)

4.1.2. Temperature
To analyse the climatic variation in the high land and low land areas of the kobo valley, Korem and Kobo
stations were selected respectively. As illustrated in the Figure 4.4, average monthly maximum temperature
of 26oc and 35oc were recorded both in June at Korem and Kobo station respectively. On the other hand,
the average monthly maximum temperature of both station show lowest record in January. The stations
have about 9oc in their maximum temperature. The maximum temperature in Korem ranges from 20oc to
26oc and in Kobo from 27oc to 35oc. The detailed record is presented in Table A.6 and Table A.8.
37
35
Max.temperature (oc)
33
31
mean monthly
29
27 kobo
25
23 Korem
21
19
17
15
Figure 4.4 Mean Monthly Maximum Temperatures of Kobo & Korem Station
Even though, the two stations have significant difference in climatic variation, the general trend of average
minimum temperature variation of both stations is similar as seen in Figure 4.5.however, the average
minimum temperature show lowest value of about 11 0C and 4 0C in the month of December for both
stations and highest values of about 19 0C in June for kobo and 12 0C in the month of July for Korem
stations. See the record in Appendix A, Table A.9 and Table A.11.
Valley, Ethiopia 15
20
15
mean monthly
min.tem.(oc)
10 kobo
Korem
5
0
Figure 4.5 Mean monthly minimum temperature (oc) of Kobo and Korem
4.1.3. Wind speed
Wind direction refers to the direction from which the wind is blowing. It is expressed by its direction and
velocity. Wind speed is the relevant variable in order to compute evapotranspiration. The mean monthly
value show lowest record of 1.01m/s in September and highest record of 2.02 m/s in March. See the record
in Appendix A,Table A.12.
2.50
2.00
Wind speed (m/s)
1.50
Kobo
1.00
0.50
0.00
Figure 4.6 Mean monthly Wind speed of kobo

4.1.4. Solar Radiation
Solar radiation changes large quantities of liquid water into water vapour through the process of
evaporation. Consequently, the evapotranspiration process is determined by the amount of energy available
to vaporize water. Mean sunshine hour for Kobo station is above 8 hours from March to May and October
to December. Minimum sunshine hour was recorded in July about 5 hours. See the other record in
Appendix A,Table A.13.
4.1.5. Evapotranspiration (ETo)

Evapotranspiration is the process in which water is returned back to the atmosphere by a combination of
evaporation and transpiration. Potential evapotranspiration is the water loss that will occur under given
climatic condition without deficiency of water supply whereas actual evapotranspiration is the amount of
water that actually returns to the atmosphere depending on the availability of water. For this study,
Penman-Monteith equation is used to estimate potential evapotranspiration from gathered weather data. It
is one of the best methods since it integrates the effect of factors such as altitude, aerodynamics, geographic
location, and solar radiation for computation. For this computation, 15 year (1996-2010) meteorological
data were used as an input for the ETo calculator Programme.The programme is developed by FAO (Food
and Agriculture Organization of the United Nations).

The estimated annual ETO for Kobo station as 1799mm as observed in Table 4.2. It is calculated based on
the Penman-Monteith method. This value is by far larger than the annual precipitation which was recorded
as 674mm. The monthly ETO at Kobo ranges from 128mm in January to 180mm in May (See Figure 4.7).
Table 4.2 Computed ETO of Kobo station using Penman-Monteith equation.
sunshine
Month Tmax(0c) T min(oc) (hour/day) Eto (mm/day) Eto (mm/month)
Jan 26.78 12.66 7.77 4.13 128.13
Feb 29.01 13.05 7.81 4.75 132.91
Mar 29.79 14.52 8.45 5.21 161.41
Apr 30.92 15.59 8.48 5.48 164.40
May 33.15 16.43 8.75 5.81 180.01
Jun 34.01 17.99 6.51 5.34 160.20
Jul 31.59 17.90 5.15 4.75 147.15
Aug 30.46 16.89 5.88 4.81 149.01
Sep 30.51 15.04 6.77 5.01 150.20
Oct 29.79 13.56 8.35 4.99 154.59
Nov 28.79 12.46 9.36 4.69 140.80
Dec 27.21 12.02 8.59 4.21 130.61
Average 30 15 8 5 1799
Perman-Montheith equation was used in to ETo Calculations with the values for Angstrom's
Coefficients: a = 0.25 and b = 0.5.
200 Precipitation Eto T average 30
180
25
160
140
20
P/ETo (mm/month)
Taverage (oc)
120
100 15
80
10
60
40
5
20
0 0
Figure 4.7 Precipitation (P), potential evapotranspiration (ETO) and average monthly Temperature (Tavg)
for Kobo station (1996 to2010)
Valley, Ethiopia 17
Actual evapotranspiration
Actual evapotranspiration is the amount of water that actually returns to the atmosphere depending on the
availability of water.
The Turc, Langbein and Wundit empirical formula was used to estimate the mean areal actual
evapotranspiration of the valley of the studied basin. A widely used formula to estimate annual values of
ETa for catchment areas and it represents all the different climates including Africa (Shaw, 1994).The
formula takes into consideration mean annual precipitation and mean annual temperature of the catchment
area. Turc showed that the formula could be applied in humid and arid climates, either hot or cold (Shaw,
1994).
E= ( )
4.2
.
[ ( )]
Where,
E: mean annual evapotranspiration (mm)
P: mean annual precipitation (mm)
t: mean annual temperature (0c)
L (t) = 300 + 25t + 0.05t3
The mean annual actual evapotranspiration of the kobo valley was calculated as 704mm according to
equation 4.2.
4.2. Groundwater recharge estimation

4.2.1. Water Balance Method
Groundwater recharge is defined as the entry into the saturated zone of water made available at the water
table surface together with the associated flow away from the water table within the saturated zone (Freeze
& Cherry, 1979). Quantifying the rate of recharge to aquifer is the most difficult of all measures in the
evaluation of groundwater resources. Estimation of groundwater recharge requires modelling of the
interaction between all the important processes in the hydrological cycle such as precipitation, infiltration,
surface runoff, evapotranspiration, soil moisture and groundwater level variations (Jyrkama and Sykes,
2007).
Meteorological data limitation together with absence of hydrological data within this basin made the
estimation of groundwater recharge very tough. In this study, recharge was estimated using water balance
method.
Thus, the groundwater recharge for the valley area can be calculated using simple water balance method;
R= P-ETa + Sr -S 4.3
Where, P: Annual precipitation (mm)

ETa: Annual Actual evapotranspiration (mm)
Sr: Annual Surface runoff (mm)
R: Recharge to groundwater (mm /year)
S: Soil moisture content (mm)
The valley floor including its escarpment receives relatively low annual precipitation (759mm) and has
annual actual evapotranspiration (704mm). Recharge in this area is generally assumed to be very minimum
from direct precipitation. However, the valley floor gets recharge from runoff along escarpments and
stream leakage which flow down the highlands. The runoff from the highland area flow out to the valley
floor along two mainstream channels, Golina and Hormate streams, and few along Keleklit stream.

Thus, the amount of recharge to the valley floor can be estimated in two ways;
The direct recharge which is surplus of evapotranspiration and soil moisture considering that
insignificant runoff over the valley plain.
The indirect recharge from runoff along the valley escarpment.
The direct recharge was calculated as 33mm/year using equation (4.2) and the soil moisture content
(22mm) was adopted from Co-SAESAR (1999) which was computed from SCS model, while the indirect
recharge was assumed 40% of the runoff from the valley escarpment considering the catchment
characteristics of the valley and previous hydrological reports of Co-SAESAR (1999).
According to RVPD (1998), the runoff coefficient for the escarpments was estimated between 0.13 to 0.22.
Assuming 15% of the precipitation as annual runoff (114mm), the indirect recharge was 46mm/year. The
total recharge for the kobo valley was the sum of the direct and indirect recharge which was resulted in
79mm/year.
4.2.2. Darcy Approach
This approach considers groundwater flux through a flow width perpendicular to the general gradient of
groundwater flow. The annual discharge of this groundwater can be estimated using the following formula.
Q = 365 ∗ T ∗ I ∗ B 4.4
Where Q = discharge (m3/year)

T = Transmissivity (m2/day)
I = hydraulic gradient
B = groundwater flow channel width (m).
Table 4.3 Darcy approach recharges estimation of Hormat-Golina and Waja-Golesha sub-basins
Slop of
Difference in water GW long Average Flow
Flow Flow line table elevation along the major Transmissivity in
Sub-basin width(m) length(m) the flow line(m) flow line (m2/day) MCM
Hormat-Golena 16680 13000 155 0.0119 600 44

Waja-Golesha 13000 14000 118 0.00843 600 24
Total 68
Source: KGVDP Hydrogeology report
Another recharge source to the valley area is direct infiltration from precipitation mainly during torrential
rain falls. The annual precipitation was calculated as 759mm. In this study the annual infiltration rate is
taken to be about 5% (38 mm) of the precipitation based on hydrology report of Co-SAESAR (1999). The
surface area of the plain area of Waja-Golesha and Hormat-Golina sub-basins is 165 and 231 km2,
respectively. The plain areas of the sub-basins were delineated by using Global Mapper software.
Therefore, the direct infiltration on these sub basins is estimated to be about 9 MCM for Hormat-Golina
and about 6 MCM for Waja-Golesha.
In conclusion, the total recharge in Waja-Golesha (30 MCM) and Hormat-Golina (53 MCM) sub-basins of
Kobo valley is estimated to amount 83 MCM.
4.3. Pump test data Analysis.

Pumping test is a scientific approach where the groundwater storage and movement is expressed on terms
of the physical and hydraulic properties of the aquifer system. Aquifers are groundwater reservoir where
the lateral continuity and vertical boundaries are often not well defined. Since direct observation of
groundwater movement is impossible, mathematical analysis offers a convenient and reliable way to
Valley, Ethiopia 19
predict what happens to water in the ground. It is therefore, imperative to derive simple mathematical
expressions for describing the flow region of water in the subsurface. Groundwater declining due to
pumping can be defined with different well flow equations which are developed for steady state and
unsteady state flows for various types of aquifers and boundary conditions. Together with the basic
assumptions and conditions for steady and unsteady state flows, the equations are presented in their final
mathematical form for practical application.
In this study, the main objective of collecting pump test data was to determine the aquifer hydraulic
parameters such as transmissivity and hydraulic conductivity, which later used as one of the model input
parameters. The aquifer parameters are important as they give an understanding of the groundwater flow in
the system. Kruseman and de Ridder (1992), suggested that generally all the analytical methods assumed
the aquifer is homogenous and isotropic, groundwater flow is horizontal and Darcy’s law is valid,
discharged at constant rate, fully penetrating well of very small diameter and geologic formations are
horizontal and have infinite horizontal extent. Geo Engineering Service (GES, 2003) and Metaferia
Consulting Engineers (MCE, 2009) were conducted some pumping test analysis for the alluvial aquifer of
the kobo valley.
The aquifers in the study area are mainly Quaternary alluvial deposits, and fractured and weathered basalts.
Totally 70 boreholes were inventoried in this study and almost all were sunk in the alluvial sediments
aquifer with the exception of very few boreholes which are located in the volcanic aquifer. However, most
of the early constructed boreholes have incomplete data. All the boreholes were sunk in the alluvial aquifer.
Among these boreholes, 10 of them have observation well (See Table B.1).
The pumping test data show that the constant test was conducted for 72 hours for most of the boreholes.
The data obtained by measuring the drawdown at a single location outside the pumping well only permit
calculation of the average permeability, transmissivity and storability (coefficient of storage) of the aquifer.
The need of two or more observation wells at different distances is to analyze the time- drawdown and
distance-drawdown relationships. The value of transmissivity and storage coefficient is important because
they define the hydraulic characteristics of a water bearing formations. The transmissivity value indicates
how much water will move through the formation and the coefficient of storage indicates how much can be
removed by pumping or draining. The distance drawdown curve helps for the determination of the effect of
pumping at any distance from the pumped well.
All the hydraulic parameters were taken from Kobo Girana Valley Development Project office (See Table
B.1).

CHAPTER 5
Groundwater Modelling of Kobo Valley
5.1. Introduction
Gathering and assemblages of relevant hydro geological data is crucial for proper groundwater modelling.
This process includes identifying hydrostratigraphic units, estimating transmisivity values, defining system
boundaries, etc.
There are two areas of hydrogeology where we need to rely on models of real hydro geological system: to
understand why a flow system is behaving in a particular observed manner and to predict how a flow
system is behaving (Fetter, 2001). There are several ways to classify groundwater flow models, models can
be either transient or steady state and one, two or three spatial dimension. Steady state flow occurs when at
any point in a flow field the magnitude and direction of the flow are constant with time (Anderson and
Woessner, 1992).
In order to have confidence in model simulation results, realistic model inputs and better understanding of
the hydrologic system of the studied area are imperative. In this chapter, the aquifer system of kobo valley
was modeled using PMWIN Pro (Chiang et al., 1998) as pre –and post – processor for MODFLOW
(McDonald and Harbaugh, 1988) assuming steady-state condition. The aquifer was modeled under
unconfined condition and confined condition represented by a two layer with varying thickness. A grid cell
size of 300m x 400m was used. Model area and the elevations of the top layer were delineated by the
ASTER DEM optimization and use of the topographic maps. Aquifer properties were adopted from the
results of the pumping test data analysis. Recharge to the major component of the system was considered to
take place as direct infiltration of precipitation for the entire model area and further inflow from the
surrounding hills. Simplified water balance method and Darcy's approach were employed to estimate the
recharge. Trial and error method was used to calibrate the model using the observed hydraulic head.
5.1.1. The modelling Process
To ensure that the modelling study is conducted correctly, it is important to use a proper modelling
methodology. This is also increase confidence in the results of the model (Anderson and Woessner, 1992).
The modelling protocol suggested by Anderson and Woessner (1992) was followed to come up with good
result (Figure 5.1).
Valley, Ethiopia 21
Figure 5.1 Steps in modelling protocol (after Anderson and Woessner, 1992).
5.2. Conceptual Model

Developing the appropriate conceptual model for a given problem is one of the most imperative steps in the
modelling process. Over simplification may lead to a model that lacks the required information, while
under simplification may result in the lack of data required for model calibration. A conceptual model
describes how water enters an aquifer system, flows through the aquifer system and leaves the aquifer
system. Briefly, it describes the hydrologic system with respect to aquifer properties, flow characteristics
and boundary conditions. According to Anderson and Woessner (1992) there are three steps in building a
conceptual model: defining hydrostratigraphic units, preparing a water budget and defining the flow
system.
Even though, there was very limited data particularly for Waja-Golesh sub-basin; a simplified conceptual
model was developed for both sub-basins for the groundwater flow system in Kobo valley. To develop the
conceptual model, some simplifying assumptions were made. The assumptions include: the model consists
of two layers, the model is two dimensional, the aquifer is unconfined and confined with varying thickness
and, the groundwater flow is horizontal.
In principle the groundwater flow and contaminant transport in porous medium domain are three-
dimensional. However, when considering regional problems, one should note that because of the ratio of
aquifer thickness to horizontal length, the flow in the aquifer is practically horizontal. The horizontal
dimension may be from tens to hundreds of kilometers with a thickness of tens to hundreds of meters (Bear,
1979).

Simplification is important because complete reconstruction of the filed system is not feasible. The
conceptual model should be simplified as much as possible while it is still remains complex enough to
represent the system behavior (Anderson and Woessner, 1992). In this study, to simplify the complex
nature of the two sub-basins, a simplified conceptual hydro geological model of the groundwater system
was developed based on information about geology, hydrogeology and hydrology.
The system is considered in a steady- state throughout the year for the modeling purpose. The simplified
conceptual groundwater system of the two-basins is shown in Figure 5.2.
Figure 5.2 Schematic diagram illustrating the simplified conceptual model.

P = Precipitation, ET = Evapotranspiration, I = Infiltration, SR = Surface runoff without defined channel,
RGF = Regional groundwater flow path, GFD = Groundwater flow direction
Figure 5.3 Simplified Conceptual model as taken from Global Mapper.
Valley, Ethiopia 23
5.2.1. Boundary Conditions
Boundary conditions are constraints imposed on the model grid that express the nature of the physical
boundaries of the aquifer being modelled. Boundary conditions have great influence on the computations of
heads within the model area. Anderson and Woessner (1992) defined three types of mathematical
conditions used to represent hydro geological boundaries:
Specified head boundaries (Dirichlet conditions)
Specified flow boundaries (Neuman conditions)
Head-dependent flow boundaries (Cauchy or mixed conditions)
Boundary conditions are mathematical statements specifying the dependent variable (head) or the
derivative of the dependent variable (flux) at the boundaries of the problem domain. In steady-state
simulation, the boundaries largely determine the flow pattern. Therefore, correct selection of boundary
conditions is a critical step in model design (Anderson and Woessner, 1992).
The lateral boundaries of the model area are: either no-flow or head dependent flux boundaries. Even
though water may enter in to the alluvial sediment from the surrounding mountainous areas at the contact,
no-flow boundaries are assigned to the model at these areas except for the fractured zones and stream bed
boundaries, assuming that minimum or no groundwater enter in to the modelled area from the ridges. The
location of head-dependent flux boundary for the study area is assumed at the localities of the fracture
zones of the surrounding ridges along valley channels and gullies. Topographically low areas along which
surface water and groundwater outflow are also considered as the head dependent flux boundary and the
model is simulated with the General-Head-Boundary (GHB) module of the MODFLOW at these localities.
Similar boundary conditions are considered for both layers of the model.
The top boundary of the model, the upper boundary of layer-1, was simulated as a free surface boundary,
which include specified-flux and head-dependent flux boundary cells. The specified-flux boundary is the
areally applied groundwater recharge and the head-dependent boundary represents springs and groundwater
seeps from river beds. Recharge was specified and simulated with the recharge (RCH) module; Golina
River was simulated with river (River) module. The bottom boundary of the model is a specified no-flow
boundary. This no-flow boundary is located where the aquifer comes in to contact with massive bedrock.
5.2.2. Stratigraphic Units
Identification of hydrostratigraphic units is crucial in determining the number of layers controlling
groundwater flow within the system. A hydrostratigraphic unit is comprised of geological units of similar
hydro geological properties. Numerous geological units may be grouped together or a single formation may
be subdivided into different aquifers and aquitards (Anderson and Woessner, 1992).
Layer-1 correspond the entire alluvial sediments which range in thickness from around 50m near the divide
of the two sub-basins near Kobo area to 270m at Central Golesha and 246m at the area downstream of
Abuare and Gedemeyu Villages where clay and silt are predominant. The weathered bedrock underlying
the alluvial sediment of the area is taken as Laye-2 of the model which has an average thickness of 40m.
(See Figure 5.6)
5.2.3. Sources and sinks of the Model area
The main groundwater source for this studied valley is direct recharge from precipitation that fall at the
highlands and the valley floor. However, the valley floor gets additional recharge from surface runoff along
the escarpments and from stream leakages that drain the highlands. The primary output or sink is
groundwater outflow in the form of base flow at Golina outlet and Selenwuha outlet for Hormat-Golina and
Waja-Golesha sub-basin respectively (See Figure 5.3).
Recharge
Groundwater recharge was estimated using water balance method and Darcy's approach which is discussed
in section 4.2. As observed from field visit, there are no recharging wells in the study area. The recharge
from excess irrigation is assumed to be negligible.

Groundwater outflow/sinks
The ways of groundwater discharge from the aquifer system is mainly by discharge to streams. The main
stream (Golina) that drains the western highlands gets its base flow from the volcanic aquifer and losses
some amount when it reaches the valley floor (the alluvial deposits). However, near the out let to Afar it
gains some amount as it was evidenced from increased flow during field visit. On the other hand, the
Hormat stream collects some seepage flows from the western highlands and loses into the alluvial
sediments along the stream course (See Figure 5.4).
Figure 5.4 Golina river near the outlet to Afar.
Evapotranspiration from the groundwater system can be assumed to be negligible, since there are no
significant groundwater discharge areas such as marshes, swamps, and/or lakes within the basin.
The river package is designed to simulate the effect of flow between rivers and aquifers based on the
following relations:
QRIV = CRIV (HRIV – h) For h>RBOT 5.1
QRIV = CRIV (HRIV – RBOT) For h<= RBOT 5.2
&'() = 5.3
*+,
-
Where:
QRIV = rate of leakage between the river and aquifer [L3T-1]
CRIV = hydraulic conductance of the river bed [L2T-1]
HRIV = head in the river [L]
h = hydraulic head in cell [L]
RBOT = elevation of the bottom of the riverbed [L]
K = hydraulic conductivity of the riverbed material [LT-1]
L = length of the river within a cell [L]
Valley, Ethiopia 25
W = width of the river [L]
M = thickness of the riverbed [L]
Data were taken for the river conductance calculation along the longitudinal of the Golina and Hormat
rivers .The hydraulic conductivity was also adopted from literature.
Groundwater has been abstracted for irrigation and water supply purposes. However, there is no recorded
data regarding the abstraction rate and duration of pumping from the wells. According to the information
from KGVDP office, Kobo town water supply office and operators, the amount of water abstracted from
the wells was estimated to be 5 Mm3y-1 (see Table 5.1).

Table 5.1 The estimated amount of water abstracted from boreholes per annual
Irrigation boreholes (water used per crop season)

Discha Jan Feb Mar Apr May Jun
Well ID rge(l/s) (24*6hrs) (24*12hrs) (24*16hrs) (24*16hrs) (24*14hrs) (24*12hrs) m3/annual m3/day
HG1 51 26438 52877 70502 70502 61690 52877 334886 917
HG2 51 26438 52877 70502 70502 61690 52877 334886 917
HG6 50 25920 51840 69120 69120 60480 51840 328320 900
HG7 50 25920 51840 69120 69120 60480 51840 328320 900
HG8 50 25920 51840 69120 69120 60480 51840 328320 900
HG9 10 5184 10368 13824 13824 12096 10368 65664 180
Hg10 34 17626 35251 47002 47002 41126 35251 223258 612
TW1 7 3629 7258 9677 9677 8467 7258 45965 126
Zeleke1 50 25920 51840 69120 69120 60480 51840 328320 900
Zeleke2 50 25920 51840 69120 69120 60480 51840 328320 900
WG1 51 26438 52877 70502 70502 61690 52877 334886 917
WG2 40 20736 41472 55296 55296 48384 41472 262656 720
WG4 52 26957 53914 71885 71885 62899 53914 341453 935
WG5 52 26957 53914 71885 71885 62899 53914 341453 935
WG14 25 12960 25920 34560 34560 30240 25920 164160 450
Kobo town water supply bore holes
K1 10 10 x 8hrs x365d 105120 288
K5 38 38 x 8hrs x365d 399456 1094
K6 38 38 x 8hrs x365d 399456 1094
Kobo Rural Water Supply bore holes
K37 2.5 2.5 x 8hrs x365d 26280 72
K38 4.2 4.2 x 8hrs x365d 44150 121
K42 4.5 4.5x 8hrs x365d 47304 130
d = average pumping days per month, hrs = pumping hours per day
Valley, Ethiopia 27
5.2.4. The Model area
The modelled area is 30 by 40 km from UTM 555000m to 585000m and 1323000m to 1363000m
Easting and Northing respectively. It contains the entire two sub-basins, Hormat-Golina and Waja-
Golesha, located in the South and North respectively ( Figure 5.5). The model uses a grid size of
300m by 400m and contains: two layers, 100 columns, 100 rows and 10,000 cells in each layer
originally. After refining the grid sizes around the main aquifer areas, it was split in to 148 columns
and 145 rows and 21,460 cells in each layer. The irregular shape and the locally bounded nature of
the aquifers of the study area reduced the number of active cells in the model. Even though the whole
catchments of the two sub-basins of Hormat-Golina and Waja-Golesha is wider, it is only the alluvial
plain and underlying weathered bedrock modelled numerically so that the modelled area is narrower
than the entire catchments area for the sub-basins ( Figure 5.5).
Figure 5.5 Plan View of the Entire Modelled Area

5.2.5. Aquifer Geometry
The aquifer was descritized vertically in to two layers (layer-1 and layer-2,Figure 5.6). Layer-1
correspond the entire alluvial sediments which range in thickness from around 50m near the divide of
the two sub-basins near Kobo area to 270m at Central Golesha and 246m at the area downstream of
Abuare and Gedemeyu Villages where clay and silt are predominant. The weathered bedrock
underlying the alluvial sediment of the area is taken as Laye-2 of the model which has an average
thickness of 40m.

Figure 5.6 Cross-Sectional View of the Model Area along Line North to South of Fig. 5-5
5.3. Numerical Model

Numerical model development allows for a detailed analysis of the movement of water through the
hydrologic units that constitute the groundwater flow system. The groundwater flow in the
unconsolidated deposit of the Kobo valley was simulated using the U.S. Geological Survey modular
three – dimensional finite- difference groundwater flow model, MODFLOW (McDonald and
Harbaugh, 1988). This numerical modelling was performed using the interface of Processing Modflow
Pro (PMWIN Pro), Version 8.0 (Chiang and Kinzelbach, 2001) as code environments for the data
input and output management. PMWIN Pro supports MODFLOW- 2000, PEST- ASP, different
packages, and models/programs. It is founded on the physical theory of groundwater movement:
Darcy’s law and the continuity equation. The steady- state groundwater flow is simulated based on the
following governing differential equation under two- dimensional aerial view (Anderson and
Woessner, 1992).
01/ 3+ 015 3 +' =0 5.4

. .2 . .2
./ ./ .5 .5
Where:
Kx and Ky = Components of the hydraulic conductivity along x, and y axes [LT-1]
R = Flux per unit volume representing sources/sinks term [T-1]
h = Hydraulic head [L]
5.3.1. Data input for the Model
The input data that were previously processed and analyzed in chapter four were simulated by the
model in accordance to how the computer code runs. The model was assumed to simulate the
conceptual model. The processed DEM (aquifer top elevation) was imported in to the model after
defining the model area and boundary conditions. The bottom elevation was taken from the cross-
sections. The cross –sections are selected, in such a way that the volcanic and the alluvial aquifers are
hydraulically connected. The General Head boundary condition was used to estimate the inflow from
the mountain to the alluvial plain. The hydraulic conductivity values from the pumping test analysis of
the alluvial aquifer were used as initial values for this aquifer in the model input. Due to data
limitation, only 35 borehole coordinates and observed water level were imported to the model for
calibration purpose.
Valley, Ethiopia 29
5.3.2. Model execution and Calibration
Model calibration is the process of making adjustments, within justifiable ranges, to initial estimation
of selected model parameters to obtain reasonable agreement between simulated and measured values.
It requires the entry of organized input data into the selected computer code, and interpretation of the
model results. These results were compared with the calibration target and if the error in the simulated
results is acceptable, the model is considered as calibrated; if the level of error is unacceptable, the
input parameter values are adjusted within a reasonable ranges and the model is run again until
acceptable results are achieved. The model runs with the interactive method.
5.3.3. Calibration target and Uncertainty
In this study, the measured hydraulic heads from the field were used as a calibration value. The main
purpose of calibration was to match the simulated head by the model with the measured head.
However, most of the measured head data and uncertainty of the model are associated with errors.
This is due to:-
Measurement errors related to measuring device and operator/user.
Errors due to averaging ground surface elevations from digital elevation models (DEM).
The water level measurements are single time measurement.
Because of data limitation on the Waja-Golesha sub-basin, the calibration mainly focuses on the
Hormat-Golina sub-basin. Due to all uncertainties, calibration becomes a challenging and tough task.
The standard deviation of the groundwater level below ground surface from measured groundwater
level showed a value of near to 7m. It was reasonable to accept a RMS error of 7m as calibration
target due to the cumulative effect of the mentioned uncertainties in the input data (see Appendix
C,Table C.1).
5.3.4. Trial and Error Calibration

Trial- and – error calibration was the first technique to be used and is still the technique preferred by
most users (Anderson and Woessner, 1992). It is the process of manual adjustment of input parameters
until the model simulates the measured heads within range of the error criteria. The model was
calibrated for steady- state conditions, assuming constant recharge and steady discharge neglecting
seasonal fluctuations. Calibration was done through trial and error by changing aquifer hydraulic
conductivity, recharge and river bed material conductance values. The steps followed for the trial-
and-error calibration is shown in Figure 5.7.
Figure 5.7 Trial and error calibration procedures (adapted from Anderson and Woessner, 1992)

5.3.5. Evaluation of calibration
The results of the calibration should be evaluated both qualitatively and quantitatively (Anderson and
Woessner, 1992). The mean of the observed and simulated heads differences was used to quantify the
average error in the calibration process. The three ways of expressing the average difference between
simulated heads (hs) and measured heads (hm) are the mean error (ME), the mean absolute error
(MAE) and the root mean square error (RMS). The main target of the calibration is to minimize these
error values.
1
78 = 9(ℎ;, − ℎA, ) 5.5
=>?
The mean difference between measured heads and simulated
1
7B8 = 9 |(ℎ;, − ℎA, ) | 5.6
=>?
The mean of the absolute value of the differences in the measured and simulated heads
1
RMSE = G (ℎI,H − ℎJ,H ) 5.7
2
9
H=1
The average of the squared differences in measured and simulated heads (See Appendix C,Table C.1).
Figure 5.8 Graphical comparison between the observed and simulated heads
The above error measures can only be used to evaluate the average error in the calibrated model. The
RMSE is usually thought to be the best measure of error if errors are normally distributed. The
maximum acceptable value of the calibration criterion depends on the magnitude of the change in
heads over the problem domain (Anderson and Woessner, 1992).
The calibrated fit between the observed and simulated heads by the model generated scatter diagram is
shown in Figure 5.9.The scatter plot has a value of RMSE of about 7m.
Valley, Ethiopia 31
Figure 5.9 The scatter diagram showing the comparison of measured and simulated heads
The differences in simulated and observed heads are tabulated in Appendix C, Table C.1.Table C.1
Comparison of the Observed and Simulated Heads and Error Calculation
. The summary of the error analysis for the calibrated model is shown in Table 5.2
Table 5.2 Error summary for the calibrated model
Type of error Value (m)

ME -1
MAE 6
RMSE 7

CHAPTER 6
Result and Discussion
6.1. Results
6.1.1. Water Budget of the model domain
The water budget for steady-state simulation is balanced (inflow minus outflow) within a percent
discrepancy of [%] 0.00. In the model area, the inflow term includes the recharge and head dependent
boundary whereas, the outflow term includes wells, river leakage and head dependent boundary. The
water budget is calculated by water budget tool in MODFLOW. The model result shows both inflow
and outflow are in balance which is consistent with the steady-state modelling theory (See Table 6.1).
The inflow-outflow balance simulated under the numerical model has some differences with that of
the conceptual model. The balance for simulated is 118MCM for inflow which is similar to the
outflow where as the balance for conceptual model is 95MCM for inflow which is equal to the
outflow. The balance difference between the two is 23MCM which favours for the simulated balance.
On the other hand, the simulated in flow 118 MCM is close to the valve of annual recharge of Kobo
valley 119MCM calculated by CO-SAERAR study which was previously discussed in Literature
review. This difference might be resulted from either from the data gap found during the analysis of
the conceptual model as the pumping test for those wells drilled at the inlet and outlet areas of the two
sub-basins is not conducted to evaluate the subsurface inflow-outflow.
Table 6.1 Water budget of the entire model domain in m3d-1
Flow term IN OUT IN-OUT

wells 0 5192 -5192
Recharge 104232 0 104232
River leakage 0 57456 -57456
Head dependent bounds 218905 260489 -41584
Sum 323137 323137 0
DISCREPANCY [%] 0.00
6.1.2. Sensitivity Analysis
Sensitivity analysis is the measure of uncertainty in the calibrated model caused by uncertainty in
aquifer parameters and boundary conditions. Sensitivity analysis was performed by systematically
changing the calibrated values of conditions (Anderson and Woessner, 1992). The main objective of a
Valley, Ethiopia 33
sensitivity analysis is to understand the influence of various model input parameters and hydrological
stresses on the aquifer system and to identify the most sensible parameter(s), which will need a special
attention in future studies. By running the calibrated model for the respective changed values of the
input parameter and comparing the result with the calibrated head, the parameter(s) sensitive to the
model was established. The parameter values were varied within a reasonable range. Thus, it is
important step in modelling studies. Accordingly, the model in this studied area is highly sensitive
with decrease of the calibrated recharge and hydraulic conductivity values and relatively less sensitive
with increasing these values which result in lower RMS error.
6.1.3. Pumping Scenario Analysis
In order to evaluate the response of the groundwater system under different groundwater abstraction
rates, pumping scenario analysis were computed. The groundwater system response was compared
with resulting changes in water level (drawdown) and groundwater outflow from the model domain.
Even though, there was limitation in recording data how the groundwater has been abstracted
currently, some estimation was done based on the information gained from project area. Three
scenarios were used for analyses of the impact of well operation on drawdown: 1) current situation: 11
wells are operated simultaneously; 2) 35 wells are operated simultaneously and 3) 70 wells are
operated simultaneously.
A total of about 27878 m3d-1 of abstracted water was used in scenario-one to observe the effect
pumping on the calibrated model. (See Appendix D, Table D.1).
In scenario-two, a total of 55825 m3d-1 groundwater was assumed to be abstracted from 70 boreholes
in order to see the effect of increased groundwater discharge over the concentrated boreholes as
compared to model result (See Table D.2).The amount of water abstracted in scenario- two was
increased by 50% from the amount used in scenario- one.
The model simulation result for scenario-one has shown that for a total of 27878 m3d-1 of abstracted
water from 35 boreholes was resulted in an average decline of groundwater level at the pumping well
by about 14m for the entire model area. However, the decline in head slightly exceeds 25m at the
borehole HG8 (See Table 6.2).
In scenario-two, for a total of 55825 m3d-1 of abstracted water from 70 boreholes, the model
simulation resulted in an average decline of water level by 32m for the entire model domain.
Similarly, the decline in head reaches 45m at borehole HG8. Generally, the effect of increased
groundwater abstraction is more pronounced in areas where there are more boreholes at close distance.
The detailed average decline in groundwater head for both scenario-one and scenario-two can be seen
in Table D.3
Table 6.2 Estimated groundwater abstraction rate and the average decline in groundwater level for
different pumping scenarios
Average decline in head (m)

Number of bore holes Total discharge (m3d-1) due to abstraction.
Current situation 11 5192 7
Scenario-one 35 27878 14
Scenario-two 70 55825 32

Figure 6.1 The model simulated groundwater heads for pumping scenario-one
Valley, Ethiopia 35
Figure 6.2 The model simulated groundwater heads for pumping scenario-two
Model simulated groundwater budget for different pumping scenarios
The model simulated groundwater budgets for different pumping scenarios were processed for Kobo
valley and the model results were shown in Table 6.3 and Table 6.4.

Table 6.3 Simulated water budget of Kobo-valley for scenario-one in m3d-1

wells 0 27878 -27878
Recharge 104232 0 104232
Head dependent bounds 227872 246929 -19057
Sum 332104 332104 0
Table 6.4 Simulated water budget of Kobo-valley for scenario-two in m3d-1

wells 0 55825 -55825
Recharge 104232 0 104232
Head dependent bounds 241064 232180 8884
Sum 345296 345296 0
Graphic comparison of the hydraulic heads simulated under different pumping scenarios.
To compare the variation in head distributions, the model generated hydraulic heads under different
scenarios were plotted together. (See Figure 6.3)
Figure 6.3 Comparison between the observed and simulated heads of different scenarios
The decline in hydraulic head under both scenarios was more pronounced Hormat-Golina sub-basin of
kobo valley, where there are relatively large numbers of discharging boreholes, which are closer to
each other.
Valley, Ethiopia 37
Scenario-one
Scenario-two
Figure 6.4 The scatter plots of simulated versus observed heads for different pumping scenarios
6.1.4. Estimation of Radius of influence and well interference
Radius of influence
The radius of influence of a pumping well can be determined in steady state for different aquifer
systems (confined and unconfined).Radius of influence is the horizontal distance from the centre of a
well to the limit of the cone of depression. It is calculated using different groundwater flow equations.
For this study, a drawdown less than 20 % of the measured drawdown in the pumping well is taken as
stabilized or insignificant drawdown based on the pump test data analysis which was conducted
previously by Geo Engineering Service (GES, 2003) and Metaferia Consulting Engineers (MCE,
2009). The radius of influence is used for future plan of boreholes drilling. From Table E.1, It was
seen that 12 wells had higher drawdown than the 20% drawdown at scenario-two based on the
calculated draw downs of Hormat-Golina sub-basin.

Well interference
Interference of the cone of depression of wells spaced closer to another will increase the drawdown in
each well and consequently decrease the discharge per well. In order to estimate the drawdown
interference of hypothetical wells for planning purpose, Muskat, (1937) formula was applied for two
identical wells at a distance B apart. Well interference is computed for two identical wells in Hormat-
MN(O P − ℎQ P )
Golina sub-basin for future well locations as follow;
L= 5.8
2.3log ( X)
U
VW
ℎQ = water column in the pumping wells (m)

Where Q = the discharge (m3/d)
ZQ = Well radius (m)

R = Radius of influence (m)
d = Saturated thickness (m)
k = Hydraulic conductivity of the aquifer (m/d)
B = well interference distance between two wells (m)
Figure 6.5 well interference in Hormat-Golina sub-basin
Valley, Ethiopia 39
Table 6.5 Well interference computation in Hormat-Golina sub-basin
Parameters Hormat-Golina sub-basin

Type of aquifer Unconfined
Radius of influence (m) * 250

Well interference distance (m) * 500
Discharge (m3/day) 4320
Radius of well (m) 0.254
Average depth (m) 142
Average SWL (m) 47
Average saturated thickness (m) 95

Average hydraulic conductivity
(m/day) 7.08
Calculated hw (m) 88.44
Calculated drawdown at 250m

from the pumping well(m) 6.56
Source: Pump test data
* Assumed values
Based on the calculated horizontal distance, three pairs of wells (K6-HG6, K6-HG1 & HG6-HG8)
have a radius of influence (209m,229m &235m) less than 250m respectively. The horizontal distance
between two wells was calculated by taking the square root of the square sum of distances in Northing
and Easting Hence; the drawdown of these wells is greater than the stabilized 20% drawdown at
scenario-two. All wells have a drawdown less than 20% in scenario-one. From the field visit, it was
mentioned that well HG8 has a problem in discharge decrease from initial 50l/s to 26 l/s. The model
simulation also indicated that HG8 had a maximum drawdown of 26m and 45m in scenario-one and
scenario-two respectively (SeeTable E.1).Therefore the well interference and the abstraction of all
wells at a time is one of the causes for discharge decline.
6.1.5. Groundwater Reserve in Kobo Valley

The total subsurface water reserve is a function of saturated thickness and storage coefficient/specific
yield. The aquifer system is generalized into water table aquifer of the sediment. The average
saturated thickness is 139 m in Waja-Golesha and 95 m in Hormat-Golina sub basins. The
groundwater reserve is computed applying the following formula.
) = Sy ∗ A ∗ H 5.9
Where V= Reserve (m3)
Sy = Specific yield (0.1 for kobo valley taken from pump test data)
A = surface area of the aquifer (m2)
H = saturated thickness (m)
The groundwater reserve in Waja- Golesha and Hormat-Golina sub basins is summarized in Table 6.6

Table 6.6 Groundwater reserve in Kobo valley
GW potential area Average saturated Allowable

Sub-basin (km2) thickness(m) extraction (MCM)
Hormat-Golena 127 95 1207
Waja-Golesha 86 139 1190
Total 213 2397
Source: Pump test data and this study.
6.1.6. Estimation of Allowable Exploitation of Groundwater
In this study, the allowable extraction of groundwater is taken to be 60% of the saturated thickness of
the sediment. Accordingly, the exploitable amount of water reaches 714 MCM and 724MCM in Waja-
Golesha and Hormat-Golina sub basins of the Kobo Valley respectively. Estimated abstraction is
summarized in Table 6.7.
Table 6.7 Allowable abstraction for 60% drawdown of the saturated thickness in Kobo Valley
GW potential area Number of existing 60% saturated Allowable

Sub-basin (km2) wells thickness (m) extraction (MCM)
Hormat-Golena 127 48 57 724
Waja-Golesha 86 22 83 714
Total 213 70 1438
Source: Pump test data and this study.
According to the availability of the exploitable groundwater amount, the number of irrigation wells to
be used in each sub basin is determined in Table 6.8 considering the following assumptions.
60% of the available groundwater amount will be abstracted on annual basis within 15 years
period.
Annual groundwater abstraction from the wells will be only for six months dry period from
January to June.
Annual groundwater abstraction from each well will be 0.8 MCM when used for six months;
which means the abstraction rate from each well is assumed to be 50 lit/sec.
The annual recharge for each sub-basin calculated by using Darcy's approach was used to estimate the
number of wells in kobo valley. This total recharge (83MCM) is less than the model calculated
recharge (118MCM).The estimation of number of wells in the valley has good safety factor since
calculation is made on the smaller recharge.
Table 6.8 Estimation of the Available Groundwater Potential and Number of Wells in Kobo Valley
Currently
15 years Annual Total Total Existing Additional
Abstraction Annual Recharge Annual number of number of Proposed
Sub-basin (MCM ) (MCM) (MCM ) (MCM ) wells wells Wells
Hormat-Golena 724 48 52 100 125 48 77

Waja-Golesha 714 48 30 78 97 22 75
Total 1438 96 223 70 152
Valley, Ethiopia 41
6.1.7. Irrigation Water use efficiency
Agriculture consumes the largest amount of water however; the water use is mostly inefficiency. With
rapid population growth, the need for food and correspondingly water for irrigation is rising. There is
an increasing demand due to urbanization and industrialization. Therefore efficient use of irrigation
water to produce more is the only option. Increasing irrigation water use efficiency is possible through
suitable crop selection, proper irrigation scheduling and effective irrigation techniques.
Strategies to optimize Irrigation water use efficiency
Farmers always ask questions how to improve irrigation water use efficiency to increase production.
Making the right decisions related to crop selection, Irrigation scheduling and methods is imperative
to improve irrigation water use efficiency. These strategies are used to reduce water and pumping
costs, increase crop yield, and maintain a higher soil quality. Crops have different daily and total
growing water needs. Maize was selected because of the sufficient added value and farmers'
preference. Proper Irrigation scheduling can eliminate too much or too little water that is applied to
crops. It integrates the time and depth of water applied to crops based on the water content in the crop
root zone, crop development stage and the amount of water used by crop since it was last irrigated. In
this studied area, the main target is to optimize water use efficiency at farm level due to high energy
cost that consumes more fuels since it is not connected to a power. Therefore producing more crops
with a drop of water is the only option. Optimum amount of water is available for plant needs through
proper irrigation scheduling. Soil enhancement measures like proper field levelling and furrow diking
is also important to improve the efficiency of irrigation practices.
6.1.8. Aqua crop Model to determine seasonal water requirement
The approach to determine irrigation needs can done by running the Aqua Crop model for the selected
crop-soil combination. The mode is determination of net irrigation water requirement along with the
output of the generation of an irrigation schedule. The Aqua Crop model can predict the net irrigation
water requirement and the water use efficiency.
Model Input
The average climatic data, soil, crop characteristics, field and irrigation management are needed to run
Aqua Crop. In this study, fifteen year average monthly rainfall, minimum & maximum air
temperature, reference crop evapotranspiration, and the default value of CO2 were used. Reference
evapotranspiration was calculated using the FAO ETo calc programme using the monthly minimum
and maximum temperature and sun shine hours (See Table 4.2). Maize was selected for the crop
characteristics since it needs more water and also longer growing period. Determination of net
irrigation water requirement for more water consuming maize is used as a maximum margin to
determine irrigation needs of other crops in the study area for water management on farm. The
growing period was started in January. The dominant soil for the study area is clay loam. All the soil
parameters were adjusted according to the soil type, clay loam. The groundwater has variable depth. In
irrigation management, the mode ' Net irrigation water requirement' was selected and the allowable
root zone depletion was 55 % Readily Available Water (RAW).Generation of irrigation schedule was
done with 55 % allowable depletion of RAW for time criteria and back to field capacity on the depth
criteria. The irrigation method was furrow. In filed management, the soil bund was taken as 0.1m.
There was no surface runoff and mulches were not used.

Model output
Net Irrigation Requirement
The seasonal water-balance components and the net irrigation requirement of maize will be extracted
from Aqua Crop simulation. It was found that 393mm net irrigation requirement, about 538mm ETo
and 145mm rainfall for the cropping season (See Figure 6.6 and Figure 6.7 ).This net irrigation
requirement exclude extra water that has to be applied to the field to account conveyance losses or
uneven distribution of irrigation water on the field. The conveyance losses can be neglected for this
study area since the means of transport of irrigation water from the well to the field is using closed
pipes.
Figure 6.6 Net irrigation requirement from Aqua Crop model.
Valley, Ethiopia 43
Figure 6.7 Soil water balance from Aqua Crop model.
Generation of Irrigation schedule.
In the mode of generation of irrigation schedule, the allowable depletion was 55% on the time criteria
and back to field capacity on the depth criteria. The soil water balance was resulted in 404mm of
Irrigation needs after generation of irrigation events. The ETo and the rain fall values were the same as
in the mode of net irrigation requirement (See Figure 6.8).
For optimization of irrigation water use efficiency, it is better to use 404mm of irrigation needs which
later used in comparison of actually used irrigation water from the wells through pumping. The water
application efficiency for maize was calculated from Figure 6.8.

Figure 6.8 Soil water balance in generation of irrigation schedules.
6.1.9. Pumping Cost.
It is crucial to encourage farmers to keep a book for registering expenditures like cost of fuel, oil (lubricant)
and spare parts to run the pump. Organizing trainings help farmers to compare their pumping cost with
irrigated land (ha) and pumping costs to their overall production to improve irrigation management. The
pumping cost and pump detail are given in Table 6.9 and Table 6.10 respectively. The existing condition
uses 730mm of irrigation water on average from a well of 50l/s average discharge to irrigate an average of
45 ha of land. This irrigation demand consumes about 36480 litres and costs about 25536 euro annually.
However ,the irrigation demand from 80% efficiency is 584mm and uses about 29472 litres and costs about
20630 euro annually (see the existing abstracted irrigation water and pumping hours in Table 5.1).The
water application efficiency of the existing condition can be estimated by dividing 404mm to 730mm
which was resulted in 55%.
Table 6.9 Cost- Analysis of pumping in well
Jan Feb Mar Apr May Jun Annual
Amount of water
pumped(m3) 25920 51840 69120 69120 60480 51840 328320
Pumping (hr) 144 288 384 384 336 288 1824
fuel
consumption(liter) 2880 5760 7680 7680 6720 5760 36480
Fuel price(Euro/liter) 0.7 0.7 0.7 0.7 0.7 0.7
Pumping cost (Euro) 2016 4032 5376 5376 4704 4032 25536
Valley, Ethiopia 45
Table 6.10 Pump detail
Pumping discharge 50l/s
Head 100m
Pump power 92kw
Generator Capacity 247kVA
Fuel consumption 20l/h
Source: KGVDP Office
The cost benefit analysis of maize production in the current condition and new condition is shown in Table
6.11. This analysis considers only costs related to the amount of water lifted from the well. The other costs
from preparation to harvesting were assumed similar in both conditions and were not included in this
analysis due to data limitations.
Table 6.11 Fuel cost-benefit analysis of Maize production
Description Current condition New condition

Yield (ton/ha) 1.8 1.8
Area (ha) 45 45
Production (ton) 81 81
Unit price(Euro/ton) 320 320
Production cost( Euro) 25920 25920
Irrigation demand(mm) 730 584
Fuel cost (Euro) 25536 20630
Net Income (Euro) 384 5290
Assuming that all other costs and benefits remain the same and only the fuel cost as variable, The net
income from 45 ha under the current condition is low at 384 Euros because of the poor irrigation water use
efficiency (55%) that leads to higher fuel cost. The net income between the current ( 55% water use
efficiency) and the new condition (80% water use efficiency) is 4906 Euros for the whole 45 ha ( see Table
6.11) or about 110 Euros/ha.
Maize is the stable cereal crop with the highest current and potential yield from available inputs, at 2.2 tons
per hectare in 2008/09 with a potential for 4.7 tons per hectare according to on-farm field trials, when
cultivated with fertilizer, hybrid seed, and farm management practices_ (Rashid, S., K. Getnet, et al.
2010).This yield is obtained by the farmers in other regions of Ethiopia. If we assume, we reach 4.7 ton/ha,
the difference in net income from 45 ha between working at 55% and 80% efficiency will be 46826 Euros (
see Table 6.12) which is about 1000Euros/ha.

Table 6.12 Fuel cost-benefit analysis of optimized maize production
Description Current condition New condition

Yield (ton/ha) 1.8 4.7
Area (ha) 45 45
Production (ton) 81 212
Unit price(Euro/ton) 320 320
Production cost( Euro) 25920 67840
Irrigation demand(mm) 730 584
Fuel cost (Euro) 25536 20630
Net Income (Euro) 384 47210
Irrigation financial management is also a key issue for sustainability of pump-fed irrigation system. It is
crucial to guide and follow up farmers to save money for the fuel and maintenance costs even though
pumps have been donated or subsidized initially by the government. Difficulty in financial management
arises when the number of farmers increases. Up to 30 farmers is usually recommended in group-based
irrigation.
6.2. Discussions
6.2.1. Water Balance Method and Darcy's Approach
Water Balance Method
Even though one of the objectives of this study was recharge estimation in the area, it is the most
difficult tough task to evaluate groundwater resources since it needs the integration of all important
processes, surface runoff, infiltration, evapotranspiration and groundwater level variations based on
the limited data and no discharge measurement even on the main river Golina outlet. In this study, the
water balance method was used to estimate groundwater recharge with very limited data. From
chapter 4, it was found a value of 79mm/year (95MCM/year).The model simulated total recharge was
resulted in 118 MCM. There is significant difference between the recharge from water balance method
and model simulated due to the horizontal flux from the mountain aquifer as an important component
of the model simulated recharge was not considered by the water balance method. Besides this, the
runoff coefficient and soil moisture content used in water balance method was adopted from previous
studies.
Darcy's Approach
Darcy's approach estimates the flux from a head gradient and transmissivity. However, the valve for
transmissivity doesn't represent the whole area of the two sub-basins of the valley rather a specific
well value. The transmissivity of the soil is poorly identified because of heterogeneity and saturation.
In spite of these limitations, the total recharge was 83MCM from Waja-Golesha about 30MCM and
Hormat-Golina sub-basin about 53MCM.This recharge calculation was used to estimate the proposed
number of wells in each sub-basin.
6.2.2. Groundwater modelling
Model calibration was achieved through trial and error approach until the simulated head fit the
observed head values to a satisfactory degree. The calibration result indicated a reasonably match
between simulated and observed heads with RMS error of 7m (See Appendix C, Table C.1).
Valley, Ethiopia 47
From the contour map of simulated heads, it was noted that a flow direction is in agreement with the
flow of conceptual model. Therefore, the calibrated groundwater flow for this study area especially the
Hormat-Golina sub-basin was able to simulate the measured head.
The hydraulic conductivity values for the kobo valley aquifer were taken from pump test data analysis.
The main target of this test was to evaluate well properties rather than aquifer properties and the
values may be in accurate to represent the whole model area. This may increase the uncertainty in the
distribution of the parameter. Even though, the hydraulic valve obtained from pump test data analysis
indicated high spatial variation, possible effort was tried to optimize the hydraulic properties during
calibration process by considering the reasonable range of valves from literature and pump test data
analysis. The hydraulic conductivity obtained from pump test data analysis ranges from 1 to 20 md-1
,on the other hand, model calibrated values mostly ranges 0.1 to 40 md-1.
6.2.3. Water budget of the model domain
The water budget of the model domain is used to quantify and identify all flows in and out of the
aquifer structure. This water budget of the model area quantitatively evaluates the amount of
groundwater through an aquifer system. Even though, the in-flow and outflow components of
groundwater system are the most difficult to calculate directly, both components were computed by
the model. The total in-flow of the entire model area at steady-state condition was 118MCM /year
(323137m3d-1) and the out flow was also 118MCM/year.
6.2.4. Pumping Scenario
In this study, the model was run for two pumping scenarios besides the current situation and the result
was interpreted for each condition. In current condition, a total of 5192 m3d-1 was abstracted resulting
in an average decline of groundwater level at the pumping well about 7m.Similarily, a total of 27878
m3d-1 and 55825 m3d-1 abstracted water resulted in an average decline of groundwater level at the
pumping well by 14m and 32m in scenario-one and scenario-two respectively. The drop in
groundwater level is more observed where the wells are large in number and located at close distance
from each other. For example in this study, wells K6 and HG6 have a close distance of 417m and their
maximum decline of groundwater level in scenario-two was about 39m and 45m respectively.
6.2.5. Radius of Influence and well interference
According to the calculated well interference of Hormat-Golina sub-basin, 12 wells had higher
drawdown than the stabilized 20% drawdown at scenario-two (see Table D.3).Concentrated well pairs
of K6-HG6, K6-HG1 and HG6-HG8 had a radius of influence 209m, 229m and 235 respectively.
Currently, HG8 well has a discharge of 26l/s from initial discharge of 50l/s as it was reported in field
visit. This decline in discharge might be due to well interference. The model reasonably finds the
actual problem in the field.
6.2.6. Groundwater reserve and allowable exploitation
The groundwater reserve of kobo valley was calculated as 2396MCM from Hormat-Golina and Waja-
Golesha sub-basins reserve of 1206MCM and 1190MCM respectively. By taking the allowable
extraction of groundwater is to be 60% of the saturated thickness of the sediment, Hormat-Golina and
Waja-Golesha sub-basins have exploitable amount of water about 714MCM and & 724MCM
respectively. Based on the availability of exploitable groundwater amount and some basic assumptions
(see Section 6.1.6), 77 additional wells for Hormat-Golina and 75 for Waja-Golesha sub-basins are
proposed considering the minimum radius of influence is 250m and a distance between wells is to be
500m in order to locate the wells.
6.2.7. Aqua Crop model
The Irrigation need of maize was determined from Aqua Crop simulation using the mode of
generation of irrigation schedule. The seasonal soil water balance was calculated as 404mm.This
irrigation amount doesn't include conveyance losses and uneven distribution of irrigation water on the
field. In this irrigation system, the conveyances and distribution losses can be neglected since the

conveyance and distribution of water from the well to field are carried by closed pipes. The other
components, ETo and rainfall were also determined as 538mm and 145mm respectively. The
production from Aqua crop simulation was about 5 ton/ha. However, the existing production ranges
from 7 to 18 quintal/ha which is similar with 0.7 to 1.8 ton/ha which was adopted from Kobo Girana
Valley Development Project office and interviewing farmers. Maize is the stable cereal crop with the
highest current and potential yield from available inputs, at 2.2 tons per hectare in 2008/09 with a
potential for 4.7 tons per hectare according to on-farm field trials, when cultivated with fertilizer,
hybrid seed, and farm management practices.(Rashid, S., K. Getnet, et al. 2010).The existing
condition uses 730mm of irrigation water on average from a well of 50l/s discharge to irrigate an
average of 45 ha of land. Therefore, water application efficiency of this irrigation (about 55%) can be
improved up to 80% by making the right decision related to irrigation scheduling and irrigation
methods simulated by Aqua Crop. The existing water application efficiency was calculated by taking
404mm irrigation water for six month to the amount of water delivered in to the farm (730mm), where
as the new water application efficiency was calculated by taking 80% efficiency which resulted in
584mm. Hence the yield of maize will be increased up to 4.7 ton/ha and there will be a reduction of
irrigation water demand from 730mm to 584mm and net income of 46826 Euros.
6.2.8. Model Limitation
Groundwater model
Simplifications and assumptions during conceptual model development made the groundwater flow
model to have limitations to represent the real world system. Due to poor quality of the existing data
and limited data, the degree of uncertainty in the model was raised. In the modelling process like
converting the real world in to conceptual and the conceptual in to numerical model may each step
bring errors. The measured hydraulic heads were used as calibration targets. No independent measured
heads were able to validate the model. Therefore the model was calibrated but not verified. Thus, the
results from the model should not be interpreted as a perfect simulation but these results can be
interpreted as a system response within reasonable and realistic model input parameters. Because of
this limitation, the model may not be mainly used for detailed groundwater resource management
uses.
Aqua Crop model
Even though Aqua crop model have many options for simulating irrigation needs like determination of
net irrigation requirement and generation of irrigation schedules depending on the irrigation
management practices, the provision for inserting the irrigation water application efficiency is not
available for the irrigation methods in Aqua crop model except the percentage of wetting of fields.
6.2.9. Groundwater model and Aqua crop model
The main aim of modelling the study area using Processing Mod-flow and Aqua Crop was to gain a
better understanding of the groundwater system and improving irrigation water use efficiency in the
studied sub-basins respectively. As it was previously mentioned in the result, there are still
uncertainties concerning groundwater flow system and few in Aqua Crop simulations. These
uncertainties can be minimized if possible removed through comprehensive data collection together
with continued development of the models.
The groundwater model can be used for analysis of contaminant transport in the future. Since
groundwater flow directions are a crucial aspect of the numerical model, more data collection is
required to expand the knowledge relating to boundary conditions of the model domain.
Since the main focus for the development of groundwater resource is the unconsolidated sediment fill
which is expected to be the main reservoir of sub-surface water, geological and hydro geological study
of the valley should be further studied through comprehensive data collection in the future.
Valley, Ethiopia 49
CHAPTER 7 .
Conclusion and Recommendation
7.1. Conclusion
The study was conducted to provide better understanding in quantifying and abstraction of
groundwater in Hormat-Golina and Waja-Golesha sub-basins of kobo valley by applying Processing
MODFLOW. Aqua Crop model is also used to improve farm-level water management and optimize
Irrigation water use efficiency. Based on the results obtained in the study, the following conclusion
can be made.
The steady state model with pumping scenarios 11, 35 and 70 wells operating simultaneously
indicated that groundwater abstraction of 5192 m3d-1, 27878 m3d-1 and 55825 m3d-1 in the
valley resulted in an average groundwater level decline due to abstraction (at the pumping
well) of about 7m, 14m and 32m respectively. Operating more than 35 wells simultaneously,
results in a negative balance between recharge and abstraction.
Assuming the pumping cost as the only variable cost and considering the current yield of 1.8
ton/ha, improving the water use efficiency from 55% to 80% increases the net income by 109
Euros/ha. If, however, the maximum maize yield of 4.7 ton/ha (this is harvested elsewhere in
Ethiopia) is reached in the study area, the net income would increase by 1000 Euros/ha.
Kobo valley has exploitable groundwater reserve of 714MCM for Hormat-Golina and
724MCM Waja-Golesha sub-basins by considering 60% of the saturated thickness. Based on
this amount of water, 77 wells for Hormat-Golina and 75 wells for Waja-Golesh sub-basins
could be added without creating negative balance between recharge and abstraction.
The steady-state groundwater model set up and calibration put a better approach how to
translate the conceptual model to the numerical model and realize the aquifer system of kobo
valley. This is also useful for transient state groundwater modelling in the future study.
The calibrated steady-state groundwater flow model was able to reasonably simulate the
hydraulic heads that match the measured heads. Moreover, the model simulated groundwater
level contours indicated that the general hydraulic gradient in the valley follows the surface
topography and agrees with the groundwater flow system defined in the conceptual model.

7.2. Recommendation
Optimum simultaneous operation of a maximum of 35 wells is recommended keeping the
minimum distance between two closer wells at 500m so as to avoid overexploitation and
ensure sustainable use of groundwater resources.
Kobo Girana Valley Development Project should create awareness to farmers about the real
threat of very low water use efficiency and the resulting high pumping costs on sustainability
of the well-based irrigation. The Project should work towards investment in power connection
to minimize the reliance on fuel.
Kobo Girana Valley Development Project should create awareness of the very low (1.8
ton/ha) maize yield in Kobo and introduce the already proven agronomic and farming
practices that have led to a maximum of 4.7 ton/ha in other regions of Ethiopia.
The output from this study can be used as a starting point for transient state groundwater
modelling for better predictions of pumping effect and for better recharge simulation since
recharge and groundwater outflow are strongly time dependent.
Groundwater abstraction from irrigation as well as from water supply boreholes should be
reported and recorded in data base periodically in order to evaluate seasonal and annual
variations.
Since there is no gauging stations along the main river like Golina and Hormat, river gauging
stations together with Meteorological stations in the west mountain should be installed in
order to improve data availability and better understanding of the sub-basins water balance.
Geological and hydro geological study of the valley should be further studied through
comprehensive data collection in the future since the main focus for the development of
groundwater resource is the unconsolidated sediment fill which is expected to be the main
reservoir of sub-surface water.
Interested professionals can use MODFLOW for similar areas of interest.
Valley, Ethiopia 51
References
Allen, G.R., Pereira, L.S., Raes, D. and Smith, M., (1998) Crop evapotranspiration-guidelines for
computing crop water requirements. FOA Irrigation and drainage paper 56. FAO, Rome, Italy
Anderson, M. P., and Woessner, W. W. (1992) Applied Groundwater Modeling. Acad. Press, San
Diego, USA.
Bear, J., (1979) Analysis of flow against dispersion in porous media- Comments. Jour. Hydrology,
40(3-4): 381- 385.
Bos, M. G., & Nugteren, J. (1974). On irrigation efficiencies (Vol. 19). Wageningen: International
Institute for Land Reclamation and Improvement.
Chiang, W.H., and Kinzelbach, W., (2001) Processing MODFLOW (Version 5.3), A simulation
system for modeling groundwater flow and pollution
Chiang, W.H., and Kinzelbach, W., (2003) Processing MODFLOW Pro (Version 7.0.17), A
simulation system for modeling groundwater flow and pollution.
Chiang, W.H., Kinzelbach, W. and Rauch, R., (1998) Aquifer simulation model for Windows
ASMWIN 6.0: Groundwater flow and transport modeling, an integrated program.Gebruder
Borntraeger Berlin etc., 137 pp.
Co-SAERAR (1997). Feasibility study of Report for the Kobo-Girana Valley Development Study
Project, Volume III: Annex C: Hydrogeology
Co-SAERAR (1997). Feasibility study of Report for the Kobo-Girana Valley Development Study
Project. Volume III: Annex B: Regional geology.
Co-SAERAR (1999). Feasibility study report for the Kobo-Girana Valley Development Study Project,
Volume III: Annex C: Hydrogeology. The Commission for Sustainable Agriculture and
Environmental Rehabilitation in Amhara Region.
Deng, X. P., Shan, L., Zhang, H., & Turner, N. C. (2006). Improving agricultural water use efficiency
in arid and semiarid areas of China. Agricultural water management.
Fetter, C.W., (2001). Applied hydrogeology + Visual MODFLOW, Flow net and Aqtesolv student
version software on CD-ROM. Prentice Hall, Upper Saddle River, 597pp.
Freeze, R.A. & Cherry, J.A., (1997) Groundwater. Prentice-Hall, Englewood Cliffs, N.J., USA. 604pp
Foster, S., (1998) Groundwater: assessing vulnerability and promoting protection of a threatened
resource. Proceedings of the 8th Stockholm Water Symposium, 10-13 August, Sweden pp.79-
90
G.Castany, E.Groba, E.Romijn A new to understanding and Estimating Natural Recharge

German Agency for Technical Co-operation (1977) Kobo-Alamata agricultural development program.
Volume II: Annexes pedology and water development. German consults.
Hagos, M. A. (2010). Groundwater Flow Modelling assisted by GIS and RS Techniques (Raya
Valley-Ethiopia).
Hsiao, T. C., Steduto, P., & Fereres, E. (2007). A systematic and quantitative approach to improve
water use efficiency in agriculture. Irrigation science, 25(3), 209-231.
Jyrkama, M.I., and Sykes, J.F., (2007) the impact of climate change on spatially varying groundwater
recharge in the grand river watershed (Ontario). Jour. Hydrology, 338(3-4): 237-250.
Kruseman, G.P. and de Ridder, N.A., (1992). Analysis and evaluation of pumping test data.
International Institute for Land Reclamation and Improvement /ILRI, Netherlands, 377pp.
McDonald, M. G., and Harbaugh, A. W. (1988) A modular three-dimensional finite-difference

groundwater flow model. Techniques of Water-Resources Investigations 06- A1, USGS, 576pp.
Metaferia Consulting Engineers (2009). Hydro geological investigation report for Kobo- Girana
Pressurized irrigation project Final report. Unpublished Report.
Muskat, M. (1937). The Flow of Homogeneous Fluids through Porous Media. New York: McGraw-
Hill Book Company Inc.
Shaw, E.M. (1994). Hydrology in practice. Third edition, Chapman & Hall.
Sileshi Mamo (2007). Raya hydrogeology and isotope hydrological investigation project, final report.
Ministry of Mines and Energy Geological Survey of Ethiopia.
Todd, D.K., (2005). Groundwater hydrology. Wiley, Hoboken, NJ, 636pp.
Rashid, S., K. Getnet, et al. (2010). "Maize Value Chain Potential in Ethiopia." Globalization and
Markets program.
RVDP (1998). Hydro geological study phase II feasibility draft report. Raya Valley Development
study Project Tigray National Regional Government
Valley, Ethiopia 53
Appendix A Metrological Data
Table A.1 Monthly Rainfall (mm) at Kobo station
Year Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Annual
1996 79.00 0.00 31.00 75.00 133.00 53.00 147.00 205.00 67.00 18.00 64.00 0.00 872.00
1997 0.00 0.00 34.20 57.60 47.80 51.50 110.60 94.40 37.40 169.30 43.30 0.00 646.10
1998 53.90 1.60 24.10 38.90 11.70 3.80 316.10 311.80 50.90 6.10 0.00 0.00 818.90
1999 20.30 0.00 2.90 48.20 11.70 2.20 234.30 315.90 0.00 0.00 0.00 0.00 635.50
2000 0.00 0.00 1.50 76.10 42.00 3.70 226.70 240.00 48.10 87.80 24.10 83.40 833.40
2001 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 21.40 0.00 178.50 0.00 199.90
2002 22.30 2.50 499.00 493.90 13.00 2.90 100.10 0.00 0.00 16.00 0.00 66.60 1216.30
2003 41.20 32.70 34.80 60.00 32.80 11.00 140.00 258.00 34.70 0.00 0.00 42.80 688.00
2004 29.90 0.00 28.10 87.90 3.00 25.90 116.20 163.40 9.20 79.20 35.50 10.20 588.50
2005 8.30 0.00 34.00 158.20 125.60 2.80 131.10 190.90 42.40 8.70 64.80 0.00 766.80
2006 0.00 10.80 56.20 56.60 16.20 3.00 81.20 222.90 75.40 12.50 0.00 0.00 534.80
2007 17.80 11.80 36.90 46.60 7.70 29.00 80.00 220.70 70.00 10.00 0.00 0.00 530.50
2008 10.00 12.00 9.00 20.00 35.00 8.00 100.00 135.00 145.40 5.40 31.90 0.00 511.70
2009 6.00 5.00 6.70 40.00 50.00 1.40 145.90 123.90 6.60 51.60 22.10 51.70 510.90
2010 0.00 10.50 4.30 63.80 64.30 1.80 239.40 316.60 33.90 14.70 0.00 0.50 749.80
mean 19.25 5.79 53.51 88.19 39.59 13.33 144.57 186.57 42.83 31.95 30.95 17.01 673.54
Table A.2 Monthly Rainfall (mm) at Alamata Station
1996 132.90 0.00 69.20 123.40 115.20 25.00 76.50 250.00 36.30 8.00 58.20 0.00 894.70
1997 46.10 0.00 125.10 26.70 28.50 22.00 87.00 54.40 72.00 191.60 139.00 0.00 792.40
1998 179.00 23.40 25.50 35.20 19.50 0.00 348.00 271.90 63.50 17.60 0.00 0.00 983.60
1999 44.30 0.00 20.80 9.00 7.00 1.00 211.40 431.80 66.70 54.50 0.00 0.00 846.50
2000 0.00 0.00 10.00 43.50 74.00 0.00 246.20 450.10 68.40 14.80 83.30 72.80 1063.10
2001 0.00 0.00 157.90 12.80 29.50 16.80 224.80 244.30 24.80 10.00 10.00 2.50 733.40
2002 98.40 0.00 18.00 112.30 8.00 3.50 72.60 213.50 46.10 13.50 0.00 89.50 675.40
2003 75.80 69.50 41.90 94.20 24.50 12.70 111.80 234.20 22.80 0.00 0.00 66.90 754.30
2004 33.00 16.00 39.60 168.00 13.50 49.50 117.00 243.00 41.10 8.20 21.00 20.00 769.90
2005 21.30 1.40 110.30 131.60 65.80 24.00 141.50 167.00 33.10 6.00 0.00 0.00 702.00
2006 0.00 0.00 215.50 176.10 4.50 10.00 123.20 192.00 54.00 2.40 0.00 23.50 801.20
2007 12.30 46.30 8.40 109.00 20.00 15.00 165.10 214.70 50.90 0.00 0.00 0.00 641.70
2008 25.80 2.90 0.00 6.60 21.10 11.00 79.20 206.20 60.80 53.10 55.00 0.00 521.70
2009 0.00 0.00 0.00 89.40 0.00 0.00 93.20 68.00 0.00 0.00 27.50 10.00 288.10
2010 0.00 41.00 170.40 29.10 56.50 28.80 229.60 320.90 27.80 26.30 0.00 0.00 930.40
mean 44.59 13.37 67.51 77.79 32.51 14.62 155.14 237.47 44.55 27.07 26.27 19.01 759.89

Table A.3 Monthly Rainfall (mm) at Zoble Station
1996 40.00 12.00 90.00 100.00 13.00 10.00 250.00 312.00 110.00 25.00 0.00 0.00 962.00
1997 20.00 8.00 15.00 10.00 4.00 30.00 180.00 320.00 60.00 6.00 0.00 0.00 653.00
1998 30.00 18.00 22.00 28.00 0.00 45.00 260.10 300.00 43.50 46.40 0.00 0.00 793.00
1999 45.10 0.00 54.50 51.50 18.80 22.40 161.70 336.60 121.60 83.00 0.00 1.80 897.00
2000 11.60 0.00 7.10 106.50 73.50 11.10 280.00 358.50 48.40 167.50 78.30 1142.50
2001 35.00 1.60 62.90 7.30 111.90 21.80 216.80 200.00 60.00 50.00 9.50 2.90 779.70
2002 131.10 7.90 23.20 147.30 45.60 6.40 285.80 285.80 90.00 20.00 0.00 5.00 1048.10
2003 38.10 71.50 122.20 38.00 45.00 60.00 234.40 387.20 71.00 30.00 10.00 0.00 1107.40
2004 20.00 33.70 24.00 220.30 0.00 30.80 106.10 235.30 45.50 99.70 69.40 4.70 889.50
2005 10.00 25.00 80.40 109.30 83.50 1.40 171.50 137.10 42.40 0.00 50.30 0.00 710.90
2006 2.90 2.00 57.90 55.20 18.50 0.00 100.20 163.50 41.60 49.00 58.70 0.00 549.50
2007 82.00 8.00 35.00 120.90 4.50 79.60 294.40 368.90 128.00 21.20 9.30 0.00 1151.80
2008 20.20 10.00 0.00 5.80 14.50 15.60 108.50 287.20 132.40 85.00 238.90 0.00 918.10
2009 10.00 0.00 9.50 0.00 22.20 0.00 156.30 117.90 32.50 21.50 0.00 67.10 437.00
2010 5.00 15.00 30.00 50.00 18.00 4.00 120.00 180.00 150.10 15.20 0.00 0.00 587.30
mean 33.40 14.18 42.25 70.01 31.53 22.54 195.05 266.00 78.47 47.97 34.96 5.82 842.17
Table A.4 Monthly Rainfall (mm) at Korem station
1996 35.90 0.60 98.10 241.80 67.90 137.10 63.10 11.40 72.70 0.00 728.60
1997 9.90 2.00 97.70 50.70 92.90 63.60 172.10 55.90 58.80 322.60 164.10 1.10 1091.40
1998 159.60 38.50 26.70 27.80 80.20 12.40 397.20 355.80 212.30 45.60 0.20 0.00 1356.30
1999 65.80 0.00 2.80 49.80 12.80 32.10 234.90 350.10 112.20 44.50 2.20 0.40 907.60
2000 0.00 0.00 3.50 48.20 76.20 9.30 317.60 321.00 90.80 133.20 65.60 93.80 1159.20
2001 1.90 3.20 130.10 22.10 36.10 50.90 282.80 380.10 62.40 13.20 2.00 13.30 998.10
2002 65.30 0.70 34.10 107.00 15.60 3.00 137.20 228.60 90.30 8.00 0.00 92.60 782.40
2003 13.00 24.00 74.20 74.80 23.70 20.10 168.00 381.10 77.50 1.80 3.90 30.10 892.20
2004 13.90 6.00 40.90 55.00 1.90 54.00 143.60 249.10 65.50 34.80 21.90 9.90 696.50
2005 8.30 0.00 106.10 223.40 163.40 28.10 253.00 297.60 40.60 36.50 0.00 0.00 1157.00
2006 9.90 1.00 182.20 96.20 44.60 6.90 149.00 307.40 51.70 60.10 0.00 0.00 909.00
2007 58.10 21.00 68.80 40.00 15.00 9.00 313.40 388.50 32.60 10.00 27.50 0.00 983.90
2008 64.20 0.00 0.00 16.60 69.80 35.80 187.10 143.30 103.40 78.00 137.00 0.00 835.20
2009 2.20 0.00 44.20 49.30 4.70 6.50 230.60 176.20 32.20 67.70 0.00 0.00 613.60
2010 0.60 5.30 42.00 111.70 61.60 11.00 311.70 427.60 64.10 7.70 0.00 27.00 1070.30
mean 33.91 6.82 63.43 80.96 49.89 27.37 229.02 290.16 77.17 58.34 33.14 17.88 968.09
Valley, Ethiopia 55
Table A.5 Monthly Rainfall (mm) at Lalibela station
1996 15.7 1.8 116.6 46.6 56.2 144.9 328.2 287.9 19.6 0.0 35.0 0.8 1053.3
1997 5.9 8.4 95.6 58.7 24.3 104.5 279.0 162.0 24.9 100.3 100.9 2.6 967.1
1998 19.3 7.7 26.6 11.8 44.0 15.8 337.0 258.6 58.1 21.5 0.0 0.0 800.4
1999 0.0 0.0 0.0 22.3 0.7 20.2 333.7 315.5 50.4 22.4 3.3 1.8 770.3
2000 0.0 0.0 25.9 79.9 13.5 16.2 213.1 206.3 71.1 80.7 0.0 0.0 706.7
2001 0.0 21.7 82.3 31.4 1.9 126.8 339.6 382.3 7.0 0.0 0.0 7.1 1000.1
2002 34.4 19.4 45.6 34.7 6.8 51.8 269.3 245.1 55.7 0.0 1.5 12.8 777.1
2003 2.0 19.7 44.2 58.1 1.8 55.5 203.9 426.7 55.2 0.0 1.6 0.0 868.7
2004 0.0 0.0 0.0 0.0 0.0 0.0 242.4 199.5 19.0 10.2 0.8 0.0 471.9
2005 6.2 21.4 53.4 27.1 57.0 36.4 328.1 165.7 40.2 0.5 0.0 0.0 736.0
2006 0.0 50.3 54.4 43.8 20.3 23.9 301.9 323.7 43.6 25.6 9.6 15.3 912.4
2007 39.5 10.7 5.3 21.7 23.3 195.7 250.4 197.9 88.7 0.3 3.8 0.0 837.3
2008 2.0 1.8 0.9 60.6 17.4 54.6 233.4 210.8 91.6 14.1 38.5 2.3 728.0
2009 0.8 2.8 29.8 10.4 2.3 5.2 239.3 203.4 20.2 24.6 22.9 0.0 561.7
2010 4.4 0.0 16.5 43.1 19.3 9.3 261.7 365.0 35.4 0.5 0.8 15.7 771.7
mean 8.7 11.0 39.8 36.7 19.3 57.4 277.4 263.4 45.4 20.0 14.6 3.9 797.5
Table A.6 Monthly Tmax (0c) at Kobo station
1996 30.40 30.10 32.10 31.70 30.20 30.40 29.70 27.10 26.10 29.76
1997 25.40 27.90 29.00 28.00 33.00 33.30 31.30 31.40 32.40 27.60 27.10 27.20 29.47
1998 25.40 30.30 28.10 32.40 33.50 36.00 30.80 28.70 29.80 29.30 28.30 27.60 30.02
1999 26.80 30.40 29.00 32.10 34.20 35.30 30.50 29.70 31.00
2000 29.40 30.40 30.90 33.80 34.70 32.10 30.30 30.50 28.60 27.40 26.10 30.38
2001 29.40 30.60 30.40 28.40 28.40 29.44
2002 26.00 28.40 29.70 31.50 34.20 34.90 34.00 30.50 29.50 26.40 30.51
2003 25.90 29.00 30.20 30.90 34.00 34.60 32.20 30.00 30.80 30.20 29.20 25.90 30.24
2004 27.50 27.60 29.20 30.50 34.60 34.00 31.00 31.50 31.90 29.90 29.70 27.10 30.38
2005 26.89 30.24 31.33 31.52 31.79 35.11 32.69 32.23 27.75 30.77 29.20 28.32 30.65
2006 28.57 29.80 30.81 30.51 33.67 35.46 32.75 31.04 30.13 30.46 27.37 30.96
2007 25.37 28.87 31.38 31.45 34.79 34.45 30.25 28.30 27.40 30.25
2008 30.01 29.40 26.88 26.78 28.27
2009 30.58 35.95 32.28 31.50 32.23 29.65 29.60 26.61 31.05
2010 26.84 28.28 29.17 31.57 32.58 35.22 31.57 29.67 30.16 30.38 28.21 26.85 30.04
mean 26.47 29.11 29.91 30.98 33.35 34.70 31.91 30.45 30.56 29.76 28.38 27.01 30.16

Table A.7 Monthly Tmax. (o c) at Alamata Station
1996 29.50 30.90 30.80 30.80 32.60 32.40 31.20 31.60 31.20 29.00 28.40 30.76
1997 27.40 29.90 28.90 30.30 34.40 34.50 33.10 33.90 33.30 29.80 28.60 28.70 31.07
1998 27.00 27.70 30.50 34.50 34.50 36.10 31.80 28.80 30.70 30.80 30.40 29.80 31.05
1999 28.80 32.40 31.00 34.10 35.80 30.90 30.10 30.10 30.10 30.00 29.20 31.14
2000 29.70 31.60 32.80 34.20 35.20 36.80 33.20 30.50 30.50 29.30 29.00 26.80 31.63
2001 27.50 30.40 30.50 31.70 34.10 33.80 31.20 28.90 29.60 30.00 27.90 27.50 30.26
2002 24.70 28.00 29.80 30.40 33.70 34.30 33.60 30.60 29.40 30.70 28.80 26.30 30.03
2003 25.80 28.50 29.00 30.10 33.20 34.20 31.40 29.40 30.10 29.90 28.40 26.00 29.67
2004 27.10 27.10 29.50 28.60 33.40 32.80 31.90 29.60 29.50 29.10 28.40 26.30 29.44
2005 25.90 29.40 29.60 29.80 30.30 33.40 31.00 29.90 30.00 29.30 28.40 27.20 29.52
2006 27.30 29.50 28.50 28.00 31.30 34.40 30.50 29.70 29.90 29.90 28.60 27.60 29.60
2007 25.10 27.60 30.60 30.50 30.30 30.60 29.80 26.30 28.85
2008 26.90 27.70 30.80 32.10 32.90 32.50 32.10 31.90 32.10 30.10 27.30 26.70 30.26
2009 26.50 29.90 31.10 30.90 31.40 32.70 33.40 32.60 32.50 31.80 31.28
2010 26.30 25.40 27.10 28.10 24.70 23.30 25.82
mean 26.90 29.23 30.25 31.14 33.15 34.01 31.54 30.18 30.47 29.99 28.42 27.15 30.20
Table A.8 Monthly Tmax. (o c) at Korem station
1996 19.00 22.10 22.40 22.30 22.40 23.70 22.80 23.50 24.50 23.90 21.20 19.90 22.31
1997 19.80 21.60 22.30 22.40 24.90 24.70 23.30 24.10 24.90 21.40 21.00 21.10 22.63
1998 19.70 20.80 22.10 24.00 24.40 26.10 22.50 21.90 23.30 21.50 20.10 19.90 22.19
1999 19.30 22.30 21.70 24.20 25.90 27.30 22.00 22.70 24.30 25.10 24.60 21.50 23.41
2000 21.60 22.40 23.00 23.50 25.30 27.20 22.90 21.60 22.20 21.40 20.00 18.70 22.48
2001 18.00 20.30 20.30 22.50 24.40 24.50 22.70 21.60 22.10 21.30 19.60 19.90 21.43
2002 17.90 20.50 21.60 22.50 25.60 26.20 25.20 22.40 21.80 22.10 21.70 20.10 22.30
2003 20.30 22.40 22.40 22.60 24.60 25.50 22.40 21.70 22.30 21.40 20.90 19.40 22.16
2004 21.10 20.80 22.20 22.80 26.10 25.10 23.50 22.60 23.00 21.40 20.90 19.50 22.42
2005 19.90 22.90 22.70 22.80 23.00 25.30 22.80 23.20 23.20 22.10 21.10 20.30 22.44
2006 21.00 22.40 22.00 21.30 24.30 25.90 23.20 22.60 23.00 22.50 21.30 20.00 22.46
2007 18.50 21.40 23.50 23.50 23.20 22.80 23.40 22.30 19.60 22.02
2008 20.10 20.00 23.10 23.80 24.10 24.60 22.80 22.00 22.20 20.70 19.70 18.90 21.83
2009 20.20 20.90 22.30 22.70 24.50 26.90 22.60 22.40 23.30 21.90 22.30 22.73
2010 20.40 21.30 21.80 22.60 23.90 25.90 23.10 21.40 21.80 22.00 22.20 21.60 22.33
mean 19.79 21.47 22.23 22.90 24.53 25.64 23.00 22.43 23.02 22.05 21.26 20.03 22.36
Valley, Ethiopia 57
Table A.9 Monthly Tmin. (0c) at Kobo station
1996 16.50 17.40 18.20 17.50 15.70 12.00 11.70 10.20 14.90
1997 13.30 12.30 15.90 16.70 17.50 18.70 18.80 17.80 16.30 15.50 15.60 11.80 15.85
1998 15.40 xx 17.20 18.10 17.90 20.70 17.90 17.20 16.10 14.50 10.10 8.40 15.77
1999 11.60 11.10 15.50 16.00 17.50 18.20 17.30 16.20 15.43
2000 10.60 14.20 16.50 15.60 15.90 14.90 11.80 7.10 9.40 7.70 7.20 11.90
2001 0.00
2002 14.50 13.00 16.00 16.40 16.90 19.20 19.60 xx xx 13.20 12.10 15.20 15.61
2003 13.50 15.00 16.30 17.00 17.70 19.30 19.40 17.30 16.80 12.60 12.50 11.50 15.74
2004 14.90 15.00 14.40 16.80 16.50 18.30 18.10 17.70 15.60 12.20 12.70 13.90 15.51
2005 13.93 14.10 13.92 17.17 18.06 18.70 18.79 17.74 12.99 13.46 12.57 9.92 15.11
2006 13.08 15.59 15.86 16.37 18.13 19.92 19.10 17.62 16.08 14.89 15.26 16.54
2007 14.34 15.73 14.09 10.98 14.33 18.95 16.92 12.50 9.20 14.12
2008 16.31 14.02 12.47 11.38 13.55
2009 16.11 20.30 18.48 18.30 16.51 14.68 12.51 15.14 16.50
2010 13.37 15.39 16.25 17.85 17.84 20.33 18.86 17.32 16.06 13.94 11.47 11.24 15.83
mean 13.79 13.78 15.48 16.35 17.04 18.92 18.29 16.95 15.05 13.37 11.99 11.57 15.21
Table A.10 Monthly Tmini. (o c) at Alamata station
1996 12.90 16.00 16.40 16.10 16.10 16.70 16.30 15.50 13.90 12.70 11.50 14.92
1997 12.90 12.90 15.80 16.20 17.60 18.60 17.60 17.50 17.00 15.90 16.20 12.80 15.92
1998 14.90 15.00 16.60 17.90 18.10 19.90 17.07
1999 12.40 11.90 13.30 17.30 20.20 18.80 16.70 10.20 5.70 3.90 1.90 12.03
2000 2.30 5.30 6.60 7.90 10.10 11.40 11.60 9.20 8.70 7.90 6.50 6.30 7.82
2001 3.90 6.10 8.30 12.20 13.90 11.80 10.70 10.40 10.80 7.80 6.90 9.35
2002 9.50 14.40 17.00 17.70 18.60 20.00 19.70 16.70 16.70 15.90 15.40 16.00 16.47
2003 14.30 15.70 17.40 18.00 20.30 20.00 19.40 17.00 17.40 15.90 15.00 12.90 16.94
2004 14.60 8.80 10.10 16.10 18.00 19.50 18.70 17.80 17.10 15.40 15.20 15.10 15.53
2005 14.40 15.30 17.20 17.70 17.90 19.10 18.90 16.30 16.60 15.60 12.00 12.10 16.09
2006 14.00 16.20 15.50 16.40 17.80 19.10 17.60 16.30 16.50 16.50 15.80 15.60 16.44
2007 14.70 16.00 16.50 16.90 17.40 16.10 15.40 13.60 15.83
2008 13.70 13.70 14.40 15.50 16.00 15.70 14.90 13.70 14.00 13.80 13.90 13.50 14.40
2009 15.70 15.10 16.00 16.00 15.80 16.20 16.70 15.30 15.60 14.10 13.80 15.48
2010 11.80 10.20 11.60 11.90 13.50 12.50 9.00 8.70 11.10 13.90 13.30 11.10 11.55
mean 12.08 12.63 14.15 15.85 16.59 17.08 16.34 14.78 14.49 13.62 12.42 11.48 14.29

Table A.11 Monthly Tmini. (oc) at Korem Station
1996 7.90 5.70 9.90 10.20 11.80 12.00 9.20 4.10 4.50 3.20 7.85
1997 5.70 5.70 9.20 8.70 8.70 11.30 12.40 11.40 7.10 8.70 10.00 4.20 8.59
1998 8.50 7.50 9.60 9.30 10.20 9.80 11.90 11.80 10.20 7.00 1.10 -0.60 8.03
1999 3.40 1.70 6.50 7.40 8.60 9.70 11.60 10.60 8.40 6.90 0.40 2.30 6.46
2000 2.50 1.80 5.60 9.50 8.90 10.80 12.30 11.50 8.50 6.80 5.30 4.40 7.33
2001 3.30 3.30 8.50 7.20 9.20 11.80 12.30 11.90 8.20 6.80 2.50 2.00 7.25
2002 6.30 4.50 8.30 8.60 7.10 11.00 11.70 11.10 8.30 4.50 3.00 7.20 7.63
2003 4.30 2.20 3.80 10.10 9.80 11.20 13.10 12.10 9.90 4.10 4.50 2.80 7.33
2004 6.70 5.80 7.40 10.60 7.80 11.30 12.50 12.10 8.00 6.00 5.00 6.20 8.28
2005 7.30 5.40 9.70 9.80 10.50 10.40 12.60 11.80 9.90 5.00 6.90 9.03
2006 4.10 6.90 8.80 9.90 10.00 11.20 12.70 11.90 9.40 8.20 5.80 9.10 9.00
2007 8.30 9.30 7.00 10.30 12.50 11.90 10.00 4.40 1.30 8.33
2008 5.30 3.20 3.40 8.30 10.10 11.00 12.40 11.80 9.10 5.50 4.90 2.60 7.30
2009 3.80 4.40 7.60 8.00 7.10 9.30 11.50 11.20 6.70 3.80 1.80 6.84
2010 4.80 6.30 7.50 9.90 10.30 11.80 12.30 12.10 9.50 5.40 2.60 4.80 8.11
mean 5.48 4.91 7.52 9.19 9.10 10.82 12.24 11.68 8.83 5.91 3.99 4.03 7.81
Table A.12 Monthly average wind speed (m/s) at Kobo Station
1996 2.60 2.00 2.10 2.00 1.60 1.80 2.10 1.70 1.20 1.30 1.50 1.60 1.79
1997 1.80 2.00 2.10 2.00 2.20 1.90 2.00 1.70 1.60 1.40 1.30 1.40 1.78
1998 1.60 1.70 2.00 2.20 2.00 2.50 2.20 1.60 1.10 1.30 1.40 1.50 1.76
1999 1.60 1.80 2.20 2.10 2.10 2.30 2.00 1.50 0.00 0.00 0.00 0.00 1.30
2000 1.60 1.80 2.20 1.80 1.70 2.20 2.20 2.00 1.50 0.00 0.00 0.00 1.42
2001 1.50 1.70 1.70 1.40 1.20 1.40 0.50 0.80 0.20 1.60 1.70 1.80 1.29
2002 1.80 1.80 1.90 1.80 1.40 2.30 2.10 1.80 0.90 1.00 1.30 1.50 1.63
2003 1.60 1.70 2.00 1.90 1.50 2.00 2.00 1.40 1.00 1.10 1.30 1.30 1.57
2004 1.70 1.90 2.00 2.00 1.80 2.10 2.20 1.50 1.10 1.20 1.30 1.60 1.70
2005 2.00 2.10 2.20 2.00 1.60 1.90 2.00 1.80 1.20 1.20 1.30 1.50 1.73
2006 1.50 1.80 2.00 1.50 1.60 2.00 2.20 1.80 1.20 1.40 1.50 1.60 1.68
2007 1.80 1.50 1.80 2.00 1.40 1.80 2.00 1.60 0.90 1.00 1.20 1.40 1.53
2008 1.70 1.80 2.00 1.80 1.80 1.50 2.10 1.40 1.00 1.40 1.40 1.60 1.63
2009 2.00 1.80 2.10 1.70 1.50 2.00 1.80 1.80 1.10 1.00 1.20 1.50 1.63
2010 1.80 2.00 2.00 1.80 2.00 1.80 2.00 1.50 1.20 1.20 1.40 1.60 1.69
mean 1.77 1.83 2.02 1.87 1.69 1.97 1.96 1.59 1.01 1.07 1.19 1.33 1.61
Valley, Ethiopia 59
Table A.13 Monthly average sunshine (hour) at Kobo station
1996 6.80 5.30 7.90 9.60 9.90 8.50 4.00 5.10 6.80 9.60 9.90 8.80 7.68
1997 8.00 7.30 9.30 6.90 9.40 6.20 3.10 5.10 7.30 9.50 9.70 8.10 7.49
1998 6.30 7.50 8.90 9.50 9.20 6.80 6.70 7.40 6.00 7.10 9.00 7.00 7.62
1999 6.90 4.40 8.70 6.60 9.10 6.30 6.10 6.50 7.00 9.60 10.10 9.30 7.55
2000 9.30 7.20 6.10 7.60 9.10 5.10 5.20 5.10 5.10 8.20 9.30 8.70 7.17
2001 8.00 7.40 8.90 7.90 9.80 8.30 3.40 6.60 7.10 7.90 8.10 7.90 7.61
2002 7.80 8.60 8.70 9.70 6.50 5.70 4.00 5.80 7.70 8.90 9.10 8.50 7.58
2003 8.70 10.20 8.10 9.00 6.80 6.40 8.30 7.10 6.70 7.30 8.60 9.00 8.02
2004 8.10 8.70 9.00 6.80 7.20 6.10 7.30 5.60 6.20 8.10 9.90 9.40 7.70
2005 8.00 8.00 9.00 9.20 9.00 6.00 5.00 5.60 7.20 9.00 9.70 8.80 7.88
2006 7.00 8.00 8.10 9.00 9.40 5.60 4.00 6.00 7.00 8.00 9.60 9.00 7.56
2007 6.50 7.50 8.00 9.60 9.10 7.00 3.50 5.50 6.80 8.40 9.00 8.50 7.45
2008 8.10 8.60 9.00 8.00 9.80 7.50 6.00 5.00 7.00 8.00 9.40 8.60 7.92
2009 9.00 10.00 8.00 9.20 9.40 6.40 4.60 6.20 6.90 7.00 10.00 9.20 7.99
2010 8.00 8.50 9.00 8.60 7.60 5.70 6.00 5.60 6.70 8.60 9.00 8.00 7.61
mean 7.77 7.81 8.45 8.48 8.75 6.51 5.15 5.88 6.77 8.35 9.36 8.59 7.65

Appendix B Pump Test Data
Table B.1 Hydraulic parameters of the alluvial sediment aquifer
Location
Well Observation Depth SWL DWL Q DD Tav Kav
ID X Y Z well (m) (m) (m) (l/s) (m) (m2/d) (m/d) S Sy
HG-1 568171 1339140 1480 112 20.59 41.49 51 20.9 225 2.3
HG-2 569552 1339024 1461 91 15.1 30.6 51 15.5 433 6.2
HG3 569659 1338130 1455 111 21 59.1 20 38.1 25.06 0.84
HG4 569354 1339493 1466 109 17.55 33.21 51 15.66 259.2 7.2
HG-5 571782 1333845 1429.0 112 19.97 34.4 50 14.43
HG-6 567804 1339909 1495.0 101 24.83 36.09 50 11.26
HG-7 568283 1340339 1487.5 105.5 20.7 52.75 50 32.05
HG-8 567346 1340-10 1502.0 110 28.73 38.4 50 9.67
HG-9 569905 1339618 1461.0 100 26.1 80.9 10 54.8
HG-10 570348 1339366 1455.0 100 24.2 61.89 34 37.69
HG11 571055 1335915 1437 116.5 14.4 35.17 50 20.77 230.5 5.52
HG12 572295 1335804 1417 110.3 16.3 30.72 50 14.42 218.5 5.2
HG13 571683 1336365 1425 110.6 18.26 31.61 50 13.35 239.6 5.7
HG14 571067 1336466 1436 108.5 16.66 29.53 50 12.87 318.2 7.58
HG-15 574995 1330597 1412.0 117 8 19.14 50 11.14
HG-16 574870 1331228 1412.0 99 8.5 44.05 25 35.55
HG-17 574671 1331878 1405.3 120 10.12 26.65 50 16.53
HG-18 574472 1332357 1399.0 119 9.25 25.05 50 15.8
Valley, Ethiopia 61
Zeleke1 570187 1338097 1452 113 20.4 37.9 50 17.5 276
Zeleke2 570658 1337490 1446 110.5 19.05 38.53 50 19.48 345
PHG1 567688 1338578 1487 PHG1 129 24.35 29.24 50 4.89 1245 13.53 0.55
PHG2 567801 1337977 1479 PHG2-OB1 150 18.92 19.41 45 0.49 3145 44.9 0.28
3.10E-
PHG3 568356 1337982 1473 PHG3 156 16.63 25.89 50 9.26 513.5 4.94 02
PHG4 566854 1339244 1507 PHG4-OB1 128 27.96 32.39 45 4.43 1412 22.75 0.31
PHG5 571398 1335248 1432 155 21.5 40.01 57 18.51 244 8.13 0.25
PHG6 571821 1334963 1426 PHG6-OB1 178 20.13 25.85 58.5 5.72 1017 23.1 0.29
PHG7 572289 1334951 1419 180 17.53 33.04 51.6 15.51 255 4.7 0.2
PHG8 570553 1334124 1447 147 20.61 37.93 46.4 17.32 267.5 5.6 0.22
PHG9 570089 1333952 1456 PHG9-OB1 158 21.67 27.25 53.5 5.58 1427 29.8 0.3
PHG10 569560 1334010 1462 146 25.1 30.88 59.5 5.78 1412 31.4 0.25
TW1 568755 1329590 1367 114 58.78 62.69 7 3.91 236
TW3 579090 1332043 1375 81 13.3 38.01 29 24.71 140
THG1 576123 1336656 1384 THG1-OB1 118 16.21 31.02 32 14.81 350.7 11.7 0.18
THG3 575801 1333260 1385 THG3-OB1 212 3.1 16.61 57 13.51 571 9.51 0.0012
THG4 575471 1331124 1408 175 5.2 12.47 62 7.27 906.5 16.8 0.0018
K37 566652 1338642 1515 57.7 35.7 36.7 2.5 1 7.9 0.2
K38 568228 1336496 1464 56.5 22.7 23.7 4.2 1 1038 6.7
K42 573764 1334973 1402 44.3 12.3 16.6 4.5 4.3 424 5.3
K1 567803 1339502 1470 120 18 10
k5 568625 1339470 1478 120 12 38

k6 567887 1339500 1474 120 11 38
TK1 570275 1355009 1418 164 19.63 61.48 75 41.85
TK7 569334 1341467 1475 TK7-OB1 160 14.27 18.96 50 4.69 486.5 3.99 0.26
PK1 568066 1340931 1491 PK1-OB1 170 25.54 34.54 55 9 280 3.04 0.22
PK2 568476 1341101 1488 PK2 137 23.56 39.13 80 15.6 341 4.06 0.57
Pk3 568204 1350350 1440 153 30.45 37.33 70 6.88
PK4 568000 1353000 1436 106 9.14 61.2 60 52.06
Pk5 568427 1351843 1428 118 17.08 50.4 30 33.32
PK6 569299 1341890 1481 PK6 145 17.52 32.55 40 15 198 2.61 0.29
PK7 569892 1341651 1474 PK7-OB1 203 20.34 23.55 40 3.21 577.5 7.41 0.06
PK8 569814 1341065 1469 PK8 181 17.98 37.16 50 19.2 168 1.43 0.37
PK9 569485 1341610 1475 PK9 145 23.85 29.76 50 5.91 768.5 8.35 0.19
WG1 563443 1355457 1505 104 13.3 51 13.4 1079 8.3
WG2 570825 1357345 1417 130 18.7 40 32.3 147 1.3
WG3 567035 1355424 1453 110 13 7 66.5 40.18 1.34
WG4 563723 1355956 1496 118 13.16 52 12.76 315.4 5.28
WG5 568109 1356681 1444 111 19.65 52 10.55 373 7.1
WG6 569681 1357025 1429 105 25.61 50 18.72 181.4 5.05
WG7 569106 1357024 1435 105.5 24.43 50 30.97 506
WG8 567608 1356024 1447 105 10.19 50 10.17 797.8 18.99
WG9 567405 1354956 1446 105 19.15 30 35.02 432 10.28
WG10 568148 1354589 1437 106 16.59 50 10.41 1041 2.47
WG11 570347 1357013 1423 104 18.7 50 20.1 227.8 5.42
WG12 573140 1357524 1399 111.3 20.3 50 25.85 1171 2.8
WG13 573250 1357002 1399 116.1 20.3 50 25.85 790.6 18.86
WG14 572214 1356017 110 27.02 56.76 25
Valley, Ethiopia 63
WG15 572027 1355422 114 30.42 54.31 10
TWJ2 573465 1355068 1398 121 23.95 77.81 15 53.86
2.00E-
TWJ3 569491 1357769 1433 TWJ3-OB1 154 16.46 23.92 60 7.46 624.3 6.24 04
TWJ4 568854 1352624 1425 199 15.58 40.49 60 24.91

Appendix C Error evaluation
Table C.1 Comparison of the Observed and Simulated Heads and Error Calculation
No well ID Hs Hm Hm-Hs |Hm-Hs| |Hm-Hs|^2

1 PK1 1465.658 1465.000 -0.658 0.658 0.432964
2 PK2 1462.638 1463.000 0.362 0.362 0.131044
3 PHG1 1460.117 1462.000 1.883 1.883 3.545689
4 PHG2 1462.117 1469.000 6.883 6.883 47.375689
5 PHG3 1457.581 1457.000 -0.581 0.581 0.337561
6 PHG4 1466.488 1478.000 11.512 11.512 132.526144
7 TK7 1460.849 1461.000 0.151 0.151 0.022801
8 PK6 1461.383 1463.000 1.617 1.617 2.614689
9 PK7 1460.036 1453.000 -7.036 7.036 49.505296
10 PK8 1459.325 1452.000 -7.325 7.325 53.655625
11 PK9 1457.864 1453.000 -4.864 4.864 23.658496
12 THG2 1386.968 1379.000 -7.968 7.968 63.489024
13 PHG6 1419.13 1409.000 -10.130 10.13 102.6169
14 PHG8 1431.608 1433.000 1.392 1.392 1.937664
15 PHG9 1435.519 1438.000 2.481 2.481 6.155361
16 PHG10 1440.718 1434.000 -6.718 6.718 45.131524
17 HG1 1458.944 1459.000 0.056 0.056 0.003136
18 HG2 1453.988 1446.000 -7.988 7.988 63.808144
19 HG4 1456.116 1448.000 -8.116 8.116 65.869456
20 HG5 1417.494 1409.000 -8.494 8.494 72.148036
21 HG6 1461.107 1470.000 8.893 8.893 79.085449
22 HG7 1460.493 1467.000 6.507 6.507 42.341049
23 HG8 1462.786 1473.000 10.214 10.214 104.325796
24 HG11 1427.591 1423.000 -4.591 4.591 21.077281
25 HG14 1428.512 1419.000 -9.512 9.512 90.478144
26 HG15 1393.873 1404.000 10.127 10.127 102.556129
27 HG16 1394.39 1404.000 9.610 9.61 92.3521
28 HG17 1395.123 1395.000 -0.123 0.123 0.015129
29 HG18 1395.97 1390.000 -5.970 5.97 35.6409
30 ZELEKE2 1435.61 1427.000 -8.610 8.61 74.1321
31 TW3 1366.793 1365.000 -1.793 1.793 3.214849
32 THG3 1386.62 1383.000 -3.620 3.62 13.1044
33 K1 1460.348 1452.000 -8.348 8.348 69.689104
34 K5 1457.681 1466.000 8.319 8.319 69.205761
35 K6 1459.998 1463.000 3.002 3.002 9.012004
sum -29.436 195.454 1541.195
sum/35 -0.841 5.584 44.034
SQRT 6.63582364
Valley, Ethiopia 65
Appendix D Groundwater Abstraction
Table D.1 The estimated amount of abstracted water used in scenario-one
Jan Feb Mar Apr May Jun

Well ID Discharge(l/s) (24*6hrs) (24*12hrs) (24*16hrs) (24*16hrs) (24*14hrs) (24*12hrs) m3/annual m3/day
HG1 51 26438 52877 70502 70502 61690 52877 334886 917
HG2 51 26438 52877 70502 70502 61690 52877 334886 917
HG6 50 25920 51840 69120 69120 60480 51840 328320 900
HG7 50 25920 51840 69120 69120 60480 51840 328320 900
HG8 50 25920 51840 69120 69120 60480 51840 328320 900
HG9 10 5184 10368 13824 13824 12096 10368 65664 180
HG10 34 17626 35251 47002 47002 41126 35251 223258 612
HG12 50 25920 51840 69120 69120 60480 51840 328320 900
HG13 50 25920 51840 69120 69120 60480 51840 328320 900
HG16 25 12960 25920 34560 34560 30240 25920 164160 450
PK2 80 41472 82944 110592 110592 96768 82944 525312 1439
PK7 40 20736 41472 55296 55296 48384 41472 262656 720
PK9 50 25920 51840 69120 69120 60480 51840 328320 900
TK1 75 38880 77760 103680 103680 90720 77760 492480 1349
THG4 62 32141 64282 85709 85709 74995 64282 407117 1115
PHG1 50 25920 51840 69120 69120 60480 51840 328320 900
PHG3 50 25920 51840 69120 69120 60480 51840 328320 900
PHG5 57 29549 59098 78797 78797 68947 59098 374285 1025
PHG7 51.6 26749 53499 71332 71332 62415 53499 338826 928
TW1 7 3629 7258 9677 9677 8467 7258 45965 126
Zeleke 1 50 25920 51840 69120 69120 60480 51840 328320 900
WG1 51 26438 52877 70502 70502 61690 52877 334886 917

WG2 40 20736 41472 55296 55296 48384 41472 262656 720
WG3 7 3629 7258 9677 9677 8467 7258 45965 126
WG4 52 26957 53914 71885 71885 62899 53914 341453 935
WG5 52 26957 53914 71885 71885 62899 53914 341453 935
WG6 50 25920 51840 69120 69120 60480 51840 328320 900
WG10 50 25920 51840 69120 69120 60480 51840 328320 900
WG14 25 12960 25920 34560 34560 30240 25920 164160 450
WG15 10 5184 10368 13824 13824 12096 10368 65664 180
PK5 30 15552 31104 41472 41472 36288 31104 196992 540
TWJ4 60 31104 62208 82944 82944 72576 62208 393984 1079
K42 4.5 4.5 x 8hrs x365d 47304 130
K5 38 38 x 8hrs x365d 399456 1094
K6 38 38 x 8hrs x365d 399456 1094
Valley, Ethiopia 67
Table D.2 The estimated amount of water abstracted from 70 boreholes in scenario-two
Jan Feb Mar Apr May Jun

Well ID Discharge(l/s) (24*6hrs) (24*12hrs) (24*16hrs) (24*16hrs) (24*14hrs) (24*12hrs) m3/annual m3/day
HG-1 51 26438 52877 70502 70502 61690 52877 334886 917
HG-2 51 26438 52877 70502 70502 61690 52877 334886 917
HG3 20 10368 20736 27648 27648 24192 20736 131328 360
HG4 51 26438 52877 70502 70502 61690 52877 334886 917
HG-5 50 25920 51840 69120 69120 60480 51840 328320 900
HG-6 50 25920 51840 69120 69120 60480 51840 328320 900
HG-7 50 25920 51840 69120 69120 60480 51840 328320 900
HG-8 50 25920 51840 69120 69120 60480 51840 328320 900
HG-9 10 5184 10368 13824 13824 12096 10368 65664 180
HG-10 34 17626 35251 47002 47002 41126 35251 223258 612
HG11 50 25920 51840 69120 69120 60480 51840 328320 900
HG12 50 25920 51840 69120 69120 60480 51840 328320 900
HG13 50 25920 51840 69120 69120 60480 51840 328320 900
HG14 50 25920 51840 69120 69120 60480 51840 328320 900
HG-15 50 25920 51840 69120 69120 60480 51840 328320 900
HG-16 25 12960 25920 34560 34560 30240 25920 164160 450
HG-17 50 25920 51840 69120 69120 60480 51840 328320 900
HG-18 50 25920 51840 69120 69120 60480 51840 328320 900
Zeleke1 50 25920 51840 69120 69120 60480 51840 328320 900
Zeleke2 50 25920 51840 69120 69120 60480 51840 328320 900
PHG1 50 25920 51840 69120 69120 60480 51840 328320 900
PHG2 45 23328 46656 62208 62208 54432 46656 295488 810
PHG3 50 25920 51840 69120 69120 60480 51840 328320 900
PHG4 45 23328 46656 62208 62208 54432 46656 295488 810

PHG5 57 29549 59098 78797 78797 68947 59098 374285 1025
PHG6 58.5 30326 60653 80870 80870 70762 60653 384134 1052
PHG7 51.6 26749 53499 71332 71332 62415 53499 338826 928
PHG8 46.4 24054 48108 64143 64143 56125 48108 304681 835
PHG9 53.5 27734 55469 73958 73958 64714 55469 351302 962
PHG10 59.5 30845 61690 82253 82253 71971 61690 390701 1070
TW1 7 3629 7258 9677 9677 8467 7258 45965 126
TW3 29 15034 30067 40090 40090 35078 30067 190426 522
THG1 32 16589 33178 44237 44237 38707 33178 210125 576
THG3 57 29549 59098 78797 78797 68947 59098 374285 1025
THG4 62 32141 64282 85709 85709 74995 64282 407117 1115
TK1 75 38880 77760 103680 103680 90720 77760 492480 1349
TK7 50 25920 51840 69120 69120 60480 51840 328320 900
PK1 55 28512 57024 76032 76032 66528 57024 361152 989
PK2 80 41472 82944 110592 110592 96768 82944 525312 1439
Pk3 70 36288 72576 96768 96768 84672 72576 459648 1259
PK4 60 31104 62208 82944 82944 72576 62208 393984 1079
Pk5 30 15552 31104 41472 41472 36288 31104 196992 540
PK6 40 20736 41472 55296 55296 48384 41472 262656 720
PK7 40 20736 41472 55296 55296 48384 41472 262656 720
PK8 50 25920 51840 69120 69120 60480 51840 328320 900
PK9 50 25920 51840 69120 69120 60480 51840 328320 900
WG1 51 26438 52877 70502 70502 61690 52877 334886 917
WG2 40 20736 41472 55296 55296 48384 41472 262656 720
WG3 7 3629 7258 9677 9677 8467 7258 45965 126
WG4 52 26957 53914 71885 71885 62899 53914 341453 935
WG5 52 26957 53914 71885 71885 62899 53914 341453 935
WG6 50 25920 51840 69120 69120 60480 51840 328320 900
WG7 50 25920 51840 69120 69120 60480 51840 328320 900
Valley, Ethiopia 69
WG8 50 25920 51840 69120 69120 60480 51840 328320 900
WG9 30 15552 31104 41472 41472 36288 31104 196992 540
WG10 50 25920 51840 69120 69120 60480 51840 328320 900
WG11 50 25920 51840 69120 69120 60480 51840 328320 900
WG12 50 25920 51840 69120 69120 60480 51840 328320 900
WG13 50 25920 51840 69120 69120 60480 51840 328320 900
WG14 25 12960 25920 34560 34560 30240 25920 164160 450
WG15 10 5184 10368 13824 13824 12096 10368 65664 180
TWJ2 15 7776 15552 20736 20736 18144 15552 98496 270
TWJ3 60 31104 62208 82944 82944 72576 62208 393984 1079
TWJ4 60 31104 62208 82944 82944 72576 62208 393984 1079
K1 10 10 x 8hrs x365d 105120 288
K5 38 38 x 8hrs x365d 399456 1094
K6 38 38 x 8hrs x365d 399456 1094
Kobo Rural Water Supply bore holes
K37 2.5 2.5 x 8hrs x365d 26280 72
K38 4.2 4.2 x 8hrs x365d 44150 121
K42 4.5 4.5x 8hrs x365d 47304 130

Table D.3 Estimating the decline groundwater level for scenario-one and scenario-two
2 2
well ID Hs Hm Hs1 Hs2 Hm-Hs1 Hm-Hs2 (Hm-Hs1) (Hm-Hs2)
PK1 1465.7 1465.000 1452.147 1434.057 12.853 30.943 165.200 957.469
PK2 1462.6 1463.000 1447.311 1427.542 15.689 35.458 246.145 1257.270
PHG1 1460.1 1462.000 1444.307 1424.826 17.693 37.174 313.042 1381.906
PHG2 1462.1 1469.000 1449.786 1432.614 19.214 36.386 369.178 1323.941
PHG3 1457.6 1457.000 1443.64 1425.378 13.360 31.622 178.490 999.951
PHG4 1466.5 1478.000 1453.099 1435.618 24.901 42.382 620.060 1796.234
TK7 1460.8 1461.000 1446.189 1425.307 14.811 35.693 219.366 1273.990
PK6 1461.4 1463.000 1446.819 1425.903 16.181 37.097 261.825 1376.187
PK7 1460 1453.000 1445.417 1424.482 7.583 28.518 57.502 813.276
PK8 1459.3 1452.000 1444.592 1423.909 7.408 28.091 54.878 789.104
PK9 1457.9 1453.000 1442.921 1422.787 10.079 30.213 101.586 912.825
THG2 1387 1379.000 1379.145 1361.212 -0.145 17.788 0.021 316.413
PHG6 1419.1 1409.000 1409.405 1389.995 -0.405 19.005 0.164 361.190
PHG8 1431.6 1433.000 1423.162 1405.347 9.838 27.653 96.786 764.688
PHG9 1435.5 1438.000 1427.456 1410.306 10.544 27.694 111.176 766.958
PHG10 1440.7 1434.000 1432.883 1416.609 1.117 17.391 1.248 302.447
HG1 1458.9 1459.000 1443.09 1423.134 15.910 35.866 253.128 1286.370
HG2 1454 1446.000 1438.24 1417.595 7.760 28.405 60.218 806.844
HG4 1456.1 1448.000 1440.312 1419.104 7.688 28.896 59.105 834.979
HG5 1417.5 1409.000 1408.456 1388.743 0.544 20.257 0.296 410.346
HG6 1461.1 1470.000 1445.133 1425.247 24.867 44.753 618.368 2002.831
HG7 1460.5 1467.000 1444.107 1423.416 22.893 43.584 524.089 1899.565
HG8 1462.8 1473.000 1447.207 1427.938 25.793 45.062 665.279 2030.584
HG11 1427.6 1423.000 1416.684 1396.496 6.316 26.504 39.892 702.462
HG14 1428.5 1419.000 1417.219 1396.825 1.781 22.175 3.172 491.731
HG15 1393.9 1404.000 1386.686 1367.127 17.314 36.873 299.775 1359.618
HG16 1394.4 1404.000 1386.686 1367.127 17.314 36.873 299.775 1359.618
HG17 1395.1 1395.000 1387.688 1367.549 7.312 27.451 53.465 753.557
HG18 1396 1390.000 1388.444 1368.26 1.556 21.740 2.421 472.628
ZELEKE2 1435.6 1427.000 1423.007 1401.851 3.993 25.149 15.944 632.472
TW3 1366.8 1365.000 1361.659 1341.15 3.341 23.850 11.162 568.822
THG3 1386.6 1383.000 1379.182 1358.752 3.818 24.248 14.577 587.966
K1 1460.3 1452.000 1444.762 1424.793 7.238 27.207 52.389 740.221
K5 1457.7 1466.000 1441.853 1421.333 24.147 44.667 583.078 1995.141
K6 1460 1463.000 1444.347 1424.307 18.653 38.693 347.934 1497.148
398.959 1085.361 6700.732 35826.753
ME 11.399 31.010 191.449 1023.622
RSME 13.837 31.994
Valley, Ethiopia 71
Appendix E Radius of influence and drawdown
Table E.1 Radius of influence and drawdown
20% 0f 20% of
well ID SWl DWl1 DWL2 Hm (SWL) DD1 DD2 DD1 DD2
PK1 1465.658 1452.147 1434.057 1465.000 12.853 30.943 2.5706 6.1886
PK2 1462.638 1447.311 1427.542 1463.000 15.689 35.458 3.1378 7.0916
PHG1 1460.117 1444.307 1424.826 1462.000 17.693 37.174 3.5386 7.4348
PHG2 1462.117 1449.786 1432.614 1469.000 19.214 36.386 3.8428 7.2772
PHG3 1457.581 1443.64 1425.378 1457.000 13.360 31.622 2.672 6.3244
PHG4 1466.488 1453.099 1435.618 1478.000 24.901 42.382 4.9802 8.4764
TK7 1460.849 1446.189 1425.307 1461.000 14.811 35.693 2.9622 7.1386
PK6 1461.383 1446.819 1425.903 1463.000 16.181 37.097 3.2362 7.4194
PK7 1460.036 1445.417 1424.482 1453.000 7.583 28.518 1.5166 5.7036
PK8 1459.325 1444.592 1423.909 1452.000 7.408 28.091 1.4816 5.6182
PK9 1457.864 1442.921 1422.787 1453.000 10.079 30.213 2.0158 6.0426
THG2 1386.968 1379.145 1361.212 1379.000 -0.145 17.788 -0.029 3.5576
PHG6 1419.13 1409.405 1389.995 1409.000 -0.405 19.005 -0.081 3.801
PHG8 1431.608 1423.162 1405.347 1433.000 9.838 27.653 1.9676 5.5306
PHG9 1435.519 1427.456 1410.306 1438.000 10.544 27.694 2.1088 5.5388
PHG10 1440.718 1432.883 1416.609 1434.000 1.117 17.391 0.2234 3.4782
HG1 1458.944 1443.09 1423.134 1459.000 15.910 35.866 3.182 7.1732
HG2 1453.988 1438.24 1417.595 1446.000 7.760 28.405 1.552 5.681
HG4 1456.116 1440.312 1419.104 1448.000 7.688 28.896 1.5376 5.7792
HG5 1417.494 1408.456 1388.743 1409.000 0.544 20.257 0.1088 4.0514
HG6 1461.107 1445.133 1425.247 1470.000 24.867 44.753 4.9734 8.9506
HG7 1460.493 1444.107 1423.416 1467.000 22.893 43.584 4.5786 8.7168
HG8 1462.786 1447.207 1427.938 1473.000 25.793 45.062 5.1586 9.0124
HG11 1427.591 1416.684 1396.496 1423.000 6.316 26.504 1.2632 5.3008
HG14 1428.512 1417.219 1396.825 1419.000 1.781 22.175 0.3562 4.435
HG15 1393.873 1386.686 1367.127 1404.000 17.314 36.873 3.4628 7.3746
HG16 1394.39 1386.686 1367.127 1404.000 17.314 36.873 3.4628 7.3746
HG17 1395.123 1387.688 1367.549 1395.000 7.312 27.451 1.4624 5.4902
HG18 1395.97 1388.444 1368.26 1390.000 1.556 21.740 0.3112 4.348
ZELEKE2 1435.61 1423.007 1401.851 1427.000 3.993 25.149 0.7986 5.0298
TW3 1366.793 1361.659 1341.15 1365.000 3.341 23.850 0.6682 4.77
THG3 1386.62 1379.182 1358.752 1383.000 3.818 24.248 0.7636 4.8496
K1 1460.348 1444.762 1424.793 1452.000 7.238 27.207 1.4476 5.4414
K5 1457.681 1441.853 1421.333 1466.000 24.147 44.667 4.8294 8.9334
K6 1459.998 1444.347 1424.307 1463.000 18.653 38.693 3.7306 7.7386

Valley, Ethiopia 73

MSc_GroundwaterModellingOptimization

Uploaded by

Copyright:

Available Formats

MSc_GroundwaterModellingOptimization

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

MSc_GroundwaterModellingOptimization

Uploaded by

Copyright:

Available Formats

Groundwater Modelling and

Optimization of Irrigation Water Use

MSc Thesis Number:WSE-HELWD-14.09

Master of Science Thesis

List of Figures vii

List of Tables viii

3. General description of the study area 9

4. Analysis and Model input data preparation 13

5. Groundwater Modelling of Kobo Valley 21

6. Result and Discussion 33

7. Conclusion and Recommendation 50

Co-SAERAR Commission for Sustainable Agriculture and Environmental Rehabilitation in

DEM Digital Elevation Model

DWL Dynamic Water Level

EIGS Ethiopian Institute of Geological Survey

ETo Potential Evapotranspiration

GES Geo-Engineering Service

GIS Geographic Information System

GPS Global Positioning System

KGVDP Kobo-Girana Valley Development Project

m.a.s.l Meter above sea level

MAE Mean absolute error

MCM Million Cubic Meters

RMS Root mean square error

RVDP Raya Valley Development Project

SCS Soil Conservation Service

SRTM Shuttle Radar Topographic Mission

SWL Static Water Level

ETo Potential Evapotranspiration (mm)

ETa Actual Evapotranspiration (mm)

k Hydraulic conductivity (m/d)

Kav Average Hydraulic conductivity (m/d)

S Soil moisture content

Sr Annual surface runoff

Tav Average Transmissivity (m2/d)

1.1. Back ground

1.2. Problem statement

1.3. Objectives of the Research

1.4. Research questions

1.5. Possible scenarios

MSc thesis, G.W. Adane 2

1.7. Outline of the thesis.

MSc thesis, G.W. Adane 4

2.1. Review of previous study

MSc thesis, G.W. Adane 6

2.2. Groundwater Modelling

MSc thesis, G.W. Adane 8

General description of the study area

Figure 3.1 Location map of the study area.

3.4. Land use and cover

3.5. Geology and Hydrogeology

MSc thesis, G.W. Adane 10

MSc thesis, G.W. Adane 12

Analysis and Model input data

4.1. Meteorological data Analysis

=Mean annual rainfall for the kobo valley (mm)

Figure 4.1 Mean monthly rainfall of the stations (1996- 2010)

Figure 4.2 Mean annual rainfalls of the stations

MSc thesis, G.W. Adane 14

Figure 4.3 Kobo station annual rainfall (1996-2010)