SPATIAL AND TEMPORAL VARIABILITY OF COFFEE (Coffea

arebica L.) WATER AND IRRIGATION REQUIREMENT MAPPING

FOR JIMMA ZONE

MSC THESIS

MINDA TADESSE BEDANE

HAWASSA UNIVERSITY, HAWASSA, ETHIOPIA

MARCH, 2019
 SPATIAL AND TEMPORAL VARIABILITY OF COFFEE (Coffea
arebica L.) WATER AND IRRIGATION REQUIREMENT MAPPING
FOR JIMMA ZONE

MINDA TADESSE BEDANE

A THESIS SUBMITTED TO SCHOOL OF WATER RESOURCE ENGINEERING

HAWASSA UNIVERSITY

HAWASSA, ETHIOPIA

IN PARTIAL FULFILLMENT OF THE REQUIREMENT FOR THE

DEGREE OF MASTER OF SCIENCE IN WATER RESOURCES

ENGINEERING AND MANAGEMENT

MARCH, 2019
 SCHOOL OF GRADUATE STUDIES

HAWASSA UNIVERSITY

ADVISORS’ APPROVAL SHEET

(Submission Sheet-1)

This is to certify that the thesis entitled “Spatial and Temporal Variability of Coffee
(Coffea arebica L.) Water and Irrigation Requirement Mapping for Jimma Zone”
submitted in partial fulfillment of the requirements for the degree of Master's with
specialization in water resource engineering and management, the Graduate Program
of the Department/School of bio-system and environmental engineering, and has been
carried out by Minda Tadesse Bedane, Id. No WREMR/018/09, under my/our
supervision. Therefore, I/we recommend that the student has fulfilled the requirements and
hence hereby can submit the thesis to the department.

Abraham Woldemichael (PhD.) _______________ ______________

Name of Major Advisor Signature Date

Tilahun Hordofa (PhD.) __________________ ________________

Name of Co-Advisor Signature Date
 SCHOOL OF GRADUATE STUDIES

HAWASSA UNIVERSITY

EXAMINERS' APPROVAL SHEET-1

(Submission Sheet-2)

We, the undersigned, members of the Board of Examiners of the final open defense by

Minda Tadesse have read and evaluated his/her thesis entitled “Spatial and Temporal

Variability of Coffee (Coffea arebica L.) Water and Irrigation Requirement Mapping for

Jimma Zone”, and examined the candidate. This is, therefore, to certify that the thesis has

been accepted in partial fulfillment of the requirements for the degree.

__________________________ __________________ _______________

Name of the Chairperson Signature Date

__________________________ __________________ _______________

Name of Major Advisor Signature Date

__________________________ __________________ _______________

Name of Internal Examiner Signature Date

__________________________ __________________ _______________

Name of External Examiner Signature Date

__________________________ __________________ _______________

SGS Approval Signature Date

Final approval and acceptance of the thesis is contingent upon the submission of the final

copy of the thesis to the School of Graduate Studies (SGS) through the

Department/School Graduate committee (DGC/SGC) of the candidate's department.

Stamp of SGS

Date: ___________________
 HAWASSA UNIVERSITY
SCHOOL OF GRADUATE STUDIES
EXAMINERS' APPROVAL SHEET-2

(Submission Sheet-3)

As members of the Board of Examiners of the final Master's degree open defense, we
certify that we have read and evaluated the thesis prepared by Minda Tadesse
Bedane under the title “Spatial and Temporal Variability of Coffee (Coffea arebica
L.) Water and Irrigation Requirement Mapping for Jimma Zone”, and recommend
that it be accepted as fulfilling the thesis requirement for the degree of Master's of
Science in Bio-system and Environmental Engineering with Specialization in Water
Resource Engineering and Management.
__________________________ __________________ _______________
Name of the Chairperson Signature Date

__________________________ __________________ _______________

Name of Internal Examiner Signature Date

__________________________ __________________ _______________

Name of External Examiner Signature Date

Final approval and acceptance of the thesis is contingent upon the submission of the
final copy of the thesis to the SGS through the DGC/SGC of the candidate's
department/School.
Thesis approved by

_______________________ _________________ ________________

DGC/SGC Signature Date
Certification of the final Thesis

I hereby certify that all the corrections and recommendation suggested by the
Board of Examiners are incorporated into the final Thesis entitled “Spatial and
Temporal Variability of Coffee (Coffea arebica L.) Water and Irrigation Requirement
Mapping for Jimma Zone” by Minda Tadesse.
______________________ __________________ _________________
Name of the Designate Signature Date

Stamp of SGS
Date: ___________________
 DEDICATION
I dedicate this thesis manuscript to my daughters Merry Minda, Sifan Minda, my mother
Ms. Warkitu Tadesse, my wife Ms. Mekdas Belay and my friend Mohammed Jihad.

i
 STATEMENT OF THE AUTHOR

The author declares that this thesis work is his bona fide work and that all sources of

materials used for this thesis have been dully acknowledged. This thesis has been

submitted in partial fulfillment of the requirements for an advanced M.Sc. Degree at the

Hawassa University and is deposited at the University Library to be made available to

borrowers under rules of the Library. The author declares that this thesis is not

submitted to any other institution anywhere for the award of any academic degree,

diploma or certificate.

Brief quotations from this thesis are allowed without special permission provided that

accurate acknowledgement of source is made. Requests for permission for extended

quotation or reproduction of this manuscript in whole or in part may be granted by

the Head of the Department of Bio-system and Environmental Engineering or the

Dean of School of Graduate Studies when in his or her judgment the proposed use

of the material is in the interest of scholarship. In all other instances, however,

permissions must be obtained from the author.

Name: Minda Tadesse Bedane Signature: ________________

Place: Hawassa University, Hawassa

Date of submission: _March, 2019_________________

ii
 ACKNOWLEDGEMENT

I am greatly thankful to the Ethiopian Institute Agricultural Research (EIAR) for

permitting me to join the School of Graduate Studies at Hawassa University to

pursue my master study in Water Resource Engineering and Management.

I want to express my sincere gratitude to my advisors, Dr Abraham Woldemichael and Dr

Tilahun Hordofa, for their intellectual stimulation, professional guidance,

encouragement, and inspiring during the project work and valuable criticism of the

thesis. I am really grateful for their friendship approach and for providing me a range of

possibilities to broaden my knowledge especially during the fieldwork at Jimma/JARC/

and thesis write-up.

I would like to extend my appreciation and sincere thanks to Mr. Robel Admasu and

Jimma Research Center staff in general for their guidance, providing me facilities

whenever required, and their cooperativeness during the author conducting research at

research center.

I also like to express my deepest gratitude to my family’s W/ro Mekdas Belay, my

Children: Merry Minda and Sifan Minda, Mr. Rata Zawde and Mr. Mohammed Jihad and

his families for their moral support and hospitality.

Above all, I would like to thank the Almighty and Merciful God, for providing him all the

patience and endurance during his study.

iii
 LIST OF ABBRIVIATIONS AND ACRONYMS

CRO Coffee Research Organization

CSA Central Statistical Authority
CWR Crop water requirement
DEM Digital elevation model
DW Dry weight
ECFF Environment and Coffee Forest Forum
ENMA Ethiopian National meteorological agency
ETc Crop Evapotranspiration
EIAR Ethiopian Institute of agricultural research
ETo Reference Evapotranspiration
FC Field Capacity
FOA Food and Agricultural Organization
FW Fresh weight
IR Irrigation requirement
JARC Jimma agricultural research center
JZIA Jimma Zone Irrigation Authority
Kc Crop Coefficient
KS Kolmogorov-Smirnov
RH Relative humidity
RMSE Root mean sum square error
SDD Spatial Dependency Degree
SNHT Standard normal homogeneity test
SSH Sunshine hours
Tmax Maximum Temperature
Tmin Minimum temperature
WS Wind speed

iv
 TABLE OF CONTENTS

CONTENTS Page
DEDICATION ................................................................................................................... i
STATEMENT OF THE AUTHOR.................................................................................... ii
ACKNOWLEDGEMENT ................................................................................................ iii
LIST OF ABBRIVIATIONS AND ACRONYMS ............................................................ iv
TABLE OF CONTENTS .................................................................................................. v
LIST OF TABLES .......................................................................................................... vii
LIST OF FIGURES ........................................................................................................ viii
LIST OF APPENDIX ....................................................................................................... ix
ABSTRACT...................................................................................................................... x
1. INTRODUCTION ..................................................................................................... 1
1.1 Background ......................................................................................................... 1
1.2 Statement of the Problem .................................................................................... 3
1.3 Objectives ........................................................................................................... 3
1.3.1 General objective ......................................................................................... 3
1.3.2 Specific objectives ....................................................................................... 3
2. LITERATURE REVIEW .......................................................................................... 4
2.1 Importance of Coffee in Ethiopia......................................................................... 4
2.2 Coffee Growing Climate and Season ................................................................... 5
2.3 Coffee Crop Water Requirement ......................................................................... 6
2.4 Coffee Crop Coefficients ..................................................................................... 7
2.5 Reference Evapotranspiration .............................................................................. 8
2.6 Coffee Irrigation Requirement ............................................................................. 9
2.7 The FAO CROPWAT Model .............................................................................. 9
2.8 Geo-statistical Interpolation and Mapping ......................................................... 11
3. MATERIALS AND METHODOLOGY .................................................................. 13
3.1 Description of Study Area ................................................................................. 13
3.1.1 Location ..................................................................................................... 13
3.1.2 Climate ...................................................................................................... 13
3.1.3 Crop productions ........................................................................................ 14
3.2 Selection of Meteorological Station ................................................................... 14

v
 TABLE OF CONTENTS (CONTINUED)
3.3 Data collection and measurement ...................................................................... 15
3.3.1 Meteorological data .................................................................................... 15
3.3.2 Crop data ................................................................................................... 18
3.4 Estimation of reference evapotranspiration ........................................................ 22
3.5 Estimation of crop water and irrigation requirement .......................................... 23
3.6 Statistical Analysis ............................................................................................ 23
3.7 Mapping of ETo, Coffee ETc and IR ................................................................. 24
4. R1ESULTS AND DISCUSSION............................................................................. 25
4.1 Climate data consistency test ............................................................................. 25
4.2 Coffee Crop Coefficient .................................................................................... 27
4.3 Reference Evapotranspiration ............................................................................ 29
4.3.1 Validation of CROPWAT 8 model ............................................................. 29
4.3.2 Explanatory analysis .................................................................................. 29
4.3.3 Annual and Monthly spatial distribution of reference evapotranspiration .... 30
4.3.4 Cross-validations of Model ........................................................................ 33
4.4 Coffee Crop water Requirement ........................................................................ 37
4.4.1 Explanatory analysis .................................................................................. 37
4.4.2 Annual and monthly spatial distribution of coffee ETc ............................... 37
4.4.3 Semi-Variogram models and parameters .................................................... 39
4.5 Cross-validations of kriging interpolations ........................................................ 40
4.6 Coffee Irrigation Requirement ........................................................................... 45
5. SUMMARY, CONCLUSION AND RECOMMENDATIONS ................................ 51
5.1 Summary and Conclusions ................................................................................ 51
5.2 Recommendations ............................................................................................. 52
6. REFERENCES ........................................................................................................ 54
APPENDICES ................................................................................................................ 65
BIOGRAPHICAL SKETCH ........................................................................................... 73

vi
 LIST OF TABLES

Table 1. Water balance and climate adjusted coffee Kc for initial, mid and late stage at
JARC station ..................................................................................... 28
Table 2. NDVI and Kc value for model development during 2017/18 years and Kc NDVI
.......................................................................................................... 28
Table 3. Coffee crop Kc and length of development stage................................................ 29
Table 4. ETo estimated from full, exported and CROPWAT estimated data at Jimma
station for 2015 ................................................................................. 33
Table 5: Statistic analysis of monthly and annual ETo for Jimma Zone ........................... 33
Table 6. Pearson Correlations of monthly ETo with metrological parameters .................. 34
Table 7. RMSE between modeled monthly ETo and observed ETo values....................... 34
Table 8. Descriptive statistics of monthly and annual Coffee ETc for Jimma Zone .......... 40
Table 9. RMSE for monthly and annual coffee ETc for ordinary and IDW ...................... 41
Table 10. Parameters of the variogram models obtained for monthly and annual values of
coffee ETc in Jimma Zone ................................................................. 42
Table 11. Statistics analysis of monthly and annual coffee IR for Jimma Zone ................ 48

vii
 LIST OF FIGURES

Figure 1.The location Map of the study area .................................................................... 13

Figure 2. Meteorological stations (A) and selected stations by Thiessen polygons (B) ..... 15
Figure 3. Classification of climate region using dendrogram ............................................ 18
Figure 4. Sketch of experimental field and sampling method ........................................... 20
Figure 5. SNHT Test results of mean annual Tmin for Jimma station 1985-2016 ............. 17
Figure 6. Double mass of Jimma station of January Tmin 1985-2016 .............................. 17
Figure 7. Corrected minimum temperature of Shebe station by double mass curve .......... 25
Figure 8. Double mass curve of annual mean rainfall of meteorological stations .............. 26
Figure 9. Mean monthly ETo for Jimma zone .................................................................. 35
Figure 10 Histogram (A) and Boxplot (B) for annual ETo ............................................... 35
Figure 11. ETo homogeneity test for August at Jimma stations in rainbow software ........ 35
Figure 12. ETo kriging maps of annual and monthly ETo for Jimma zone ....................... 36
Figure 13: Histogram (A) and box-plot (B) for annual coffee ETc (mm/ year) ................. 42
Figure 15. Semi-variogram and fitted model for mean annual coffee ETc in Jimma ......... 43
Figure 14: Histogram (A) and box-plot (B) for annual coffee ETc (mm/ year) ................. 43
Figure 16. ETc kriging maps of monthly coffee crop ETc for Jimma zone ....................... 44
Figure 17. Monthly coffee ETc, IR and effective rain fall for Jimma zone ....................... 49
Figure 18. IR kriging maps of Monthly and annual coffee IR for Jimma zone.................. 50

viii
 LIST OF APPENDIX

Appendix Table 1. Monthly Tmax data of Jimma stations prepared for MICE imputation 65
Appendix Table 2. Descripitve statistics of climate dataset for different stations .............. 66
Appendix Table 3. Areal mean climate dataset estimated by Thiessen polygon method for
Jimma zone ........................................................................................ 67
Appendix Table 4. KS test values calculated for 80% non- exceedance probability for
monthly ETo (1985 to 2016) for each station...................................... 68
Appendix Table 5. ETo (mm/day) values of the different stations .................................... 69
Appendix Table 6. Coffee ETc (mm) estimated for different stations............................... 70
Appendix Table 7. coffee irrigation requirement (mm) estimated for different stations .... 71

ix
 ABSTRACT
Coffee is one of the dominant cash crops grown in Ethiopia, particularly in Jimma zone. Absence
of rainfall during four to six months of the year affects coffee production. Study was undertaken in
Jimma zone, Oromia regional state Southwest of Ethiopia, with the objective of evaluating and
characterizing the water requirement of coffee (ETc) and mapping for planning and management
options for Jimma Zone. The study was undertaken at 22 sites of which 14 being from Jimma zone
and the remaining eight sites were from around the vicinity of study area. The reference
evapotranspiration (ETo) was estimated using Penman–Monteith method, with monthly datasets of
22 stations for the period of 31 years. An adjustment was made to Kc values to account for
deviations from standard conditions. The coffee ETc and irrigation requirement (IR) were
determined using CROPWAT 8.0 software. The results of ETo, ETc and IR were further analyzed
by means of geo-statistical tools in R software using gstat package. Maps were created using
ordinary kriging combined with linear regression method. The results have shown that the ETo,
coffee ETc and IR have shown moderate to greater variability, i.e the CV for annual ETc and IR
ranged from 5.5 to 31.8%, respectively. The high annual and monthly coffee ETc values were
observed around the extreme eastern tip, extreme southern tip and north western tip. In contrary
low ETc values were observed in southern and western regions. April (131.1 mm) and November
(80.9 mm), respectively, presented the highest and the smallest coffee ETc. The high annual coffee
IRs locations include: northern and eastern parts whereas the low in western and southern parts.
The highest IR values was predicted in February (77.7mm) whereas no IRs from June to August.
The coffee ETc for Jimma zone is a good indicator and benefits to the coffee growers to help in
planning the water resources of the area for best agricultural water management practices.

Key words: Coffee, ETc, irrigation requirement, ETo

x
 1. INTRODUCTION
1.1 Background

Water is one of the most important inputs essential for the crop production because plants
require water for growth and tissue expansion. Plants need water continuously during their
life and in huge quantities (Nidhi et al, 2012). However, over 90% of the water required by
plants is lost through transpiration (Morison, 2008). Thus, the evapotranspiration is an
important hydrological process (Lu et al., 2005). Hence improved water management in
agriculture depends on reliable estimates of water use by plants (Pereira, 2011).

The first step in estimation of crop water requirement is determination of evaporative

demand of atmosphere, which is expressed as the reference evapotranspiration (ASCE,
2005). Its accurate estimates were important in irrigation management. However, direct
measurement is both inconvenient and expensive. For this reason, many methods have
been developed for estimating ETo from climate data (Doorenbos and Pruitt, 1977; Allen
et al., 1998). However, despite difference in their climate data requirement, most of the
methods lack required accuracy in estimating the value of ETo for any particular area. But,
Penman Monteith method provides more accurate estimation of ETo (Allen et al., 1998).

The crop coefficients are the 2nd important parameter that plays an essential role in
estimating the crop ETc in general (Pereira et al., 1999) particularly for coffee. Allen et al.
(1998) suggested Kc values for many crops including coffee under different climatic
conditions which are commonly used in places where the local data is not available.
However, there is a need for local calibration of the Kc under given climatic conditions
(Kashyap and Panda, 2001). Thus, the coffee Kc obtained from FAO should be adjusted to
local condition especially for study areas by combining NDVI and soil water balance since
both are related to fractional ground cover (Rouse et al., 1974; Peirera et al., 1957).

Crop water requirement (ETc) is the amount of water required to compensate the
evapotranspiration loss from the cropped field during a specified period of time
(Todorovic, 2005). The requirement is ether applied by rainfall or irrigation. The major
environmental factors affecting ETc are climate and water which limiting plant growth
(Sijali, 2001). ETc can be either directly measured or in-directly estimated. Direct
measurement is remains challenging, because it is expensive and time consuming (Amatya
et al., 2016) particularly for developing country like Ethiopia. Among in-direct methods,
1
ETc estimated by combining ETo and Kc, is the widely used method. Like annual crops,
several authors directly measure coffee ETc using Bowen ratio-energy balance technique
(Gutierrez and Meinzer, 1994; Fabio et al., 2005). But it is impractical for large areas, to
resolve this problem, coffee ETc estimates from ETo and Kc is among best alternatives at
stations (Lemos Filho et al., 2010). Accurate estimates of ETc for coffee are crucial
because too little water is substantially reducing growth without wilting (Pereira, 1994).

Irrigation plays important roles in agriculture production (Sax, 2000). Adequate and timely
irrigation leads to high yield of crops in general particularly for coffee. It raises the
productivity of land through increase of crop productivity per hectare (Ali, 2010). The
productivity of coffee can be enhanced with irrigation. Coffee irrigation is a promising
technique that may provide both yield increase and expansion of coffee plantations in areas
considered unsuitable due to the occurrence of water shortage (Silva et al., 2008). Brazilian
can obtain a potential yield of 6 to7 ton ha-1 under irrigation (Naan, 2009). Even in
Ethiopia, Tesfaye et al. (2013) found in a coffee growing area in Jimma zone that the
increase in coffee yield by about 24% from fully irrigated compared to rain-fed coffee
trees. However, the Ethiopian coffee production systems particularly for Jimma zone are
remained traditional till now, where coffee productivity is severely affected by water
shortage especially during critical stage of flowering and fruit set.

In recent year, drought severity happed frequently in Jimma zone. For example, in 2016/17
during dry years, hundreds hectares of coffee plantation were devastated due to the lack of
supplemental irrigation (ECFF, 2017). This is happed because of farmer practiced poor
irrigation management; they just traditionally wait for rain. One of the main problems of
the irrigator is not know the amount of water that has to be applied to the field to meet the
water needs of the crops. In addition, most farmers in study area have not an appropriate
plan for irrigation (JZIA, 2017).

As Jimma is a large zone also in practice estimating of ETo and ETc at every spatial
location is difficult from sparsely distributed data. Geo-statistical interpolation methods
such as ordinary kriging are a promising tool to overcome this problem. It generates both
spatial autocorrelation and prediction map depending up on semi-variogram and distance
(Li and Heap, 2008). However, ordinary kriging method is not capable of using secondary
information as predictor. Thus, combined method of ordinary kriging with linear

2
regression helps to solve this problem (Knotters et al., 1995). For areas like Jimma, which
is characterized by diverse topography and limited number of stations, using auxiliary
information such as elevation plays significant role in improving interpolation quality.

1.2 Statement of the Problem

Coffee plays a significant role in the economy of Ethiopia, contributing over 60% of
the nation’s foreign exchange earnings, 30% of the government's direct revenue, 8%
output o f the agricultural sector and 4% of the gross domestic (Girma, 2011). Coffee
is the dominant cash crop cultivated in Jimma zone, covering an area of about 17.8% and
contributing 19.5% to the countries' national coffee product. In spite of the importance of
the crop in the country’s economy, it is characterized by low productivity, which is about
76.92% yield reduction as compared to potential yield recorded at JARC station (CSA,
2016; Tadesse, 2017). The low yield may due to traditional production technologies.
Seasonal water stress and recurrent drought are among the major factors which account for
low yields of the crop. In addition, the coffee production systems in the zone are largely
rain-fed and thus, their success is sensitive to climate variability.

The temporal and spatial distribution of rainfall throughout the year is un-evenly
distributed in Jimma zone. Over 71.1% of rainfall is received during the rainy summer
season (June to September) while the rest 28.9 % received during eight months (October to
May). This means atmospheric water demand is greater than rainfall during dry spells.

To achieve effective planning on irrigation water management, accurate information is

needed for crop water requirement. The availability of spatial and temporal coffee ETc and
IR information is limited in study area and yet not estimated. In addition, the shortfalls of
data at a regional scale and for irrigation project design engineers.

1.3 Objectives
1.3.1 General objective
The general objectives of this study were to analyze and characterize the water requirement
of coffee crops and mapping for planning and management options for Jimma zone.
1.3.2 Specific objectives
To assess the variability of water requirement for coffee crop in Jimma zone,
throughout the year
To map spatial and temporal ETc and IR requirement of coffee in Jimma zone.
3
 2. LITERATURE REVIEW

2.1 Importance of Coffee in Ethiopia

Coffee production is important to the Ethiopian economy with about 15 million people
directly or indirectly deriving their livelihoods from coffee. Coffee is also a major
Ethiopian export commodity generating about 25% of Ethiopia’s total export earnings.
Ethiopia is the fifth largest producer of coffee in World, Almost two percent of the world’s
coffee comes from Ethiopia (Tefera and Tefera, 2013; Chauhan et al., 2015 ).

Coffee Arabica is dominant cash crops produced in Jimma zone. Arabica coffee is native
to southern Ethiopia and was cultivated in the Gibe valley for many centuries. Southwest
coffee growing areas include Jimma, Limmu, and Mizan Tafari, because these areas enjoy
the best soil and climatic combinations (Yonas, 2002). Coffee production systems in
Ethiopia are generally categorized into four areas i.e. forest coffee, semi-forest coffee,
garden coffee, and plantation coffee. Forest coffee is a wild coffee grown under the shade
of natural forest trees. Semi-forest coffee farming is a system where farmers thin and select
forest trees to let sufficient sunlight to the coffee trees and to provide adequate shade.
Garden coffee is normally found in the vicinity of a farmer’s residence with other crops.
Plantation coffee is planted by the government or private investors for export purposes
(Tefera and Tefera, 2013).

In Ethiopia, flower and fruit development are phased to maximize the likelihood that the
fruit will expand during the rains and after a flush of new leaves. Hence floral initiation
occurs during the cool, dry winter period; the flowers then remain dormant during the dry
season and blossom after the first showers that invariably precede the main rains. The
`pinhead fruits remain dormant before expanding after the beginning of the rains by which
time the new flush of leaves, triggered by the same `blossom' showers, have expanded.
According to (Wintgens, 2009) reports the coffee dry periods were determined when the
monthly rainfall amount was less than approximately 60 mm. The Flowering is assumed
to start one week after the first heavy rainfall following the dry season (DaMatta et al.,
2007). Intense rainfall throughout the year (without dry seasons) can lead to scattered
harvests and low yield (Cannell, 1985).

4
2.2 Coffee Growing Climate and Season

The coffee growth in Ethiopia particularly for Jimma zone is significantly affected by
environmental conditions. Coffee Arabica can successfully provide yield within narrow
range of environmental conditions. The temperature, rain fall and variation of season of
the year are the major factor governing the coffee growth and yields. Lack of rain fall and
increase of temperature can aggravate the soil moisture depletion and coffee getting
stressed. The variation of temperature through the year is due to annual weather patterns
or seasonality (ECFF, 2017).

Coffee Arabica is successfully grown in the altitude range of 1000-1500m, optimum

temperature ranges from 18 to 25°C. Require fertile soil with optimum pH range from 5.4
to 6.0, annual rainfall ranges from 1200 to 2000 mm is required and especially during
flowering and maturity, but regulated moisture deficit during dormancy period.
Flowering is initiated after the first rains showers during dry season probably in January,
February or early March and maturity of cherries requires a dry period that can be up to 5
months (Kuit et al., 2004). The climate of Ethiopia’s coffee lands is tropical, but due to
the large central highland area (mostly above 1000 m) much of the country can be
classified as cool-tropical. The seasons are largely defined by rainfall. There are three
main seasons: Bega, Belg, and Kiremt (Robel, 2012).

Bega (October to January/February) the long dry season, this is when coffee harvesting
and processing takes place. The severity of the main dry season depends largely on
location but also on other physical characteristics, including altitude, slope and slope
aspect. There may be slight to moderate rainfall or barely any at all. Belg (February to
May): the first rains before the main wet season, which may either represent the early part
of the wet season, or a distinct short wet season followed by a short dry period before the
main rains. This is the main period for coffee flowering, fruit initiation and early
development. Kiremt (June to September): the main wet season. This is the period for
final coffee fruit development and ripening. Around the end of September there is a sharp
decline in rainfall. Throughout the coffee lands including Jimma zone, temperatures are
generally higher during the dry season and lower in the wet seasons (Robel, 2012; ECFF,
2017). Thus, the productions of coffee in Jimma zone are greatly dependent on climate
variability especially rainfall and temperatures.

5
2.3 Coffee Crop Water Requirement

Crop water requirements are defined here as "the depth of water needed to meet the water
loss through evapotranspiration (ETc) of a disease-free crop, growing in large fields
under non-restricting soil conditions including soil water and fertility and achieving full
production potential under the given growing environment, Crop water requirement refers
to the amount of water that needs to be supplied (Doorenbos and Pruit, 1977).

ETc can also be derived from meteorological and crop data by means of the Penman-
Monteith equation. The Penman-Monteith method is used for the estimation of
atmospheric water demand or ETo (Allen et al., 1998). The accuracy of determination of
ETc will be largely dependent on the type of the climatic data available and the accuracy
of the method chosen to estimate the evapotranspiration (Nuha and Henery, 2000).

There are various methods to calculate ETc of coffee crop. These are direct and indirect
methods. The indirect methods involve calculating ETc from meteorological data, soil
water balance, and pan evaporation method (Blore 1966; Pereira, 1957; Wallis, 1963).

Allen et al. (1998) developed general guideline that used to calculate ETo from
meteorological data using Penman-Monteith equation and finally multiply it by crop
coefficient to obtain crop water requirement.

Lemos Filho et al. (2010) analyzed ETc for coffee grown at Minas Gerais State, Brazil
from 17 years (1961-1978) meteorological data. They reported annual coffee ETc values
varied from 798 to 1510 mm/year. However, they did not adjust coffee Kc obtained from
FAO. But this method is relatively simple and cost effective because coffee ETc is easily
calculated at already established meteorological stations. That is whys this method was
selected to analyze coffee ETc and IR in Jimma zone.

Silva et al. (2009) measured coffee ETc using soil water balance with neutron probe; the
water balance was performed at 14 days interval for about 2 years (720 day) in Brazil.
They reported coffee ETc varied from 1.83 to 6.4 mm /day with average of 4.01 mm/day.

The direct method such as Bowen ratio-energy balance technique, which directly
measure ETc of coffee as evaporative heat flux (Gutierrez and Meinzer, 1994).
However, this method requires advanced material and skilled expertise; besides this it’s

6
impractical to measure directly coffee ETc especially for large areas like Jimma zone
where duplicated experiment is needed that represents the whole area.

2.4 Coffee Crop Coefficients

The crop coefficient (Kc) is a factor indicating the ratio of ETc to the ETo. There are
various factors that determining Kc are crop type, climate, soil evaporation and crop
growth stage. The variability of actual Kc to tabulated data further increases when
regarding not only climatic differences but also variation in soil type, seeds and
management practices (Neugebauer, 2013).

There are different methods to calculate coffee Kc. These methods include water balance,
NDVI and Bowen ratio-energy techniques. Silva et al. (2006) measured coffee Kc using
water balance method combined with ETo. Gutierrez and Meinzer (1994) directly
measured coffee Kc with help of Bowen ratio-energy method in combination with local
ETo. However, Almhab (2011) derived the coffee Kc from satellite images; normalized
difference vegetation index was compared with FAO Kc to determine actual Kc at local
level. Blore (1966) attempted to measure the coffee Kc of weed clean coffee by
measuring soil water contents and combine with local climate data.

The most commonly used and standard coffee Kc value is the one that published on FAO
irrigation and drainage paper 56. The tabulated Kc values presented for coffee are in the
range of 0.9-0.95 for a clean weeded crop, and 1.05-1.10 for a crop with weeds. These
values are for well managed crops, 2-3 m tall, grown in a sub-humid climate (Allen et al.,
1998). However, the coffee Kc should be adjusted to the local conditions.

Depletion of soil-water content between p and the permanent wilting point will result in a
proportional reduction of actual evapotranspiration (FAO, 2012). The coffee water
extraction rate depends on amount of water available in the soil, soil depth and rooting
depth. Maximum rooting depth is defined as the deepest depth attained by a crop for any
specific soil situation (Johnson, 2007). The maximum rooting depth for coffee ranges from
0.9-1.5 m depth (Allen et al., 1998). Root concentration of coffee is in the 30-60 cm depth
(CRO, 2006). According to Pereira (1957) the most active portion of coffee root depth is 1
m where coffee gets most of their water. Water at depth greater this may be lost to deep
percolation.

7
2.5 Reference Evapotranspiration

The reference surface is a hypothetical grass reference crop with an assumed crop height
of 0.12 m, a fixed surface resistance of 70 s m-1and an albedo of 0.23. The reference
surface closely resembles an extensive surface of green, well-watered grass of uniform
height, actively growing and completely shading the ground (Allen et al, 1998).

Various methods were developed to estimate ETo, amongst the FAO four methods are
the commonly utilized methods namely Blaney Criddle , radiation, pan evaporation and
Penman methods .However, these method have their own shortcoming, their prediction
value are normally deviated from actual values (Doorobos and prutt 1977). The selection
of each methods are greatly depends on availability of climatic data and the level of
accuracy required. The inputs required for estimation of ETo are climate, altitude and
cultural conditions and related data. Some of the methods require limited input, where as
others require the extensive measured data, such T-mean, T-max, T-min, radiation,
relative humidity, wind speed at 2 m, sunshine hours, and different constants values.
Blaney-Criddle method uses mean temperature (T) and percentage (p) of total annual
daylight hours occurring during the period being considered (Blaney and Criddle, 1950 ;
Allen et al., 1998; Doorenbos and Pruit, 1977).

Radiation method uses measured air temperature and sunshine, cloudiness or radiation to
estimate ETo, however this method has its own shortcoming. Modified Penman utilizes
measured energy balance (radiation) and aerodynamic including wind speed and
humidity to estimate ETo, the shortcoming of this method is the requirement of local
calibration of wind function to obtain best result. In Pan Evaporation, ETo is the product
of both evaporation from pan surface and pan coefficients. The measurement of ETo are
primary affected by different factors such as radiation, wind, temperature and humidity.
The pan transfers heats to waters and facilitates evaporation process, and winds takes
evaporated water, subjected to local microclimate and produce more error when it is not
properly installed. The mentioned four methods are subjected short coming and deviation
from normal value. To overcome such defect of four methods, experts and researchers
were developed more accurate method called Penman Montieth (Allen et al, 1998).

8
2.6 Coffee Irrigation Requirement

Irrigation water requirements are the quantity of irrigation water in addition to precipitation
required to produce the desired crop yield. The amount and timing of precipitation strongly
influence irrigation water requirements (SCS, 1993). Irrigation aims to replace the water
used by the crops or lost by evaporation or drainage through the soil with objects of
producing the optimum yield and to apply water efficiently (Luke, 2006).

Mitchell (1988) summarized the merit of irrigating coffee as follows as:

 Reduces drought period in coffee;

 Provides a reliable insurance cover against crop failure for the coming year;
 Increases the beneficial microbial content of the soil;
 Increases the organic matter decomposition in soil in-situ because of the
prevailing high temperatures;
 Enables uniform fruit set and uniform berry size; and
 Improves the nutrient uptake.

Fernandes et al. (2009) concluded that, the productivity of drip irrigated coffee plant is
95% higher than non-irrigated ones. This suggests that coffee irrigation can significantly
increase the productivity of coffee crops. According Tesfaye et al. (2008) study
conducted in Jimma zone indicated that, supplemental irrigation is required as the annual
rainfall is less than the seasonal coffee ETc and its distribution is usually erratic.

2.7 The FAO CROPWAT Model

The FAO CROPWAT is a computer program that can calculate ETo, ETc and IR from
climatic and crop data. The program is interactive in nature. In addition, the program
allows the development of irrigation schedules for different management conditions and
the estimation of scheme water supply for varying cropping patterns. The CROPWAT
model is based on a water balance model where the soil moisture status is determined on
a daily basis from calculated evapotranspiration and inputs of rainfall and irrigation uses
the sole recommended FAO Penman-Monteith method for estimating ETo (FAO, 2009).

The climatic data required for input into CROPWAT are temperature, relative humidity,
wind speed and sunshine hours for the calculation of ETo. It has also four different

9
methods to calculate effective rainfall (FAO, 2009). Through the input of crop data
(growth stages, Kc value, root zone depth and allowable soil moisture depletion factor),
the programme calculates the ETc on a decade (10-day) basis. The decade values of ETc
summed up to obtain monthly values (Savva and Frenken, 2002).

Several researchers have been used the CROPWAT model for analysis of ETo, ETc and
irrigation requirements of crops including coffee during different growth stages
(dormancy, bud initiation, flowering, fruit set and maturity) (Naik et al., 2015; Surendran
et al., 2015; Lemos Filho et al., 2010).

Wubengeda et al. (2014) calculated ETo by CROPWAT 8.0 model using Penman -
Monteith equation for Sub-humid areas of Arsi zone, Ethiopia at two meteorological
stations. They reported monthly ETo values ranged from 3 to 5mm/day.

Fitsum (2017) analyzed temporal variability and trend of ETo in Sinana District, south
eastern Ethiopia. This study, considered Tmax, Tmin, RH, wind speed and sunshine hours
data with respect to station’s locations to calculate monthly ETo values for sub-humid
areas. The calculation of ETo was performed in CROPWAT 8.0 model using Penman –
Monteith equations. He reported ETo values varied from 3.22 to 4.34 mm/day.

Patel et al. (2017) carried out similar work on ETo estimation using CROPWAT model at
Ludhiana district (Punjab), north India. In this study, ETo was calculated using Penman -
Monteith equations from climate data record from 1970-2012. They detected that Penman-
Montieth method is the best method to estimate ETo. They reported average ETo values of
3.65 mm/day.

Mahtsente (2017) studied water demand analysis and irrigation requirement for major
crops at Holetta catchment, Awash Sub-basin, Ethiopia. She determined ETo, ETc and IR
for major crops for humid areas using CROPWAT 8 model based up on penman Monteith
method. The reported ETo values ranged from 3 to 5 mm/day.

Trivedi et al. (2018) estimated ETc of crop using CROPWAT 8.0 model for Shipra River
Basin in Madhya Pradesh, India. They estimated ETc of Soybean and other major crop
from long-tem meteorological data including rainfall, Tmax, Tmin, relative humidity, wind
speed and sunshine hour of stations as input data in CROPWAT Model 8.0. ETo was
calculated using Penman Monteith method. They reported average annual ETo values of
10
4.14 mm/day. However, the problem of this study was, they used average long-term ETo
values instead of 80% probability of non-exceedance ETo values to calculate ETc.
Doorenbos and Pruitt (1977) recommended 75 to 80% probability of non-exceedance ETo
values for calculation of the crop ETc and IR especially for design purpose.

2.8 Geo-statistical Interpolation and Mapping

The potential application of geostatistical interpolation in water resource and in particular

in irrigation management has been widely acknowledged by several authors and many
interpolation methods such as ordinary kriging, simple kriging and universal kriging for
derive climate related parameters such as temperature, rainfall, ETo and crop water
requirement had been successfully tested in different locations ((Fonteh and Podmore,
1994; Smith et al., 2007; Lemos Filho et al., 2010; Agnew and Palutikof, 2000).
However, the real applications of geostatistical interpolation method in irrigation
management are few in Ethiopia particularly for study areas.

Geo-statistical interpolation such as ordinary kriging play an important role in preparing

regional maps from sparsely distributed data; because it is not always possible to sample
the entire study area. Thus, unknown values must be estimated from data collected from
particular location that can be sampled (Burrough and McDonnell, 1998; Sahoo, 2012).

Several authors tested the performance of ordinary kriging methods for analyze spatial
variability of precipitation, evapotranspiration, and temperature. Cross-validation along
with applicable conditions of different models is applied to obtain the best-fit
interpolation model (Mardikis et al., 2005; Yang et al., 2011). However, ordinary kriging
does not take into account secondary information which has strong relationship with
primary variable (Li and Heap, 2008). For large areas like Jimma zone that consists of
diverse topography it is difficult to accurately estimate ETo and ETc without considering
topographic effect.

Hence, to improve interpolation results, combined utilization of ordinary kriging with

linear regression helps to overcome these problems. The auxiliary variables that affect the
spatial distribution of ETo and ETc at regional scale particularly for study area are many
uncertain factors, mainly elevation. Several authors used multiple regression model

11
combined with interpolation method, they obtained better results (Agnew and Palutikof,
2000; Martinez, 1996).

Agnew and Palutikof (2000) used multiple regression models that were refined by kriging
of the residuals to develop seasonal maps of temperature and precipitation in the
Mediterranean basin using latitude, elevation and distance from the sea as predictor and
they obtained better estimation results of ETo and precipitation.

Smith (2014) compared estimation methods of ETo at un-sampled locations. He

compared different interpolation methods in interpolating climate parameter and ETo.
Among them, Kriging with covariates of elevation and temperature had resulted in
smaller residuals.

Martinez (1996) carried out works on multivariate geo-statistical analysis of

evapotranspiration and precipitation in mountainous terrain. The ETo and precipitation
data was extended to 5-km grid for cokriging purpose. He pointed out that Co-kriging
improved ETo estimation uncertainty by about 11.5% as compared to ordinary kriging.
The assumption of co-kriging and combined use of ordinary kriging with linear
regression is more or less similar. Thus combined utilization of ordinary kriging with
linear regression play important role in minimizing ETo and ETc estimation error
particularly for topographically diverse area like Jimma zone.

Lemos Filho et al. (2010) analyzed spatial distribution map of coffee ETc using 45
meteorological stations data using ordinary kriging in R with GeoR package and obtained
good results. However, they didn’t take into account the topographic effect in
interpolation process. But, for study area it is important to consider topographic effect in
interpolation process to obtain best results because elevations have strong correlation
with ETo and ETc. For interpolation purpose, the computer program like R-software can
have ability to generate spatial surface prediction map using ordinary kriging method
combined regression method (Team, 2013)

12
 3. MATERIALS AND METHODOLOGY

3.1 Description of Study Area

3.1.1 Location

The study was conducted in one of coffee growing areas of Jimma zone, Oromia region
state, south west of Ethiopia. It is geographically located approximately between 7° 5'
38"N - 8° 54' 19" N latitude and 34° 49' 34" E - 38° 39' 17" E longitude with maximum
and minimum altitude of 3231 and 859 m above sea level, respectively. Due to altitudinal
differences, the zone has great physiographic diversity. The zone has a total area of about
19,293.5 km2. It has 20 districts and one administrative town with a total of 545 kebeles
(lowest administrative units) of which 515 of them are rural and 30 are urban. According
to CSA (2013) reports, the total projected population of the zone for 2017 is 3,209,127.

Figure 1.The location Map of the study area

3.1.2 Climate

The agro-climate condition of Jimma zone has been classified under sub-humid agro-
ecological zone. From long-term (1985-2016) records of meteorological data of 20
stations, the areal mean climate time series for the study area were estimated using the
Thiessen polygon method procedures (Fiedler, 2003).

The average areal annual rainfall was 1728.7 mm, about 71.5% of the annual total
amount of rainfall occurs between May and September and the rest 28.5%
13
from October to April, with a peak in the months of July and August. The mean
areal minimum and maximum monthly rainfall values were 31.2 and 280 mm occurring
during January and August, respectively. The average areal monthly Tmin varied from
10.7 to 12.9℃, with the annual mean of 11.9℃. The average areal monthly Tmax of the
zone ranged from 23.0 to 28.2℃ with the annual mean of 25.8℃, for March and
July, respectively. The average RH is varied from 55.5 to 82.3. The average mean
sunshine hour durations of zone varied from 3.56 to 8.16 hours (Appendix Table 3).

3.1.3 Crop productions

Farmers grow crops twice a year, one during the rainy season (June-September) with
rainfall, the other during dry season (December-May) using irrigations. The farmers grow
different crops in the study area. The major annual crops grown include teff, maize,
sorghum, wheat, and barley during rainy season; and tomato, onion, carrot, cabbage and
pepper during off-season. Farmers also produce stimulant perennial cash crops like coffee
and Khat. Of these, major crops coffee takes the lion shares. Coffee is mainly produced
under rain-fed conditions.

3.2 Selection of Meteorological Station

The Thiessen polygon method (Fiedler, 2003) was used to select spatially representative
stations and clipped to study area’s boundary in ArcGIS 10.1. Accordingly, 20 stations
were clipped to study area as of polygon. It includes Chida, Gojeb, Shebe, Dedo, Jimma,
Yebu, Bedelle, Arjo, Ejajji, Atinago, Limmu-genet, Natri, Sokoru, Yaya, Wolkite,
Botorbacho, Gatira, Chira, Agaro and Asendabo. Additionally 2 stations were added to
increase the extent of data to cover the entire area of zone; it includes Gore and Bonga.
Hence a total 22 stations were used for this study, fourteen being in Jimma zone and
remaining eight from around the vicinity of study area (Figure 2).

14
 A
B

Figure 2. Meteorological stations (A) and selected stations by Thiessen polygons (B)

3.3 Data collection and measurement

3.3.1 Meteorological data

The daily data for the period 1985 through 2016 (31 years) were collected from ENMA.
These collected climate data series includes maximum temperature, minimum
temperature, and wind speed at 2 m height, relative humidity, sunshine hours and rainfall.

Climate data processing: The climate time series at stations is require careful analysis for
their missing value, validity and homogeneity test before use. The monthly climate data
were tabulated in “Table” for each month starting from 1985 to 2016.

Filling missed climate time series: These were entered in R-software with MICE package
(Van Buuren and Groothuis-Oudshoorn, 2011). The monthly missed climatic variables
including rain fall were entered from the same months in the given time interval (1985-
2016) using multivariate imputation and chained equations method.

Homogeneity test of time series: Standard normal homogeneity test (SNHT) method
proposed according to Alexandersson (1986) was used for detecting inhomogeneity of
annual mean or total climate time series of Tmin, Tmax, RH, WS, sunshine hours and
rainfall data from 1985-2016 at gauging stations . SNHT performed using XLSTAT 2017
software (Addinsoft, 2017). Change in average or a presence of trend was detected.

The breaks in a certain year were corrected by double mass slopes with help of
Hydrognomon software (Kozanis et al., 2010). In this method a group of 5 to 10
neighboring stations are chosen in the vicinity of doubtful station. The mean annual time

15
series values are serially arranged in reverse chronological order to determine relative
consistency.

From figure 3, it was observed that the slight shift of mean in annual minimum temperature
from 11.24 to 11.96℃ around 1990, two means were detected one 11.24℃ and the other
11.96℃, about the deviation of 0.72℃ from the 1st mean. Which showed the data before
1990 was inconsistent and corrected by double mass curve of Figure 4. Similar to this one,
for the other twenty-one stations, the consistency of their climate time series (Tmin, Tmax,
WS at 2 m, RH, SH, and rainfall) were checked using similar procedure.

The double mass curve of Tmin of annual for Jimma station was presented as an example
in figure 4. From figure 4, a slight break in the slope of the double mass curve of Tmin
was observed around 1990; the software corrected Tmin data of Jimma station as follows:

 Ratio of the consistent period (2016-1990 or left line slope = 2 =

∑ ,
∑ ,
) was (Jimma station cumulative / other stations cumulative).

 Ratio of the in-consistent period (1989-1985 or right line slope = 1 =

∑ , ∑ ,
∑ , ∑ ,
) was (Jimma Station cumulative / other stations

cumulative).
 Therefore, the correction factor was (consistent period ratio / inconsistent period
.
ratio) , = .
= 1.12, and then Tmin data from 1989 to 1985 was:

= × 1.12 (Figure 4).

 Finally, the corrected annual Tmin of Jimma station was presented as of Figure 7

16
 13

Annual mean minimum temperature

12.5

12

11.5

11

10.5
1980 1985 1990 1995 2000 2005 2010 2015 2020

Year

mean mu1 = 11.238 mu2 = 11.957

Figure 3. SNHT Test results of mean annual Tmin for Jimma station 1985-2016

Figure 4. Double mass of Jimma station of January Tmin 1985-2016

Importing monthly Wind speed, relative humidity and sunshine hours: Agglomerative
hierarchical clustering classification method (Everitt et al., 2001) was used to classify
meteorological stations into similar climate regime with an input of parameter latitude,
longitude, altitude, Tmin, Tmax, and precipitation. The similarity between meteorological

17
stations was calculated using Ward's method (Ward, 1963). The Euclidian distance
was used to determine the distance between stations as dendrogram height.

From figure 5, it was seen that the 22 meteorological stations considered for this study
were grouped into three broad clusters, when we cut the dendrograms at the threshold of
10. Namely, cluster 1: Agaro, Atinago, Bonga, Chida, Gojeb, Jimma, Limmu-genet, Natri,
and Shebe; cluster 2: Arjo, Bedelle, Chira, Dedo, Gatira and Yebu; and cluster 3:
Asendabo, Botorbacho, Ejajji, Sokoru, Wolkite and Yaya. First individual cluster average
was calculated. Therefore, these climate time series data were exported from their
respective cluster.

Figure 5. Classification of climate region using dendrogram

3.3.2 Crop data

The collected crop data were includes planting date, harvesting date, dormancy period,
flowering date, pinhead date, rapid pinhead expansion, maturity date, Kc values, and
growth stages, rooting depth, critical depletion and crop height.

Coffee crop coefficient (Kc): The crop coefficient value obtained from published FAO
56 paper was adjusted to local conditions (Allen et al., 1998).

18
Adjustment of Kc initial: In this study, the Kc initial was derived both from water
balance and satellite images of the same coffee field (About 6 ha) found at JARC station.
A water balance was estimated from five years old coffee stands (74110 coffee variety).
The spacing between rows and trees are 2 and 1.5 m, respectively. One plot of about 40
× 40 m was selected from center of coffee field for water balance establishment. Three
coffee trees were sampled diagonally from selected plot for soil moisture sampling and
plant height measurement; all measurement was performed on these selected trees. Soil
water balance was performed taking soil samples for water content down to the 1 m depth
below tree canopies and about 20 cm from the tree trunk. Prior to weighing and oven
drying, the soil sample taken from 1m depth was thoroughly mixed. Finally, soil samples
were oven dried for 24 hours at 105 ℃. The gravimetric soil moisture content was
determined by using equation (3.8).

( )
MC (mm) = 1000 ∗ ∗ ∗ 3.8

where, FW is fresh weight (gm), DW is dry weight (gm), BD is bulk density (gm cm-3),
MC is moisture content and Zr is root depth and is 1m, most feeder roots of coffee founds
up to 1m depth, beyond this depth water is lost as deepercolation. Once soil moisture
content of a period was determined, change in soil water content ∆S of a root zone of a
coffee crop is equal to the difference between the amount of water added to the root zone,
Qi, and the amount of water withdrawn from it, Qo (Hillel, 1998) in ten day interval was
expressed by equation (3.9).

∆S = − 3.9

Where, ∆S is Change in soil water content (mm), is moisture content after ten days
mm, and is moisture content at start time (mm).

Bulk density: The soil sample taken by core sampler was oven dried at 105 ℃ until
constant weight attained. According Miller and Donahue (1990) Soil bulk density is:

p = 3.10

where: is soil bulk-density (g/cm3), Wd is weight of dry soil (g), and is volume
of core sampler (cm3).

19
Coffee ETc was determined using Equation (3.11) from measurement of gravimetric soil
sampling in ten-day interval during the initial season. Deepercolation and runoff values
were estimated by AquaCrop (Steduto et al., 2009).

ETc = + − DP − RO ± ∆ 3.11

where, P is rainfall (mm), DP is deep percolation (mm), RO is runoff (mm) and ∆ is

change in soil moisture (mm).

Figure 6. Sketch of experimental field and sampling method

Finally, the initial Kc value was determined from calculated ETc and ETo during
respective days. The daily ETo was calculated using standard method with decade
climatic data of JARC stations, situated about 212 m away from measuring site.

Kc = 3.12

The calculation was done for each interval separately and finally the Kc values of three
intervals were used for Kc-NDVI modeling.

To obtain more accurate information, the Kc initial was also estimated by satellite image
of landsat 8 with 30 m × 30 m pixel size for field under study. Four sky clear images
were downloaded from (http://earthexplorer.usgs.gov). The dates were 03 December
2017, 19 January 2018, 12 February 2018 and 12 June 2018. The downloaded images
were clipped to coffee field boundary. Prior calculation of NDVI the red and near-
inferred bands were converted into TOA. The NDVI was calculated using Rouse et al.
(1974) method.

NDVI = 3.13

20
where, NIR is near-infrared band, R is red band and NDVI is normalized difference
vegetation index.

The Kc initial was modeled using NDVI, climate corrected Kc FAO, and field measured
Kc data following (Kamble et al., 2013) methods using linear regression equation. The
Equation 3.15 was derived using Kc of water balance and climate adjusted Kc FAO with
its corresponding values of NDVI close to the water balance plot in December, January,
February and June.

Kc = 1.4402x − 0.14 3.15

After determining the Kc for all pixels within an image the Kc for a field can be
determined by calculating the arithmetic mean of all pixels within its boundary:

, = ∑ , 3.16

where, Kc field is field the arithmetic mean Kc for a crop field, Kc, iis the crop coefficient
for a crop image pixel and n is the number of pixels within a field. Finally, the arithmetic
mean of field Kc value derived from NDVI was used as input for initial stage in
CROPWAT 8.0.

Adjustment of Kc Midseason and late season: For specific adjustment in climates

where RHmin differs from 45% or where u is larger or smaller than 2.0 m/s, the Kc mid
and late values (Allen et.al., 1998) were adjusted by equation 3.17.

.
Kc = Kc + ⌊0.04(u − 2) − 0.004(RH − 45⌋ 3.17

where, Kc (tab) is the value of Kc taken from Table 12 of Allen et al. (1998), u is the
long-term mean value for daily wind speed at 2 m during the mid-season or late season
stages (m/s), RH - mean value of daily minimum relative humidity during mid-season
and late season growth stage, h is the mean plant height during both season.

The long-term daily RH and wind speed were collected from JARC stations during
respective growth stage of coffee crops. The adjusted Kc mid-season and late season
values were finally used for CROPWAT 8.0 input. However, Kc adjustment using water

21
balance is impractical during these periods due to progressive rainfall. Three plants were
diagonally selected; their heights were measured using tape and finally, mean value used.

Coffee growth stage: The development stage length data were collected from JARC
field observation report (Yilma et al., 1997) and interviewing coffee agronomy
researchers. Having these data the lengths of coffee development stage were determined
based on coffee phonological stages. Coffee was harvested from October to December.

3.4 Estimation of Reference Evapotranspiration

The CROPWAT 8.0 models based on Penman -Monteith equation was used for estimation
of ETo for each meteorological stations: The ETo was calculated using equation 3.3:

. ∆( ) ( )
ET = (3.3)
∆ ( . )

where, ET ∶ Reference evapo transpiration [mm day ], R ∶Net radation at crop

surface [JM m day ], G: soil heat flux [JM m day ], T: mean daily temperature
[℃ ], :Wind speed at 2m height [m s ], ( − ): Saturation vapor pressure [kPa],
∆: Slope vapor pressure curve [kpa ℃ ], : Psychrometric constant [kpa ℃ ]

The monthly values of climate parameters T max (℃), T min (℃), RH (%), wind speed (m
s-1) at 2 m height and sunshine hours of each year was fed into a computer-based
CROPWAT 8.0 for windows. The monthly ETo values for each year from the long-term
meteorological data were obtained from CROPWAT 8.0 (FAO, 2009).

The accuracy of ETo estimation through CROPWAT 8.0 from limited climatic data was
evaluated with that of full climatic data. The accuracy of exported data from neighboring
stations was also accessed with that of ETo estimated from full set of climatic data.

The monthly long-term from 1985 to 2016 (31 years) ETo data obtained were entered and
fitted to different standard frequency distribution models using a computer-based routine,
Rainbow software (Raes et al., 2006). The distribution that best fitted the data was used to
work out ETo occurrence at 80 percent probability of non-exceedance level. Kolmogorov-
Smirnov (KS) test was used to test monthly ETo (185-2016) normality for individual

22
station separately in rainbow software. Homogeneity test of monthly ETo was performed
based on the cumulative deviations from the mean (Buishand, 1982) in rainbow software.

Once the normality and homogeneity of ETo data was checked for individual station, the
ETo at 80 percent probability of non-exceedance was taken out for all stations. Then the
extracted regional monthly ETo values was again subjected to normality test using
Shapiro-Wilk test.

3.5 Estimation of crop water and irrigation requirement

Having values of Kc and ETo at 80 percent probability of non-exceedance, a ETc was

calculated according to Allen et al. (1998) using equation 3.4.

ETc = ETo x Kc 3.4

where, ETo is ETo (mm/day), ETc is crop water requirement (mm/day) and Kc is crop
coefficients given as fraction.

The irrigation requirements was estimated as difference between crop water

requirement and effective rainfall. It was calculated using equation (3.5):

= − 3.5

where, IR is irrigation requirement (mm) and Pe is effective rain fall (mm)

The effective rainfall was calculated considering losses and dependable rainfall. This
formula may be used for design purposes where 80% probability of exceedance is required.
The effective rainfall was computed following procedure given in FAO (2009) from:

P = 0.6 ∗ P − 10 for Pmonth <= 70 (3.6)

P = 0.8 ∗ P − 24 for Pmonth > 70 (3.7)

where, Peff is effective rainfall (mm) and P is monthly rainfall (mm).

3.6 Statistical Analysis

Having ETo, coffee ETc and irrigation requirement calculated at station level, exploratory
data analysis were performed using R (version 3.4.1) software. The distributions of data
were analyzed using classical statistics. To protect unwanted data outliers that affect the
23
descriptive statistics and characterization of spatial variation, histograms and Box-plots
were used for each parameter that requires inspections.

3.7 Mapping of ETo, Coffee ETc and IR

To improve the accuracy of estimation, a linear regression equation was developed

between observed data and station’s elevation. Using the linear regression equations the
residuals of ETo and ETc were estimated as the difference between observed and predicted
values at gage locations. Then the residuals of ETo and ETc value were interpolated using
ordinary Kriging over the entire region. Using DEM and regression equation, station based
data was further extended to 5 km spatial grids. The experimental semi-variogram was
calculated and variogram models were fitted for extended dataset of ETo and ETc. Next
ETo and ETc for each month starting from January to December were interpolated to
1km×1km spatial resolution by ordinary kriging using gstat package (Graler et al., 2016) in
R software (Team, 2013).

The final estimation of ETo and ETc at any location were obtained by:

Final prediction = residual interpolated map + regression kriged map

However, the IR was directly interpolated without considering elevation effects, because
weak correlation were observed between stations’ elevation and IR during dry season
(Table, 12).

The performances of the each fitted model prediction at un-sampled location were checked
by RMSE. The measured data at 22 stations were used to cross-validation by living one out
method; the model with smallest values of RMSE was selected for final interpolation
(Pebesma and Bivand, 2005)

∑ ( )
RMSE = 3.8

where RMSE is root mean square error, Oi is obseverd value, Pi is predicted value, n is
number of samples

24
 4. RESULTS AND DISCUSSION

4.1 Climate data consistency test

The annual mean rainfall of all the stations were found to be consistent for the period of
1985 to 2016 with strong determination coefficient of R2 value greater than 0.9983. The
graphs of the cumulative annual rain fall of tests stations versus the cumulative rainfall of
the other station were found to be a straight line or perfectly fitted (Figure 8)

From figure 7 it was observed that the annual minimum mean temperature of Jimma
station was homogeneous.

Figure 7. Corrected minimum temperature of Jimma station by double mass curve

25
Figure 8. Double mass curve of annual mean rainfall of meteorological stations
26
4.2 Coffee Crop Coefficient

From three water balances during three months during initial stage, mean Kc values ranged
from 0.54 to 0.88, with the average of 0.70 at JARC station (Table 1). The mean Kc-
NDVI varied between 0.61 and 0.84, with the mean of 0.69 (Table 2) during initial stage.
The climate adjusted Kc for both mid-season and late-season stage was found to be 0.99.
For all growth stages, the variability of Kc value was large, ranging from 0.54 to 0.99
during initial and mid-season, respectively (Table 3). The Kc value during initial stage was
much low as compared to midseason stage due to water stress as the result of limited
precipitation during the dormancy period. During initial stage, Coffee plants are often
intentionally water stressed to reduce vegetation growth and to encourage development of
berries, under these condition, Kc values from FAO should be reduced (Allen et al., 1998).
Thus, the coffee Kc values were more probably related to seasonal rainfall fluctuation and
canopy cover.

Furthermore, the estimated Kc was observed to be high between wet months of March and
October during pinhead expansion and fruit set; and low during dry month of the
December to 1st week of February, corresponding to dormancy period.

The Kc was linearly decreased from December to 1st week of February as soil moisture
getting low due to limited rainfall and coffee plants were under water deficit, consequently,
losing leaves. Due to water stress and climatic condition adjusted Kc values were lower
than reported Kc for coffee that ranges between 1.05 to maximum 1.10 under shade
condition (Doorenbos and Pruitt, 1977). However, this result was found within reasonable
coffee Kc ranges suggested by different authors. Marin et al. (2016) reported that the
coffee Kc ranges between 0.6 (dry period) and 1.9 (wet period), with a mean of 0.99, when
average ETo is 3.2 mm day-1 in Brazil. Carr (2001) found that the coffee Kc values range
from 0.7 to 0.8. In a similar way the result was consistent with Oliveira et al. (2003) who
reported that Kc values varied from 0.7 to 1.5 using water balance method in Minas Gerais,
Brazil. Silva et al. (2009) reported that Kc varies between 0.75 and 1.43 at Piracicaba,
Brazil. Also Marin et al. (2005) presented coffee Kc values varies from 0.6 to 1.6 at Brazil.

The coffee Kc values was fluctuated with season, the lower Kc value was recorded during
dry season than wet season, which is consistent with Gutierrez and Meinzer (1994) who
reported much lower Kc during dry season than wet-season, suggesting that a seasonal
27
reduction in water use by coffee may have occurred independently of seasonal variation in
evaporative demand. Coffee stomata have been exhibited a strong closing response to
reduced atmospheric humidity (Fanjul et al., 1985).

Table 1. Water balance and climate adjusted coffee Kc for initial, mid and late stage at
JARC station
ETo
P ETc
Month DAP DP (mm) RO (mm) ∆S (mm) mm/ Kc
(mm) mm/dec
day
11-Dec
20-Dec 1-10 0.00 0.00 0.00 -29.75 3.40 29.75 0.88
9-Jan
19 -Jan 31-40 11.50 2.40 0.00 -11.9 3.37 22.8 0.68
29-Jan
8-Feb 51-60 10.20 1.10 0.00 -9.95 3.51 19.05 0.54
Kc initial mean 0.70
Long-term mean monthly climate, PH and climate adjusted Kc during mid-season and late season

Mid-season Late-season
Variables Apr May Jun Jul Aug Sep Mean October
RHmin 50.90 56.7 63.81 70.40 70.36 65.91 63.01 61.00
WS
(m/s) 1.01 1.02 1.01 1.10 1.03 1.00 1.03 1.02
PH 2.81 2.85 2.87 2.93 2.96 2.98 2.93 3.12
Kc adjusted 0.99 0.99
Qo = previous soil moisture content, Qi= current soil moisture content, P= rainfall, DP=deep percolation,
RO= run-off, ∆S= change in soil moisture, DAP= days after harvesting, PH (mm) = plant height, WS= wind
speed (m/s), RHmin = minimum relative humidity (%).

Table 2. NDVI and Kc value for model development during 2017/18 years and Kc NDVI
Kc Kc
measur NDVI
Image ed at
acquir Pixel that
ed date NDVI pixel Coffee field canopy characteristic
3-Dec 0.697 0.88 0.84 Moderate canopy cover
19-Jan 0.58 0.68 0.62 sign of leaf wilt for coffee and shed trees
12-Feb 0.46 0.54 0.61 leaf fall for coffee and shed trees, weeds between rows dried
9-Jun 0.79 0.99* Full effective canopy cover
Kc Mean initial stage 0.69
‘*’ climate adjusted FAO Kc value for mid-season

28
Table 3. Summary of coffee Kc and length of development stage
Development FAO Kc Develop Root P Description
stage Kc adjusted ment depth
stage (m)
Dormancy period, from
Initial season 1.05 0.69 96 0.9 0.40 harvesting to flowering. From
01 November to 04 February
From flowering to fast pinhead
>> 0.40 expansion. From 05 February to

>>

>>
Development 66
11 Apri1
From fast pinhead expansion to
maturity.
Mid-season 1.10 0.99 168 1.5 0.40 April,12 to September, 26

From maturity to harvesting,

Late-season 1.10 0.99 35 1.5 0.40 from 27 September to 31
October.
Total 365
Plant height (average) 3.25 m

4.3 Reference Evapotranspiration

4.3.1 Validation of CROPWAT 8 model

The results revealed that the ETo estimated from exported data from similar climate
regime have good agreement with that of ETo estimated from full set of climatic data at
Jimma stations. The root mean square error (RMSE) value was 0.14 mm day-1. It was also
exhibited by strong determination coefficient of R2 value of 0.88. While, the ETo estimated
from limited data have lower accuracy to estimate ETo, because when compared to ETo
estimated with full set of climatic data have RMSE value of 1.24 mm day-1 and R2 value of
0.6. The result show that the monthly ETo estimated from exported data were better than
ETo estimated from limited data (Table 2).

4.3.2 Explanatory analysis

Homogeneity test of ETo: From figure 11, it can be clearly seen that deviations line were
not crossed horizontal lines of probabilities, the monthly ETo values of August for Jimma
station was homogenous. Similar to Jimma station, the monthly ETo values for 22 stations
are homogenous. When the deviation crosses one of the horizontal lines the homogeneity
of the data set is rejected with respectively 90, 95 and 99% probability (Dirk, 2013).

29
Normality test: The Kolmogorov-Smirnov values of long-term monthly and annual ETo
value for each station considered varied from 0.06 to 0.18, which showed that the data was
normally distributed, because estimated KS values was greater than significance level α at
p = 0.05. The distribution accepted with the significance level of 0.05 probabilities
(Appendix Table 4). With a significance level of α = 0.05, if KS ≥ 0.05, the data can be
deemed to be in normal distribution (Molin and Abdi, 1998).

Having the normal distributions of long term ETo at each stations. The regional estimated
annual and monthly P-value of Shapiro-Wilk test varied from 0.11 to 0.95, these values
was found within normal ranges. Therefore, regional monthly and annual ETo values were
normally distributed (Table 5). In Shapiro-Wilk test, when the P-values is less than or
equal to 0.05, the data does not fit the normal distribution with confidence level of 95%
(Royston, 1982). Furthermore, the box-plot revealed there was no outlier both for mean
annual and monthly ETo values (Figure 10 B). The histogram reveals some data dispersion
around the mean value and its frequency distribution (Figure 10 A). The data presented
with acceptable CV, however, relatively high CV was recorded during summer season due
to rainfall variation while low during dry season (Table 5).

4.3.3 Annual and monthly spatial distribution of reference evapotranspiration

Annual ETo values varied from 3.53 to 4.20 mm day-1 and averaged 3.84 mm day-1
when all 22 weather stations were combined (Table 5). Figure 12 M clearly indicates
that the maximum annual ETo value were located in southern tip (Shebe-Sombo, Dedo and
Omonada), northeastern (Botor-Tollay) and eastern tip (Sokoru) regions. Of the locations,
highest annual ETo value (4.5 mm day -1) was found in extreme eastern tip. The minimum
annual ETo value in Jimma zone were located in the western highlands (Gera, Setama,
Sigmo and Gatira), southern (Dedo and Seka-Chekorsa) and northeastern regions. From
these areas, lowest ETo value (3.2 mm day-1) was located in southeast region. The lowest
ETo had a reduction of 25.1 % as compared highest ETo predicted in extreme eastern tip.

In general, ETo value was smaller in highland areas because it is associated with colder
climate and higher precipitation. However, ETo value was higher in low altitude areas,
hence it is associated with warmer climates. Thus, the spatial variability of predicted ETo
value caused by varied elevation. The decrease of ETo with altitude is in accordance with
the decrease of air temperature with altitude. Because an altitudes have significant strong
30
negative correlation with annual ETo in Jimma zone (Table 6). The lower ETo in the
highland areas may contributed from altitudinal difference, the highland areas
characterized by high altitudes as compared to lowland areas. This means the greater
altitude of an area is the smaller temperatures, as altitudes increase the temperature
decreases because the temperature is the major climatic factor, which affects ETo (Allen et
al., 1998). With increasing elevation, the air temperature decreases and RH increase; thus,
the ETo generally decreases (Um et al., 2017). ETo will vary with altitude, every 1000 m
rise in altitude will result in 10% ETo reductions (Doorenbos and Pruitt, 1977)

The monthly ETo spatial distribution presented in Figure 12 has shown some variations
through the year. The spatial distribution was more or less similar for monthly and annual
ETo values. The monthly ETo increases from the southern highland areas towards the
extreme southern tip lowlands and again increase from eastern highland to extreme eastern
tip lowland areas along Gibe river basin, whereas ETo had decreasing trend from central
region to western highland regions of Jimma zone (Figure 12 A-L).

There were a significant (P<0.05) temporal variations among monthly ETo values (Table
5). The graph shows a decrease from March to July and increase from December to March
(Figure 9). The high ETo values were predicted from January to May and the values
ranged from 3.94 to 4.97 mm day-1, corresponding to dry season. Of all months, the
highest ETo value was predicted in March and the value varied from 4.12 to 4.97 mm day-1
(Table 5). The highest ETo value in March might due to warmer air and lower humidity.
During dry season, high amount of energy was available when temperatures are higher.

ETo values during summer season (July to September) was greatly very low and the values
ranged from 2.24 to 3.99 mm day-1 with the mean of 3.18 mm day-1. Summer season is a
main rainy season in Jimma zone. Of these months, the smallest ETo value was predicted
during the rainy month of July and the values varied from 2.34 to 3.43 mm day-1 (Table 5).
ETo during July had a reduction of 35.9% as compared to march. This low ETo value
during rainy season may due to lower air temperature and higher humidity as the result of
progressed rain fall. According to Singo et al.(2016) who reported that the
evapotranspiration rates tended to drop to low levels when the air around the plants
was too humid during rainy season.

31
The temporal variation of ETo was observed due to changes in solar radiations as a result
of a change in the season, statistically a significant strong correlations were observed
between monthly ETo values and climatic parameters at p<0.05 (Table 6). Low ETo
values were obtained during the summer when it was cool, cloudy and humid, while the
high during winter when it was hot, sunny and dry, (Table 5). The monthly ETo values
varies with same pattern with maximum air temperature and sunshine hours, but varies in
opposite direction with rainfall and relative humidity. Sunshine hour duration had strong
positive significant correlations with monthly ETo of November, December, January,
April, and May with the correlation values ranged from 0.79 to 0.89 (Table 6). Usually the
rises of temperature are happen, due to clear sky during dry season, and consequently
increasing ETo.

Maximum air temperatures had positively strong significant correlation (0.94***) with
monthly ETo values. Implying seasonal ETo values was mainly affected by maximum air
temperature (Table 6), as air temperature increases ETo will increase. This phenomenon
was confirmed by Rasul and Farooqi (1993) stated that an increase in temperature will
directly influence the evapotranspiration rate of atmosphere. The rainfall was negatively
correlated with ETo especially during the rainy month of July (-0.50), i.e. as rainfall
increases ETo become decreased (Table 6). Summer rainfall plays a vital role in reducing
the evaporative demand to some extent (Naheed and Rasul, 2010). Relative humidity had
strong negative correlation with monthly ETo values, Thus the relative humidity was the
dominant factor that determine the temporal fluctuation of monthly ETo values. In
equatorial countries like Ethiopia, the difference in ETo is due to cloud cover, temperature
and relative humidity (Adem et al., 2017).

This result was in agreements with Savva and Frenken (2002) who reported that ETo value
of 4.2 mm day-1 for April at Jimma zone. Similar finding was reported by De Bruin et al.
(2010) who reported that ETo values varied from 3 to 5 mm day-1 for February at Jimma
zone areas. Also similar results of monthly ETo values ranged from 4.01 to 5.24 mm day -1
were reported by Fitsume et al. (2017) at Akaki areas. They found out that lower ETo
values during peak rainy season and maximum during a dry season.

32
4.3.4 Cross-validations of Models

Among the fitted models, the exponential model has superior performance in producing
lower RMSE in seven situations including March, April, May, June, August, November,
and annual. The spherical model produced smaller RSME during the month of February,
October and December, whereas, the gaussian model produced lower RMSE during the
months of January, July and September (Table 7).

Table 4. ETo value estimated from full, exported and CROPWAT estimated data at Jimma
station for 2015
Month CROPWAT (full data) exported data CROPWAT estimated
January 3.52 3.66 5.37
February 4.15 4.25 5.88
March 4.47 4.65 6.02
April 4.35 4.52 5.44
May 4.29 4.05 5.08
June 3.46 3.45 4.65
July 2.71 3.07 4.06
August 3.18 3.27 4.48
September 3.78 3.62 4.78
October 3.88 3.81 4.73
November 3.75 3.75 4.44
December 3.74 3.78 4.93
Annual 3.77 3.82 4.99
RMSE 0.14 1.26
Adjusted R² 0.88 0.60

Table 5: Statistic analysis of monthly and annual ETo for Jimma zone

Station ETo Model predicted ETo

Mean Max Min Shapiro-Wilk Min mean max
Period (mm) (mm) (mm) SD CV (%) P-value (mm) (mm) (mm)
January 3.91de 4.23 3.70 0.15 3.84 0.51 3.49 3.90 4.22
February 4.43b 4.79 4.10 0.18 4.18 0.95 3.94 4.42 4.79
March 4.61a 4.99 4.20 0.20 4.44 0.77 4.12 4.60 4.97
April 4.47b 4.89 4.11 0.20 4.50 0.92 4.17 4.46 4.67
May 4.12c 4.47 3.66 0.19 4.62 0.91 3.63 4.10 4.48
June 3.59g 3.94 2.98 0.26 7.22 0.33 2.90 3.57 4.10
July 2.97j 3.25 2.57 0.19 6.53 0.21 2.34 2.95 3.43
August 3.04i 3.34 2.46 0.23 7.64 0.12 2.24 3.02 3.63
September 3.42h 3.70 2.88 0.23 6.82 0.11 2.64 3.40 3.99
October 3.95d 4.40 3.39 0.23 5.80 0.75 3.37 3.93 4.37
November 3.85e 4.30 3.43 0.23 5.89 0.92 3.16 3.84 4.36
December 3.78f 4.14 3.46 0.20 5.29 0.54 3.13 3.76 4.26
Annual mean 3.84 4.20 3.53 0.19 5.05 0.85 3.17 3.77 4.23
LSD ( 0.05) 0.06

33
Means followed by the same letters in a column are not significantly different from each other at a 5%
probability level, CV= coefficient of variation, SD=standard deviation

Table 6. Pearson Correlations of monthly ETo with metrological parameters

Monthly Metrological parameters

Period Altitude Tmin Tmax RH Sunshine Rainfall
spatial correlations
January -0.69** 0.43* 0.69** -0.49 0.88* -0.11
February -0.64** 0.47* 0.65** -0.64 0.83* -0.07
March -0.58** 0.38 0.61* -0.72 0.86* -0.21
April -0.34 0.26 0.41 -0.43 0.89* -0.43
May -0.62** 0.39 0.57* 0.36 0.83* -0.32
June -0.65** 0.29 0.66** -0.69 0.52 -0.42
July -0.78*** 0.24 0.77*** -0.69 0.54 -0.41
August -0.84*** 0.33 0.8*** -0.72 0.39 -0.50*
September -0.81*** 0.39 0.76*** -0.64 0.44 -0.34
October -0.61** 0.34 0.59** -0.44 0.68 -0.22
November -0.75** 0.37 0.76*** -0.70 0.79* -0.33
December -0.79*** 0.43* 0.77*** -0.62 0.83* -0.12
Mean -0.78** 0.42* 0.75*** -0.52 0.79* -0.44*
Temporal correlation - 0.10 0.94*** -0.86** 0.81** -0.66*
*, ** and *** = significant at 0.05, 0.01 and < 0.0001 level, respectively, WS= wind speed, RH=relative humidity

Table 7. RMSE value between modeled monthly ETo and observed ETo values

Ordinary kriging models

Month Spherical Exponential Gaussian
January 0.16 0.16 0.15*
February 0.18* 0.19 0.22
March 0.22 0.19* 0.20
April 0.23 0.20* 0.28
May 0.20 0.19* 0.23
June 0.26 0.21* 0.22
July 0.22 0.17 0.16*
August 0.24 0.22* 0.23
September 0.25 0.25 0.24*
October 0.24* 0.24 0.24
November 0.23 0.22* 0.24
December 0.21* 0.21 0.22
Annual 0.19 0.18* 0.20
RMSE=Root mean square error, “*” = is selected model for interpolation.

34
Figure 9. Mean monthly ETo for Jimma zone

A B

Figure 10 Histogram (A) and Boxplot (B) for annual ETo

Figure 11. ETo homogeneity test for August at Jimma stations in rainbow software

35
Figure 12. ETo kriging maps of annual and monthly ETo for Jimma zone

36
4.4 Coffee Water Requirement

4.4.1 Explanatory analysis

The Shapiro-Wilk test results indicated that the monthly and annual coffee ETc was
normally distributed. Thus the calculated P-values ranged from 0.12 - 0.98, which was
found within acceptable ranges, which is suitable for statistical analysis without needing to
be processed (Table 8). If the p-value is greater than the chosen alpha level (0.05), the data
tested come from normally distributed populations (Royston, 1982). The histogram shapes
of annual coffee ETc looked like bell-shaped with the highest point in the middle or
symmetrical slopes in either side, which showed that the data are normally distributed
(Figure 13 A). Similar behavior is observed for monthly coffee ETc. The box-plot of mean
annual coffee ETc showed that there were no data outliers, similar patterns were
detected for mean monthly coffee ETc (Figure 13 B). This condition plays a significant
role in variogram analysis and kriging interpolations (Graler et al., 2016).

Coefficient of variation was found to be the greatest from June to September, and ranged
from 6.61 to 7.25 %, probably due spatial variation of rainfall in study areas. However,
relatively low CV (5.43 % on average) during the rest months (Table 8).

4.4.2 Annual and monthly spatial distribution of coffee water requirement

Spatially, the annual ETc predicted for coffee varied from 1048 to 1377 mm year-1 over the
entire study areas (Table 8). The spatial distributions of annual coffee ETc are more or less
similar with annual ETo values. The upper most annual coffee ETc value is with the
narrow range of 1300 to 1377 mm year-1 which predicted in extreme southern tip,
easternmost tip and extreme northeastern regions. Of the locations, the highest coffee ETc
value of 1377 mm year -1 was found in extreme eastern tip. The central region had medium
annual ETc values covering some parts of Serbo, Mana, Gomma, Limmu-genet, and
Asendabo Woredas (Figure 16). The least coffee ETc value was predicted in the western,
southern and northeastern regions, ranged from 1048 to 1150 mm year -1. Of locations, the
lowest ETc value of 1048 mm was predicted in southeastern part of region (Figure 16 M).

The apparent spatial variation of annual coffee ETc values might be due to ETo variation
as affected by elevation. The lower ETc values were predicted in highland areas, whereas
higher in lowland areas of Jimma zone. Generally, the possibility of an increasing in the

37
ETc values from the southern highland to extreme southern tip lowland and again increase
from eastern highland to extreme eastern lowland areas, whereas decrease from central
plain to western highland parts. The reason is that the effluence of altitudinal differences,
the highland parts of study area is characterized by higher elevation as compared to
lowland areas (Figure 16). This condition was justified by altitude was significantly
negatively strong correlated with ETo in the study area, as altitudes increase ETo will
decreases, consequently the ETc decreases (Table 6). Complimentary to this result, at
elevation above 2300 m, the annual evapotranspiration values have been gradually reduced
by the low atmospheric demand because of low temperatures as the elevation increases
(Kiptala et al., 2013). With increasing height, air temperature drops uniformly with altitude
at the rate of approximately 6.5 ℃ per 1000 meters (Jocik, 2004). An altitude is the major
factor which affect the ETc of the crops because it directly affects temperature. This was
also verified by Rasul and Farooqi (1993) who reported that increase in temperature will
directly influence the evapotranspiration rate and water requirement of crop.

The spatial distributions pattern of monthly coffee ETc was similar to that of annual coffee
ETc (Figure 16 A-L). Statically significant variation were observed among monthly ETc
values at P<0.05 (Table 8). The upper most range of monthly ETc value is within the
narrow range of 119.01 to 131.11 mm during the months of March, April, and May. Even
though statistically non-significant between them, the highest coffee ETc was observed in
April, during this month the coffee plants found at stage of fast pinhead expansions (Table
8). This highest ETc during April might be due availability of high evaporating energy,
optimum soil moisture and coffee is existed at full canopy cover and then it can able to
transpire at higher rate.

The least ETc value was recorded during the months of November, December, and
February, with the narrow ranges of 80.9- 91.84 mm, with the lowest in the November, but
non-significant between them, which corresponding to the dormancy period. The minimum
ETc value during these months might be due to limited soil water content and reduction of
transpiring areas as the result of leaf fall (Figure 16 K, L and A).

In general, ETc of coffee was lowest during the initial stage and highest during mid-season
stage. However, moderate during summer rainy season (June to September) due to lower

38
ETo. During summer rainy season the amount of energy available to make
evapotranspiration was quite low, consequently ETc values was suppressed.

This value was in agreement with Lemos Filho et al. (2010) who reported monthly coffee
ETc values for all months ranged from 37.63 to 152.79 mm, with annual ETc ranges from
798.25 to 1510.33 mm year-1 and mean of 1123.3 mm year-1. Pereira et al. (2011) who
reported that coffee ETc value ranges from 3.68 to 4.04 mm day-1 in Brazil, when ETo is
3.53 to 3.88 mm day-1. The current annual mean ETc value was consistent with Allen et
al. (1998) who reported that coffee ETc values ranged from 800 to 1200 mm year-1 for
sub-humid climates. Oliveira et al. (2003) reported coffee ETc value varied from 2.52 to
3.50 mm day-1 in Brazil. Gutierrez and Meinzer (1994) found that coffee ETc values vary
from 2.04 to 5.23 mm day-1. Also FAO (2018) recommended coffee ETc values varies
from 2.86 to 3.57 mm day-1 (20 to 25 mm per week) in Lao in Southeast Asia.

4.4.3 Semi-Variogram models and parameters

For all of variograms models fitted, the monthly and annual mean values of ETc revealed
moderate to strong spatial dependence degree, presenting SDD smaller than 27.1%. The
spatial SDD ranged from 0 to 27.1%. The strong spatial dependency was detected for all
months except for March in case of gaussian. The exponential model (0%) presented
strong spatial dependency degree, followed by spherical (2.2%) and gaussian (18.2%)
models. This means that the three models tested were characterized by strong spatial
dependency. The average SDD considered for all months varied from 0 to 18.23% (Table
10). Similar results were reported by Lemos Filho et al. (2010) who reported the spatial
dependency degree for monthly coffee ETc ranges from 0 to 30.42%, which exhibit
moderate to strong SDD. According to Cambardella et al. (1994), spatial dependency
degree has classified into: strong spatial dependency (SDD≤25%), moderate spatial
dependency (25%≤SDD≤75%), and weak (SDD≥75%).

In all months of the year; there is considerable variation of ranges, which varied from 9 to
30 km. The spherical model produced the range varied from 26 to 30 km, with a mean of
26.5 km. which means autocorrelation existed over quite a large range or sill will not reach
quickly. Whereas, exponential model produced ranges varied from 13.6 to 14.2 km, while
the gaussian model varied from 8.8 to 14.2 km, with the mean of 12.1 km. comparatively
spherical model produced the longer mean range, which is approximately 54.2 and 48.2%
39
of exponential and gaussian models, respectively; both exponential and gaussian models
produced approximately similar results (Table 10). The geographic diversity might cause
spatial heterogeneity of coffee ETc in Jimma zone within shorter range. Thus, the values
beyond this range distance are not spatially auto-correlated or no longer relationship.

4.4.4 Cross-validations of interpolation models

Among the fitted models, the exponential model was superior over other models followed
by gaussian model; however, the spherical model produced the best result only during the
month of March with lower RMSE (5.18). The exponential model produced the smallest
RMSE during February (3.64), May (5.94), June (7.04), September (6.46), October (6.60),
November (4.90) and annual (59.02). However, gaussian models produced the lowest
RMSE in January (3.46), April (5.19), July (4.77), August (6.01) and December (4.47)
(Table 9). There was strong determination coefficient R2 of 0.98 between exponential
model predicted annual ETc and observed ETc (Figure 14).

Table 8. Descriptive statistics of monthly and annual Coffee ETc for Jimma zone

Station ETc Model predicted ETc

Shapro-
Month Mean SD CV Max Min wilk Min Mean Max
mm mm % mm mm P-value mm mm mm
January 85.98efg 3.41 3.96 93.00 80.9 0.61 75.94 85.73 93.4
February 91.84def 3.76 4.10 99.30 85.2 0.94 81.7 91.59 99.31
March 124.26ab 5.37 4.32 134.50 113.5 0.86 111.4 123.9 133.7
April 131.11a 5.48 4.18 143.00 120.5 0.98 121.1 130.9 138.5
May 125.86ab 5.85 4.65 136.70 112.3 0.79 111.4 125.5 135.5
June 106.56c 7.21 6.77 116.50 89.5 0.19 86.56 106.1 121.3
July 93.29de 6.17 6.61 102.20 80.5 0.12 73.34 92.81 108.0
August 94.21de 6.83 7.25 103.30 77.4 0.20 70.78 93.65 111.5
Septembe 101.79cd 6.69 6.57 110.00 86.5 0.12 79.55 101.3 118.2
October 119.07b 6.72 5.64 132.40 103.1 0.81 101.5 118.6 132
November 80.92g 4.70 5.80 90.10 72.5 0.88 66.8 80.58 91.35
December 82.42fg 4.33 5.25 90.50 75.6 0.49 68.98 82.09 92.34
Annual 1237.22 60.1 4.85 1350.10 1129.4 0.57 1048 1233 1377
CV (%) 16.8
LSD@5% 10.2
Note, SD=standard deviation, CV= coefficient of variation, means with the same later are not significantly
different from each other at 0.05.

40
Table 9. RMSE value for monthly and annual coffee ETc for ordinary models

Months Spherical Exponential Gaussian

January 3.60 3.59 3.46*
February 4.04 3.64* 3.68
March 5.18* 5.25 5.25
April 5.40 5.92 5.19*
May 6.13 5.94* 6.32
June 7.31 7.02* 7.47
July 5.41 5.49 4.77*
August 6.86 6.61 6.01*
September 6.96 6.46* 7.14
October 6.82 6.60* 6.86
November 5.09 4.90* 5.16
December 4.74 4.47 4.47*
Annual 61.75 59.02* 63.29
*= selected model for interpolation, RMSE=root mean sum square of error

41
Table 10. Parameters of the variogram models obtained for monthly and annual values of coffee ETc in Jimma zone

Month model Partial sill Nugget SDD Range Model Partial sill Nugget SDD Range Model Partial sill Nugget SDD Range
Jan Exp 7.9 0.0 0.0 13.9 Sph 7.2 0.0 0.0 25.8 Gau 6.3 0.9 12.2 10.8
Feb Exp 8.1 0.0 0.0 13.8 Sph 7.3 0.1 1.5 26.1 Gau 6.3 1.1 15.1 12.0
Mar Exp 13.2 0.0 0.0 13.7 Sph 11.8 0.3 2.6 26.4 Gau 8.6 3.2 27.1 14.4
Apr Exp 8.2 0.0 0.0 13.6 Sph 7.1 0.4 5.8 27.5 Gau 6.4 1.1 14.9 10.3
May Exp 62.6 0.0 0.0 14.0 Sph 51.6 7.0 12.0 30.0 Gau 44.1 13.3 23.1 13.3
Jun Exp 31.5 0.0 0.0 14.1 Sph 28.5 0.2 0.6 26.1 Gau 23.1 5.3 18.5 13.1
Jul Exp 31.2 0.0 0.0 14.1 Sph 28.3 0.1 0.5 26.1 Gau 25.6 2.9 10.2 8.8
Aug Exp 43.0 0.0 0.0 14.2 Sph 39.0 0.1 0.3 26.1 Gau 33.6 5.3 13.7 11.8
Sep Exp 38.7 0.0 0.0 14.1 Sph 35.2 0.1 0.2 26.0 Gau 27.7 7.0 20.2 13.5
Oct Exp 24.3 0.0 0.0 14.1 Sph 21.9 0.3 1.3 26.4 Gau 19.5 2.6 12.0 10.5
Nov Exp 15.6 0.0 0.0 14.1 Sph 14.2 0.1 0.6 26.1 Gau 11.3 2.7 19.4 13.3
Dec Exp 14.2 0.0 0.0 13.9 Sph 12.8 0.1 1.0 26.0 Gau 10.1 2.7 21.0 13.5
Annual Exp 2798.0 0.0 0.0 14.1 Sph 2546.8 0.0 0.0 26.0 Gau 1995.7 514.6 20.5 13.5
Exp=exponential, Gau=Gaussian, Sph=Spherical, =( / + ), SSD=spatial degree of dependency, Co= nugget, C= partial sill

A
B

Figure 13: Histogram (A) and box-plot (B) for annual coffee ETc (mm/ year)

42
 1350

1300
y = 0.837x + 201.0
Predicted ETc mm

R² = 0.986
1250

Predict…
1200

1150

1100
1100 1150 1200 1250 1300 1350 1400
Measured ETc mm

Figure 14: Annual ETc validation using linear regression between actual and predicted data

Figure 15. Semi-variogram and fitted model for mean annual coffee ETc in Jimma

43
Figure 16. ETc kriging maps of monthly coffee crop ETc for Jimma zone

44
4.5 Coffee Irrigation Requirement

4.5.1 Explanatory analysis

The Shapiro-Wilk test result showed that monthly IR was normally distributed for January,
March, April, October, November, and December and annual with the p-values ranged
from 0.07 to 0.63 (Table 11). The higher CV of monthly IR was recorded for months of
May, June and September with the range of 167-268%, which is contributed by higher
spatial variability of rainfall in corresponding months. While monthly IR for other months
founds within acceptable CV range i.e. 13.23 to 62.7 %, and was considered to exhibit
moderate to higher heterogeneity (Table 11).

4.5.2 Spatial distributions of monthly and annual coffee Irrigation Requirement

The annual coffee IR for all stations ranged from 300.6 to 842.9 mm year-1. However the
predicted value ranged from 300.7 to 723.3 mm year-1 within outline of study areas (Table
12). The spatial distributions of annual coffee IR are heterogeneous in study area with CV
of 31.05 %, which is contributed by higher spatial variability of rainfall. The upper most
annual IR was predicted around regions of extreme eastern tip (Sokoru lowlands),
southeastern and northeastern regions, and the values ranged from 650 to 750 mm. In
contrast, the lower coffee IR, was predicted around locations of western (Sigmo, Gera,
Setama and Gatira), southeastern (Omonada) and southern (Dedo and Seka-Chekorsa),
with the values ranges from 300.7 to 450 mm year-1. However, the some regions of
Limmu-Kosa and Limmu-Seka experienced moderate coffee IR, with the values
approximately varied from 500 to 550 mm year-1 (Figure 18 H).

Spatially, the annual coffee IR was experienced increasing trends from western parts to the
eastern and again increases from southern to northern parts of the study area. The reason
of increment in both directions might be due to spatial variation of rainfall. The rain fall
had significant strong negative correlation with both latitude and longitude during dry
periods (Table 12). This means the amount of rain fall had significantly decreasing from
west to east and from south to north in study areas.

The spatial variations of monthly coffee IR were heterogamous with over all CV of 88.1%,
which is contributed by higher spatial variability of rainfall in corresponding months. The
spatial variation distributions of monthly IR for January, February, March and April have

45
indicated more or less similar distributions pattern. The high IR during these months was
indentified around region of eastern (Sokoru) and northern (Limmu-Kosa and Limmu-
Seka). Whereas the low coffee IR was predicted in western (Gera, Setama, Sigmo and
Gatira), southern (Seka-Chekorsa and Dedo) and eastern (around Natri) regions (Figure 18
A, B, C and D). The spatial distribution of coffee IR during May was similar with
September. During these months the least IR were predicted for western and southern
regions. However, the extreme eastern tips and southeastern region require high irrigation
relatively (Figure 18 E and F). The spatial distributions of coffee IR during October are
more or less similar with annual IR (Figure 18 G).

Statistically significant variations were observed among monthly coffee IR at p<0.05. The
upper most monthly value of coffee IR was predicted in November, December, January,
February and April, with range of 53.4 to 77.7 mm, and the highest predicted during
February (77.7 mm) (Table 12). This highest IR during period might be due to smaller
rainfall and high evaporative demand of atmosphere during these dry spells. The minimum
coffee IR was predicted from June to September (Figure 17) due to probability occurrence
of high rain fall during summer season (June to September). There is no irrigation
requirement during months of June, July and August in study areas, with mean values 0 to
1.7 mm, which approaches to ~ 0 mm; the ETc demand during these three months, were
completely satisfied by rain fall (Table 12).

In General, the coffee IR amount considerably decreased with increasing rate of

precipitation. Thus, IR has an inverse relationship with rainfall; it decreased with rise in
rainfall but increased with reduction fall of rainfall. The monthly IR exhibited the
decreasing trends from March to August. Thus, the main causes for temporal variation of
coffee IR might be due to fluctuation of rain fall and ETo through the year in Jimma zone.
The spatial variation of coffee IR in study area might be due to differences in climatic
parameters such as changes in temperature, solar radiation and percentage change in
effective rainfall (Nkomozepi and Chung, 2012).

During the months of November to end of December the coffee plants stay under
dormancy period or flower bud undergo dormant for about two months (November to end
of December), even though the requirement is there, the plants will remain un-irrigated for
about 60-70 days (Carr, 2013) to promote bud development, because continued rain or

46
irrigation would delay imposition of the drought (Astegiano et al., 1988). After optimum
impositions of soil moisture stress, the dormancy should be broken out by showers of rain
or irrigations, which is approximately around end of December, any delay rainfall during
this period, would results in prolonged dormancy period. If severely high air temperatures
are coupled with moisture stress, it lengthens the dormancy of some crops including coffee
(Allen et al., 1998). Much supply of water also resulted in delaying of coffee flowering.
Providing irrigations during dormancy periods will leads to physiological imbalance in the
flowering of coffee (Titus and Pereira, 2005). Similar recommendation was forwarded by
FAO (2018) coffee needs to be water-stressed for about four to eight weeks before
flowering to give a strong uniform flowering.

The winter dry period may extends from October to April/May in study area, and during
this period, coffee plants are subject to higher water deficit. High amount of leaves
dropped from coffee trees as physiological response. Winter irrigations (End of December
to 1st week of February) used to shorten dormancy periods, similar suggestions were
forwarded by Nkomozepi and Chung (2012), winter irrigations during December helps to
reduce drought period. Winter irrigation in study area during February and March will
helps to supplement showers rain fall deficit to obtain a complete blossom. Similar studies
conducted by Titus and Pereira (2005) indicated that winter irrigations helps to overcome
the inadequate natural blossom showers in the month of February and March essential for
flowering.

Tesfaye et al. (2013) conducted research at JARC; they clearly demonstrated that the
need of supplemental irrigation when ETc greater than seasonal rainfalls for coffee
during dry spells in study area, accordingly, supplemental full irrigation significantly
increased coffee yield, Supplemental full irrigation had 24 % yield advantage over rain
fed treatment. Eriyagama et al. (2014) mapped global coffee irrigation requirement, they
indicated positive coffee IR, which is varied from 0-200 and >200 mm year-1 for south
southwester Ethiopia (Jimma zone) and central Ethiopia, respectively. Peasley and Rolfe
-1
(2003) who reported that 484 mm year of coffee IR, at Newrybar site (Australia) for
low stressed coffee plants per annum. The values are almost consistent with my study.

47
Table 11. Statistics analysis of monthly and annual coffee IR for Jimma zone

Stations IR Model Predicted IR

Mean SD CV Max Min SW Min Mean Max
MONTH (mm) (mm) (%) mm mm P-value mm mm mm
Jan 76.19a 10.08 13.23 92.1 57.2 0.53 57.2 74.8 86.9
Feb 77.70a 13.00 16.73 96.1 47.6 0.35 47.6 76.2 89.7
Mar 72.36ab 24.77 34.23 112.2 28.7 0.24 32.4 69.1 98.6
Apr 53.44b 29.67 55.52 112.5 4.5 0.63 4.5 49.7 94.2
May 13.48c 22.63 167.94 79.4 0.0 0.00 0.0 10.5 54.1
June 1.70c 4.55 268.62 14.9 0.0 0.00 0.0 1.1 13.0
Jul 0.00c 0.00 0.00 0.0 0.0 NA 0.0 0.0 0.0
Aug 0.00c 0.00 0.00 0.0 0.0 NA 0.0 0.0 0.0
Sep 5.93c 12.09 204.04 41.1 0.0 0.00 0.0 4.9 31.8
Oct 52.49b 32.92 62.71 116.7 7.5 0.10 16.6 50.5 99.0
Nov 60.98ab 15.67 25.70 90 38.3 0.25 41.4 67.0 82.1
Dec 70.18ab 13.44 19.16 90.4 43.7 0.33 43.7 75.3 83.5
Annual 484.45 150.48 31.06 842.9 300.6 0.07 300.7 466.2 723.3
CV 88.1
LSD@5% 21.1
SW= Shapiro-Wilk

Table 12. The Pearson correlation between Rainfall and geographic parameter

Month Longitude latitude Altitude

Jan -0.62** -0.78** -0.08
Feb -0.33 -0.83** -0.13
Mar -0.46* -0.73** 0.04
Apr -0.64** -0.69** 0.24
May -0.79** -0.16 0.53*
Jun -0.47* 0.40 0.67**
Jul -0.19 0.48* 0.68**
Aug -0.29 0.46* 0.73**
Sep -0.61** 0.16 0.53*
Oct -0.83** -0.22 0.23
Nov -0.77** -0.59** 0.15
Dec -0.63** -0.76** -0.03
Annual -0.70** -0.04 0.58**
*, ** and *** = significant at 0.05, 0.01 and < 0.0001 level, respectively, WS= wind speed, RH=relative humidity

48
Figure 17. Monthly coffee ETc, IR and effective rain fall for Jimma zone

49
Figure 18. IR kriging maps of Monthly and annual coffee IR for Jimma zone

50
5. SUMMARY, CONCLUSION AND RECOMMENDATIONS

5.1 Summary and Conclusions

The coffee crop water requirements, as expressed by the crop evapotranspiration, vary
considerably spatially from location to location in study area. In this study, an
attempt was made to analyze spatial and temporal dynamics of coffee ETc in Jimma zone
of Oromia region. The aim was to estimate and to map coffee ETc and IR.

From the results maximum ETo was observed around the extreme southern tip and
extreme eastern tip region. Whereas smaller values were founds around western and
southern part. The temporal variation of ETo was significant, the highest ETo value was
recorded in March (4.61mm day-1) and the smallest in July (2.91 mm day-1).

The seasonal coffee Kc varied from 0.69 to 0.99 for Kc-NDVI and climate adjusted,
respectively. The lowest Kc values during initial stage whereas highest during midseason
stage. From monthly and seasonal maps of coffee ETc, the high annual coffee ETc were
observed in the extreme eastern, southern tip of Jimma zone, the values ranged from
1129.4 to 1350.1 mm year-1. In contrary the western and southern parts indicated by
small ETc requirement. The monthly coffee ETc has exhibited similar spatial distribution
with that of annual ETc. The monthly mean coffee ETc ranged from 80.9 (November) to
131.11 mm (April).

The maps of monthly and annual coffee IR exhibited that there were greater spatial
variability distributions in study areas, because the mean CV for all months and annual
IR varied from 31.1 to 86.8%. The higher coffee IR was found to be over the
northeastern, southeastern and central regions and some tip areas of the southwestern
area. Whereas the western and southern half region of study areas have been indicated by
smaller IR. Temporally, the high IR were occurred from January (76.19 mm) to April
(53.44), with the highest in the month of February (77.70 mm). In contrast, the coffee
plants did not need irrigation from June to September; thus the water requirements were
completely satisfied by rainfall. Naturally, coffee plants need dormancy period of about
28-70 days from November to end of December. Otherwise, a plant continues to grow in
vegetative as irrigation is continuous during dormancy period instead of flowering.

51
Therefore, even though the requirement is there, the coffee plants will not be irrigated
during the dormancy period.

Based on the results the following conclusion was drawn:

1. The high ETo was observed around areas of extreme southeastern tip and western
tip, while lower over areas of Western and half of the southern area.
2. Highest monthly ETo value was found during dry season of April
(4.61mm), while the lowest during wet rainy season of July (2.91 mm day-1).
3. Maximum monthly and annual coffee ETc was indicated in the regions of extreme
western, extreme southern tip and north western regions. In contrast, minimum
requirement exhibited around western and half of southern regions as well as
some pocket areas of Natri and Botorbacho areas. Temporally, minimum coffee
ETc recorded from November to February was due to critical soil moisture stress.
While greatest ETc value was seen from April to June, during midseason stage.
4. Highest irrigations requirement was observed during the month of February
(77.70 mm), while slightly nil from June to September.

5.2 Recommendations

In this study, Coffee ETc and IR were analyzed and mapped with the input of climate and
crop data. The Kc values were adjusted to local condition by water balance, satellite image
and long term climatic data at JARC station. Therefore, Further studies on predicting
coffee ETc at different locality using models and supplemented with replicated field
experiments (Bowen-ratio and lysimeter ) are essential to accurately estimate and map
coffee ETc in study area.

Based on the results the following recommendations were suggested:

1. Results of this study may literally help for designing purposes for engineers and
irrigation practitioners in the absence of data.
2. The results of this study will also help both experts and producers in order to
develop more efficient irrigation strategies in study areas.
3. Based upon long-term irrigation requirement estimation, coffee plants will
not require irrigation during June to August, since existing rainfall is adequate.

52
4. Due to limited rainfall during the extended dry period, the farmers are advised to
irrigate their coffee crop during dry months.
5. Generally it is recommended to create awareness for those who are participated in
coffee production systems, even though Jimma is situated in sub-humid areas it is
characterized by dry period for about 4-6 months, during these months severe soil
moisture deficit will need to compensate with artificial water application.
6. In view of the explicit distinction made by the geo-spatiality of the coffee crop
water requirement in Jimma zone, a correct understanding of the ETc value for
each locality will be most useful to coffee growers, where irrigation
management is concerned

53
 6. REFERENCE

Addinsoft, S.A.R..L. 2017. XLSTAT Premium version 19.5: A complete statistical add‐in
for Microsoft Excel. Addinsoft SRL, Paris, France.

Adem Anwar, Aynalem Dessalew, Tilahun Seifu and T.S. Steenhuis. 2017. Predicting
reference evaporation for the Ethiopian highlands. J Water Reso and Protection 9:
1244-126.

Agnew, M.D. and J.P. Palutikof. 2000. GIS-based construction of baseline climatologies
for the Mediterranean using terrain variables. Climate Reso 14: 115-127.

Alexandersson, H. 1986. A homogeneity test applied to precipitation data. J Clim 6: 661-

675.

Ali, M.H. 2010. Fundamentals of irrigation and on-farm water management. Springer
science and business media. Bangladesh, 511 pp.

Allen, R.G., L.S. Pereira, D. Raes and M.Smith. 1998. Crop evapotranspiration. Guidelines
for computing crop water requirements. FAO Irrig. Drain. Paper No. 56. FAO.
Rome, Italy.

Almhab, A.A. 2011. Assessment of Crop Water Requirement Using Landsat ETM and
MODIS Remote Sensing Data in Arid Area. Inter-Islamic Network on Space
Sciences and Technology, Royal Jordanian Geographic Centre, Jordan, 60 p.

Amatya, D.M., S. Irmak , P. Gowda, G. Sun, J.E. Nettles, K.R. Douglas-Mankin. 2016.
Ecosystem evapotranspiration: challenges in measurements, estimates, and
modeling. Trans. ASABE 59: 555–560.
ASCE (American Society of Civil Engineers). 2005. The ASCE Standardized Reference
Evapotranspiration Equation. American Society of Civil Engineers, Reston, VA,
USA.
Astegiano, E.D., M. Maestri and M.D.M. Estevao. 1988. Water stress and dormancy
release in flower buds of Coffee Arabica: water movement into the buds. Journal of
Horticultural Science 63(3): 529-533.

Blaney, H.F. and W.D. Criddle. 1950. Determining water requirements in irrigated areas
from climatological and irrigation data. Soil Conservation Service, Washington,
D.C, USA 96: 48 p.

54
Blore, T.W.D. 1966. Further studies of water use by irrigated and unirrigated Arabica
coffee in Kenya. Journal of Agricultural Science 67(2): 145-154.

Buishand, T.A. 1982. Some methods for testing the homogeneity of rainfall records.
Journal of hydrology 58(1-2):11-27.

Burrough, P.A. and R.A. McDonnell. 1998. Creating continuous surfaces from point
data. Principles of Geographic Information Systems. Oxford University Press,
Oxford, UK.

Cambardella, C. A., T. B. Moorman, J. M. Novak, T. B. Parkin, D. L. Karlen, R. F. Turco

and A. E. Konopka. 1994. Field scale variability of soil properties in central Iowa
soils. Soil science society of America journal 58(5): 1501-1511.

Cannell, M.G.R., 1985. Physiology of the coffee crop. . pp. 108-134. In: Clifford, M.N.
and K.C. Willson. (ed.). Coffee. Springer, Boston, USA.

Carr, M.K.V. 2013. The Water Relations and Irrigation Requirements of Macadamia
(Macadamia Spp.): A Review. Journal of Experimental Agriculture 49(1): 74-90.

Carr, M.K.V., 2001. The water relations and irrigation requirement of coffee. Expe
Agri, 37(1): 1-36.

Chauhan, R., M.S. Hooda and A.A. Tanga. 2015. Coffee the Backbone of Ethiopian
Economy. International Journal Economic Plants 1(2): 82-86.

CRO (Coffee Research Organization). 2006. Arabica and Robusta coffee, best climate for
growing coffee beans. [http://www.coffeeresearch.org/agriculture/climate.htm.]
Site visited on 15/10/2017.

CSA (Central Statistic Agency). 2013. Population Projection of Ethiopia for All Regions

At Wereda Level from 2014 – 2017. Addis Ababa, Ethiopia. 118 p

CSA (Central Statistic Agency). 2016. The Federal Democratic Republic of Ethiopia
Central Statistical Agency report on area and production of major crops. Addis
Ababa, Ethiopia. 178 p.

DaMatta, F.M., C.P. Ronchi, M. Maestri and R.S. Barros. 2007. Ecophysiology of
coffee growth and production. Brazilian Journal of Plant Physiology 19: 485–510.

55
De Bruin, H.A.R.., I.F. Trigo, M.A.Jitan, Temesgen Enku, C. Tol and A.S.M. Gieske.
2010. Reference crop evapotranspiration derived from geo-stationary satellite
imagery: a case study for the Fogera flood plain, NW-Ethiopia and the Jordan
Valley, Jordan. Hydrology and earth system sciences 14(11): 2219-2228.

Dirk, R.A.E.S. 2013. Frequency analysis of rainfall data. In: Soil Physics 30th Anniversary
(1983-2013), 25 February - 1 March, 2013, Pp. 244-254,The Abdus Salam
International Centre for Theoretical Physics, Trieste, Italy.

Doorenbos, J, and W.O. Pruitt, 1977. Crop water requirements. FAO Irrigation and
Drainage Paper 24, Rome. Italy.

ECFF (Environment and Coffee Forest Forum). 2017. Coffee farming and climate change
in Ethiopia. Impacts, forecasts, resilience and opportunities summary report.
Environment Management 24 (1): 39-54.

Eriyagama ,N., Y. Chemin, and R. Alankara. 2014. A methodology for quantifying global
consumptive water use of coffee for sustainable production under conditions of
climate change. Journal Water and Climate Change 5 (2): 128-150.

Everitt, B.S., S. Landau and M. Leese. 2001. Cluster analysis (4th edition). Arnold, A
member of the Hodder Headline Group, London. United Kingdom.

Fabio, R.M., R.A. Luiz, Z.R. Evandro and C.S. Paulo. 2005. Evapotranspiration and
irrigation requirements of a coffee plantation in southern Brazil. Experimental
Agriculture 41: 187–197.
Fanjul, L., R. Arreola-Rodr1guez, and M.P. Mndez-Castejn. 1985. Stomatal responses to
environmental variables in shade and sun grown coffee plants in Mexico. Journal
of Experimental Agriculture 21:249–258.

FAO (Food Agriculture and Organization). 2018. Coffee tree pruning and management.
[http://www.fao.org/docrep/008/ae939e/ae939e07.htm] site assessed on June, 8,
2018.

FAO (Food and Agricultural Organization). 2009. CROPWAT 8.0 for windows user guide.
Rome. Italy.

FAO (Food and Agricultural Organization). 2012. Crop yield response to water. FAO
Irrigation and Drainage Paper No. 33, Rome. Italy.

56
Fernandes, A.L.T., P.V. Rabelo, L.C.D. Drumond, and R.O. Silva. 2009. Evaluation of
productivity and quality of coffee growing in Cerrado (Brazilian savannah)
conditions and irrigated using the drip system. : pp. 1210-1214. 22nd International
Conference on Coffee Science, September 14-19, 2008. ASIC, Campinas, SP,
Brazil.

Fiedler, F.R. 2003. Simple, practical method for determining station weights using
Thiessen polygons and isohyetal maps. Journal of Hydrological Engineering 8(4):
219- 221.

Fitsum Bekele. 2017. Analysis of Temporal Variability and Trend of Reference

Evapotranspiration in Sinana District, South Eastern Ethiopia. Imperial Journal of
Interdisciplinary Research 3 (3): 1747-1756.

Fitsume Yemenu, Kidist Abera, Michael Eshetu, Richard Koech and Molla Mekonnen
Alemu. 2017. Irrigation water planning for crops in the central highlands of
Ethiopia, aided by FAO CROPWAT Model. African Journal of Agricultural
Research 12(28): 2329-2335.

Fonteh, M.F. and T. Podmore. 1994. Application of geostatistics to characterize spatial

variability of infiltration in furrow irrigation. Agricultural water management
25(2): 153-165.

Girma Nigussie. 2011. Marketing information operation in Ethiopia with special reference
to the Ethiopia Commodity Exchange Coffee Trading. MSc. Thesis. Swedish
University of Agricultural sciences Uppsala, Sweden. Department of Urban and
Rural development, 75 p.

Graler,B., E. Pebesma and G. Heuvelink, 2016. Spatio-Temporal interpolation using

gstat. The RFID Journal 8(1): 204-218.

Gutierrez, M.V. and F.C.Meinzer. 1994. Estimating water use and irrigation requirements
of coffee in Hawaii. Journal of American Soc for Horti Sci 119(3): 652-657.

Hillel, D. 1998. Environmental soil physics. Academic press. Elsevier, USA. 771 pp.
Jocik, A.A. 2004. Estimate ambient air temperature at regional level using remote sensing
techniques. Msc Thesis. Department of sustainable agriculture. International
institute for Geo-information science and Earth observation. Nederland. 99p.

57
Johnson, G. 2007. Effective rooting depth for row crops. [http://kentag extension
.blogspot.com/2007/12/effective-rooting-depth-for-row-crops. html] site visited on
22/10/2017.

JZIA (Jimma Zone Irrigation Authority). 2017. Annual report of Jimma Zone irrigation
development, Jimma Zone. Un-published report.

Kamble, B., A. Kilic and K. Hubbard. 2013. Estimating Crop Coefficients Using Remote
Sensing-Based Vegetation Index. Remote Sensing 5(4): 1588–1602.

Kashyap, P.S. and R.K. Panda. 2001. Evaluation of evapotranspiration estimation methods
and development of Kc for potato crop in a sub-humid region. Agricultural Water
Management 50: 9-25.

Kiptala, J.K., Y. Mohamed, M.L. Mul and P. Zaag. 2013. Mapping evapotranspiration
trends using MODIS and SEBAL model in a data scarce and heterogeneous
landscape in Eastern Africa. Water Resources Research 49: 8495–8510.

Knotters, M., D.J. Brus and J.H. Voshaar. 1995. A comparison of kriging, cokriging and
kriging combined with regression for spatial interpolation of horizon depth with
censored observations. Geoderma 67: 227-246.

Kozanis, S., A. Christofides, N. Mamassis, A. Efstratiadis and D. Koutsoyiannis. 2010.

Hydrognomon open source software for the analysis of hydrological data. In:
European Geosciences Union General Assembly, 2-5 May, 2010, Geophysical
research abstracts (Vol. 12), Vienna, Austria, 10p.

Kuit, M., D.M. Jansen and V.T. Nguyen. 2004. Manual for Arabica cultivation. Tan Lam
Agricultural Product Joint Stock Company, Khe Sanh, Vetenam, 213 pp.

Lemos Filho, L.C.D.A., C.R.D. Mello, M.A.D. Fariaand and L.G.D. Carvalho. 2010.
Spatial-temporal analysis of water requirements of coffee crop in Minas Gerais
State, Brazil. Brazilian Journal Agricultural and Environmental Engineering 14(2):
165-172.

Li, J. and A.D. Heap. 2008. A Review of Spatial Interpolation Methods for Environmental
Scientists. Geoscience Australia, Australia, 137 pp.

58
Lu, J. B., G. Sun, S. G. McNulty and D. M. Amataya. 2005: A comparison of six potential
evapotranspiration methods for regional use in the southeastern United States.
Journal of American Water Resource Association 41: 621-633.

Luke, G. 2006. Irrigation scheduling – how and why. [http://www.agric.wa.gov.au

/content/LWE/WATER/IRR/F02390.HTM] site visited on 17/10/2017.

Mahtsente Tibebe. 2015. Water Demand Analysis and Irrigation Requirement for Major
Crops at Holetta Catchment, Awash Sub-basin, Ethiopia. Journal of Natural
Sciences Research 5(15): 117-127.

Mardikis ,MG, DP. Kalivas and VJ. Kollias. 2005. Comparison of interpolation methods
for the prediction of reference evapotranspiration an application in Greece. Water
Resource Management 19(3): 251–278.

Marin, F.R., Angelocci, L.R., Nassif, D.S., Costa, L.G., Vianna, M.S. and Carvalho, K.S.,
2016. Crop coefficient changes with reference evapotranspiration for highly
canopy-atmosphere coupled crops. Agricultural Water Management 163: 139-145.

Marin, F.R., L.R. Angelocci, E.Z. Righi and P.C. Sentelhas. 2005. Evapotranspiration and
irrigation requirements of a coffee plantation in southern Brazil. Journal of
Experimental Agriculture 41(2): 187-197.

Martinez, A. 1996. Multivariate Geostatistical Analysis of Evapotranspiration and

Precipitation in Mountainous Terrain. Journal of Hydrology, 174:19–335.

Miller, R.W. and R.L. Donahue .1990. Soils, an introduction to soils and plant growth,
Sixth Edition. Prentice-Hall Inc., Englewood Cliffs. New Jersey. USA. 60p.

Mitchell, H.W. 1988. Cultivation and harvesting of Arabica coffee. pp. 43- 90. In: Clarke,
R.J. and R. R. Macnae (eds.). Coffee Agronomy. Elsevier applied science. London
and New York.

Molin, P., H. and Abdi. 1998. New Tables and numerical approximation for the
Kolmogorov Smirnov/Lillierfors/Van Soest test of normality. University of
Bourgogne, Digon, France, 12 pp.

Morison, J.I, N.R. Baker, P.M. Mullineaux and W.J. Davies. 2008. Improving water use in
crop production. Philosophical Transactions of the Royal Society B: Biological
Sciences 363(1491): 639-658.

59
Naan, D., Drip Irrigation for Coffee Plantations: feasible and profitable.
[www.naandanjain.com/uploads/catalogerfiles/coffee-2/Coffee-dripirrigation.pdf]
site visited on 05/12/2018.
Naheed, G. and G. Rasul. 2010. Projections of crop water requirement in Pakistan under
global warming. Pakistan Journal of Meteorology 7(13): 45-51.

Naik, B.R., T.H. Latha and C.M. Babu. 2015. Command area development by using FAO
Cropwat 8.0 model and impact of climate change on crop water requirement-a case
study on Araniar reservoir basin (Pichatur dam). IJAR 1(13: 142-155.

Neugebauer, N. 2013. Estimation of Crop Water Requirements based on Remote Sensing

Data, a Case Study of the Marchfeld Region in 2010. MSc Thesis. Institute of
surveying, remote sensing and land information, University of Natural Resources
and Applied Life Sciences, Vienna, Austria. 78 P.

Nidhi, J.S. and H.M. Petel. 2012. Estimation of water requirement in mahi right bank
canal command area. India.
Nkomozepi1, T. and S. Chung. 2012. Assessing the trends and uncertainty of maize
net irrigation water requirement estimated from climate change projections for
Zimbabwe. Agri. Water Manage 111: 60–67.

Nuha, H. and F.M. Henery. 2000. Rain water harvesting for natural resource
management: a planning guide for Tanzania. RELMA technical paper No. 22.
Sida’s Regional Land Management Unit, Nairobi, Kenya, 156 pp.

Oliveira, P.M., A.M. de Silva, and N.P. Castro (2003). Estimate of coffee crop Coffee
Arabica Coefficient and evapotranspiration. [http://www.cab
abstractsplis.org/google/abstract.asp?AcNo=2004304951] site visited on
10/10/2017.

Patel, A., R. Sharda, S. Patel and P. Meena.2017. Reference evapotranspiration estimation

using cropwat model at ludhiana district (punjab). International Journal of Science,
Environment 6(1): 620 – 629.

Peasley, D. and C. Rolfe. 2003. Developing irrigation strategies for coffee under sub-
tropical conditions. RIRDC, Barton, Australia. P.11- 52.

Pebesma, E. and R.S. Bivand. 2005. Classes and Methods for Spatial Data: the sp
Package. R news 5(2): 9-13.

60
Pereira, A.R., M.B.P.D. Camargo, V. Nova and N. Augusto. 2011. Coffee crop coefficient
for precision irrigation based on leaf area index. Bragantia 70(4): 946-951.

Pereira, H.C., 1957. Field measurements of water use for irrigation control in Kenya
coffee. Journal of Agricultural Science 49(4): 459-466.

Pereira, L.S., A. Perrier, F.G. Allen and I. Alves.1999. Evapotranspiration: Concepts and
future trends. Journal of Irrigation and Drainage Engineering 125: 45-51.

Raes, D., A. Timmerman, M. Hermy, and J. Mentens. 2006. Green roof water balance
model. K.U.Leuven University, Faculty of Bioscience Engineering, Division of
Soil and Water Management, Leuven, Belgium.

Rasul, G. and A. B. Farooqi. 1993: Water Requirement of Cotton Crop in Pakistan.

Journal of Engineering and Applied science 4 (2): 154-165.

Reas, D., P. Willems, F. Gabaguidi. 2006. Rainbow software package for analyzing of
climatological and hydrological data frequency analysis- test of Homogeneity. K.U.
Leuven University. Lueven. Belgium.

Robel Getachew. 2012. The effect of El Niño-Southern oscillation ( ENSO ) on the

Ethiopian Rainfall Pattern. PhD Dissertation. Department of Physics, school of
graduate studies (education). Addis Ababa University, Ethiopia. 73p.

Rouse, J.W., R.H. Haas, J.A. Schell, D.W. Deering. 1974. Monitoring vegetation systems
in the Great Plains with ERTS. pp. 309–317. In: Fraden S.C., Marcanti E.P., Becker
M.A. (eds.), Third ERTS-1 Symposium, 10–14 Dec. 1973, NASA SP-351,
Washington D.C, USA.

Royston, J.P. 1982. An extension of Shapiro and Wilk's W test for normality to large
samples. Applied Statistics. pp.115-124.

Sahoo, P.M. 2012. Data models, basics of GIS and GPS. Indian agricultural statistics
research institute. New Delhi, India.

Savva, A.P. and K. Frenken. 2002. Crop water requirements and irrigation scheduling.
FAO Sub-Regional Office for East and Southern Africa, Harare. Zimbabwe. 132p.

Sax, D. 2000. The shortage of water resources in 21 century. Express Water Resource and
Hydropower Information 21(8): 1-5.

61
SCS (Soil Conservation Service). 1993. Irrigation Water Requirements chapter 2. In:
Derrel, L.M., R.G. James and W. B. Otto (eds) Part 623 National Engineering
Handbook. Soil Conservation Service. Washington, DC,USA. pp. 1-302.
Sijali, I.V., 2001. Drip irrigation: Options for smallholder farmers in Eastern and Southern
Africa (No. 24). Regional land management unit. Nairobi, Kenya, 70 pp.

Silva, A.L.D., I.P. Bruno, K. Reichardt, O.O. Bacchi, D. Dourado-Neto, J.L. Favarin, F.M.
da Costa and L.C. Timm. 2009. Soil water extraction by roots and Kc for the coffee
crop. Brazilian Journal of Agricultural and Environmental Engg 13(3): 257-261.

Silva, A.L.D., R. Roveratti, K. Reichardt, O.O.S. Bacchi, L.C. Timm, I.P. Bruno, J.C.M.
Oliveira and D.N. Dourado. 2006. Variability of water balance components in a
coffee crop in Brazil. Journal of Agricultural science 63(2): 105-114.

Silva, C. A ., R.E. F. Teodoro, B. Melo, 2008. Productivity and yield of coffee plants
submitted to irrigation slides. Brazilian Agricultural Research 43: 387-394.
Singo, L.R., P.M. Kundu and F.I. Mathivha. 2016. Spatial variation of reference
evapotranspiration and its influence on the hydrology of Luvuvhu River
Catchment. Journal of Agriculture and Environmental Management 5(6) 187-196.

Smith, K. 2014. Estimation Methods of Reference Evapotranspiration at Un-sampled

Locations. MSc Thesis. Department of Mathematics, New Mexico Institute of
Mining and Technology. Mexico. 78 p.

Smith, M.J., M.F. Goodchild and P.A. Longley. 2007. Geospatial analysis.3rd ed.
Winchelsea Press. Beijing, China, 55 pp.

Steduto, P., Hsiao, T.C., Raes, D., Fereres,E., 2009. AquaCrop, the FAO crop model to
simulate yield response to water: I. Concepts and underlying principles. Agro.
Journal 101(3): 426-437.

Surendran, U., C.M. Sushanth, G. Mammen and E.J. Joseph. 2015. Modeling the crop
water requirement using FAO-CROPWAT and assessment of water resources for
sustainable water resource management: A case study in Palakkad district of humid
tropical Kerala, India. Aquatic Procedia 4: 1211-1219.

Tadesse Banti. 2017. Progress in Arabica Coffee Breeding in Ethiopia: Achievements,

Challenges and Prospects. International Journal of Sciences: Basic and Applied
Research 33(2): 15-25.

62
Team, R.C. 2013. R: A language and environment for statistical computing. R Foundation
for Statistical Computing, Vienna, Austria.

Tefera Abu and Tefera Teddy, 2013. Ethiopia: Coffee Annual Report. Gain report ET-
1302. Foreign agricultural service, Washington, D.C, United states, 52 pp.

Tesfaye Shimbir Gesesse, M.R. Ismail, H. Kausar, M. Marziah and M.F. Ramlan. 2013.
Plant water relations, crop yield and quality of Arabica coffee as affected by
Supplemental deficit Irrigation. International Journal of Agriculture and Biology
15(4): 665-672.

Tesfaye Shimbir, I.M. Razi and M. Maziah. 2008. Effects of deficit irrigation and partial
root zone drying on growth, dry matter partitioning and water use efficiency in
young coffee plants. Journal of Food, Agriculture and Environment 6: 312-317.

Titus, A. and G.N. Pereira. 2005. The Fine art of irrigation in Robusta coffee plantations.
[https://ecofriendlycoffee.org/the-fine-art-of-irrigation-in-robusta-coffee-
plantations] site visited on June 25, 2018.

Todorovic, M. 2005. Crop water requirements. Water Encyclopedia 3: 557-558.

Trivedi, A., S.K. Pyasi and R.V. Galkate. 2017. Estimation of Evapotranspiration using
CROPWAT 8.0 Model for Shipra River Basin in Madhya Pradesh, India.
International Journal of Current Microbiology and Applied Sciences 7(5): 1248-
1259.

Um, M.J., Y. Kim and D. Park. 2017. Spatial and Temporal Variations in Reference Crop
Evapotranspiration in a Mountainous Island, Jeju, in South Korea. Water 9(4): 261.

Van Buuren, S. and K. Groothuis-Oudshoorn . 2011. MICE: Multivariate Imputation by

Chained Equations in R. Journal of Stat Software 45(3): 1-67.

Wallis, J. A. N. 1963. Water uses by irrigated Arabica coffee in Kenya. Journal of

Agricultural Science, Cambridge 60: 381-388.

Ward J.H. 1963. Hierarchical groupings to optimize an objective function. Journal of the
American Statistical Association 58: 238-244.

Wintgens, J.N. 2009. Coffee: growing, processing, sustainable production. Guides book
for growers, processors, traders and researchers. Wiley-Vch.Wormer Publisher.
weinheim. Germany 1000p.
63
Wubengeda Admasu, Kassu Tadesse, Tilahun Hordofa and Dawit Habte. 2014.Sensitivity
analysis of climatic parameters on Reference Evapotranspiration (ETo) calculated
by FAO Penman - Monteith Equation (CROPWAT MODEL 8.0) for Sub-humid to
humid areas of Arsi Zone, Ethiopia. International Journal of Agricultural Science
Research 3(4): 039-051.

Yang, G.G., J. Zhang and Y.Z. Yang. 2011. Comparison interpolation methods for typical
meteorological factors based on GIS- a case study in Jitai basin, China. pp. 1-5. In:
Wei, G. and B. Shuming (eds). Proceeding of the 19th international conference on
geoinformatics. June 24-26, 2011, cpgis.org, Shanghai, China.

Yilma Alemsegad, Y. Edjamo and Tesfaye Shimbir. 1997. Studies on Arabica coffee fruit
phenology and bean size at Jimma research center. Jimma, Ethiopia.

Yonas Seifu. 2002. A historical survey of Jimma town (1936-1974). MSc Thesis.
Department History, Addis Ababa University, Addis Ababa, Ethiopia. 166p

64
APPENDICES

Appendix Table 1. Monthly Tmax data of Jimma stations prepared for MICE imputation

year Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
1985 29.7 29.3 30.3 27.9 26.3 25.4 23.4 23.9 25.3 26.5 27.5 28.2
1986 29.5 29.3 28.9 27.6 27.8 24.8 24.6 25.4 25.3 26.8 28.8 28.3
1987 28.8 29.6 27.8 28.5 26.9 26.3 25.9 25.6 26.9 27.2 28.2 27.8
1988 28.1 28.2 30.3 28.9 28.1 25.8 23.3 24.3 24.7 26.0 27.1 27.8
1989 28.0 27.6 27.9 26.4 27.5 25.5 24.2 25.2 25.1 26.3 26.7 26.1
1990 26.9 26.4 27.6 27.7 27.3 25.5 24.4 24.8 25.3 27.0 27.5 27.9
1991 28.0 28.2 28.3 28.6 27.9 25.9 23.8 24.1 25.6 26.8 27.9 26.6
1992 27.4 27.8 29.6 28.9 28.2 26.8 24.5 24.1 26.1 26.3 27.0 27.9
1993 27.4 27.4 29.7 27.5 27.2 26.1 24.8 25.1 26.3 27.1 28.2 29.3
1994 30.2 30.9 29.9 28.8 27.0 25.3 23.4 24.0 25.5 28.0 28.0 29.1
1995 30.0 30.5 31.1 28.8 28.2 27.4 24.8 25.5 26.6 28.0 28.0 27.8
1996 27.0 29.7 28.9 27.8 27.3 25.2 24.5 24.9 26.1 27.5 27.8 27.3
1997 27.8 30.2 30.9 27.6 27.4 26.6 25.3 25.7 27.5 27.2 27.1 27.8
1998 28.1 29.2 29.6 31.0 28.8 27.6 24.9 25.5 26.4 27.1 28.1 29.2
1999 29.9 32.6 30.8 30.7 28.2 27.2 24.8 25.4 27.5 26.6 28.1 29.4
2000 30.8 32.5 33.2 29.7 28.4 26.8 25.7 25.3 27.2 27.1 28.1 28.7
2001 28.9 31.3 29.7 30.0 28.8 26.6 25.3 25.4 27.9 28.1 28.5 29.7
2002 28.6 31.2 29.9 30.6 29.2 26.1 25.8 25.4 26.8 27.7 28.2 27.6
2003 27.9 30.6 29.9 29.2 30.7 27.1 24.6 25.1 26.5 27.9 28.6 28.5
2004 29.7 30.0 30.3 29.1 29.1 25.9 25.0 25.3 26.3 27.0 28.0 27.8
2005 28.3 31.6 30.5 29.3 27.6 26.4 25.1 25.9 26.1 27.3 27.4 29.3
2006 30.2 31.2 29.9 28.7 28.3 27.2 25.5 24.6 25.9 27.6 27.7 27.5
2007 27.9 28.7 30.5 28.9 27.9 26.6 25.5 25.1 26.5 28.3 29.8 30.6
2008 31.3 31.3 32.5 30.1 28.3 27.1 24.8 25.6 25.8 26.8 26.6 28.0
2009 28.4 29.7 30.4 29.1 28.9 27.4 24.9 24.9 26.4 26.8 28.5 26.7
2010 28.3 28.5 29.0 29.3 27.6 26.7 25.0 24.9 25.9 27.8 28.3 27.2
2011 29.4 31.2 30.0 30.3 28.3 26.4 26.0 24.8 26.4 28.0 27.5 28.2
2012 30.3 31.6 31.5 29.5 29.0 26.3 24.9 24.9 25.8 27.8 27.9 28.1
2013 29.3 30.9 30.2 30.9 27.5 26.4 24.4 24.8 26.6 26.9 27.5 28.1
2014 28.7 30.2 30.4 NA NA 27.5 25.7 25.4 25.9 NA NA 28.0
2015 29.8 31.6 31.7 29.9 NA NA NA 26.6 27.6 NA 27.7 NA
2016 29.5 NA NA 28.5 NA NA NA 25.8 NA NA NA NA
NA is missed data

65
Appendix Table 2. Descriptive statistics of climate dataset for different stations

Tmin (oC) Tmax (oC) Rain fall (mm) RH (%) WS (m/s) SH (hrs)

Stations mean Max Min Mean Max Min Sum Max Min mean Max Min mean Max Min Mean max min
Agaro 12.16 13.04 10.79 28.69 30.78 26.12 1621.0 244.9 29.7
Arjo 10.88 11.87 10.06 20.30 23.03 17.11 2364.2 417.6 20.0 73.08 89.16 55.09 1.53 1.89 1.12 6.48 8.41 2.94
Asendabo 12.49 14.24 9.42 27.15 29.56 23.79 1198.1 195.5 21.1
Atnago 13.96 14.94 13.09 27.81 30.51 24.89 1664.6 310.7 13.9
Bedelle 12.81 14.24 11.58 25.55 28.45 22.23 1886.0 346.9 14.5 73.07 86.97 59.54 1.08 1.29 0.94 6.98 8.74 3.47
Bonga 11.99 13.10 10.15 27.26 28.67 25.62 1603.4 200.5 42.1
Botorbacho 2.50 3.33 0.14 24.61 26.91 21.48 1487.1 299.8 15.8
Chida 15.04 15.79 14.45 26.73 29.41 22.99 1535.2 201.8 39.3
Chira 11.70 12.23 10.99 23.80 25.64 21.34 1890.7 262.3 50.8 65.04 81.35 48.84 1.00 1.09 0.89 6.26 8.09 3.27
Dedo 12.58 12.84 12.09 23.14 24.94 20.83 2047.8 333.7 37.4
Ejajji 13.91 15.75 11.41 28.06 31.46 23.71 1367.4 274.2 11.0 69.72 84.61 55.02
Gatira 11.53 12.15 11.00 22.15 24.44 19.74 2048.4 341.1 36.9 70.08 82.84 60.17 1.51 1.79 1.32 5.93 7.58 3.07
Gojeb 14.88 16.60 12.45 30.66 33.21 28.20 1668.6 257.8 40.2
Jimma 11.69 13.86 7.92 26.66 30.04 24.83 1541.2 224.6 34.2 73.53 83.00 64.39 0.81 0.97 0.71 6.45 8.05 3.73
Limu-genet 13.57 14.93 11.83 27.14 30.00 23.99 1831.31 318.3 23.1 66.39 82.17 49.05 0.96 1.15 0.84 7.10 8.99 3.88
Natri 12.74 14.43 10.98 26.81 29.81 23.53 1583.17 265.7 21.3
Shebe 13.30 13.90 12.74 25.74 28.34 23.03 1600.41 235.1 40.4
Sokoru 13.34 14.46 12.16 26.13 28.59 23.06 1350.53 233.6 24.9 65.04 81.35 48.84 0.93 1.01 0.86 7.10 8.70 4.06
Wolkite 12.99 13.49 12.29 28.34 30.52 25.53 1133.3 233.8 7.5
Yaya 13.42 15.50 10.05 30.20 32.73 25.77 1025.05 219.2 11.5
Yebu 12.88 13.51 12.35 25.83 27.73 23.63 2067.74 324.0 42.5

66
Appendix Table 3. Areal mean climate dataset estimated by Thiessen polygon method for Jimma zone

Variables Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Annual Min Max
Rain fall (mm) 31.17 31.52 94.28 132.52 202.23 255.26 278.86 279.96 219.09 116.57 48.29 33.19 1728.67 31.17 279.96
Tmin (oC) 11.13 11.17 12.54 12.85 12.80 12.45 12.34 12.23 12.04 11.47 10.86 10.65 11.96 10.65 12.85
Tmax (oC) 27.19 27.25 28.21 27.34 26.32 24.66 23.03 23.21 24.28 25.44 26.20 26.52 25.84 23.03 28.21
RH (%) 61.97 56.56 57.36 63.60 73.71 78.87 83.47 84.09 81.50 71.90 66.32 62.21 70.13 84.09 56.56
WS (m/s) 1.07 1.19 1.27 1.30 1.26 1.15 1.09 1.07 1.05 1.01 0.96 0.97 1.11 1.30 0.96
sunshine hour (hrs) 7.97 8.03 7.50 7.23 6.89 5.69 3.56 3.61 5.06 7.69 8.16 7.96 6.61 8.16 3.56
Tmin is minimum temperature, Tmax is maximum temperature, RH is relative humidity, WS is wind speed

67
Appendix Table 4. KS test values calculated for 80% non- exceedance probability for monthly ETo (1985 to 2016) for each station

Stations Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec annual

Agaro 0.10 0.07 0.08 0.08 0.10 0.11 0.10 0.09 0.08 0.12 0.09 0.08 0.11
Arjo 0.15 0.11 0.11 0.12 0.13 0.14 0.18 0.08 0.13 0.18 0.09 0.12 0.10
Asendabo 0.07 0.07 0.12 0.06 0.14 0.11 0.14 0.12 0.11 0.08 0.10 0.15 0.08
Atinago 0.10 0.07 0.09 0.15 0.07 0.09 0.11 0.07 0.08 0.09 0.09 0.13 0.09
Bedelle 0.11 0.17 0.15 0.10 0.10 0.13 0.14 0.14 0.11 0.10 0.10 0.13 0.15
Bonga 0.08 0.08 0.10 0.10 0.06 0.07 0.11 0.08 0.10 0.07 0.08 0.12 0.14
Botorbacho 0.11 0.16 0.12 0.13 0.08 0.18 0.17 0.12 0.13 0.14 0.13 0.14 0.14
Chida 0.11 0.12 0.12 0.10 0.12 0.16 0.06 0.14 0.15 0.18 0.09 0.10 0.12
Chira 0.07 0.12 0.10 0.11 0.13 0.15 0.14 0.07 0.07 0.12 0.07 0.15 0.07
Dedo 0.09 0.13 0.09 0.08 0.14 0.12 0.12 0.10 0.14 0.18 0.15 0.06 0.11
Ejajji 0.10 0.15 0.12 0.07 0.11 0.13 0.15 0.14 0.10 0.12 0.12 0.09 0.14
Gatira 0.13 0.14 0.08 0.11 0.16 0.08 0.21 0.16 0.09 0.11 0.08 0.09 0.08
Gojeb 0.11 0.06 0.09 0.17 0.06 0.08 0.17 0.11 0.13 0.07 0.11 0.10 0.10
Gore 0.08 0.08 0.13 0.09 0.17 0.14 0.10 0.16 0.16 0.13 0.18 0.06 0.12
Jimma 0.17 0.08 0.06 0.11 0.12 0.06 0.08 0.08 0.12 0.16 0.09 0.12 0.10
Limmu-genet 0.06 0.15 0.10 0.09 0.18 0.14 0.19 0.15 0.14 0.12 0.08 0.06 0.06
Natri 0.11 0.09 0.14 0.15 0.12 0.09 0.13 0.16 0.11 0.14 0.14 0.11 0.12
Shebe 0.06 0.06 0.06 0.11 0.06 0.11 0.09 0.09 0.12 0.08 0.06 0.10 0.13
Sokoru 0.07 0.15 0.06 0.08 0.14 0.16 0.09 0.13 0.15 0.10 0.08 0.12 0.07
Wolkite 0.08 0.10 0.09 0.10 0.07 0.15 0.08 0.16 0.09 0.11 0.09 0.16 0.06
Yaya 0.08 0.13 0.09 0.11 0.08 0.12 0.17 0.11 0.08 0.07 0.12 0.14 0.13
Yebu 0.11 0.08 0.06 0.08 0.08 0.12 0.12 0.06 0.13 0.09 0.15 0.11 0.12

68
 Appendix Table 5. Reference evapotranspiration (mm/day) values of the different stations

Months
Stations Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Agaro 3.95 4.45 4.66 4.47 4.31 3.80 3.16 3.30 3.70 4.01 3.96 3.87
Arjo 3.70 4.19 4.44 4.43 3.93 3.20 2.57 2.46 2.88 3.61 3.43 3.46
Asendabo 4.03 4.56 4.72 4.57 4.15 3.74 3.04 3.13 3.50 4.15 4.04 3.94
Atnago 3.91 4.46 4.70 4.50 4.24 3.94 3.11 3.25 3.65 3.96 3.90 3.81
Bedelle 3.89 4.39 4.66 4.72 4.15 3.35 2.79 2.85 3.44 4.02 3.83 3.77
Bonga 3.87 4.31 4.48 4.26 4.06 3.63 3.10 3.23 3.63 3.99 3.91 3.83
Botorbacho 3.75 4.17 4.38 4.30 3.95 3.51 2.97 2.98 3.28 3.82 3.71 3.61
Chida 3.90 4.40 4.40 4.30 4.00 3.50 2.97 3.12 3.48 4.10 3.94 3.91
Chira 3.78 4.10 4.20 4.11 3.83 3.35 2.77 2.80 3.10 3.67 3.57 3.61
Dedo 3.80 4.26 4.40 4.32 4.02 3.28 2.73 2.81 3.24 3.71 3.64 3.58
Ejajji 4.13 4.66 4.88 4.77 4.34 3.77 3.01 3.07 3.46 4.09 4.01 3.95
Gatira 3.70 4.22 4.39 4.24 3.91 3.36 2.71 2.78 3.17 3.82 3.57 3.49
Gojeb 4.08 4.57 4.83 4.52 4.30 3.81 3.19 3.34 3.69 4.09 4.08 4.07
Gore 3.74 4.36 4.52 4.42 3.66 2.98 2.62 2.67 2.98 3.39 3.49 3.48
Jimma 3.86 4.33 4.50 4.34 4.16 3.67 3.07 3.21 3.67 3.94 3.84 3.76
Limmu-genet 4.05 4.72 4.92 4.67 4.24 3.77 3.08 3.07 3.39 3.97 3.98 3.85
Natri 3.80 4.42 4.67 4.41 4.22 3.78 3.03 3.04 3.39 3.84 3.74 3.57
Shebe 3.84 4.34 4.52 4.21 4.02 3.55 2.99 3.22 3.53 3.87 3.79 3.76
Sokoru 4.02 4.56 4.71 4.58 4.19 3.64 3.01 3.06 3.43 4.07 4.01 3.91
Wolkite 4.14 4.65 4.82 4.70 4.29 3.93 3.22 3.26 3.59 4.34 4.24 4.09
Yaya 4.23 4.79 4.99 4.89 4.47 3.94 3.25 3.28 3.67 4.40 4.30 4.14
Yebu 3.95 4.47 4.56 4.51 4.11 3.45 2.86 2.90 3.32 3.95 3.82 3.70

69
Appendix Table 6. Coffee ETc (mm) estimated for different stations

Stations Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Annual
Agaro 87.7 92.5 125.3 131.8 131 112.4 99.6 102.3 109.6 121.9 83 84.5 1281.5
Arjo 80.9 87 119.8 128.9 119.6 95.4 80.5 77.4 86.5 110.1 72.5 75.6 1134.4
Asendabo 88.6 94.5 127.1 133.7 127.3 110.1 95.9 97.1 104.8 124.8 84.8 86 1274.9
Atnago 86 92.5 126.3 132.3 129.9 115.3 96.4 100.7 108.1 120.3 81.8 83.2 1272.7
Bedelle 85.6 91.2 125.9 136.9 126.4 100.3 87.9 90.4 101.4 120.7 80.8 82.3 1229.7
Bonga 85 89.4 120.4 125.6 123.8 107.5 97.3 100.1 107.6 121 82.2 83.4 1243.3
Botorbacho 82.2 86.7 118.1 125.8 120.8 103.9 93 92.5 98.3 115 77.9 79 1192.9
Chida 85.9 90.8 119.1 126.1 121.9 103.8 93.5 96.6 104.3 123.2 83.1 84.9 1233.1
Chira 82.5 85.2 113.5 120.5 116.7 99.2 87 86.9 93 110.4 75.3 78.7 1148.7
Dedo 83.1 88.3 118.7 126.6 121.7 98.1 85.8 87.6 96.4 112 76.4 78.4 1173.2
Ejajji 90.6 96.7 131.5 139.4 132.6 111.3 95.1 95.4 103.6 123.1 84.3 86.4 1289.9
Gatira 81.1 87.4 118.2 124.3 119.1 99.5 85.6 86.5 95 114.2 75.4 76.4 1162.7
Gojeb 89.8 95 129.4 133.4 130.9 112.7 100.5 103.3 109.8 124.2 85.7 88.4 1302.9
Gore 82.2 90.1 121.5 128 112.3 89.5 81.7 82.9 88.9 103.1 73 76.3 1129.4
Jimma 85.4 89.9 121.2 127.8 126.5 108.7 96.7 99.7 108.3 119.6 80.7 82.1 1246.7
Limmu-genet 88.3 97.5 132.2 136.9 129.9 111.2 96.8 95.4 101.6 120 83.3 84.2 1277.1
Natri 83.6 91.6 125.3 130.1 128.6 111.7 95.5 94.6 101 116 78.4 78.5 1234.6
Shebe 83.6 91.6 125.3 130.1 128.6 111.7 95.5 94.6 101 116 78.4 78.5 1234.6
Sokoru 88 94.3 127 134.1 127.9 107.9 94.7 95.1 102.8 122.6 84.1 85.3 1263.7
Wolkite 91.1 96.4 130 137.5 131.6 114.9 101.2 101.1 108 130.4 88.9 89.2 1320.2
Yaya 93 99.3 134.5 143 136.7 116.5 102.2 102 110 132.4 90.1 90.5 1350.1
Yebu 87.3 92.5 123.4 131.6 125.1 102.7 89.9 90.4 99.3 118.6 80.2 81.4 1222.5

70
Appendix Table 7. Coffee irrigation requirement (mm) estimated for different stations

Stations Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Annual
Agaro 79.7 82.2 76.2 64.5 6.2 0.0 0.0 0.0 0.0 27.4 64.8 75.6 476.7
Arjo 78.9 81.8 66.8 34.0 0.0 0.0 0.0 0.0 0.0 18.5 50.6 64.2 394.9
Asendabo 83.3 84.1 81.1 68.7 43.7 0.0 0.0 0.0 31.8 98.9 82.1 81.2 655.0
Atnago 85.9 84.7 87.8 71.1 6.9 0.0 0.0 0.0 0.0 51.0 68.5 80.4 536.3
Bedelle 85.4 86.9 94.4 79.8 1.7 0.0 0.0 0.0 0.0 31.1 70.6 76.5 526.4
Bonga 66.4 74.2 57.4 18.0 0.0 0.0 0.0 0.0 0.0 37.9 43.6 61.0 358.4
Botorbacho 81.4 81.7 83.8 75.1 4.2 0.0 0.0 0.0 9.4 90.1 67.3 78.8 571.8
Chida 72.2 61.1 28.7 6.7 0.0 14.8 0.0 0.0 4.6 41.6 63.7 50.7 344.1
Chira 57.2 63.9 43.0 19.0 0.0 0.0 0.0 0.0 0.0 16.6 42.7 58.1 300.6
Dedo 69.5 69.6 40.9 4.5 0.0 0.0 0.0 0.0 0.0 42.5 45.2 65.9 338.1
Ejajji 90.5 95.9 107.6 95.3 41.4 0.0 0.0 0.0 5.9 81.8 75.8 86.3 680.4
Gatira 65.6 74.8 77.1 28.0 0.0 0.0 0.0 0.0 0.0 21.7 48.3 64.2 379.6
Gojeb 70.6 76.9 69.5 51.1 2.8 0.0 0.0 0.0 0.0 56.6 60.4 74.2 462.2
Gore 74.5 84.4 89.8 66.8 0.0 0.0 0.0 0.0 0.0 7.5 38.3 69.3 430.6
Jimma 74.1 79.3 72.3 44.2 3.8 0.0 0.0 0.0 0.0 58.0 53.7 68.7 454.2
Limmu-genet 84.3 88.2 89.5 66.1 3.8 0.0 0.0 0.0 0.0 20.1 68.3 80.4 500.4
Natri 62.9 47.6 42.5 45.6 4.8 0.0 0.0 0.0 0.0 58.4 49.8 43.7 355.4
Shebe 62.9 47.6 42.5 45.6 4.8 0.0 0.0 0.0 0.0 58.4 49.8 43.7 355.4
Sokoru 81.5 85.8 92.5 69.0 34.0 0.0 0.0 0.0 6.7 86.7 79.1 80.1 615.4
Wolkite 90.3 85.5 101.7 90.8 59.0 7.6 0.0 0.0 30.9 116.7 87.5 89.1 759.2
Yaya 92.1 96.1 112.2 112.5 79.4 14.9 0.0 0.0 41.1 114.2 90.0 90.4 842.9
Yebu 66.9 77.0 34.7 19.2 0.0 0.0 0.0 0.0 0.0 19.0 41.4 61.5 319.8

71
Appendix Figure1. The DEM of study area

Appendix Figure 2.The point elevation extracted from DEM of study area for interpolation

72
BIOGRAPHICAL SKETCH

Minda Tadesse was born on September 07, 1987 in Dedo Wereda, Oromia region,

Ethiopia. He Attended his elementary school at Dedo Elementary School and Dedo high

school. He joined Alage Agricultural College in 2004 and graduated in 2006 with

Diploma in Natural Resource and immediately after his graduation, he was employed

by the Oromia Agricultural Bureau in Dedo wereda as supervisor and expert from 2007

to 2009 and after that he joined Jimma University in 2010 and graduated in 2012 with BSc

degree in Natural Resources Management soon after his graduation, he was employed by

the Ethiopia Institute of Agricultural Research (EIAR) at Jimma Agricultural

Research Centre and served as Assistant researcher I of Irrigation and Drainage researcher

from 2013 to 2016.

The author joined the School of Graduate Studies at Hawassa University 2016 to pursue

his MSc degree in Water Resources Engineering and Management in the School of

Bio-system and Environmental Engineering

73