Girma Ketsela thesis

Identification of potential rain water harvesting areas in the Central

Rift Valley of Ethiopia using a GIS�based approach

Master of Science Minor Thesis

Girma Moges Ketsela

August, 2009

Girma Moges Ketsela

MSc Minor Thesis

Identification of potential rain water harvesting areas in the

Central Rift Valley of Ethiopia using a GIS�based approach

Master of Science Minor Thesis

Name : Girma Moges Ketsela Registration No. : 771224430080 Program : Agricultural and Bioresource Engineering Course code : Number of Credits : 24 ects Examiner : prof. dr ir PWG Groot Koerkamp Supervisor : Willem Hoogmoed, Herco Jansen and Huib Hengsdijk Chair Group : Farm Technology Bornsesteeg 59 6708 PA Wageningen P: 03176482980

F: 03176484819

Dedication

To my wife, Nitsuh Girma (Baricho) to her love and dedication

i

ACKNOWLEDGEMENT

Firstly, I am sincerely grateful to the Netherlands Fellowship Program

(NFP/NUFFIC) for their financial support in the form of a scholarship to

undertake this MSc study in The Netherlands. Secondly, I like to thank the

project ‘Improving livelihoods and resource management in the Central Rift

Valley of Ethiopia (ILCE)’ of the partnership program of the Dutch Ministry

of Foreign Affairs – Development Cooperation and Wageningen University

and research centre on Globalization and sustainable rural development within

theme 2 -- Competing claims on natural resources, for funding this MSc work.

I would like to thank my supervisors, Dr. Willem Hoogmoed, Ir. Herco Jansen

and Dr. Huib Hengsdijk for the guidance, constructive comments throughout

this work and for the supervision during fieldwork.

I am thankful to the staff of Melkassa Agricultural Research Center for

providing me the necessary material. My very special thanks go to Tewodros

Mesfin for the logistic support, Yusuf Kedir for allowing me to use his

literature collected for years and Mezgebu Getnet for sharing data. I would

like to thank three of you for the scientific discussions.

A lot of thanks to my friends: Akililu Alem, Minwiyelet Nigatu, Urulo

Kebede, Birtukan Kebede, Getahun Tolla and Yonas Berhane for the

enjoyable time spent together.

I would like to extend my heartfelt thanks and love to my parents, my father

Moges, my mother Sintayehu, my brother Alemayehu (Chuchu) and my

sisters Bezawit, Selamawit and Rediet. Their continuous encouragement to be

successful from my early childhood has given me the energy to overcome the

difficulties I faced.

Last but not least, great thanks and love to my wife, Nitsuh Girma (Baricho)

for her care and affection in particular and for her love and endurance at large.

I would also like to thank her for being patient when I was busy with my

work.

ii

TABLE OF CONTENTS

Acknowledgement ......................................................................................................... i

Table of Contents.......................................................................................................... ii

List of Figures .............................................................................................................. iv

List of Tables .................................................................................................................v

List of Abbreviations and Acronyms........................................................................... vi

1. Introduction................................................................................................................9

1.1 Background ..........................................................................................................9

1.2 Problem statement..............................................................................................11

1.3 Objectives ..........................................................................................................13

1.4 Outline of the report...........................................................................................14

2. Literature review......................................................................................................15

2.1 RWH definition and classification.....................................................................15

2.2 History of RWH in Ethiopia ..............................................................................18

2.3 RWH in Ethiopia................................................................................................19

2.4 Evaluation of RWH in Ethiopia.........................................................................21

2.5 Experiences of other countries...........................................................................24

2.6 Critical factors for RWH site selection..............................................................27

3. Material and methods...............................................................................................30

3.1 Study area...........................................................................................................30

3.2 Identifying and assessing existing RWH structures...........................................31

3.3 Methodology of RWH potential mapping .........................................................32

3.3.1 General approach ........................................................................................32

3.3.2 Criteria selection and assessment of suitability level .................................32

3.3.3 Establishing the criteria weights .................................................................37

3.3.4 GIS Database ..............................................................................................40

3.3.5 GIS Analysis ...............................................................................................43

3.3.6 Evaluation ...................................................................................................44

4. Results and Discussions...........................................................................................46

4.1 Survey results.....................................................................................................46

4.1.1 RWH practiced............................................................................................46

4.1.2 Current state of affairs ................................................................................47

4.1.3 Use of harvested water................................................................................48

iii

4.1.4 Productive purpose of RWH.......................................................................49

4.1.5 Operation and maintenance.........................................................................50

4.1.6 The best option?..........................................................................................51

4.1.7 Lifting mechanism ......................................................................................53

4.1.8 Farmers’ involvement during implementation............................................53

4.1.9 Community ponds.......................................................................................54

4.2 Identification of potential RWH areas in the CRV............................................55

4.3 Validation...........................................................................................................62

5. Conclusions and recomendations.............................................................................64

References....................................................................................................................66

Appendices...................................................................................................................70

iv

LIST OF FIGURES

Figure 1: Classification of water harvesting techniques ...................................... 18

Figure 2: The Continuous Rating Scale developed by Saaty (1977). ................... 38

Figure 3: Textural map.......................................................................................... 42

Figure 4: Soil depth map....................................................................................... 42

Figure 5: Rainfall surplus map.............................................................................. 42

Figure 6: Slope map .............................................................................................. 42

Figure 7: Groundwater depth map ....................................................................... 43

Figure 8: Land cover map..................................................................................... 43

Figure 9: Flow chart for identification of potential sites. ..................................... 44

Figure 10: ponds lined by combining cement, sand and ‘kuyissa’ ....................... 48

Figure 11: The corrugated sheet were stolen ........................................................ 50

Figure 12 the owner used the corrugated sheet for other purposes and left the

tank open................................................................................................ 50

Figure 13: this pond was left unprotected and livestock was drinking directly

from the pond.........................................................................................51

Figure 14: pond at Edo kejele kebele with no water............................................. 53

Figure 15: RWH used for raising pepper seedling at Bulbula Wereda, Aleaku

Gubantaboke kebele............................................................................... 53

Figure 16: Percent of the study area per each suitability level for pond............... 56

Figure 17: Percent of the study area per each suitability level for in-situ ............ 56

Figure 18: Distribution of pond suitability level per land cover type................... 59

Figure 19: Distribution of in-situ suitability level per land cover type................. 59

Figure 20: Percent of very high and high suitability per Woreda area. ................ 60

Figure 21: Percent of very high to high suitability for each Woreda per total

very high and high suitability area......................................................... 61

Figure 22: Pond suitability map............................................................................ 62

Figure 23: In-situ suitability map.......................................................................... 62

Figure 24: In-situ suitability map with criteria given equal weights .................... 62

v

LIST OF TABLES

Table 1: Suitability rank for soil texture ............................................................34

Table 2: Suitability rank for soil depth ..............................................................35

Table 3: Suitability rank for rainfall surplus...................................................... 36

Table 4: Suitability rank for slope ..................................................................... 36

Table 5: Suitability rank for groundwater depth................................................ 37

Table 6: Suitability rank for land cover ............................................................. 37

Table 7: Pair-wise comparison matrix for ponds............................................... 39

Table 8: Pair-wise comparison matrix for in-situ.............................................. 39

Table 9: Weight (Percent of Influence). ............................................................ 39

Table 10: Very high and high suitability level per Woreda............................... 60

vi

LIST OF ABBREVIATIONS AND ACRONYMS

ADLI Agricultural Development-Led Industrialization

AHP Analytical Hierarchy Process

asl. above sea level

CA Catchment Area

CBO Community-Based Organizations

CRV Central Rift Valley

DEM Digital Elevation Model

ETB Ethiopia birr (currency 1birr = 0.08 USD (August, 2009))

FAO Food and Agricultural Organization

FDRE Federal Democratic Republic of Ethiopia

GPS Global Positioning System

GIS Geographic Information System

HGL Halcrow Group Limited

GIRD Generation Integrated Rural Development

HH Household

Kebele smallest administrative unit in the Ethiopian government structure

MCE Multi Criteria Evaluation

MoARD Ministry of Agricultural Rural Development

NGO Non Governmental Organization

OIDA Oromia Irrigation Development Authority

RWH Rainwater Harvesting

SNNPR Southern Nations Nationalities and Peoples Region

WLC Weighted Linear Combination

WOP Weighted Overlay Process

Woreda Administrative boundary comprising various Kebeles

vii

ABSTRACT

The Central Rift Valley (CRV) in Ethiopia is characterized by erratic rainfall

and high evapotranspiration. Total annual rainfall is sufficient for crop

production but the highly variable distribution in time and space frequently

threatens crop production and contributes to food insecurity. Rainwater

harvesting (RWH) for supplementary irrigation has been proposed in the

literature to mitigate dry spells. During the last decade, the Ethiopian

Government and various civil society organizations have supported the

implementation of RWH interventions throughout the country, including the

CRV. A systematic identification of areas suitable for different forms of RWH

may contribute to more effective interventions and targeted investments in

RWH.

The objective of this study is to identify potential suitable areas for RWH in

the CRV using Geographic Information Systems (GIS). In addition, a survey

has been conducted in selected Woredas of the CRV to identify and evaluate

qualitatively recent RWH interventions, and to support the prioritization of

suitability criteria and the validation of the developed suitability maps. A GIS-

based model has been developed to generate suitability maps for ponds and

in-situ RWH by using multi criteria evaluation. Six suitability criteria were

identified, i.e. soil, climate, topography, land cover, climate and groundwater

depth for RWH ponds while for in-situ RWH the same criteria were used

except groundwater depth. For each criterion five suitability levels were

identified, i.e. ‘very high’, ‘high’, ‘moderate’, ‘low’ and ‘very low’. Weights

were assigned to the criteria based on their relative importance for RWH

using an analytical hierarchical process. Using ArcGIS, maps for each

criterion were prepared and suitability maps for RWH ponds and in-situ

RWH.

The field survey showed that the majority of the RWH interventions in the

CRV consist of household ponds and tanks, and community ponds of which the

majority is not functioning satisfactorily. Reasons for the poor performance

are manifold but in general relate to improper site selection and RWH design

viii

and to the lack of ownership. The suitability maps show that the area with a

very high and high suitability for RWH ponds is about 4994 km2, which is

about 49% of the total study area; For in-situ RWH this area is about 6145

km2, or 60% of the total study area. The suitability maps provide an easy to

understand source of information to quickly identify areas that are more

promising than other areas for RWH intervention. The applied method can be

modified easily to incorporate other criteria or information with other spatial

resolution. It is concluded that suitability mapping using GIS is a flexible,

time-efficient and cost-effective method to screen large areas for RWH-

suitability facilitating decision-making for investments in RWH.

Keywords: Rainwater harvesting, GIS, Central Rift Valley, Ethiopia

9

1. INTRODUCTION

1.1 Background

Ethiopia is located in the Horn of Africa with an estimated population of

approximately 80 million. Agriculture is the mainstay of the Ethiopian economy

providing 60% of the Gross Domestic Product. It generates about 88% export

earnings and provides the majority of the employment. Many of the Ethiopian

smallholders depending on rainfed agriculture are food insecure. In many places, the

amount of rainfall and the duration of the rainy season are highly variable frequently

resulting in low crop yields and associated low incomes.

Surprisingly, Ethiopia is well-endowed with water resources (Sileshi et al. 2003). The

total annual runoff is estimated at 110 billion m3 however, much of which are carried

away by trans-boundary rivers. Groundwater reserves are estimated at 2.6 billion m3.

These natural resource bases have a potential for supporting a far greater number of

people than the current population (Seyoum 2003). However, current use of the

available water resources is very limited. For example, only 0.2 million ha is currently

irrigated (Alem 1999) which is less than 5% of the total land suitable for irrigation

(Dubale 2003). Large-scale dam and irrigation projects have not been widely

implemented in Ethiopia as they have often proved to be too expensive and

demanding in terms of construction and maintenance (Rämi 2003). Moreover, the

country is unable to utilize its large rivers for irrigation development due to various

reasons (Desta 2004). Therefore, rainfed agriculture is still the backbone of the

country’s economy and rural livelihoods.

Rainfed agriculture in semi-arid areas of Ethiopia is suffering from moisture stress

(Temesgen 2007), which is a major limiting factor for successful crop production.

The Central Rift Valley (CRV), 150 km southeast of Addis Ababa, is one of those

environmentally vulnerable areas in Ethiopia where poverty and natural resource

degradation are intertwined. The predominant livelihood strategy for the majority of

10

the population in the CRV (about 1.5 million) is the small mixed rainfed farming

system comprising cereal and livestock production. Because of large differences in

rainfall distribution between years and within years coupled with short rainy seasons,

rainfed agriculture is very susceptible to water shortage (Jansen 2009). The average

annual rainfall recorded in most of meteorological stations in the CRV is well above

700 mm yr-1 (Jansen et al (2009). In theory, this amount would be sufficient to grow

crops, yet large areas do not achieve food self-sufficiency. The underlying reason for

low crop yields might be that a high proportion of the rainfall is not available to the

crops because of excessive surface runoff and unproductive soil evaporation. Hence

the water availability to crops can be improved if the rainwater is retained in the area.

Improving the performance of rainfed agriculture is key to improve the livelihoods of

rural poor people. Rainwater harvesting (RWH) can improve agricultural production

by making water available during the time of dry spells. RWH is the deliberate

collection of rainwater from a surface known as catchment and its storage in physical

structures or within the soil profile (Mati et al. 2006). A small pond (~1000 m3) filled

by runoff can provide about half the water requirements of a half hectare plot to

overcome a mid season dry spell (Senay and Verdin 2004).

Rainwater harvesting can be practiced to provide water for irrigation, domestic water

and water for livestock. It can also serve as a way to replenish groundwater. The

rainwater harvesting techniques most commonly practiced in Ethiopia are runoff

irrigation (run-off farming), flood spreading (spate irrigation), in-situ water harvesting

(ridges, micro basins, etc.) and roof water harvesting (Alem 1999). Traditionally,

ponds are the main RWH structures in the Ethiopian Rift Valley where groundwater is

deep and other sources of water are not available (Alem 1999). Mostly, water

collected in ponds is used for growing vegetables and fruits around homesteads for

markets and home consumption.

Rainwater harvesting and storage has been recognized by the Ethiopian government

as a promising way for improving the water availability for crop production, domestic

11

use and water for livestock. To mitigate the erratic nature of rainfall in the arid and

semi-arid parts of the country, a national food security strategy based on the

development and implementation of rainwater harvesting technologies either at a

village or household level was adopted (Amha 2006). Accordingly, the Federal

Government allocated a budget for food security programs, ETB 100 million and

1,000 million ETB1 for 2002 and 2003, respectively. The majority of the budget was

used by regional states for the construction of RWH schemes including household

ponds (Rämi 2003).

1.2 Problem statement

Crop production in Ethiopia is mainly practiced under rainfed conditions and this

sector is the back bone of the country’s economy. Major part of this crop production

is in semi-arid areas such as in the CRV. To increase crop yields and improve food

security, effective planning and development of water resources in the CRV is

critically important.

Over the years, the Ethiopian government together with non-governmental

organizations has been involved in the development of RWH to enhance water

availability for crop production, drinking water for humans and livestock. The

government identified 315 very highly to moderately vulnerable Woredas (districts)

according to a number of criteria that includes drought risk, probability of extreme

weather conditions and past emergency needs (Profile 2005) and excavated many

ponds since 2003 mainly in four regions (Amhara, Tigiray, Oromia, SNNPR).

Although governmental and non-governmental organizations have been advocating

the use of RWH to improve the livelihoods of rural people, the implementation has

been confined with a range of problems and its adoption is low. A number of studies

(Alamerew 2006; Chala et al. 2004; Chala et al. 2003; Rämi 2003) suggested that

1 ETB= Ethiopian birr, 1birr=0.08 USD (August, 2009)

12

most of the constructed RWH are not performing as anticipated in terms of harvesting

and storing adequate amounts of runoff to meet the water demands particularly for

crop production due to various reasons.

One of the reasons is that there is the lack of scientific information to properly

allocate and plan RWH interventions. Currently, most RWH interventions are planned

on ad-hoc basis without much knowledge about the location-specific conditions. A

more systematic approach to the selection of feasible sites for RWH interventions

may improve their performance and rate of adoption.

The selection of potential areas depends on a multitude of factors including

biophysical and socio-economic conditions. Different studies used different

parameters. For instance FAO (2003) as cited by Kahinda et al. (2008) lists six key

factors to be considered when identifying RWH sites: climate (rainfall), hydrology

(rainfall–runoff relationship and intermittent water courses), topography (slope),

agronomy (crop characteristics), soils (texture, structure and depth) and socio-

economic criteria (population density, work force, people’s priority, experience with

RWH, land tenure, water laws, accessibility and related costs). Rao et al. (2003)

identified land use, soil, slope, runoff potential, proximity, geology, and drainage as a

criteria to identify suitable sites for RWH. Kahinda et al. (2008) used physical (land

use, rainfall, soil texture and soil depth), ecological (ecological importance and

sensitivity category) and socio-economic factors.

Ground survey is the best technique to identify suitable areas for RWH in relatively

small areas as it gives detailed information and does not need any interpolation

between points resulting in less variation from the actual nearby estimate. However

for larger areas like the CRV, ground survey is difficult and time consuming. The

application of GIS can be helpful for a first screening and identification of areas

potentially suitable for RWH (Prinz et al. 1998). However, the application of GIS for

identification of RWH potential areas in Ethiopia is almost not existing and there is

no documented work in this regard. GIS is a powerful set of tools used to collect,

13

store, retrieve, transform and display spatial data from the real world for a particular

purpose (Burrough 1986). GIS has been recommended as a decision-making and

problem-solving tool in RWH during the decision-making process (Mbilinyi et al.

2005). The purpose of this investigation is to identify potential areas suitable for

RWH in the CRV. For this purpose the combined effort of GIS analysis and field

surveys is used. The data from field surveys on current RWH interventions in the

CRV provide and supplement information to fine tune and calibrate the used GIS-

based methodology.

1.3 Objectives

• To identify and evaluate qualitatively selected existing rainwater harvesting

interventions in the CRV.

• To identify potential areas for different types of rainwater harvesting in the

entire CRV

14

1.4 Outline of the report

This report is structured in five sections. In section 1, a general introduction of the

study area was given and the problem of water shortage in Ethiopia particularly in

CRV is addressed. The need for more understanding of RWH to mitigate dry spells

along with the importance of using GIS in identifying potential areas is also

presented.

In section 2, a literature review on RWH is presented along with the general

classification of RWH interventions including a historical background of RWH in

Ethiopia.

In section 3, starting with the description of the biophysical characteristics of the

CRV, the data collection and research methodology are presented.

In Section 4, the results are presented and discussed. First, the result from field survey

in section 4.1 and the result from the GIS analysis for the identification of potential

areas in the sections 4.2 and 4.3.

Section 5 summarizes the overall conclusions based on the results followed by

recommendations for further studies.

15

2. LITERATURE REVIEW

2.1 RWH definition and classification

As water harvesting is an ancient tradition and has been used for millennia in most

drylands of the world, many different techniques have been developed. However, the

same techniques sometimes have different names in different regions and others have

similar names but, in practice, are completely different (Oweis 2004). Consequently,

there are a dozen of different definitions and classifications of water harvesting

techniques and the terminology used at the regional and international levels has not

yet been standardized (Nasr 1999).

Kahinda et al. (2008) defined RWH as the collection, storage and use of rainwater for

small-scale productive purposes. Critchley (1991) defined it as the collection of

runoff for productive use. Oweis (2004) defined it as the concentration of rainwater

through runoff into smaller target areas for beneficial use. Mati et al. (2006) defined

RWH as the deliberate collection of rainwater from a surface known as catchment and

its storage in physical structures or within the soil profile.

Runoff may be harvested from roofs and ground surfaces as well as from intermittent

or ephemeral watercourses and thus water harvesting falls into two broad categories:

Water harvesting techniques which harvest runoff from roofs or ground surfaces

named RWH and all systems which collect discharges from water courses named

flood water harvesting (Critchley et al. 1991). RWH technologies and systems can be

classified in several ways, mostly based on the runoff generation process, size of the

catchment and type of storage. Runoff generation criteria yields two types of systems

i.e. runoff based systems (runoff concentrated from a catchment) and in-situ water

conservation (rainfall conserved where it falls). The runoff storage criteria yield two

categories, i.e., storage within the soil profile and storage structures. The size of

catchment yields two categories, i.e., macro catchments and micro catchments (within

field). A simplified and practical classification of the various water harvesting

16

techniques and their characteristics and uses which was established by the World

Bank within the context of the “sub- Saharan Water Harvesting Study” in 1986- 1989

is shown in Figure 1.

In general, RWH systems for crop production are divided into three different

categories basically determined by the distance between catchment area (CA) and

cropped basin (CB) (utilization area): In-situ RWH, internal (Micro) catchment RWH

and External (Macro) catchment RWH (Hatibu and Mahoo 1999). To give the general

overview of the three categories, a short summary extracted from Hatibu and Mahoo

(1999) for each is presented below.

A. In-situ RWH

The first step in any RWH system involves methods to increase the amount of water

stored in the soil profile by trapping or holding the rain where it falls. This may

involve small movements of rainwater as surface runoff in order to concentrate the

water where it is wanted most. In-situ RWH is sometimes called water conservation

and is basically the prevention of net runoff from a given cropped area by holding

rainwater and prolonging the time for infiltration. This system works better where the

soil water holding capacity is large enough and the rainfall is equal or more than the

crop water requirement. Essentially, it includes all conventional approaches to soil

and water conservation designed to enhance rainwater infiltration. Examples of in-situ

RWH techniques include deep tillage, dry seeding, mixed cropping, ridges and

borders, terraces (“fanya juu” and “fanya chini”) and trash lines (Mbilinyi et al.

2005), vegetative / stone contour barriers, contour trenching, contour farming and tie

ridging methods (Sivanappan 2006).

B. Internal (Micro) catchment RWH

This is a system where there is a distinct division of CA and CB but the areas are

adjacent to each other. This system is mainly used for growing medium water

17

demanding crops such as maize, sorghum, groundnuts and millet. The major

characteristics of the system include small semi-circular pits, strip catchment tillage,

contour bunds, semi-circular bunds, meskat-type system and land conservation

aspects.

C. External (Macro) catchment RWH

This is a system that involves the collection of runoff from large areas which are at an

appreciable distance from where it is being used. This method is sometimes applied

with intermediate storage of water outside the CB for later use as supplementary

irrigation.

This system involves harvesting of water from catchments ranging from 0.1 hectare to

thousands of hectares either located near the cropped basin or far away. The

catchment areas usually have slopes ranging from 5-50%, while the harvested water is

used on cropped areas which are either terraced or on flat lands. When the catchment

is large and located at a significant distance from the cropped area the runoff water is

conveyed through structures of diversion and distribution networks. The most

important systems included in this category area hillside sheet/rill runoff utilization,

floodwater harvesting within the stream bed and ephemeral stream diversion.

18

Figure 1: Classification of water harvesting techniques (FAO, Critchley et al.(1991))

2.2 History of RWH in Ethiopia

The history of RWH practices in northern Ethiopia dates back as early as 560 BC,

during the Axumite Kingdom. In those days, rainwater was harvested and stored in

ponds for agriculture and domestic use (Seyoum 2003). Other evidences include the

remains in one of the oldest castles in Gondar (Fasiludus) from the 17th century which

used to have a sophisticated RWH system with a flume used for transporting water to

the palace pool used for swimming and religious rituals. In the south of the country,

the Konso people have a long and well-established tradition of building terraces to

harvest rain water for producing sorghum under extremely harsh environmental

conditions, i.e., low, erratic and unreliable rainfall (Alem 1999).

Water Harvesting

Rainwater Harvesting

(Local source)

Floodwater Harvesting

(Channel flow)

Runoff Harvesting Rooftop harvesting

Soil storage Deep ponding

Water

supply

Deep ponding

Plant

production

Water

supply

Deep ponding Soil storage

Water

supply

Plant

production

Micro-catchment

systems

Runoff Farming Floodwater farming

External catchment

systems

Category of

WH by source

Storage

Productive

use

Main plant

production categories

Sub division

19

Despite its long history, only a few decades ago RWH has received renewed attention

from policy makers. According to Promotion and application of RWH techniques

addressing water scarcity began through the government-initiated soil and water

conservation programs as response to the 1971-1974 (during Derg regime) drought in

Tigray, Wollo and Hararge (Seyoum 2003). However, the intervention was limited

because of the low level of community participation and declining attention by the

government.

After the fall of the military Derrg regime, both the Transitional Government of

Ethiopia (TGE), established in 1991, and the Federal Democratic Republic of

Ethiopia (FDRE), established in 1995, adopted the Agricultural Development-Led

Industrialization (ADLI) strategy, which emphasises improvement in agricultural

productivity to achieve food security and sustainable development. Besides,

recognizing the problem of variability in the rainfall distribution across the country,

the strategy advocates water-centered sustainable rural development (Desta, 2004).

Based on this, many different RWH technologies have been developed by regional

states, NGOs, communities, and individual farmers throughout the country.

2.3 RWH in Ethiopia

Ethiopia comprises of three main agro-climatological zones, i.e., Wet, Dry and

Pastoral. The dry zone comprises about 68% of the total land, 45% of the total arable

land and over 25% of the population. Food security in these areas is tied to the small

farmers, who rely heavily on rainfed agriculture. In many places of this zone, the

amount of rainfall and the duration of the rainy season are variable resulting in low

crop yields and associated low incomes. These areas are often food insecure.

Water is considered as one of the three pillars (land, labour and water) for the

development in the Ethiopia’s Agricultural Development led Industrialization (ADLI)

policy and food security programs. The Ethiopian Government has committed

financial resources to increase the irrigated area (Soriano 2007). Rainwater harvesting

technologies at the village or household level are proposed by the government of

20

Ethiopia as a practical and effective alternative to improve the livelihoods of rural

people at little cost and with minimal outside inputs. The Ministry of Agricultural

Rural Development (MoARD) and respective regional Bureaus planned and

implemented aggressive and ambitious water harvesting programs along the country's

food security programs (Desta 2006).

The Ethiopia government, prior to the large-scale implementation of RWH

technologies, conducted a study/survey in most parts of the country and in some other

countries having a longer RWH experience. This resulted in a “water harvesting

technologies package” including household-based RWH systems providing water for

humans, livestock and home garden horticultural crops (Desta 2004).

The RWH techniques most commonly practiced in Ethiopia are run-off irrigation

(run-off farming), flood spreading (spate irrigation), in-situ water harvesting (ridges,

micro basins, etc.), roof water harvesting (Alem 1999), birkas2 in Somalia region and

different runoff basins in Konso (Amha 2006) shallow wells (Soriano 2007), Ellea

and Haffirs3 (Kedir 2009)

Like other regions, Oromiya regional state where most CRV located has started the

implementation of water harvesting technologies to overcome problems related to

food security and poverty. Based on this, several RWH technologies have been

constructed by the regional Government, NGOs, communities, and individual farmers

throughout the region within the last few years and more are planned. In 2002/2003,

83,400 ponds, 500 underground tanks and 6,100 hand dug wells were planned and

49% of the ponds, 102% of the underground tankers and 64% of the wells were

completed in the same year (Chala et al. 2003).

2 Birka a traditional way to harvest rainwater in Somalia region. 3 Ella - Traditional deep water well in Borena zone.

Haffirs – are earthen embankments constructed with the aid of heavy machinery.

21

Different types of in-situ RWH have been used in different parts of Ethiopia. In

Tigray, micro-basins (roughly 1 m long and < 50 cm deep) are often constructed

along retention ditches for tree planting. The major conservation structures, meant for

erosion control commonly practiced in Ethiopia also conserve water in-situ and

include soil bunds, stone bunds, fana yaju and grass strips Alem (1999). These are

constructed in contour or graded depending on the rainfall of the area. For high runoff

areas graded structures are used. In low rainfall areas of southern Ethiopia, farmers

have developed a highly specialized water harvesting system. The cropland is

prepared in multitudes of circular depressions (3-4 m in diameter and < 1 m deep)

where a variety of crops are inter-cropped (Rockström 2000). Tied ridging are

traditionally used by small farmers in the eastern Hrarghe area as in- situ RWH

technique in sweet potato system using hand hoe (EARO 2000).

2.4 Evaluation of RWH in Ethiopia

The implementation of thousands of RWH structure has been confined with a range of

problems. Most people working in the field of RWH argue that most of the

constructed RWH structures do not perform as planned. A number of studies have

been conducted to evaluate the performance of the ambitious plan of the government

to develop RWH. The studies were conducted by the government, NGOs or

Academia. Some studies assess financial benefit, some technical viability and others

focused on assessing past experiences and identifying ways forward in order to

facilitate Government policies.

According to the progressive evaluation report on the implementation of RWH in

Oromiya (Chala et al. 2003), 98% of sampled beneficiaries responded high seepage.

The amount of collected water was not sufficient to meet the intended purpose

according to 53% of the farmers in East Shoa and 22% in Welega. All the

beneficiaries indicated that the catchment for collecting runoff was sufficient but not

the size of the RWH structure. The report concluded that the status of the constructed

ponds was not good owing to various problems like the unavailability of plastic sheet

22

to reduce seepage losses, lack of coordination and facilitation during implementation,

while community and land holding size were not taken into account during the design

and implementation.

The evaluation of RWH ponds that were constructed in 2005 and 2006 in northern

Ethiopia, Tigray, showed that the large majority failed because of insufficient water

collection or leakage problems due to poor construction (Segers et al. 2007). In

addition, a considerable number of ponds suffered from lack of maintenance

contributing to the poor performance. Some RWH ponds were silted up completely

and remain as gentle depressions in the landscape. The other reason for failures was

that households did not construct or maintain the diversion channels and inlets that are

needed to harvest runoff water.

Similar conclusions were drawn by Alamerew (2006) who summarized the various

constraints in implementing the RWH projects including inadequate public awareness

and ownership of local communities, lack of adequate knowledge and skills in

management of RWH schemes, insufficient involvement of communities in planning

and implementation processes and lack of facilitation for establishment/strengthening

of community-based organizations (CBOs). In addition, the author pointed out that

the various RWH efforts lacked research, for example, on indigenous knowledge and

best practices in RWH, improving traditional practices and/or adapting new

technologies to local conditions which constitute among the critical inputs for a

successful intervention.

Field assessments by the Oromiya Irrigation Development Authority (OIDA)

identified problems with the implementation of RWH technology related to inefficient

runoff collection and unwise use of the harvested runoff (Chala et al. 2004). This is

due to poor catchment selection and characterization in relation to the structure design

and lack of information on the utilization of the system (like family drip irrigation).

The evaluation also collected opinions from non-target farmers (who don’t have

RWH), other OIDA staff and development agents (DAs) on to the issue “why RWH

23

structures don’t retain or hold water”. The answer from the three groups pointed at

high seepage losses attributed to either poor compaction, cracks in tanks, and poor site

selection due to a lack of experience.

A RWH impact assessment at Alaba Woreda showed a positive effect on agricultural

productivity (Amha 2006). The cropping pattern in the studied area changed and farm

households started to grow cash crops which were not previously grown in the area.

However, benefits depend on market and infrastructure access and crop

diversification to minimize risk. Despite its potential, adoption of RWH technology is

slow. Some reasons for the low adoption were poor quality of the construction

resulting in cracks in the cemented floor and loss of water, improper site selection

(insufficient runoff) and fear of malaria spread.

The RWH structures in Amhara and Tigray face many problems, many of which

originate from the speed and scale of implementation (Rämi 2003). Among the

identified problems were (1) poor site selection; it was done hurriedly and without

experience, with the consequence that many tanks do not hold water. Often the level

of poverty was used as the most important criteria for selection of target beneficiaries,

while the technical criteria of runoff and water collecting potential were neglected.

Poor site selection is the most important reason for failures. (2) Leakage; Most of the

tanks (cemented) were leaking due to cracks in the walls of the structures. This was

attributed to the lack of skilled labour during construction, the poor quality of work as

result of the quota system imposed by the regional Government that put junior experts

and development agents under pressure to construct tanks quickly. (3) siltation and (4)

wastage and uneconomical use of water.

In the semi-arid areas of Ethiopia, tied ridging as in-situ water harvesting have been

found to be very efficient in storing rain water particularly in drier seasons and lead to

substantial increase of yields in some of the major dry land crops including maize,

sorghum wheat and mung bean regardless the different planting patterns used, i.e.

24

planting in the furrow or on top of the ridge compared to the flat seed-bed (farmers

practice) (W/Giorgis 2002).

Promotion of RWH in Ethiopia has given more emphasis on structural storage (ex-situ

deep ponding) than in-situ RWH (Desta 2004). However, in-situ RWH is preferred as

it does not require water lifting from the pond and water application, is more closely

linked to traditions, and the costs are lower (Desta 2004). Ex-situ RWH is not suitable

for staple crops such as cereals, which are needed by the farmers to secure their food

needs as it is evaluated from cost benefit analysis. Mesfin (2004) also concluded that

the most efficient and cheapest way of conserving water is in-situ RWH. Evaporation

losses can be reduced greatly if rainfall is stored in the soil rather than in an open

structure. Gebre and Giorgis (1980) and Beyu and Alemu (1998) further concluded

that in-situ RWH practices are also more economically feasible to resource poor

farmers than ex-situ RWH methods.

Different studies in Ethiopia showed the potential of different in-situ RWH methods

to enhance soil moisture storage and rainfall use. Mesfin (2004) studied different in-

situ RWH methods on soil water storage and on the growth, grain yield and water use

efficiency of sorghum in the CRV and he concluded that soil moisture increased

compared to conventional tillage methods. Temsgen (2007) showed that tied-ridging

performed better in retaining water than the local tillage practice (with Maresha plow)

and inverted broad beds (with broad bed maker) providing more water to crop

production in a semi-arid region where rainfall is erratic.

2.5 Experiences of other countries

Water harvesting is an ancient technology practiced in many parts of world such as

North America, Middle East, North Africa, China, and India. Different indigenous

RWH techniques and systems have been developed and remain an important source

of water for agriculture in different parts of the world. The Middle-East and North-

Africa have a long tradition in RWH as one of the methods for survival in the area.

25

According to an assessment of desertification and RWH the first RWH system in

history was built in the Middle-East and North-Africa (Nasr 1999). Researchers have

found signs of early RWH structures constructed over 9000 years ago in the Edom

Mountains in southern Jordan. In Israel complete RWH systems have been found in

the Negev Desert, which were about 4000 years old.

Remnants of other RWH installations have also been discovered in Iraq and in the

Arabian Peninsula along the routes used by caravans. The RWH installations

consisted mainly of means to collect rainwater and divert it into natural and/or

artificial ponds and reservoirs. Other evidences of RWH have been found in Yemen,

Palestine, Morocco and Egypt. At present, all countries in the Middle-East and North-

Africa region practice intensively one or more RWH techniques to collect and store

rainwater for use in meeting crop, human and animal needs.

A study by Ngigia (2005) in the Laikipia district, Kenya showed that improved farm

ponds provide one of the feasible options of reducing the impacts of water deficit that

affect agricultural productivity in semi-arid environments in Sub-Saharan Africa. The

field evaluation revealed that on-farm RWH systems are common in the study area

with sizes ranging from 30 to 100 m3 and catchment areas varying from 0.3 to 2 ha.

The hydrological evaluation of the farm ponds revealed that one of the challenges was

how to reduce the seepage and evaporation water losses. He reported significant water

losses through seepage and evaporation, which accounted on average for 30–50% of

the stored runoff. The high losses are one of the factors that affect the adoption and

up-scaling of on-farm water storage systems. If seepage loss is reduced with lining

material and if RWH is combined with drip irrigation on-farm storage systems can be

economically viable and farmers are able to recover the full investment costs within 4

years

Another study in Kenya of the Mwala division on the impact of run-off water

harvesting for dry spell mitigation in maize showed that harvesting runoff water for

supplemental irrigation is a risk-averting strategy, pre-empting situations where crops

26

depend on rainfall that is highly variable both in distribution and amounts (Barron and

Okwach 2005 ). Water harvested and stored in an earthen dam provided a technically

feasible option to supplement crop water demand. By using underground spherical

tanks having a combined capacity of 60 m3, seasonal water for supplemental irrigation

for an area about 400 m2 was guaranteed. With RWH, farmers diversified production

through horticultural cash crops and dairy resulting in higher earnings compared to

those from rainfed maize. Barron and Okwach (2005) also showed that combining

supplemental irrigation and low N fertilizer inputs resulted in a yield increase of more

than 50%.

An evaluation study of small basin water harvesting in Jordan showed that a

‘Negarim’ micro-catchment of 25 m2 can provide enough water to support crop water

requirements if properly designed and if water holding capacity of the root zone is

adequate (Oweis and Taimeh 1996). Another study from Northern Burkina Faso

showed that RWH has potential for supplemental irrigation to ensure self sufficiency

of staple cereals for the small-holder farmer (Fox and Rockström 2000).

Fox et al. (2005) studied the risk associated with and economic viability of RWH for

supplemental irrigation in semi-arid Burkina Faso and Kenya and suggested that

supplementary irrigation can generate economic benefits and improve long-term food

self-sufficiency compared to rainfed agriculture. However, they stressed that the

investment in supplementary irrigation is economically viable, but only if it is

combined with growing a cash crop during the winter (dry) season.

Hatibu et al. (2004) tried to evaluate farmer-initiated and managed RWH systems on

farmers’ income and living standards using two districts in Tanzania, i.e. Maswa and

Same. They compared the performance of the RWH for maize in terms of yields,

gross margins and return to labor for the different levels of rainwater harvesting. The

levels of water availability were divided into four main categories: (1) rainfed system

where the farmer only captured and conserved the rainwater falling directly on the

field without additional water from external sources; (2) poor RWH in which the

27

farmers capture and conserve all direct rainfall (in-situ RWH) and also irregularly

obtain some extra run-off from external (micro-catchment) sources with reliability of

less than 25%; (3) medium RWH is where the reliability of obtaining runoff from

external (micro and macro-catchment) sources is above 25% but less than 75%; and

(4) a good RWH where the availability of runoff from external (macro-catchment)

sources was above 75% - and where storage ponds are used. Good RWH increases

yield of maize (in Same area) by four fold of rainfed yield level, and two fold for rice

(in Maswa area). They compared the performance of good RWH harvesting across

four different crops and they found that using rainwater harvesting for vegetable

production is consistently very beneficial to the farmer with returns to labor

exceeding US$ 10 per person day and in some years reaching nearly US$ 200 per

person day.

2.6 Critical factors for RWH site selection

Although the government and non-governmental organizations have been advocating

the use of RWH, its performance and adoption rate is not as much as it was

anticipated as different evaluation studies have shown and discussed above (section

2.4). It must be underscored, however, that the technology by itself is often not the

problem for the low performance and adoption, but rather the poor implementation.

Proper implementation including area selection and design could improve the

performance of RWH and improve the livelihoods of many poor.

The identification of potential areas suitable for RWH is therefore the key for a

successful RWH intervention. One of the main reasons for failure of RWH structures

is the lack of scientifically verified information which could be used to identify areas

where RWH can be applied and for which type of RWH techniques.

The potential of areas for RWH depends on a multitude of parameters, either physical

factors like rainfall, land use, soil and topography and/or the combination of the

physical factors and socio-economic factors.

28

FAO (2003) by Kahinda et al. (2008) lists six key factors to be considered when

identifying RWH sites: climate (rainfall), hydrology (rainfall–runoff relationship and

intermittent watercourses), topography (slope), agronomy (crop characteristics), soils

(texture, structure and depth) and socio-economic (population density, work force,

people’s priority, experience with RWH, land tenure, water laws, accessibility and

related costs). Rao et al. (2003) use land use, soil, slope, runoff potential, proximity to

the utility points (like irrigation and drinking water supply schemes), geology, and

drainage as a criteria to identify suitable sites for RWH. To develop a GIS-based

RWH model (RSM) that combines a Multi-Criteria Evaluation (MCE) process,

Kahinda et al. (2008) used physical (land use, rainfall, soil texture and soil depth),

ecological (ecological importance and sensitivity category) and socio-economic

factors. Ramakrishnan et al. (2008) used slope, porosity and permeability, runoff

potential, stream order and catchment area as criteria to select suitable sites for

various RWH /recharging structures in the Kali watershed, Dahod district of Gujarat,

using remote sensing and GIS techniques.

United Nations Environment Programme (Mati et al. 2006) conducted a study to

determine if RWH technologies can be mapped at continental and country scales. The

project utilized a number of GIS data sets including rainfall, land use, land slope, and

population density to identify four major commonly adaptable RWH technologies:

roof top RWH, surface runoff collection from open surfaces into pans/ponds, flood

flow storages and sand/sub-surface dams and in-situ RWH.

Another study deepened the scholarly understanding of the role of indigenous

knowledge to identify potential sites for RWH in Tanzania (Mbilinyi et al. 2005).

This study concluded that farmers have a substantial amount of knowledge on RWH

systems and identification of potential sites for different RWH systems. Most of

indigenous knowledge (although it varies among locations) is based on biophysical

factors, including topography, soil type and distance from water sources.

29

For relatively small areas, the critical factors can be assessed by field surveys

However, for larger areas the application of GIS can be helpful for a first suitability

screening with less time, cost and labour.

30

3. MATERIAL AND METHODS

3.1 Study area

The CRV in Ethiopia (38°00’-39°30’ E and 7°00’-8°30’ N) covers about 1 million ha

and is part of the Great African Rift Valley. The study area is in the centre of the

Ethiopian Rift, 150 km southeast of the Addis Ababa (Figure 2). CRV encompasses

parts of Oromiya Regional State and Southern Nations Nationalities and Peoples’

Regional State (SNNPR). The total population of CRV is approximately 1.5 million

with an average population density of 1.5 p ha-1 (Jansen 2009). The CRV is a closed

river basin with elevations ranging from 1600 m asl. in valley floor to about 3000 m

to the east and west. Annual rainfall in CRV ranges from about 600 mm near the lakes

at the valley floor up to 1250 mm in the higher elevations near the borders of the

basin. About 70% of the rainfall precipitates in the short rainy season (July to

September). The soils in the western part of the CRV are mainly Cambisols and

Luvisols in the hills and foot slopes, and Vertisols in parts of the flat plains.

The CRV is a mixture of moderately to intensively cultivated land, open bush, open

woodland, lakes and forest. Rainfed agriculture (moderately and intensively

cultivated) dominates the land cover in the CRV with an estimated area of 742600 ha

(more than 70% of the total). This rainfed agriculture depends much on the water

availability which is highly variable in terms of temporal and spatial distribution.

Irrigated agriculture is now increasing in the CRV especially around Meki town using

shallow groundwater and Lake Ziway as a source of water; however it is still minor as

compared to rainfed agriculture.

31

Figure 2: Location of study area with Woreda boundary

3.2 Identifying and assessing existing RWH structures

An inventory of past and recent RWH interventions in the study area was conducted

through a review of the literature, interviews with key stakeholders and a field survey

of RWH farmers. During the field survey data were collected through observations

and interviews with farmers using a semi-structured questionnaire (Appendix I). The

field survey focused on the type of RWH practiced, farmers’ experiences and

constraints, duration of water storage, water uses and application methods. This

helped to understand the bottlenecks and successes of RWH interventions in the CRV.

The qualitative assessment of the problems associated with RWH interventions

helped to identify the type of RWH for which the suitability map(s) were developed.

In addition, the collected field data were used to generate information to fine tune and

calibrate the methodology and validate the derived suitability maps.

32

3.3 Methodology of RWH potential mapping

3.3.1 General approach

The objectives and associated technologies for runoff harvesting are highly location-

specific, and depend on physiographic, environmental, technical and socio-economic

conditions. Therefore appropriate technologies are developed for particular regions

and cannot simply be replicated in other areas (Winnaar et al. 2007).

The identification of suitable areas for RWH is a multi-objective and multi-criteria

problem. The field survey indicated that most RWH interventions in the CRV

focus(ed) on (i) surface runoff collection from open areas and storage in ground-based

structures (ponds) and (ii) in-situ RWH and storage in the soil profile for crop

production Therefore, the objective of this report is to map potential areas for both

RWH interventions, i.e. for ponds and in-situ RWH. Six and five suitability criteria

for ponds and in-situ respectively were identified. This multi-objective multi-criteria

methodology involved the following major steps: (i) Selection of criteria; (ii)

assessment of suitability level for criteria for ponds and in-situ RWH; (iii)

assignments of weights to these criteria; (iv) collection of spatial data for the criteria

including GPS survey to supplement and generating maps for each using GIS tools;

(v) developing a GIS-based suitability model which combines maps through MCE

process; and (vi) generate suitability maps.

3.3.2 Criteria selection and assessment of suitability level

From the literature review (section 2.6) and information obtained from field survey

supported by expert judgment, six criteria were selected for the identification of

potential areas for ponds, i.e. (soil texture, soil depth, rainfall surplus (precipitation-

evapotranspiration), topography, groundwater depth and land cover. For in-situ the

same criteria were considered except groundwater depth. The criteria used in the GIS

analysis are presented and discussed below.

33

Because of the different scales on which the criteria are measured, MCE requires that

the values contained in the criterion map are converted into comparable units.

Therefore, the criteria maps was re-classed into five comparable units i.e suitability

classes namely; 5 (very high suitability), 4 (high suitability), 3 (medium suitability), 2

(low suitability), and 1 (very low suitability). The suitability classes are then used as

base to generate the criteria maps (one for each criterion).

3.3.2.1 Soil texture

The suitability of a certain area either as catchment or as cropping area in RWH

depends strongly on its soils characteristics. Soils with high infiltration rates, such as

sandy soils, are not favorable for RWH structures. A challenge in design and

construction of on-farm water storage structures, such as farm ponds, is how to

minimize water losses (mainly due to seepage and evaporation) (Ngigia et al. 2005).

Therefore, runoff harvesting into unlined ponds depends on soil type especially to

avoid seepage problems (Mati et al. 2006). For in-situ RWH, the actual suitability

will depend on the capacity of the soil to retain as well as to deliver water to the root

zone. Both soil texture and depth determine the total soil water storage capacity,

which controls the amount of water that is available using in-situ RWH for crops

during the dry periods (Oweis 2004).

Farm ponds are suitable in areas with low soil permeability (Rao et al. 2003). Based

on the survey results and the literature review, almost all constructed ponds in the

study area have seepage problem. Therefore, finding suitable areas for ponds in this

study area should be able to focus in locating areas with good potential in retain the

harvested water. Clay soils have low permeability( high hydraulic resistance) and can

hold the harvested water, and therefore they are the best soils for water storage

(Mbilinyi et al. 2005).

Regarding in-situ RWH, soils with high water holding capacity are suitable for ridges

and borders (in-situ) where as sandy soils are not suitable (Mbilinyi et al. 2005).

34

Therefore loamy soils are most suitable for in-situ RWH whereas clay soils are less

suitable because of their low infiltration capacity and risk of water logging.

Table 1: Suitability rank for soil texture

No. Soil textural class Pond Suitability In-situ Suitability

1 fine 5 2

2 Fine and medium 4 3

3 Medium 3 5

3 Medium and coarse 2 4

4 coarse 1 2

3.3.2.2 Soil depth

Soils should be deep enough to allow excavation to the prescribed depth for farm

ponds. One of the reason for the poor performance of RWH in Tigray was that the

prescribed pond depth had not been reached during construction (Segers et al. 2007).

According to another study (Rämi 2003), one farmer in Warkaja Kebele, Wollo was

advised to dig out a dome shaped underground-structure in order to collect roof water.

At a depth of three meters he found groundwater and was forced to stop digging.

Therefore, soil depth is considered as one of the criteria for pond suitability in this

study.

Soil depth is also important for identifying potential areas for in-situ RWH (for crop

production) as it ensures adequate rooting development and storage of the harvested

water. Critchley et al. (1991) consider soil depth as one criteria and suggest deeper

soil depth as suitable for various micro RWH methods. Kahinda et al.(2008) also use

soil depth as one criterion for selecting potential areas for in-field RWH.

35

Table 2: Suitability rank for soil depth

No. Soil depth class Depth(m) Pond In-situ

1 Very deep >1.5 5 5

2 Deep 1.0 - 1.5 2 5

3 Moderately deep 0.50 - 1.0 1 5

4 shallow 0.25 - 0.5 1 3

5 Very shallow <0.25 1 1

3.3.2.3 Rainfall surplus

The magnitude of harvestable rainfall plays a significant role in assessing the

suitability of RWH for a given area (Kahinda et al. 2008). Because of the very high

variability in distribution and amount of rainfall in the study area, it is very important

to consider the rainfall erraticness/ distribution, evapotranspiration and availability of

harvestable runoff. RWH is particularly relevant if rainfall is irregular than if rainfall

is evenly distributed over the year (Jansen 2009). The rainfall surplus is therefore

considered as criteria to account for the spatial distribution of harvestable runoff

availability. The logic behind calculating rainfall surplus relies on assumptions that the value of

difference between rainfall and evapotranspiration indicates runoff availability.

The rainfall surplus is calculated by subtracting long-term average monthly values of

the evapotranspiration from the precipitation (P-ET) for seven meteorological

stations. The annual rainfall surplus was calculated at each metrological station by

adding only the positive values of the difference (P-ET). Then the spatial distribution

of the rainfall surplus was generated by interpolating the point values. High suitability

rank was given for areas with large rainfall surplus as it ensures the availability of

runoff to be harvested.

36

Table 3: Suitability rank for rainfall surplus

No. Rainfall surplus class Values(mm) Pond In-situ

1 Very large deficit < 75 1 1

2 Large deficit 75 - 150 2 2

3 Medium deficit 150 - 225 3 3

4 Small surplus 225 – 300 4 4

5 Large surplus > 300 5 5

3.3.2.4 Topography

The slope of land is important in site selection and implementation of all ground

based RWH systems, especially ponds, pans, weirs and also in-situ RWH (Mati et al.

2006). In-situ RWH is not recommended for areas where slopes are greater than 5%

due to uneven distribution of run-off and large quantities of earthwork required which

is often costly (Critchley et al. 1991). Farm ponds are generally more appropriate in

areas having a rather flatter slope however a slight slope is needed for better

harvesting of the runoff. Therefore flat areas with a slope less than 2% were assigned

a higher suitability rank for in-situ RWH whereas for ponds areas with slope ranging

from 2 to 8% were given higher suitability rank.

Table 4: Suitability rank for slope

No. Slope class Slope (%) Pond In-situ

1 flat <2 3 5

2 undulating 2-8 5 4

3 rolling 8-15 4 3

4 Hilly 15 – 30 2 2

5 mountainous >30 1 1

3.3.2.5 Groundwater depth

Ponds only make sense if any other alternative is not available. Investment into other

schemes (if any) like river diversion, hand dug and deep wells are generally favored

37

(Rämi 2003). Therefore, the depth of the groundwater layer was used as criterion for

the suitability of ponds; i.e. less weight was given to areas with groundwater

potential.

Table 5: Suitability rank for groundwater depth

No. Groundwater class Pond

1 <40m 2

2 40-120m 4

3 high relief area 5

3.3.2.6 Land cover

As this study focused on RWH for crop production, both in-situ and ponds should be

located close to agricultural areas. Therefore, the land cover was used as one criterion

for in-situ and pond RWH to identify potential areas on agricultural land.

Table 6: Suitability rank for land cover

No. Land cover class Land cover types Pond In-situ

1 Very high Intensively cultivated 5 5

2 high Moderately cultivated 5 5

3 medium Forest, exposed surface 2 1

4 Low/ restricted mountain 2 1

5 Very low/restricted Water body, urban area restricted restricted

3.3.3 Establishing the criteria weights

Since not all the criteria are equally important for the identification of potential RWH

areas, different weights were assigned to the criteria. For the development of weights,

the pair-wise comparison known as the Analytical Hierarchy Process (AHP)

developed by Saaty (1977) was used. Pair-wise comparison concerns the relative

importance of two criteria involved in determining the suitability for a given

objective. The rating between two criteria is provided on a 9-point continuous scale

(Figure 2) ranging from extremely less important to extremely more important. The

38

comparison is done for every possible pairing of criteria and the rating is entered into

a pair-wise comparison matrix. Only the lower triangular half of the matrix needs to

be filled in as the matrix is symmetrical.

1/9 1/7 1/5 1/3 1 3 5 7 9

extremely

very

strongly

strong ly

modera tely

equa lly

m odera tely

strong ly

very

strong ly

extremely

Less important More important

Figure 2: The Continuous Rating Scale developed by Saaty (1977).

During pair-wise comparison, criteria were rated based on the literature review, information from the field survey and discussions with people working and having experience with RWH. The relative weights for each criterion and suitability rank for classes are assigned for the two categories of RWH. The result from the pair-wise comparison is presented in The final weight calculation requires the computation of the principal eigenvector of

the pair-wise comparison matrix to produce a best-fit set of weights. For this

calculation the WEIGHT module of Idrisi software was used and the result is

summarized in Table 9. The Consistency Ratio (CR) of the matrix, which shows the

degree of consistency that has been achieved during comparing the criteria or the

probability that the matrix ratings were randomly generated, was 0.03 and 0.02 for

pond and in-situ, respectively, which is acceptable as the values are less than or equal

to 0.1 (Saaty 1977).

Table 7 and Table 8 for ponds and in-situ respectively.

39

The final weight calculation requires the computation of the principal eigenvector of

the pair-wise comparison matrix to produce a best-fit set of weights. For this

calculation the WEIGHT module of Idrisi software was used and the result is

summarized in Table 9. The Consistency Ratio (CR) of the matrix, which shows the

degree of consistency that has been achieved during comparing the criteria or the

probability that the matrix ratings were randomly generated, was 0.03 and 0.02 for

pond and in-situ, respectively, which is acceptable as the values are less than or equal

to 0.1 (Saaty 1977).

Table 7: Pair-wise comparison matrix for ponds.

Texture Depth Rainfall surplus Groundwater Land cover Slope

Texture 1 4 3 4 6 5

Depth 1/4 1 1 1 3 2

Rainfall surplus 1/3 1 1 3 4 3

Ground Water 1/4 1 1/3 1 3 2

Land cover 1/6 1/3 1/4 1/3 1 1/2

slope 1/5 1/2 1/3 1/2 2 1

Table 8: Pair-wise comparison matrix for in-situ.

Texture Depth Rainfall surplus Land cover Slope

Texture 1 2 3 7 4

Depth 1/2 1 2 5 3

Rainfall surplus 1/3 1/2 1 4 3

Land cover 1/7 1/5 1/4 1 1/2

Slope 1/4 1/3 1/3 2 1

Table 9: Weight (Percent of Influence).

Weights (%) No. Criteria Ponds In-situ

1 Soil texture 43.2 42.6

2 Soil depth 13.6 26.2

40

3 Rainfall surplus 19.8 17.8

4 Groundwater depth 11.5 Not used

5 Topography/slope 7.2 8.5

6 Land cover 4.6 4.9

Sum 100 100

3.3.4 GIS Database

The GIS dataset of the criteria required for the identification of suitable areas were

derived from available data sets most of them provided by Halcrow Group Limited

and Generation Integrated Rural Development (HGL and GIRD) Consultants, and

supplemented with information from a GPS survey. The GIS database required for

identifying RWH potential areas was developed using ArcGIS software, by utilizing

both vector and raster databases (Figure 3 - 8).

The soil textural map was derived from the land suitability dataset developed by HGL

and GIRD. This dataset has five textural classes namely coarse, medium and coarse,

medium, fine and medium and fine. The soil texture layer was clipped to the study

area and reclassified into five numerical categories and assigned different suitability

rankings for ponds and in-situ RWH. The suitability ranking is made on a scale from

1 to 5 with 5 implying most suitable (Table 1).

The soil depth layer was derived from the land suitability dataset of HGL and GIRD. This dataset has

five depth classes namely very deep (> 150 cm), deep (100-150 cm), moderately deep (75-100 cm),

shallow (50-75 cm) and very shallow (<50 cm). Then the soil depth layer was clipped to the study area

and reclassified to numeric values and assigned different suitability rankings for ponds and in-situ

RWH (

41

Table 2).

The rainfall surplus map was generated by interpolating seven rainfall surplus point

values. The values were calculated by subtracting evapotranspiration from

precipitation (P-ET) using data of seven meteorological stations in and near the study

area. The calculation was done for each month and for each station and the

accumulated rainfall surplus was calculated by adding the positive values of the

difference (P-ET) starting from rainy season. To get a surface map of rainfall surplus,

the calculated values were interpolated using the Inverse Distance Weight (IDW)

interpolation method of ArcGIS. The new data were clipped to the study area and re-

sampled to 90 m. The rainfall surplus map comprises five classes, i.e. large surplus,

small surplus, moderate surplus, small deficit and large deficit (Table 3).

A slope map, expressed in percentage, for the study area was derived from the DEM

(elevation dataset) with 90 m resolution obtained from HGL and GIRD. The slope

map was reclassified into five classes based on the FAO classification (FAO 2002 as

cited by Meti et. al.(2006)) namely 0-2% is flat; 2-8% is undulating; 8-16% is rolling;

16-30% is hilly; > 30% is mountainous and assigned different suitability rank for

ponds and in-situ RWH (Table 4).

The groundwater depth layer was generated by digitizing (on screen) the hydrological

map of the region which was obtained from the Ministry of Water Resource. The

dataset has three classes namely <40 m, 40-120 m and high relief area and these

values were converted into to numeric values (Table 5).

The land cover map was generated from landuse/cover (year 2006) dataset from HGL

and GIRD. The generated land cover map has 11 classes namely intensively

cultivated (52%), moderately cultivated (20%), water body (7.6%), shrubland (5.4%),

forest (4%), afro-alpine (4%), (1%), woodland (2%), marshland (1.6%), exposed

surface (1.4%) and urban area (0.1%). The land cover map was reclassified and

assigned numeric values (Table 6).

42

Figure 3: Textural map

Figure 4: Soil depth map

Figure 5: Rainfall surplus map

Figure 6: Slope map

43

Figure 7: Groundwater depth map

Figure 8: Land cover map

3.3.5 GIS Analysis

All the processing in finding RWH suitability map has been implemented in a suitability

model developed in the model builder of ArcGIS 9.3. The suitability model generates

suitability maps for RWH by integrating different input criteria maps using Weighted

Overlay Process (WOP) also known as Multi-Criteria Evaluation (MCE). MCE can be

achieved by a weighted linear combination (WLC) wherein continuous criteria (factors)

are standardized to a common numeric range, and then combined by means of a weighted

average. With a weighted linear combination, criteria are combined by applying a weight

to each followed by a summation of the results to yield a suitability map using the

following equation:

ii xwS ∑=

Where S = suitability wi = weight of factor i xi = criterion score of factor i

44

A number of tools of ArcGIS were incorporated in the model to solve various spatial

problems, i.e. calculating slope, reclassifying values, clipping, re-sampling, reprojecting,

over laying, etc. (Appendix II). All source maps were in vector type formats, each

containing their related attribute files. These have been converted into raster datasets and

then re-sampled to the same cell size (90 m) to enable the ArcGIS overlay operation. The

conceptual framework is shown in Figure 9.

3.3.6 Evaluation

Validation of the suitability maps was done by cross-checking the suitability map with

existing RWH structures. Global positioning system (GPS) readings were taken on

existing RWH structures and incorporated in the ArcGIS environment for analysis.

During the GPS survey, readings were taken from both successful and failed

interventions.

Figure 9: Flow chart for identification of potential sites.

45

46

4. RESULTS AND DISCUSSIONS

4.1 Survey results

The size of the survey on RWH structures is not large enough to represent all RWH

technology present in the study area. However, most RWH technologies in the area share

the same features, i.e design type, method of implementation, year of construction and

more or less similar social setup and results therefore point at some general success and

failure lessons. A total of randomly selected 30 households with different RWH

technologies were interviewed using a semi-structured questionnaire. Moreover, informal

group discussions were organized with farmers to discuss the various RWH interventions.

Additionally visual observations were made on existing RWH interventions to assess

their current status.

4.1.1 RWH practiced

The field study revealed that different types of RWH systems exist in the study area.

Many RWH systems at household level have been developed with support of the

Government, and especially ponds and concrete tanks can be found in almost every

Kebele. The household (HH) ponds have a trapezoidal shape and are 8-12 meters wide

and 2-3 meters deep. Tanks have a hemispherical shape with a capacity ranging from

approximately 40 m3 to 60 m3 and almost all surveyed tanks are cemented and roofed.

Most of the surveyed farmers with HH ponds and tanks have started to use the technology

since the year 2003/04 during which the government started extensive implementation of

RWH at HH level. None of the surveyed structures were built by farmers’ own capital but

mainly by government fund (MoRAD) and only few by NGOs. Farmers contributed

labour during the construction.

Community managed ponds are present in some Woredas. The community ponds in

Adamitulu woreda are constructed during Hailessilese (at least 35 years ago) and the

Derge regime (about 20 years ago) mainly as a source of drinking water for domestic and

47

livestock. These community ponds were constructed well prior to the current massive

implementation of RWH.

The surveyed RWH schemes mainly harvest runoff from either natural catchment located

adjacent to the ponds or from roads, footpaths and cattle-tracks.

4.1.2 Current state of affairs

According to the surveyed farmers, most of the observed HH ponds were not performing

as intended in terms of storing/retaining harvested runoff. Farmers argued that the poor

performance was caused by high water losses mainly through seepage and evaporation.

Almost all surveyed farmers with unlined ponds reported seepage losses from ponds as

the critical issue. The unlined ponds retain the harvested water for up to one to two

months after the main rainy season (around October). Worst cases were reported by the

development agent of the Bulbula Kebele, Ziway woreda where the area is dominated by

coarse textured soils, most of the harvested water was lost almost immediately after the

rainy season. By contrast, concrete tanks and ponds lined with plastic were found

relatively effective in holding the harvested water for two to three months longer than the

than the unlined ones. Only few farmers reported that the poor performance was

attributed to the improper siting of the ponds which led to poor runoff harvest.

To reduce seepage losses, only few farmers were provided with the plastic sheet

promised by MoRAD. Surveyed farmers strongly suggest that water losses can be

reduced by lining the ponds either with plastic or cement. One innovative and skillful

farmer at Jewe Bofu Kebele with experience in both lined tanks and unlined ponds said

that the lined one could manage to hold the harvested water two to three months longer

than the unlined one. This farmer tried to reduce the seepage with some success in one of

his ponds by lining with what he called a ‘cost-effective method’, by combining cement,

sand and ‘kuyissa’(soil of excavated and piled by termite mounds) with the ratio 1, 2 and

5, respectively (Figure 10).

48

Figure 10: ponds lined by combining cement, sand and ‘kuyissa’

Also improper utilization of the plastic provided to farmers to reduce seepage losses was

observed, e.g. poor handling, use for other purposes, damage by animals and theft. One

farmer reported that his plastics were stolen twice.

Farmers in the Bulbula kebele, Ziway Woreda, indicated during informal discussion that

the high water losses through seepage, which they observed in the pond of the farmer

training center (FTC) for demonstration purpose contributed to their reluctance to adopt

the technology.

4.1.3 Use of harvested water

Though the use of harvested water varied from place to place depending on the household

priority; farmers in general use it for various purposes including drinking water for

animals, watering vegetable and only some use it for watering for trees. Some farmers

use the water for washing cloths, cooking and for making mud blocks. Some farmers use

the water for raising pepper seedling as in Aleaku Gubantaboke Kebele.

49

4.1.4 Productive purpose of RWH

In crop production RWH can serve two purposes, i.e. to raise horticultural seedlings

during the dry period preceding the main growing season and to provide water as

supplementary irrigation whenever there is a shortfall in water during the growing season,

especially near crop maturity (Desta 2004). It is with these above objectives that RWH

has especially been implemented in the study area for production of high value crops like

vegetables, cash crops and fruits.

Farmers indicated that the harvested water was not adequate to meet the crop water

requirements either to mitigate the dry spells or off-season irrigation. Farmers with

unlined ponds outlined that most of the harvested water in ponds is lost through seepage,

while farmers with concrete tanks indicated that the small storage capacity (i.e. 40–60

m3). The low performance of the RWH systems resulted in poor interest for adequate

maintenance which further reduces the already low storage capacity of the structures.

Farmers used to clean concrete tanks till two to three years after construction. One

woman farmer at Edo Kejele Kebele reported that she abandoned her concrete tank built

by the MoARD because the harvested water did not last until the crop was ready to be

harvested despite her tank holds water and is in a good condition with no crack during the

time of survey.

The type of water application method farmers used, applying the water either via unlined

canals or directily applying to the crops using cans, resulted in unnecessary water losses

which further aggravate the insufficiency of the harvested water. To improve the

performance of RWH in terms of water use efficiency, drip irrigation kits have been

promoted by the government. However, none of the surveyed farmers were provided with

the drip irrigation kits.

Because of the problems discussed, most surveyed farmers were disappointed in RWH

and abandoned the technology for supplementary irrigation which in turn has led to low

adoption by non-beneficiary farmers. Most beneficiary farmers have shifted from using

50

the water for supplementary irrigation to other uses like drinking water for animals and

domestic use.

4.1.5 Operation and maintenance

Once the RWH physical structure is in place, it is normally the responsibility of the

owner to carry out operation and maintenance. Many of the surveyed ponds function

below their potential only because the owner failed to accomplish his responsibility. This

is more serious in community owned ponds. The observed problem in this regard is that

ponds and tanks were not repaired and maintained, tanks were not any longer covered

because roofs were stolen (Figure 11) or used for other purposes leaving the tank open

(Figure 12) and ponds were left unprotected allowing livestock to drink directly from the

pond (Figure 13).

The beneficiaries still consider that the one who constructed the system (Government or

NGOs) is responsible to repair and maintain the system. One good example in this regard

is that many of the treadle pumps (used as lifting device) supplied to farmers were

malfunctioning only because farmers were waiting for lubricating oil from the

government or NGOs while it would literary cost them few birr to buy oil from a nearby

town.

Figure 11: The corrugated sheet were stolen

Figure 12 the owner used the corrugated sheet

for other purposes and left the tank open

51

Figure 13: this pond was left unprotected and

livestock was drinking directly from the pond

4.1.6 The best option?

Obviously, RWH is one option to irrigate and produce high value crops to reduce poverty

and food insecurity. The assumption is that producing high value crops enables farmers to

get returns from selling the product and thus increasing the ability of farmers to generate

income. However, form this survey it is not clear under which conditions/scenarios RWH

can contribute to income generation. From the start, farmers were advised to use the

harvested water for supplementary irrigation to produce high value crops like vegetables

so that the family could benefit from selling the product. However, most surveyed

farmers are disappointed in the RWH interventions as their expectations were not

fulfilled. Only few farmers still use the RWH structure to irrigate tiny plots to produce

vegetables for home consumption and to raise pepper seedlings. The main reasons are

that much of the harvested water is lost through seepage and thus it is not sufficient and it

is very labour intensive to irrigate the whole fields by pumping the water manually from

the pond and applying directly to the crop.

There are, however, some positive experiences. For example, in Aleaku Gubantaboke

Kebele (Bulbula Wereda) some farmers have clearly confirmed the importance of RWH

as best option for pepper (seedling) production. Realizing the importance of pepper in

52

terms of cash return, some farmers have already shifted from maize, teff and wheat to

pepper. Three surveyed farmers used the harvested water to raise pepper seedlings which

they later transplant to the main field when the main rainy season starts. They reported a

net return of 10,000 birr per quarter hectare. They also sell seedlings to other non RWH

farmers if they have more seedlings than they need for themselves. They get 150 to 200

birr per seedling bed with a size of approximately 1 by 1.5 m. Pepper production is only

possible when there is rain in April to raise seedlings, otherwise famers shift to either

maize, teff or wheat. It is often rare for the area to have rain in April according to the

surveyed farmers. During the survey year (2009), too, there was not sufficient rain in

April as confirmed during the survey. Thus RWH may play a role by storing water till

April either the rainwater that came during last rainy season (implying a storage period of

six month) or capture the early rains in April (not frequently happening). The problem,

however, is that the ponds with water from the last rainy season are generally dried up

when it is needed for preparing the seedling. The problem is aggravated by farmers’ use

of the harvested water for other purpose (washing clothes, cattle drinking etc).

One industrious farmer from Jewe Bofu Kebele has devised an exemplary scenario. He

uses the harvested water for producing ‘Gesho’ (Rhamnus Pridoides), Pigen pea

(Cajanus cajan) and Alfalfa (Medicago sativa) seed. He gets a lucrative income from

selling Gesho by taking it directly to the market and the Alfalfa seed to NGOs. He got

500 birr from a kilo of Alfalfa seed. He also uses the water for sheep fattening and

poultry. But the two plastic sheets to cover his pond which were denoted by the Woreda

Agricultural office and NGO were stolen. He is now using his two RWH structures for

some time after the rainy season is stopped and both his RWH structures were almost

empty at the time of survey (Figure 14 and Figure 10).

Therefore, finding options to avoid/minimize water loss through seepage and evaporation

and encouraging farmers to develop innovative uses of the harvested water such as the

described examples should be the way forward to make better use of the the potentials

that RWH provide.

53

Figure 14: pond at Edo kejele kebele with no

water

Figure 15: RWH used for raising pepper seedling

at Bulbula Wereda, Aleaku Gubantaboke kebele

4.1.7 Lifting mechanism

Harvesting and storing the rainwater is not sufficient if the harvested water is meant for crop

production. The harvested water should be pumped from the storage to make it available for

the crop. Most surveyed RWH farmers were provided with a foot operated treadle pump by

MoRAD and different NGOs, to pump the water from ponds and tanks. The water is applied

either via unlined furrows to the crop or pump the water to small surface depression and

then directly apply the water to the crop by cans. However, much of the treadle pumps are

not functioning due to poor manufacturing and shortage of lubricants. Those farmers who

still use RWH structure are forced to use hand watering using old cans and applying the

water directly to the crop. This type of water application as reported by the farmers is

laborious and time-consuming. Moreover, it is inefficient in terms of using the scarce water

with much loss which consequently led to poor water management. Few farmers use chain

and washer type of pumps.

4.1.8 Farmers’ involvement during implementation

During the recent extensive implementation program of RWH, farmers were not consulted

on their needs and preferences and about the benefit to be generated. During field surveys,

farmers were seen short of proper understanding of the technology and some farmers

54

became skeptic regarding the benefits they could generate using RWH. Moreover, the

pressure from the implementers’ side to adopt the technology further aggravates the

farmers’ reluctance towards the technology. The pressure to use RWH resulted in a low

motivation of farmers to use RWH despite some of structures (most tanks) were in good

condition during the time of visit. For instance one innovative woman farmer was told by

Woreda agricultural office to construct a tank and she was provided with the necessary

material like cement, corrugated sheet and wood. She used the structure for three years to

produce vegetables like cabbage and onion and also to water trees. However, she lost

interest since a nearby farmer was provided with a shallow well by MoARD. She prefers a

well over a RWH tank because its supply of water is more reliable. Therefore, farmers’

preference must be assessed and farmers should thoroughly be told the befits as the success

often depends on it.

4.1.9 Community ponds

Community managed ponds are common in some Woredas such as Bulbula, Suro, Edo

Kejela, Zeleku Golanita Kebele. The community ponds in Suro and Zeleku Golanita Kebele

were constructed during Hailessilese (at least 35 years ago) and Derge regime (about 20

years ago) and the other were constructed around four years ago mainly for drinking water

of livestock and people. Most of them have a serious siltation problem because of improper

operation and maintenance.

Two nearby community ponds in Suro Kebele, one for domestic use and the other for

livestock, suffered both from a serious siltation problem. The community used to clean the

ponds at least every year; however they stopped around 10 years ago. This has led to almost

complete siltation of the pond for livestock. The silt traps provided with the ponds are not

serving any longer because they are too small in relation to the amount of runoff and easily

fill up with silt. In such community ponds equal importance in the design and

implementation must be given to the socioeconomics and physical structure. As well-

functioning of such ponds requires full participation of the entire community and thus strong

community organization is required to mobilize labour for operation and maintenance. In

55

general, farmers are easily mobilized for the initial construction as most felt the problem and

being excited to be relived from the problem. However, it is difficult to motivate and

mobilize farmers for operational and maintenance work. Thorough understanding of the

socioeconomics of the community is required to device mechanisms and instruments to

organize communities (local association) which will enable the sustainable management of

such ponds.

4.2 Identification of potential RWH areas in the CRV

The process of identifying suitable RWH was implemented in the ArcGIS model

environment using the model builder of ArcGIS 9.3. The suitability model generated

suitability map for RWH by integrating different input criteria maps using MCE. Different

spatial analysis tools were incorporated in the model to solve various spatial problems in the

process of identifying suitable areas. The identification process in this study was considered

as a multi-objective and multi-criteria problem.

The survey results and literature review indicated that most of the constructed ponds have

considerable seepage problems. In addition, most surveyed ponds could be filled with two or

three showers confirming that the catchment is no problem. Therefore, it was concluded

that, soil texture being responsible for seepage, was more important than the other criteria

which in turn result in a higher weight for soil texture.

The suitability model generated two suitability maps for ponds and in-situ RWH each with

five suitability classes, i.e. Very high, High, Moderate, Low and Very low suitability. The

spatial distribution of the suitability map for ponds (Figure 22) showed that the eastern and

western part of the study area dominated by cultivated land cover types is in very high (red)

and high (green) suitability category. The central and northern part is mostly moderate

suitable whereas the northwestern and some central part surrounding Lake Shala and

Abiyata has a low suitability. According to their acreage, 4 and 44% of the study area has

very high and high suitability for ponds, respectively, while 25, 18 and 1% of the area has a

moderately, low and very low suitability, respectively (Figure 16). The majority of the areas

56

with very high to high suitability have slopes between 2 and 8% and with an intensively

cultivated land cover. The major soil type in the very high and high suitable area is clay,

clay loam and loam with fine and medium texture and the rainfall ranges from 800 up to

1000 mm.

On the in-situ side, the spatial distribution of potential areas (Figure 23) showed that there is

no any very high suitable area whereas high suitability occurred in most part of the study

area except southern and western part. The northern and southern part is mostly moderately

suitable whereas the central part has a low suitability at few places. The area with very low

suitability for in-situ RWH can be neglected. According to their acreage, 60% of the study

area has a high suitability followed by 32 and 1% of the area with a moderate and low

suitability, respectively (Figure 17). The majority of the areas with high suitability have

slopes between 2 and 8% and the land cover is intensively and moderately cultivated. The

major soil type in this area is sandy loam, clay loam and loam with fine to medium and

medium to coarse texture and the rainfall ranges from 700 up to 1000 mm.

Pond

Suitability level

Very low Low Moderate High Very high

% o

f the

tota

l are

a

0

10

20

30

40

50

60

Figure 16: Percent of the study area per each suitability

level for pond

In-situ

Suitability level

Very low Low Moderate High Very high

% o

f the

tota

l are

a

0

10

20

30

40

50

60

Figure 17: Percent of the study area per each

suitability level for in-situ

One of the criteria for identifying potential areas for RWH was land cover type. Not all land

cover types are suited for RWH as RWH system meant for crop production should be close

to agricultural areas. It was with this reason that cultivated lands (intensive and moderate)

57

were given higher ranks than the other land cover types. This resulted in higher percentage

of very high to high suitability level to be under intensively and moderately cultivated areas

for both pond and in-situ RWH. The classification of suitability level per land cover type

expressed as a percentage of each land cover type is presented in Figure 18 and Figure 19

for pond and in-situ respectively.

To identify which Woreda is more suited for RWH, the pond and in-situ suitability map was

overlayed with the Woreda boundary map. The majority of the very high and high suitable

areas for pond RWH are in Degeluna Tijo, Tiyo, Munessa, Limuna Bilbilo, Ziway dugeda

and Meskan Woreda sharing 21, 16, 15, 13, 9 and 9% of the total very high to high suitable

areas , respectively. Kondaltiti, Kokir Gedbano Gutazer, Ezha, Akililna Mohr, Adami Tulu

Jido Kombolcha, Gumer, Seden Sodo, Dalocha and Alicho Woriro Woredas have almost no

land which is very high to high suitable for RWH ponds (

58

Table 10 and Figure 20). With respect to suitability level within the Woreda, the majority of

the land of Tiyo (99%), Hitosa (93%), Degeluna Tijo (93%), Limuna Bilbilo (85%) and Silti

(83%) Woreda are very high to highly suitable followed by Kofele (71%), Meskan (69%),

Lanfero (66%), Mareko (63%), Munessa (60%) and Alicho Woriro (58%) (

59

Table 10 and Figure 21).

The majority of very high and high suitable areas for in-situ RWH are in Degeluna Tijo,

Ziway Dugda, Adami Tulu Jido Kombolcha, Tiyo and Limuna Bilbilo Woreda sharing 18,

15, 14, 12 and 11% of the total very high to high suitable areas, respectively. Alicho Woriro,

Seden Sodo, Dalocha, Akililna Mohr, Ezha, Gumer, Kokir Gedbano Gutazer and Kondaltiti

Woredas have almost no land classified as very high to highly suitable (

60

Table 10 and Figure 20). With respect to suitability level within the Woreda, the majority of

the land of Gumer (100%), Tiyo (99%), Hitosa (94%), Degeluna Tijo (98%), Hitosa (91%)

and Limuna Bilbilo (90%) Woreda is very high to highly suitable for in-situ RWH followed

by Ziway Dugda (75%), Mareko (64%), Dugda(64%), Alicho Woriro (61%), Adami Tulu

Jido Kombolcha (61%), Lanfero (59%) and Kofele (55%) (

61

Table 10 and Figure 21).

The very high and high suitability for both ponds and in-situ RWH occurred together in 39%

of the study area mainly in eastern and southern part.

62

Land cover types

Shrublan

d

Moder

ately

Cul

ti

Inte

nsive

ly Cu

lti

Grass

land

Marshlan

d

Fores

t

Expos

ed S

urfac

e

Woo

dland

Afro-A

lpin

e

% o

f sui

tabi

lity

leve

l

0

20

40

60

80

100

Very lowLowModerateHighVery high

Figure 18: Distribution of pond suitability level per land cover type

Land cover types

Shrub

land

Ž�de

ratel

y Cult

i

Grass

land

Marsh

land

Forest

Expos

ed S

urface

Woo

dland

Afro-A

lpine

% o

f sui

tabi

lity

leve

l

0

20

40

60

80

100

Very lowLowModerateHigh

Very High

Figure 19: Distribution of in-situ suitability level per land cover type

63

Table 10: Very high and high suitability level per Woreda

Pond RWH In-situ RWH Woerda name Area

(km2) Very high and high suitability area (km2)

% of Woreda area

% of total very high and high suitability

Very high and high suitability area (km2)

% Woreda area

% of total very high and high suitability

Kofele 257 183 71 4 141 55 2 Limuna 630 536 85 11 568 90 9 Arsi Negele 1282 156 12 3 455 35 8 Munessa 1044 627 60 13 484 46 8 Dalocha 12 5 45 0 3 26 0 Degeluna Tijo 942 872 93 18 922 98 15 Lanfero 139 91 66 2 82 59 1 Tiyo 649 645 99 13 644 99 11 Silti 343 285 83 6 129 38 2 Adami T. Jido 1153 0 0 0 707 61 12 Gumer 0 0 0 0 0 100 0 Alicho Woriro 32 19 58 0 20 61 0 Mareko 323 202 63 4 207 64 3 Hitosa 245 229 93 5 226 91 4 Meskan 547 375 69 8 190 35 3 Ezha 5 0 0 0 0 2 0 Akililna Mohr 19 0 0 0 0 3 0 Ziway Dugda 1071 380 35 8 801 75 13 Dugda 724 111 15 2 461 64 8 Kokir 17 0 0 0 0 0 0 Sodo 735 265 36 5 92 13 2 Kondaltiti 40 0 0 0 0 0 0 Seden Sodo 57 13 23 0 11 20 0

Woreda

Kofele

Limun

a Bilb

ilo

Ars i N

egel

e

Mun

essa

Daloch

a

Degelu

na T

ijo

Lanfer

oTi

yo Silti

Adami T

. Jid

o K.

Gum

er

Alich

o W

oriro

Mar

eko

Hitosa

Mes

kan

Ezha

Akilil

na M

ohr

Ziway

Dugda

Dugda

Kokir

G. G

utaz

er

Sodo

Konda

ltiti

Seden

Sod

o

% fr

om th

e to

tal s

tdy

area

0

2

4

6

8

10

12

14

16

18

20

PondIn-situ

Figure 20: Percent of very high and high suitability per Woreda area.

64

Woreda

Kofele

Limun

a B

ilbilo

Arsi N

egele

Mun

essa

Daloch

a

Degelu

na T

ijo

Lanfe

roTiyo Sil t

i

Adam

i T.

J ido K

.

Gumer

Alich

o W

oriro

Marek

o

Hitosa

Meska

nEzh

a

Akililn

a Moh

r

Ziway

Dug

da

Dugda

Kokir G

. Guta

zer

Sodo

Konda

ltiti

Seden

Sod

o

% fr

om W

ored

a ar

ea

0

20

40

60

80

100

120

PondIn-situ

Figure 21: Percent of very high to high suitability for each Woreda per total very high and high suitability

area.

Since in most areas of the study area in-situ RWH is not practiced, criteria weight

assignment for in-situ was not supported with information from the field survey and only

based on the literature. Therefore another suitability map was generated with criteria

given equal percent of influence (Figure 24). This suitability map shows that 1 and 65%

of the study areas have very high and high suitability, respectively, while 24, 3 and

almost 0% of the study area has a moderate, low and very low suitability, respectively. In

general the same areas that appear as suitable for in-situ RWH in Figure 23 (using

different weight for the criteria) also appear in Figure 24 (with equal weight for the

criteria).

65

Figure 22: Pond suitability map

Figure 23: In-situ suitability map

Figure 24: In-situ suitability map with criteria given equal weights

4.3 Validation

Validation of the surveyed RWH was done using information obtained from the field

survey and the generated suitability map. The validation consisted of comparing the

generated suitability map and the location of the surveyed RWH structures using

proximity analysis tool of ArcGIS 9.3. For the purpose of the validation, the surveyed

RWH were first rated in two categories; successful and unsuccessful. From the result of

the field survey, most of the surveyed RWH structures were proved as having various

66

problems and categorize in the unsuccessful category. The assumption made during

validation was that if these RWH structures which were categorized as unsuccessful

category is found in non suitable areas in the derived suitability map, the result from the

suitability model would be proofed as good.

From the proximity analysis result, most of exiting RWH structures categorized as

unsuccessful (53%) were within the moderately suitable areas followed by low suitable

(43%). Only 3 % are with highly suitable areas. The fact that most of exiting RWH

structures categorized as unsuccessful are not found in the very high and high suitability

level in the derived suitability map indicated that the generated suitability map indeed

identified reliably the potential areas for RWH technologies. The validation results

showed that the database and methodology used for developing the suitability model

including the suitability levels of the criteria and the criteria’s relative importance

weights have given good results.

67

5. CONCLUSIONS AND RECOMMENDATIONS

The adoption and widespread replication of any RWH is extremely risky without any ex-

ante assessment or screening of the possible effects under the new physical conditions.

The failure or poor performance of many RWH structures in Ethiopia and in the study

area in particular can be partly explained by the improper selection of intervention areas.

A number of manuals for RWH implementation are available to assist in site selection

and the type and design of RWH in relation to the physical criteria. This has given

important lesson that suitability assessment should be given more attention. Maybe due to

the hasty implementation of many RWH interventions, often driven by the need to satisfy

Woreda specific RWH quota, the guidelines of such manuals are not followed. This has

been confirmed by the field survey and literature review. The results are an improper site

selection locating ponds in areas with less suitable soils, improper design (only one type

of tanks and ponds are constructed everywhere), lack of farmer involvement during the

planning and implementation, lack of materials like plastic sheet to reduce seepage in

ponds and lack of maintenance. This report, however would like to stress that the RWH

technology by itself may be the suitable but that the bottleneck is the planning,

implementation and management. RWH has brought many success stories from some

parts of the country (Chala et al. 2004; Rämi 2003) and in other parts of the world (Li et

al. 2000; Ngigia et al. 2005). The government and NGO’s appear to stimulate RWH with

unbelievable speed but forgot that proper interventions need time for location-specific ex-

ante assessments, consultation with of stakeholders and training of farmers in

management and governance (especially for community based ponds).

Utilizing runoff in an efficient and sustainable manner is crucial to improve the

performance of rainfed farming in the CRV. RWH is one option to use runoff better by

capturing and storing when rainfall is abundant for periods when water is scarce.

Providing information on the suitability of areas for different types of RWH interventions

is an important step prior to the actual planning and implementation. Identifying potential

areas for RHW requires spatial knowledge on a number of critical physical factors such

as soil, climate, topography and landuse. In this study the identification of potential areas

68

was done using GIS-based suitability model using the ArcGIS 9.3 model builder. The

suitability model used a MCE process that combined different biophysical factors: soil

texture, soil depth, climate, slope, land cover and groundwater depth. However,

socioeconomic factors (e.g. market access, infrastructure, population density) which are

also necessary for a complete assessment of the suitability of land for RWH were not

considered due to lack of readily available data for this large area. It is therefore

recommended to include such socio-economic factors in future studies to improve the

suitability assessment.

The suitability model generated two suitability maps; one for ponds and one for in-situ

RWH. Based on the pond suitability map, 4, 44, 25, 18 and 1% of the study area has very

high, high, moderate, low and very low suitability, respectively. For in-situ, 60, 32 and

1% of the study area has high, moderate and low suitability, respectively. The result from

the suitability model was validated using information obtained from field survey. The

validation results showed that the database and methodology used for developing the

suitability model including the suitability levels the criteria and the criteria’s relative

importance weights have given good results.

In this study, GIS was proved to be a flexible, time-saving and cost-effective tool to

screen large areas for their suitability of two types of RWH intervention. The suitability

maps provide an easy to understand source of information to quickly identify areas that

are more promising than other areas for RWH intervention. Such information is helpful

for decision-makers and planners, but one should be careful in the interpretation of the

generated information. Actual RWH implementation should always be preceded by a

field survey as the spatial resolution of the analysis does not guarantee that every location

in an area classified as highly suitable is indeed highly suitable for RWH. Vice versa, also

in low suitable areas there may be spots that can be suitable for RWH. Apart from this,

on-the-ground work is also needed for getting the socio-institutional setting of the area to

help complete the work as indicated before. The analysis as presented, however, provides

a first valuable screening of large areas and can be modified easily to incorporate other

criteria or information with other spatial resolutions.

69

REFERENCES

Alamerew E (2006) Towards widespread rainwater harvesting in Ethiopia: ERHA’s experience as a successful Rainwater Harvesting Capacity Centre (RHCC) In '10th Southern and Eastern Africa Rainwater Network (SearNet) International '. Mombasa,Kenya. (Ed. KG Sectetariat) p. 90.

Alem G (1999) Rainwater Harvesting in Ethiopia: An Overview. In 'Integerated

Development for Water Supply and Sanitation'. Addis Ababa. Amha R (2006) Impact Assessment of Rainwater Harvesting Ponds:The Case of Alaba

Woreda, Ethiopia. M.Sc thesis, Addis Ababa university Barron J, Okwach G (2005 ) Run-off water harvesting for dry spell mitigationin maize

(Zea mays L.): results from on-farm research in semi-arid Kenya. Agricultural Water Management 74, 1-21.

Bayu W, Alemu G (1998) Soil Management and Crop Technologies for improved Crop

production in marginal rainfall areas of Wello. In 'Agricultural Research and Technology Transfer attempts and achievements in Northern Ethiopia. ' (Eds B Seboka, A Deressa)Bahir Dar, Ethiopia).

Burrough P (1986) 'Principles of Geographical Information Systems for Land Resources

Assessment.' (Clarendon Press: Oxford). Chala A, Demisie A, et al. (2004) Evaluation Report on Benfits of Water Harvesting

Schemes Implemenetes in Nine Zones Of Oromia Region. Oromia Irrigation Developmenet Authority.

Chala A, Etissa E, Lemecha G, Hunde H, Desta L, Jateno W, Ayele W (2003)

Progressive Evaluation on Water Harvesting Schemes Implementation in Oromia Region(part I). Rural and Agricultural Development supreme Office.

Critchley W, Siegert K, Chapman C (1991) A Manual for the Design and Construction of

Water Harvesting Schemes for Plant Production In 'AGL/MISC/17/91, FAO'. (FAO: Rome).

Desta L (2004) Concepts of Rainwater Harvesting and its Role in Food Security – The

Ethiopian Experience. Paper presented on a National Water Forum. Addis Ababa. Desta L (2006) Keynote Speech. In 'Workshop on the Experience of Water Harvesting

Technology in East Shewa and Arsi Zones'. Melkassa. (Eds Y Kedir, K Shiratori) p. 67.

70

Dubale P (2003) Present and future trends in natural resources management in agriculture: An overview. In 'Integrated water and land managment research and capacity building priorities for Ethiopia'. Addis Ababa. (Eds M P.G., K A.B., G Tadesse).

EARO (2000) Dry land research strategy document. (Ed. E Ethiopia Agricultural

Research Organization. Addis Ababa). Fox P, Rockstrijm J, Barron J (2005) Risk analysis and economic viability of water

harvesting for supplemental irrigation in semi-arid Burkina Faso and Kenya. Agricultural Systems 83, 231-250.

Fox P, Rockström J (2000) Water-Harvesting for Supplementary Irrigation of Cereal

Crops to Overcome Intra-Seasonal Dry-Spells in the Sahel. Phys. Chem. Earth 25, 289-296.

Gebre H, Giorgis K (1980) Sustaining crop production in the areas of Ethiopia. Ethiopia

J. Agric. Sci 10, 99-107. Hatibu N, Mahoo H (1999) Rainwater harvesting technologies for agricultural

production: A case for Dodoma, Tanzania. Sokoine University of Agriculture Department of Agricultural Engineering and Land Planning, Morogoro, Tanzania.

Hatibu N, Senkondo EM, Mutabazi K, Msangi ASK (2004) Economics of Rainwater

Harvesting for Crop Enterprises in Semi-Arid Areas. In 'New directions for a diverse planet: Proceedings of the 4th International Crop Science Congress'. Brisbane, Australia.

Jansen H (2009) Personal Communication. Kahinda JM, Lillie ESB, Taigbenu AE, Taute M, Boroto RJ (2008) Developing

suitability maps for rainwater harvesting in South Africa. Physics and Chemistry of the Earth 33 788-799.

Kedir Y (2009) Personal Communication. Li F, Cook S, Geballe GT, Jr WRB (2000) Rainwater Harvesting Agriculture: An

Integrated System for Water Management on Rainfed Land in China’s Semiarid Areas. Ambino 29.

Mati B, Bock TD, Malesu M, Khaka E, Oduor A, Nyabenge M, Oduor V (2006)

Mapping the Potential of Rainwater Harvesting Technologies in Africa: A GIS overview on development domains for the continent and ten selected countries p. 116. ( World Agroforestry Centre (ICRAF), Netherlands Ministry of Foreign Affairs.: Nairobi, Kenya).

71

Mbilinyi BP, Tumbo SD, Mahoo HF, Senkondo EM, Hatibu N (2005) Indigenous knowledge as decision support toolin rainwater harvesting. Physics and Chemistry of the Earth 30 792-798.

Mesfin T (2004) Effect of In-Situ Water Harvesting on the Growth, Yield and Water Use

Efficiency of Sorghum (Sorghum bicolor (L.) Moench) Under Semi-Arid Conditions of Ethiopia. Alemaya University.

Nasr M (1999) Assessing Desertification and Water Harvesting in the Middle East and

North Africa: Policy Implications, ZEF – Discussion Papers On Development Policy No. 10, . Center for Development Research, Bonn.

Ngigia SN, Savenije HHG, Thome JN, Rockstrom J, Vries FWTPd (2005) Agro-

hydrological evaluation of on-farm rainwater storage systems for supplemental irrigation in Laikipia district, Kenya. Agricultural Water Management 73 21-41.

Oweis TY (2004) Rainwater Harvesting for Alleviating Water Scarcity in the Drier

Environments of West Asia and North Africa. In 'For presentation at the International Workshop on Water Harvesting and Sustainable Agriculture'Moscow, Russia).

Oweis TY, Taimeh AY (1996) Evaluation of a Small Basin Water-Harvesting System in

The Arid Region of Jordan. Water Resources Management 10, 21-34. Prinz D, Oweis T, Oberle A (1998) Rainwater Harvesting for Dry Land Agriculture -

Developing a Methodology Based on Remote Sensing and GIS. In 'Proceedings, XIII International Congress Agricultural Engineering'. Rabat, Morocco.

Profile BIP (2005) Water Harvesting and Small–Scale Irrigation p. 47. (GOVERNMENT

OF THE FEDERAL DEMOCRATIC REPUBLIC OF ETHIOPIA. Ramakrishnan D, Rao KHVD, Tiwari KC (2008) Delineation of potential sites for water

harvesting structures through remote sensing and GIS techniques: a case study of Kali watershed, Gujarat, India. Geocarto International 23, 95-108.

Rämi H (2003) Ponds filled with challenges. Water harvesting - experiences in Amhara

and Tigray. Rao KHV, Durga, Bhaumik MK (2003) Spatial Expert Support System in Selecting

Suitable Sites for Water Harvesting Structures — A Case Study of Song Watershed, Uttaranchal, India. Geocarto International 18, 43 - 50.

Rockström J (2000) Water Resources Management In Smallholder Farms In Eastern and

Southern Africa : An Overview. Phys. Chem. Earth (B) 25, 275 - 283.

72

Saaty TL (1977) A Scaling Method for Priorities in Hierarchical Structures. Mathematical Psychology 15, 234-281.

Segers K, Dessein J, Nyssen J, Hailec M, Deckers J (2007) Wanted by the state: farmers

making headway - a case study in Tigray, northern Ethiopia. In 'States at work: African public services in comparative perspective'. African Studies Centre, Leiden, The Netherlands. (AEGIS European Conference on African Studies).

Senay GB, Verdin JP (2004) Developing Index Maps Of Water−Harvest Potential In

Africa. Applied Engineering in Agriculture 20, 789−799. Seyoum M (2003) Overview of Ethiopia Rainwater Harvesting Association(ERHA). In

'Integrated water and land managment research and capacity building priorities for Ethiopia'. Addis Ababa, Ethiopia. (Eds M P.G., K A.B., G Tadesse) p. 267.

Sileshi Z, Tegegne A, Tsadik GT (2003) Water resources for livestock in Ethiopia:

Implications for research and development. In 'Integrated water and land management research and capacity building priorities for Ethiopia.' Addis Ababa, Ethiopia. (Eds M P.G., K A.B., G Tadesse) p. 267.

Sivanappan RK (2006) Rain Water Harvesting, Conservation and Management Strategies

for Urban and Rural Sectors. In 'National Seminar on Rainwater Harvesting and Water Management 11-12'Nagpur, India).

Soriano JF (2007) Financial Investment Appraisal of Shallow Wells in Dugda Woreda,

Ethiopia. MSc thesis, Cranfield University. Temesgen M (2007) Conservation Tillage Systems and Water Productivity Implications

Smallholder Farmers in Semi-arid Ethiopia., Delft University of Technolog. W/Giorgis K (2002) Water Harvesting, the key to successful crop production and

resource management in Water Stress Environments: Eastern African Experience. Mekelle, Ethiopia.

Winnaar Gd, Jewitt GPW, Horan M (2007) A GIS-based approach for identifying

potential runoff harvesting sites in the Thukela River basin, South Africa Physics and Chemistry of the Earth 32, 1058-1067.

73

APPENDICES

I. Questionnaire on assessments of Rain Water Harvesting (RWH) in CRV

1. General Information

1.1. Date of interview: ______________________________ 1.2. Name of the farmer: ____________________________ 1.3. Village_________________ Woreda________________ 1.4. Administrative Zone________________ 1.5. Altitude:___________Location (Coordinate):__________

2. Household characteristics

2.1. Household head:

2.1.1. Male____ Female______ 2.1.2.Age______ 2.1.3 Level of formal education:____________ 2.1.4. Number of household members__________ 2.1.5 Number of adults (> 17 Yr.):_______ Children_________

2.1. Total farm size___________ 2.2. Land tenure of farm Own_____ Family_______ Own/ Family_____ Share cropping______ Rent______ Borrow______

74

3. Water harvesting practices and status

3.1. What type of water harvesting structure do you have? _____________________ 3.2. By whom it was constructed? ________________________________________ 3.3 When did you start to use the structure?_________________________________ 3.4 For what purpose do you use the structures?

Crop production ________________

Animal fattening _______________

Drinking water ________________

Others _______________________

3.4 Size of the structure (appx.)?_____________________________ 3.5 Have you experienced problem while you use it? Yes:____ No:____

If yes, The Problems ________________________________________________ ____________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

3.7 Do you have success story?

Yes:____ No:____ If yes ______________________________________________________________

________________________________________________________________________________________________________________________________________________________________________________________________________________________ 3.8 Is the harvested water is sufficient for supplementary irrigation?_________________

Yes:____ No:____

If No_______________________________________________________________ _______________________________________________________________

3.9 What type of water lifting mechanism you used? ____________________________

75

3.10 Initial investment cost(if you know)? _____________________________________ 3.11 Have you ever made maintenance work?

Yes:____ No:____ If Yes how what type and how often ____________________________________ ___________________________________________________________________

If No why?_______________________________________________________ ___________________________________________________________________

3.12 Any maintenance cost you incurred? ______________________________________ 3.12 Do you experience any difficulties to market your products? 3.13 General remark __________________________________________________________ ________________________________________________________________________________________________________________________________________________________________________________________________________________________

76

II. Suitability Model A/ Soil texture

77

B/ Soil depth

78

C/ Slope

79

D/ Rainfall surplus

80

E/ Land cover

81

F/ Groundwater depth

82

G/ Overlay operation for ponds

83

H/ Overlay operation for in-situ