irrigation potential kenya

FutureWater

Costerweg 1G

6702 AA Wageningen

The Netherlands

+31 (0)317 460050

[email protected]

www.futurewater.nl

Assessment of the Irrigation Potential in

Burundi, Eastern DRC, Kenya, Rwanda,

Southern Sudan, Tanzania and Uganda

Inception Report

June 2011

Client

Nile Basin Initiative

NELSAP Regional Agricultural Trade and Productivity Project

Authors

P. Droogers

W. Terink

J. Brandsma

W.W. Immerzeel

Report FutureWater: 100

2

Table of contents

1 Introduction 7

1.1 Contract Details 7

1.2 Inception Report 7

1.3 Acknowledgements 7

2 General Information 8

2.1 Background 8

2.1.1 Nile Basin Initiative 8

2.1.2 The Nile Equatorial Lakes Subsidiary Action Program (NELSAP) 8

2.1.3 Regional Agricultural Trade & Productivity (RATP) Project 8

2.2 Project objectives 9

3 Research Area and Main Issues 10

3.1 Overview, 10

3.2 Irrigation potential 11

4 Data and Tools 17

4.1 Data and Interpretation 17

4.1.1 Current studies 17

4.1.2 Climate 18

4.1.3 Streamflow 20

4.1.4 Digital Elevation Models 22

4.1.5 Soils 23

4.1.6 Land cover 24

4.1.7 Vegetation indices 26

4.1.8 Evapotranspiration 27

4.1.9 Groundwater 28

4.1.10 Large scale irrigation practices in the Nile basin 28

4.1.11 Other data sources 30

4.2 Water resources assessment: PCRaster 31

4.3 Irrigation water requirements and potential crop yields 31

5 Methodology 33

5.1 Overview 33

5.2 Work Packages 34

5.3 Phase 1 36

5.3.1 Collection and review of existing information 36

5.3.2 Land suitability assessment 37

5.3.3 Water resources assessment 38

5.3.4 Assessment of irrigation water requirements 39

5.3.5 Potential crop yield assessment 40

5.3.6 Environmental, socio-economics 40

5.3.7 Institutional and legal framework 41

5.3.8 Integration 41

5.4 Phase 2 42

5.4.1 Collection and review of existing information 42

5.4.2 Land suitability assessment 42

3

5.4.3 Water resources assessment 43

5.4.4 Assessment of irrigation water requirements 43

5.4.5 Potential crop yield assessment 44

5.4.6 Environmental, socio-economics 44

5.4.7 Institutional and legal framework 45

6 Staffing and Management 46

6.1 Consultant Staff 46

6.2 Management and meetings 49

6.3 Travel 49

7 Deliverables 51

7.1 Overall 51

7.2 Draft Report Phase 1 51

7.3 Draft Report Phase 2 52

7.4 Final Report 53

8 References 54

9 APPENDIX: Literature Review 57

9.1 Introduction 57

9.2 Irrigation related studies 57

9.3 Other relevant studies 63

9.4 Country specific studies 67

9.4.1 Burundi 67

9.4.2 Eastern DRC 67

9.4.3 Kenya 67

9.4.4 Rwanda 68

9.4.5 Sudan 69

9.4.6 Tanzania 69

9.4.7 Uganda 70

10 Appendix: Summary Countries 72

10.1 Burundi 72

10.1.1 General 72

10.1.2 Socio-economy 72

10.1.3 Relief, climate, and hydrography 72

10.1.4 Main crops and land use 73

10.1.5 Agriculture 73

10.1.6 Summarized facts 77

10.2 Eastern DRC 78

10.2.1 General 78


10.2.3 Climate 78

10.2.4 Agriculture and main crops 79

10.2.5 Summarized Facts 81

10.3 Kenya 82

10.3.1 General 82


10.3.3 Climate and hydrography 82



4

10.4 Rwanda 87

10.4.1 General 87



10.4.4 Agriculture, land use, and main crops 88


10.5 Sudan 91

10.5.1 General 91





10.6 Tanzania 95

10.6.1 General 95





10.7 Uganda 99

10.7.1 General 99





11 Appendix: PCRaster-NELmod 103

11.1 Hydrological modeling 103

11.2 PC-Raster 104

11.2.1 Introduction 104

11.2.2 Previous applications 105

11.2.3 Discretization 106

11.2.4 Model concept 106

11.2.5 Model data 111

11.3 Conclusions 112

12 Appendix: Facilitator Contract 113

13 Appendix: Data Form 115

14 Appendix: Minutes Validation Workshop 118

15 Appendix: Local expert selection 127

5

Tables Table 1. World’s Major River Systems2 ....................................................................................10

Table 2. Countries in Nile Basin ..............................................................................................11

Table 3. Breakdown of the consumptive use in the irrigation sector by country (Source: WaterWatch, 2009). ................................................................................................................29

Table 4. Actually irrigated areas in the Nile Basin according to different sources (Source: WaterWatch, 2009). ................................................................................................................29

Table 5: Soil and terrain suitability for surface irrigation by country. .........................................59

Table 6: Nile basin, irrigation potential, water requirements, water availability and areas under irrigation ..................................................................................................................................60

Table 7: Overview of irrigable and irrigated areas in Burundi (Niyongabo, 2007). .....................74

Table 8: Area equipped for irrigation in Burundi. ......................................................................77

Table 9: Agro-climatic zones in DRC (Iessime, 2007). .............................................................79

Table 10: Area equipped for irrigation in DR Congo. ................................................................81

Table 11: Agricultural land sizes in Kenya (Isaya, 2007). .........................................................84

Table 12: Area equipped for irrigation in Kenya. ......................................................................86

Table 13: Agro-climatic zones in Rwanda (AQUASTAT, 2005). ...............................................87

Table 14: Area equipped for irrigation in Rwanda. ...................................................................90

Table 15: Annual projection of water consumption (in BCM) in Sudan in 2003 (Salih, 2007). ...91

Table 16: Total arable land by farming sector in Sudan (Dawelbeit, 2008). ..............................92

Table 17: Area equipped for irrigation in Sudan. ......................................................................94

Table 18: Area equipped for irrigation in Tanzania. ..................................................................98

Table 19: Area equipped for irrigation in Uganda. ..................................................................102

Figures Figure 1: Overview Nile Basin. ................................................................................................13

Figure 2: Overview study area. ................................................................................................14

Figure 3: Sub-basins in NELarea. ............................................................................................15

Figure 4: Rainfall estimate obtained from FEWS-NET (24/11/2000). ........................................19

Figure 5: Overview of UNH discharge stations in the study area (http://www.grdc.sr.unh.edu/html/Stn/B2.html). ........................................................................21

Figure 6: SRTM 90 m Digital Elevation Data of the world. ........................................................22

Figure 7: Harmonized World Soil Database (HWSD). ..............................................................24

Figure 8: Two phases approach in the project and associated Work Packages. .......................33

Figure 9. Time planning. (T) indicates travel to the region (note that one travel can serve more than one Work Package, details in section 6.3). ......................................................................36

Figure 10: Assessment of irrigation potential. ..........................................................................58

Figure 11: Internal renewable water resources by country (in km3). .........................................59

Figure 12: Schematic balance of Lake Victoria, Kyogo, and Albert (km3/year) (Source: Sutcliffe and Parks, 1999).....................................................................................................................62

Figure 13: Map of Burundi with the Nile basin. .........................................................................76

Figure 14: Agricultural area and arable land in Burundi. ...........................................................77

Figure 15: Agricultural production in Burundi. ..........................................................................77

6

Figure 16: Map of Eastern DRC with the Nile basin. ................................................................80

Figure 17: Agricultural area and arable land in DR Congo. ......................................................81

Figure 18: Agricultural production in DR Congo. ......................................................................81

Figure 19: Map of Kenya with the Nile basin. ...........................................................................85

Figure 20: Agricultural area and arable land in Kenya. .............................................................86

Figure 21: Agricultural production in Kenya. ............................................................................86

Figure 22: Map of Rwanda with the Nile basin. ........................................................................89

Figure 23: Agricultural area and arable land in Rwanda. ..........................................................90

Figure 24: Agricultural production in Rwanda...........................................................................90

Figure 25: Map of Sudan with the Nile basin. ...........................................................................93

Figure 26: Agricultural area and arable land in Sudan..............................................................94

Figure 27: Agricultural production in Sudan. ............................................................................94

Figure 28: Map of Tanzania with the Nile basin. ......................................................................97

Figure 29: Agricultural area and arable land in Tanzania. ........................................................98

Figure 30: Agricultural production in Tanzania. ........................................................................98

Figure 31: Map of Uganda with the Nile basin........................................................................101

Figure 32: Agricultural area and arable land in Uganda..........................................................102

Figure 33: Agricultural production in Uganda. ........................................................................102

Figure 34: Relation between spatial scale and physical detail. The green ellipses show the position of different models....................................................................................................103

Figure 35: Example of output of PCR-GLOBWB hydrological model output, in this case the internal renewable water resources based on the 2000-2009 climatology in the MENA region (Immerzeel et al., 2011). .......................................................................................................105

Figure 36: Hydrological model concept as will be used in this study. ......................................107

7

1 Introduction

1.1 Contract Details

The Nile Basin Initiative (NBI), under the Nile Equatorial Lakes Subsidiary Action Program (NELSAP) and the project Regional Agricultural Trade and Productivity Project (RATP) has announced a Request for Proposals (RFP) entitled “Assessment of the Irrigation Potential in Burundi, Eastern DRC, Kenya, Rwanda, Southern Sudan, Tanzania and Uganda” in July 2010 (RATP/CONSULTANCY/04/2010). FutureWater, in association with WaterWatch, has submitted a proposal in response to this RFP. Based on an independent Technical and Financial evaluation FutureWater, in association with WaterWatch, has been selected to undertake the study. The consulting services contract was signed between the “Nile Basin Initiative / The Regional Agricultural Trade and Productivity Project” and “FutureWater in association with WaterWatch” entitled “Consulting Services for Assessment of the Irrigation Potential in Burundi, Eastern DRC, Kenya, Rwanda, Southern Sudan, Tanzania and Uganda”. This contract was dated at 5-Feb-2011 and total project duration is 16 months. The Contract Reference Number is: NELSAP CU/RATP2/2011/01

1.2 Inception Report

This Inception Report contains a summary of the background of the project, objectives and key issues of the assignment, overview of the study area, a description of the tasks to be executed, the proposed approach, a work program and time schedule and a report on the organizational structure and staffing. This Inception Report has been discussed during the Validation Workshop in Nairobi on 28-29 March, 2011. This Inception Report has been approved during the meeting under the condition that recommendations from the Validation Workshop were included. This version of the Inception Report includes these comments.

1.3 Acknowledgements

The Consultants wish to acknowledge the support, fruitful discussions and useful comments from all NBI-RATP staff and stakeholders in the countries. In particular Dr. Innocent Ntabana and Dr. Gabriel Ndikumana are acknowledged for starting this initiative and their support and advice on the study.

8

2 General Information

2.1 Background

2.1.1 Nile Basin Initiative

The Nile Basin Initiative (NBI) is a partnership of the riparian states that seeks to develop the river in a cooperative manner, share substantial socioeconomic benefits, and promote regional peace and security through its shared vision of “sustainable socioeconomic development through the equitable utilization of, and benefit from, the common Nile Basin water resources”. NBI’s Strategic Action Program is made up of the Shared Vision Program (SVP) and Subsidiary Action Programs (SAPs). The SAPS are mandated to initiate concrete investments and action on the ground in the Eastern Nile (ENSAP) and Nile Equatorial Lakes sub-basins (NELSAP). This study falls under NELSAP.

2.1.2 The Nile Equatorial Lakes Subsidiary Action Program (NELSAP)

The Nile Equatorial Lakes Subsidiary Action Program has its Coordination Unit (CU) based in Kigali, Rwanda and reports to the Nile Equatorial Lakes Technical Advisory Committee (NELTAC) and the NBI Secretariat for strategic guidance. The NELTAC reports to the Nile Equatorial Lakes Council of Ministers (NELCOM). The Nile Basin Initiative (NBI) through the Nile Equatorial Lake Subsidiary Action Program (NELSAP) seeks to promote a productive water use in Nile basin agriculture. The NELSAP through its sub basin programs implements pre-investment programs in the areas of power trade and development and natural resources management. The NELSAP-CU in partnership with the countries carries out selected preparatory initiatives that have trans-boundary implications and helps the countries to mobilise resources for project development including planning, data collection, surveys and feasibility studies. Pre-investment programs comprise specific studies of the various users of the water resources, formulation of options for water resources development taking in to account various intervening factors and users, identification of specific water resources developments integrating options, preliminary design of each project, cost benefit evaluation, preliminary Environmental Impact Assessment, comparative studies based on technical, socio-economic and environmental criteria, selection of priority projects and comparison with other sectoral possibilities. Within the pre-investment framework, the Regional Agricultural Trade and productivity Project, in concert with the NELSAP, will promote irrigation development as a contribution towards agricultural development in the NEL Countries.

2.1.3 Regional Agricultural Trade & Productivity (RATP) Project

RATP is a technical assistance project financed by Canadian International Development Agency (CIDA) through a recipient-executed trust fund. The project is managed by a Project Management Unit (PMU) based in Bujumbura-Burundi, and is administratively linked to the NBI’s Subsidiary Action Program for the Nile Equatorial Lakes (NELSAP), which has a

9

coordinating unit (NELSAP-CU) based in Kigali. Although the activities of the proposed project focus on the Nile Equatorial Lakes sub-basin area, it supports generation of agricultural knowledge that is basin-wide, in line with the aims of the NBI’s Institutional Strengthening Project (ISP) and NELSAP’s Subsidiary Action Program.

2.2 Project objectives

With a rapid rate of population increase and high pressure on arable land, increased food production is one of the main concerns and priorities of the governments of the seven countries involved in the Irrigation Potential study. Improved irrigation technology and better water resources management have been suggested as mechanisms for increased production. One of the constraints identified is the reliance on rain fed agriculture as well as low mechanization. The goal of the study is to ensure household food security, improve farmers’ income and alleviate poverty through increase in agricultural production and productivity resulting from accessibility to irrigation water; and as such, it will contribute to NBI’s overall objective of achieving sustainable socio economic development through equitable utilization of and benefits from the common Nile Basin water resource. Within the NELSAP, Planning for water use is carried out on the basis of river basins or sub basins. On the other hand, land use is usually computed or planned according to political boundaries. The study will therefore determine the irrigation potential of the proposed countries considering the physical resources of 'soil' and 'water', combined with the irrigation water requirements as determined by the cropping patterns and climate. This will inform the subsequent preparation process and resource mobilization for the preparation phase. The general objective of the study is to assess the irrigation potential of seven Nile Countries (Burundi, Eastern DRC, Kenya, Rwanda, Southern Sudan, Tanzania and Uganda) in order to fill gaps in the NBI and member country information bases on agriculture water use. This assignment will be carried out under the RATP project, with the support of NELSAP and the Directorate of Irrigation in the Ministries in charge of Water and Irrigation in the seven countries. The specific objectives of the study are to: (i) determine the irrigation potential of the proposed countries considering the physical resources of 'soil' and 'water', combined with the irrigation water requirements as determined by the cropping patterns and climate; (ii) provide a preliminary assessment of probable environmental and socioeconomic constraints to be considered to ensure sustainable use of physical resources within the Nile basin, as well as (iii) an indication of required resources for the preparation and investment phase. The study can be categorized as preparation for a development program.

10

3 Research Area and Main Issues

3.1 Overview 1,2

The Nile River Basin is probably one of the world’s most famous river basins (Figure 1). There is a fascination about the Nile River which has captured human imagination throughout history. Some five thousand years ago a great civilization emerged depending on the river and its annual flooding cycle. The Nile is one of the world longest rivers, flowing south to north 6,850 kilometers, over 35 degrees of latitude. Its catchment basin covers approximately 10% of the African continent, with an area of about 3,100 km2, and spreads over 10 countries (Table 2)2. The Nile is distinguished from other great rivers of the world by the fact that half of its course flows through countries with no effective rainfall (Table 1). Almost all the water of the Nile is generated on an area covering only 20 percent of the basin, while the remainder is in arid or semi-arid regions where the water supply is minimal and where evaporation and seepage losses are very large. The shape of the Nile we know today is complex and is the result of the interconnection of several independent basins by rivers which developed during the last wet period which affected Africa after the retreat of the ice of the last glacial age, some 10,000 years ago. The basins which constitute part of the present river were disconnected, forming internal lakes. At times when the climate was wet, they overflowed their banks and became connected to other basins. At other times, when the climate was very dry, they ebbed, shrank into saline pools or dried altogether. The basins stand out in the longitudinal section of the river, as flat stretches or landings with very little slope, which are connected today with rivers, which have considerably steeper slopes (Sutcliffe, 2009) Table 1. World’s Major River Systems 2

Source: UNEP, 2000.

1 This section is derived from various sources and is included as generic background 2 Numbers mentioned in this Inception Report are all taken from existing references. Some inconsistency in numbers

might therefore occur, which will be clarified and verified during the project.

11

Table 2. Countries in Nile Basin 1

1Source: CIA World Factbook, 1999. 2Source: FAO, 1997. The basin of the Nile is characterized by the existence of two mountainous plateaus rising some thousands of meters above mean sea level. The Equatorial or Lake Plateau in the southern part of the Nile basin (Figure 3), situated between the two branches of the Great Rift, is at a level of 1,000 to 2,000 meters and has peaks of 5,100 and 4,300 meters. This plateau contains Lakes Victoria, George, Edward (Mobutu Sese Seko) and Albert, which slope gently toward the north at an average rate of one meter for every 20 to 50 km of stretch. In contrast the rivers which connect these lakes fall at an average rate of one meter every kilometer or less of length. The Ethiopian or Abyssinian Plateau, which forms the eastern part of the basin, has peaks rising to 3,500 meters. North of the Lake plateau the basin descends gradually to the Sudan plains where the Nile runs at altitudes lower than 500 m in its northerly direction. About 200 km south of the Egyptian border the river cuts its channel in a narrow trough bounded from each side by the contour line of 200 m ground surface level. Almost 200 km before discharging into the sea, the river bifurcates and its two branches encompass the Nile Delta. The enormous Sudd and Central Sudan basins extend for a distance of 1,800 km from Juba to Khartoum and form a gently sloping region with a small rate of slope of one meter for every 24 kilometers of stretch. The basin of the present-day Nile can be divided into six major regions: the Lake Plateau, the Sudd, the White Nile, the Ethiopian Plateau, the Main Nile and the Nile Delta.

3.2 Irrigation potential 2

Both Burundi and Rwanda are characterized by a rolling topography with a continuous pattern of hills and valleys, with lakes and marshy lowlands at the bottom of the valleys. Improving the drainage network in part of the swamp areas, combined where possible with an irrigation 1 Note that figures might vary slightly from source to source and here the original numbers are presented. 2 Summarized from FAO 1997

12

network, would allow year-round cultivation, which is important for these small, but very densely populated countries. The total area of these valley bottoms in the Nile basin is estimated at 105,000 ha for Burundi and 150,000 ha for Rwanda (FAO, 1997). For Tanzania the irrigation potential has been estimated at 30,000 ha, but this would require the construction of considerable water conveyance works. In addition to this, at the beginning of the century settlers from Germany, the then colonial power in the country, proposed a plan to transfer water from Lake Victoria to the Vembere Plateau in the Manonga River basin in central Tanzania to irrigate between 88,000 and 230,000 ha of cotton. Though this project is still on the table, it would be very expensive. The transfer would be affected by gravity as the plateau lies below the water level of the lake (FAO, 1997). The Lake Victoria basin in Kenya covers only 8.5% of the total area of the country but it contains over 50% of the national freshwater resources. The national water master plan identified an irrigation potential of 180,000 ha based on 80% dependable flow. As part of the plan, dams and water transfers to other (sub) basins are proposed. At present only about 6,000 ha are irrigated. Moreover, in Kenya there has been lengthy debate as to whether, given adequate technology, Lake Victoria basin water should be transferred to arid areas of the country for irrigation. It is considered that perhaps the most appropriate location for such an experiment would be the Kerio Valley (located in the Rift Valley), for which a special development authority has been established by the Kenyan Parliament. The feasibility of such a project is a question of engineering and several observers consider it possible. Such an undertaking would use significant quantities of water (FAO, 1997). The Nile basin in DRC covers less than 1% of the area of the country. The area is hilly and does not really lend itself to irrigation. This area is rather densely populated with most people engaged in cattle rearing and fishery activities around Lake Albert. It is considered that about 10,000 ha could be developed for irrigation (FAO, 1997). Uganda has large swamp areas covering about 700,000 ha. The irrigation potential is estimated at 202,000 ha, requiring, however, major works such as storage, river regulation and large-scale drainage. At present only 5,550 ha are irrigated (FAO, 1997). Irrigation potential in Sudan has been estimated at over 4.8 million hectares, but this figure does not take into consideration the available water resources. The irrigated area was about 1.6 million hectares in 1979 and 1.9 million hectares in 1990. There are plans to increase irrigation to about 2.8 million hectares by the year 2000, almost all to be irrigated by Nile water (FAO, 1997). A more detailed description of the individual countries is provided in the Appendix.

13

Figure 1: Overview Nile Basin.

14

Figure 2: Overview study area.

Tanzania

Kenya

Southern Sudan

Uganda

Rwanda

Burundi

Eastern DR Congo

±0 100 200 300 400 500

Kilometers

Research area

Height in m

High : 5882

Low : -13

Legend

Water

Nile

Nile basin

Borders

Major town

15

Figure 3: Sub-basins in NELarea.

17

4 Data and Tools

4.1 Data and Interpretation

4.1.1 Current studies

Best Practices for Water Harvesting and Irrigation

Many studies have already been performed on the subject of irrigation potential. An overview of the studies found so far, are shown elsewhere in this report. The studies of Best Practices for Water Harvesting and Irrigation, performed for each of the seven NEL countries, provide knowledge and information on water harvesting, community managed irrigation, and public/private managed irrigation in these countries. These reports are relevant for the assessment of the irrigation potential in each of the seven NEL countries. These reports are part of the Efficient Water Use for Agricultural Production (EWUAP) project, which is one of the eight projects of the Nile Basin Initiative’s (NBI) Shared Vision Program (SVP). These reports give us useful information on the irrigation potential in each of the seven NEL countries, as evaluated so far. They also provide information regarding defined agro-climatic zones and country specific climate information. Rapid Baseline Assessment of the Agricultural Secto r For five of the seven countries in the NEL region a “Rapid Baseline Assessment of the Agricultural Sector” report was made. These reports were also written in the context of EWUAP. The objective of these studies is to give a quick assessment of the agricultural sector in the NEL countries. Unfortunately, this study has not been undertaken for Uganda and DRC. These reports are useful in evaluating the current agricultural sector, e.g. in numbers of farm sizes, irrigated areas, rainfed areas etc. FAO studies The Food and Agriculture Organization (FAO) has undertaken many studies related to irrigation potential. The FAO 1997 study focuses on a physical assessment of the irrigation potential in Africa, with a focus on the Nile basin. They assessed the irrigation potential based on soil and terrain suitability, and water resources. Although the importance of this study, it currently considered as somewhat outdated in terms of changes occurred in the countries since the study was released. Moreover, methodological development has also taken place over the last ten years, making the current study more precise. Peer-reviewed articles Peer-reviewed articles are an important source of information, because they are built on scientific knowledge. So far, some peer-reviewed articles have already been found. These articles provide us with useful information regarding climate (change), hydrological impacts, hydrological extremes, crop evapotranspiration, irrigation water requirements, crop yields, hydrological modeling, water resource development, and the socio-economic effects of irrigation.

18

Country studies A summary of the relevant available country studies can be found in the appendix. The EWUAP reports, as mentioned before, are an important source of information for these countries. In addition to this, the ministries of some of these countries have performed agriculture related studies in these countries. These studies give us numbers regarding the development of drainage and irrigation, projections of population growth, development of the agricultural sector, poverty, and other development plans.

4.1.2 Climate

Local data In terms of climate data required to undertake the study as much as possible local data will be used. An intensive and shared effort, using support from various NBI sections, Ministries and the NBI National Liaison Officers (NLOs), has been started on 29-Mar-2011. The assignments signed mutually on this can be found in Appendix 0. CRU

One source of meteorological data at a relatively high spatial and temporal resolution is available from the Climatic Research Unit of the University of East-Anglia. The CRU TS 2.1 dataset comprises 1224 monthly grids, for the period 1901-2002, and covers the global land surface at 0.5° × 0.5° resolution (Mitchell et al., 2004). The dataset comprises: cloud cover, diurnal temperature range, precipitation, temperature and vapor pressure (note that no wind speed data are available in the CRU dataset). The dataset is based on raw station data. Since data can be scarce in some regions or periods, a method called 'relaxation to the climatology' was used to create continuous grids. This implies that, for some areas or regions, data are less accurate. More information on the CRU gridded datasets can be found at http://www.cru.uea.ac.uk/cru/data/hrg/. TRMM Rainfall can be obtained from TRMM1 (Tropical Rainfall Measuring Mission). TRMM is a joint space project between NASA and the Japanese Aerospace Exploration Agency (JAXA). TRMM is designed to measure tropical precipitation. The TRMM satellite has a latitudinal range from 50°S – 50°N. The TRMM satellite has been launched November 27, 1997 and the mission has recently been extended to 2009. The sensors to measure rainfall consist of precipitation radar, a multi-frequency microwave radiometer and a visible and infrared (VIS/IR) radiometer. These sensors are used in a complementary way to deduct rainfall. Several orbital and gridded data products are available for download at the Goddard Distributed Active Archive Centre (DAAC). The 3B-42 product is considered as the most applicable for hydrological studies. The output is rainfall for 0.25x0.25 degree (~ 25x25 km) grid boxes on a 3-hourly basis. For the current study this will be accumulated to a daily time step.

1 http://trmm.gsfc.nasa.gov/

19

Further information can be obtained at: http://disc.sci.gsfc.nasa.gov/data/datapool/TRMM/01_Data_Products/index.html and Tropical Rainfall Measuring Mission (TRMM): http://trmm.gsfc.nasa.gov/. FEWS-NET One day estimates of precipitation for the African continent are prepared operationally at the Climate Prediction Center (CPC) for the United States Agency for International Development (USAID) as a part of the Famine Early Warning System Network1 (FEWS-NET). The algorithm for the rainfall estimates uses Meteosat 7 geostationary satellite infrared data that are acquired in 30-minute intervals, and areas depicting cloud top temperatures of less than 235K are used to estimate convective rainfall. Two other satellite rainfall estimation instruments are incorporated into the algorithm, being the Special Sensor Microwave/Imager (SSM/I) on board of the Defense Meteorological Satellite Program satellites, and the Advanced Microwave Sounding Unit (AMSU). All satellite data are first combined using the maximum likelihood estimation method, and then GTS station data are used to remove bias. Warm cloud precipitation estimates are not included in the algorithm. The most recent version available is the RFE2.0 version, which is compared to version RFE1.0 more accurate. This version produces daily precipitation output on a 0.1 degree spatial resolution, and on a spatial extent from 40°S-40°N and 20°W-55°E. An example of one day output of RFE2.0 is shown in Figure 4. The Famine Early Warning Systems Network (FEWS-NET) RFE2.0 version can be downloaded at: http://www.cpc.ncep.noaa.gov/products/fews/data.shtml.

Figure 4: Rainfall estimate obtained from FEWS-NET (24/11/2000). Data archives There are a couple of data archives available that store and distribute local data. In principle, these archives store only data that have been obtained from national weather services and one would expect that those data is already possessed by the national NBI offices. However, if the national NBI offices fail to gather these data, one could try the public domain data archives. The most relevant are:

1 http://www.fews.net/Pages/default.aspx

20

• GSOD database NOAA/WMO: o http://www.ncdc.noaa.gov/cgi-bin/res40.pl?page=gsod.html

• Weather underground: o http://www.wunderground.com

Recommendation It should be emphasized that climate data, and especially precipitation data is of major importance for assessing the irrigation potential in the NEL countries. It is recommended to follow a three level approach in information. The most reliable information is data from local stations. If these data are not available (or are not resealed by national weather agencies) for a certain area or period one could go for FEWS-NET as far as precipitation concerns. TRMM would be an option as well, but because FEWS-NET provides data on a higher resolution it is preferred to use FEWS-NET. A final option would be to use CRU. This would be less desirable, because it only provides data on a monthly resolution. During Phase 1 of the project a comparison of these various data product will be made, including availability, resolution and accuracy.

4.1.3 Streamflow

Local data In terms of climate data required to undertake the study as much as possible local data will be used. An intensive and shared effort, using support from various NBI sections, Ministries and the NBI National Liaison Officers (NLOs), has been started on 29-Mar-2011. The assignments signed mutually on this can be found in Appendix 0. GRDC The Global Runoff Data Center1 (GRDC) holds the Global River Discharge Database. GRDC's role is to serve as a facilitator between data providers and data users. It serves under the auspices of the World Meteorological Organization (WMO) and has been established at the Federal Institute of Hydrology (BfG), Germany. The GRDC provides river discharge data for stations all over the world. For the NEL countries, data from 77 stations are available. During Phase 1 the full set of the seven countries will be requested. Discharge data from the GRDC can be ordered at: http://www.bafg.de/cln_007/nn_294112/GRDC/EN/02__Services/01__RiverDischarge/riverdischarge__node.html?__nnn=true UNH The University of New Hampshire developed a global runoff database. The database is developed using a global water balance model and fine-tuned using observations. These observations can be downloaded, but includes only monthly averages from historic data and are identical as the ones from the GRDC (http://www.grdc.sr.unh.edu).

1 http://www.bafg.de/cln_015/nn_293894/GRDC/EN/Home/homepage__node.html?__nnn=true

21

Figure 5: Overview of UNH discharge stations in the study area (http://www.grdc.sr.unh.edu/html/Stn/B2.html). River Discharge Database The Center for Sustainability and the Global Environment (SAGE) of the University of Wisconsin-Madison holds a database with global river discharges. Data is identical as the GRDC, but with a few less stations. In contrast to the GRDC, data can be downloaded directly from their website: http://www.sage.wisc.edu/riverdata/ RivDis database The Global River Discharge Database development represents the first step in a continually evolving compilation of river discharge information. One of the primary sources of information for the database development was the UNESCO river archives and the series of publications entitled "The Discharge of Selected Rivers of the World" which was provided, in book form from 1969 through 1984. The series served as an important source of information on approximately 1000 stations. RivDis v1.0 provides discharge data from the original UNESCO publication series in a digital format that can be easily acquired and analyzed by researchers and planners in the water sciences community. For Africa data of 269 stations are available. More details can be obtained from: http://www.rivdis.sr.unh.edu/ LEGOS

TANZANIA

SUDAN

KENYA

UGANDA

BURUNDI

RWANDA

DR CONGO

Height

High : 5882

Low : -13

Legend

Major towns

Discharge stations

Nile basin

Water

Borders

±0 150 300 450 600

Kilometers

Discharge stations

22

The GOHS (Géophysique, Océanographie et Hydrologie Spatiales) group of LEGOS (Laboratoire d’Etudes en Géophysique et Océanographie Spatiales) in Toulouse collects and distributes river, lake and reservoir levels based on satellite observations. The water level time-series are based on altimetry measurements from Topex/Poseidon, Jason-1, ERS-2, ENVISAT and GFO satellites. The database includes water levels for over 130 lakes and man-made reservoirs, 250 virtual stations on rivers and about 100 sites on flooded areas. Most river water level time series are based on Topex/Poseidon observations and start in January 1993. The lake water level time-series are multi-satellite data combinations. They also start in January 1993. The time series are regularly updated and the number of sites increases regularly. Recommendation Based on the available sources of discharge data, all sources will be used for the current study. First of all, the local data as collected during the study will be used. If these data are insufficient other date sources as presented in this section will be used.

4.1.4 Digital Elevation Models

SRTM

The most commonly public domain Digital Elevation Model (DEM) dataset is the SRTM dataset which is acquired with the space shuttle during an 11-day mission in 2002. More information on the SRTM can be found at: http://www2.jpl.nasa.gov/srtm/. The dataset can be downloaded from various sources. Probably the most user-friendly one is provided by the CGIAR (version 4). This DEM has been obtained at a spatial resolution of 90 m. This dataset has been slightly modified from the original version provided by the USGS (seamless.usgs.gov) and data gaps are filled using an automated procedure. For more information reference is made to: http://srtm.csi.cgiar.org/. An example of this DEM for the entire world is shown in Figure 6.

Figure 6: SRTM 90 m Digital Elevation Data of the w orld. ASTER GDEM ASTER GDEM is an easy-to-use, highly accurate DEM covering all the land on earth, and available to all users regardless of size or location of their target areas. Anyone can easily use the ASTER GDEM to display a bird's-eye-view map or run a flight simulation, and this should realize visually sophisticated maps. ASTER GDEM provides elevation data on a spatial

23

resolution of 30m. An advantage of this DEM with respect to the SRTM DEM is that the ASTER GDEM provides elevation data in topographically steep areas, where the SRTM sometimes does not provide these data. More information regarding the ASTER GDEM can be found at: http://www.ersdac.or.jp/GDEM/E/index.html Recommendation For this project the SRTM DEM will provide enough detail. For the high elevation areas, however, it may happen that the SRTM does not provide the necessary data. For these regions we will use the ASTER GDEM.

4.1.5 Soils

Local data In terms of soil data required to undertake the study as much as possible local data will be used. An intensive and shared effort, using support from various NBI sections, Ministries and the NBI National Liaison Officers (NLOs), has been started on 29-Mar-2011. The assignments signed mutually on this can be found in Appendix 0. FAO/UNESCO The most commonly used global dataset on soil is the FAO/UNESCO Soil Map of the World at scale 1:5,000,000 (FAO, 1974). Based on this map the Digital Soil Map of the World (DSMW) was published. One of the drawbacks of this dataset is that the information is qualitative rather than quantitative. However, procedures have been developed to transform the original classes into information needed by models (Droogers et al., 2001). More information regarding the Soil Map of the World can be found at: http://www.fao.org/nr/land/soils/digital-soil-map-of-the-world/en/ e-SOTER e-SOTER’s objective is to characterize all soils from all corners of the world under a single set of classification rules. As there was no universally accepted system for world-wide classification of terrain, e-SOTER has designed its own system (Van Engelen et al., 1995). Regarding the Nile Basin Countries, however, only for Kenya the e-SOTER soil map is completed. e-SOTER has also published a CD-ROM with “Soil and Terrain Database for northeastern Africa”, but this dataset has the same soils information as the FAO/UNSECO global dataset. More information considering the e-SOTER available data can be found at: http://www.esoter.org/ Harmonized World Soil Database In 2008 a new global dataset was developed under the name “Harmonized World Soil Database” (HWSD) (FAO, 2009) existing out of a 30 arc-second (~1 km resolution) raster database with over 15000 different soil mapping units that combines existing regional and national updates of soil information worldwide (SOTER, ESD, Soil Map of China, WISE) with the information contained within the 1:5 000 000 scale FAO-UNESCO Soil Map of the World (FAO, 1974). The resulting global raster database consists of 21600 rows and 43200 columns, which

24

are linked to harmonized soil property data. The use of a standardized structure allows for the linkage of the attribute data with the raster map to display or query the composition in terms of soil units and the characterization of selected soil parameters (organic Carbon, pH, water storage capacity, soil depth, cat-ion exchange capacity of the soil and the clay fraction, total exchangeable nutrients, lime and gypsum contents, sodium exchange percentage, salinity, textural class and granulometry). The HWSD of the world is shown in Figure 7. The HWSD can be obtained at: http://www.iiasa.ac.at/Research/LUC/External-World-soil-database/HTML/

Figure 7: Harmonized World Soil Database (HWSD). IUSS The International Union of Soil Sciences is starting a new initiative to develop an updated digital soil map, with special emphasis on the need of the modeling community. It will take however years before the first results will emerge. More details considering this source of data can be found at: http://www.globalsoilmap.net/. Recommendation Based on the previous sections of various sources of soil data it is clear that locally collected data will be used as much as possible. If these data are insufficient the Harmonized World Soil Database can be considered as the most appropriate source to be used for the study.

4.1.6 Land cover

Local data In terms of land cover data required to undertake the study as much as possible local data will be used. An intensive and shared effort, using support from various NBI sections, Ministries and the NBI National Liaison Officers (NLOs), has been started on 29-Mar-2011. The assignments signed mutually on this can be found in Appendix 0. USGS The United States Geological Survey (USGS) has been active in developing and distributing land cover information. The Africa Land Cover Characteristics Data Base Version 2 is a subset of a global land cover characteristics data base that was developed on a continent-by-continent

25

basis and are based on 1-km AVHRR data spanning April 1992 through March 1993 (Loveland et al., 1999). There are seven different legends available depending on the need of the user. A drawback of this dataset is that it is based on information from over 15 years ago. More information regarding the USGS land cover data can be found at: http://landcover.usgs.gov/usgslandcover.php. GLCF The Global Land Cover Facility1 (GLCF) is a center associated to the University of Maryland. GLCF provides earth science data and products to help everyone to better understand global environmental systems. In particular, the GLCF develops and distributes remotely sensed satellite data and products that explain land cover from the local to global scales. The University of Maryland’s Land Cover Classification was based on imagery from the AVHRR satellites acquired between 1981 and 1994. The database includes fourteen land cover classes and is available at three spatial scales: 1 degree (~100 km), 8 km and 1 km pixel resolutions. A drawback of this dataset is that it is based on information from over 15 years ago. GlobCover GlobCover Land Cover v2 is a global land cover map at 10 arc second (300 meter) resolution. Its 23 global land cover classes are defined within the UN Land Cover Classification System (LCCS). GlobCover LC v2 was developed as part of the GlobCover project, a European Space Agency (ESA) initiative in partnership with JRC, EEA, FAO, UNEP, GOFC-GOLD, and IGBP. GlobCover products are based on the ENVISAT satellite mission's MERIS sensor (Medium Resolution Image Spectrometer) Level 1B data acquired in Full Resolution (FR) mode with a spatial resolution of 300 meters. GlobCover LC v2 was derived from an automatic and regionally-tuned classification of a time series of MERIS FR composites covering the period December 2004-June 2006. GLC2000 The Land Cover map of Africa is one regional component of the GLC2000 exercise, conceived and coordinated by the European Commission’s Joint Research Centre. The GLC2000 maps are based on daily observations made from 1st November 1999 to 31st December 2000 by the VEGETATION sensor on the SPOT 4 satellite. The Africa map’s legend pays special attention to the forest and savannah biomass. The map shows specific land-cover features as the irrigated agriculture, the ribbons of secondary forest of the swamp forests at a spatial detail never achieved before. The current version is 05 (Mayaux et al., 2004). GLC2000 has a legend with 27 classes and can be considered as the best land cover dataset currently available for the entire continent. The spatial resolution is 1 x 1 km. More information regarding this source of land cover can be found at: http://bioval.jrc.ec.europa.eu/products/glc2000/glc2000.php. AfriCover The AfriCover Project2 developed a combined approach to promote the sustainable use of natural resources. The purpose of the AfriCover Project is to establish a digital geo-referenced database on land cover and a geographic referential for the whole of Africa including:

• Geodetically homogeneous referential

1 http://www.landcover.org/index.shtml 2 http://www.africover.org/index.htm

26

• Toponomy • Roads • Hydrography

The Multipurpose AfriCover Database for the Environmental Resources (MADE) is produced at a 1:200,000 scale (1:100,000 for small countries and specific areas). The Eastern Africa module is the first operational component of the AfriCover Project. It was formulated to meet several African countries request for assistance in the set-up of reliable and geo-referenced data-bases on natural resources. It is part of FAO assistance to the Nile Basin countries. The Project has been operational in the period 1995-2002 and was signed by ten countries, including the seven NEL countries. For these seven NEL countries the map scale is 1:100,000. Global Irrigated Area Map The Land and Water Division of the Food and Agriculture Organization of the United Nations and the Johann Wolfgang Goethe Universität, Frankfurt am Main are co-operating in the development of a global irrigation mapping facility. The first global digital map of irrigated areas on the basis of cartographic information and FAO statistics has a resolution of 0.5 degree and was developed in 1999. Since 1999 the methodology to produce the map has been improved, which made it possible to increase the spatial resolution of the map to 5 minutes (about 10 km at the equator). The objective of the co-operation between the Johann Wolfgang Goethe Universität and FAO, is to develop global GIS coverage of areas equipped for irrigation and to make it available to users in the international community. The data collected through the AQUASTAT surveys was used to improve the overall quality and resolution of the information (Siebert et al, 2006). The latest version of the "Global Map of Irrigation Areas" is version 4.0.1. More information regarding this map of global irrigated areas can be found at: http://www.fao.org/nr/water/aquastat/irrigationmap/index10.stm Recommendation The most useful source of land cover will be the AfriCover Project. It’s most reliable, because it was formulated to meet the countries requests for a reliable geo-referenced database on natural resources.

4.1.7 Vegetation indices

MODIS

The Moderate Resolution Imaging Spectro radiometer1 (MODIS) is a sensor aboard the Terra (EOS AM) and Aqua (EOS PM) satellites. Terra's orbit around the Earth is timed so that it passes from north to south across the equator in the morning, while Aqua passes south to north over the equator in the afternoon. Terra MODIS and Aqua MODIS are viewing the entire Earth's surface every 1 to 2 days, acquiring data in 36 spectral bands. Various vegetation characteristics are derived from MODIS. The most relevant products are:

• MOD 13 - Gridded Vegetation Indices (Max NDVI & Integrated MVI) • MOD 15 - Leaf Area Index & FPAR

1 http://modis.gsfc.nasa.gov

27

• MOD 17 - Net Photosynthesis and Primary Productivity Most of these products are available at spatial resolution of 250, 500 and 1000 m and at temporal resolutions of 16 days and one month. Various sites are available for downloading data. The primary site to obtain MODIS data is LP DAAC (Land Processes Distributed Active Archive Center) at: https://lpdaac.usgs.gov/. Data can be obtained using three tools:

• Data Pool: The Data Pool is the publicly available portion of the LP DAAC online holdings. Data Pool provides a more direct way to access files by foregoing their retrieval from the near line tape storage devices;

• WIST: The Warehouse Inventory Search Tool (WIST) is a web-based client to search and order earth science data from various NASA and affiliated centers;

• GloVis: The USGS Global Visualization Viewer (GloVis) is an online search and order tool for selected satellite data;

Landsat The Landsat program is the longest running enterprise for acquisition of imagery of Earth from space. The first Landsat satellite was launched in 1972; the most recent one, Landsat 7, was launched on April 15, 1999. Landsat 7 data has eight spectral bands with spatial resolutions ranging from 15 to 60 meters. Since May 2003 Landsat 7 has a scan line failure, but images are still useable. Original Landsat images can be downloaded from various sites. For the purpose of the current project it is recommended to use the so-called GeoCover2000 images. Specific detailed studies might benefit from the original high-resolution Landsat images. The GeoCover2000 images are distributed by various sites, but most recommended is the Earth Science Data Interface of the Global Land Cover Facility (http://glcfapp.umiacs.umd.edu). Spot SPOT (Satellite Pour l'Observation de la Terre) is a high-resolution, optical imaging Earth observation satellite system operating from space. It is run by Spot Image based in Toulouse, France. SPOT4 was launched in 1998 and SPOT5 in 2002. SPOT5 has two high resolution geometrical instruments and has a resolution of 2.5 to 5 meters in panchromatic mode and 10 meters in multispectral mode. A drawback of SPOT images is the very high price. Depending on the resolution, areal extent and date of images various prices are asked. Purchasing images for the entire NEL countries is therefore no option and only for specific detailed projects one could consider purchasing images. SPOT offers also free NDVI images as 10 days composites at a resolution of 1 km. These images are already geo-referenced and cloud-filtered. Data can be obtained from http://free.vgt.vito.be/. Recommendation Vegetation indices are primarily used to support land cover classification and as input to hydrological models. For visualization purposes vegetation maps (sometimes combined with elevation) are used as well. For the research area it is recommended to use MODIS images as these provides also derived products as LAI and NPP. Landsat GeoCover2000 images can be used for visualization using three bands.

4.1.8 Evapotranspiration

WaterWatch, together with partners, has developed the SEBAL, and the derived ETLook algorithms. Both actual and potential evapotranspiration can be calculated on a pixel-by-pixel

28

basis by solving the energy balance at the earth surface using satellite images. The actual and potential evapotranspiration will be used as input for our hydrological modeling. Various SEBAL/ETLook analyses have been undertaken for the Nile basin over the last couple of years and should be included in the current Irrigation Potential study. More information regarding SEBAL/ETLook is described in the following chapter.

4.1.9 Groundwater

Groundwater information can be divided into aquifer characteristics (location, KD values) and groundwater levels. Public domain datasets and remote sensing can hardly be used to get this information. There are however interesting developments on two fronts. First of all groundwater, or better phrased soil moisture contents, can be observed using various satellites. GRACE (Gravity Recovery And Climate Experiment) is a twin-satellite mission, developed to measure changes in the Earth's time-variable gravity field (Tapley et al., 2004). GRACE data are available since May 2002. GRACE data products are expressed in mm equivalent water. Two factors are important when evaluating results. Firstly, no distinction between snow cover, soil moisture and deep groundwater storage can be made. Secondly, results are given relative to the long-term average from Apr-2002 to Apr-2006. This means that no absolute values of water storage can be provided and that no spatial differences in water stored can be observed. In other words GRACE detects only changes in stored water. GRACE data is resolution independent, but data can only be used reliable at resolutions of at least 100 x 100 km2 or larger. Since groundwater irrigation should be based on sustainable principles, abstraction should always be smaller than recharge (assuming no significant inflows from other areas) at the sub-basin level. The PCRaster tool, as discussed in the following paragraphs, will provide also estimates of groundwater recharge rates.

4.1.10 Large scale irrigation practices in the Nile basin

An extensive study on the inventory of large scale irrigation schemes in the Nile Basin was accomplished by WaterWatch in 2009. Results and data from this study will be used in the current project. The WaterWatch study focused only on the Large Scale Irrigated (LSI) areas of eight Nile Basin countries. The various Nile Basin countries use different definitions for LSIs. In the study a minimum irrigated area of 200 ha was used to comply with the minimum size of the Nile country definitions. This means that all areas smaller than 200 ha are disregarded. The study concluded that there is an international demand for food production in Africa. Several international water programs such as FAO and NGO’s promote the development of irrigation systems and efficient use of irrigation water. The national agricultural politicians embrace this development, and most Nile basin countries embarked on preparing national irrigation plans. Ethiopia has for instance prepared a Nile basin irrigation and drainage plan. Egypt has prepared its National Water Resources Plan etc. Irrigation growth should not necessarily be realized by horizontal expansion, but alternatively by (i) improved irrigation management, (ii) a higher irrigated/irrigable area fraction, and (iii) improved annual irrigation intensity. While most scenarios by the Governments are based on a larger irrigated area, it is not unlikely that solutions need to be sought within the boundary conditions of reduced diversions.

29

Prior to any future development the water resources in the current LSIs need to be estimated in a standard way so that countries can be compared. The annual total crop evapotranspiration of all 4.9 million ha of irrigated land in the Nile Basin has been calculated. The total crop water consumption in irrigated agriculture of the Nile Basin is 36.9 billion m3 (BCM). A breakdown is provided in Table 3. This number is considered to be on the low side but without proof, the study used this 36.9 BCM tentatively as a working number to base planning upon. Table 3. Breakdown of the consumptive use in the ir rigation sector by country (Source: WaterWatch, 2009).

Table 4. Actually irrigated areas in the Nile Basin according to different sources (Source: WaterWatch, 2009).

The net irrigated area is 4.9 million ha. Hence the annual average crop water consumption per unit irrigated land is 753 mm. This average crop water consumption of 753 mm (7531 m3/ha/yr) varies considerably across the basin. The minimum values are 100 mm and the maximum values as much as 1400 mm/yr. The total crop land area in the Nile Basin is 23.7 million ha, out of which 4.9 million ha (21 %) is irrigated. The remaining part (18.8 million ha) is thus rainfed land. The total crop water consumption in agriculture is 184.6 billion m3/yr, and the 36.9 BCM accounts for 20 % of the agricultural water consumption in the Nile Basin. The 36.9

30

billion m3 is however also renewable and thus manageable, which justifies further investigations. The majority of this water is consumed in Egypt (65 %) and Sudan (30 %). Ethiopia appeared to be the 3rd largest consumer of irrigation water.

4.1.11 Other data sources

Other data sources not belonging to one of the previous categories are mention here. These references might be useful for some specific needs during the study. Land cover

• IGBP: o IGBP includes the various land cover types: forest, shrub land, savanna,

grassland, cropland/natural vegetation mosaic, wetland, urban and built-up, snow and ice, barren or sparsely vegetated, and water bodies.

• Crop land and other land uses: http://duckwater.bu.edu/lc/mod12q1.html • Food crops vs. non-food crops:

o http://www-tem.jrc.it/glc2000/ o http://www.geog.umd.edu/landcover/1km-map/download.html

• Irrigated vs. rain-fed crop land: http://www.geo.uni-frankfurt.de/ipg/ag/dl/forschung/MIRCA/index.html

• Current vs. future: o FAO Nile and FAO statistics: FAOSTAT, AQUASTAT, FAO world soil map o National and regional agricultural statistics

Irrigated and rainfed crop area and actual and pote ntial yield in a baseline year

• FAO Nile and other FAO estimates in 1997 (data available at the country level) • IWMI- PODIUM (IWMI, 2010) (data available at the country level) • Country and regional agricultural statistics

Crop ET: Potential and actual for the baseline year and other selected years

• EWUAP remote sensing-based estimates • International database: University of Washington and University of Montana, historical

and real-time assessment Water supply and water use Infrastructures: water s torage (reservoirs), irrigation systems, and rainfall harvesting systems

• ENDIS • FAO AQUASTAT • IWMI – PODIUM

• IFPRI – IMPACT-WATER (Rosegrant et al., 2009) Water productivity

• Challenge Program of Water and Food (CPWF), CGIAR, Basin Focus Project for Nile: http://cpwfbfp.pbworks.com/

• EWUAP remote sensing-based estimates

Crop pattern and farming systems

31

FAO/World Bank report (2001) Agricultural planning data (used for future food pr oduction projection) including crop land and yield change, irrigation planning International report from FAO (e.g., FAO Water Report for sub-Saharan Africa and similar reports from IFPRI) Crop prices - producers and consumers prices in the baseline year; international food trade prices IFPRI, FAO Trade and Food Security Database (2005)

4.2 Water resources assessment: PCRaster

Water resources assessment is a factor often getting less attention in irrigation potential studies, where focus is often on soil and land quality. The current study will ensure that availability of water resources and the potential irrigation source (stream, groundwater, reservoir) will be quantified at an very high spatial detail (250 m) using the state-of-the-art PCRaster approach. For this purpose PCRaster will be tailor made to the need of the current project, resulting in a model referred to as NELmod. Details of NELmod can be found in the Appendix.

4.3 Irrigation water requirements and potential cro p yields

The interaction between crop water needs, rainfall during the cropping season, and the water balance throughout the year determines the water shortage and thus irrigation water requirements. The SEBAL based ETLook remote sensing technique will be used to access this irrigation water requirement. ETLook is an algorithm for processing remote sensed data which was specifically developed to compute the (i) reference evapotranspiration (ET), (ii) potential ET, (iii) actual ET and (iv) ET deficit. Since ETLook is based on radar satellite information it is applicable under all weather conditions. Its main driving force is soil moisture, derived from passive microwave sensors. ETLook provides 8-day estimates of actual evapotranspiration, evapotranspiration deficit and biomass production and these will be accumulated to monthly values. This will provide a first estimate of the water requirements of the area. ETLook analysis indicates water shortage under current conditions. To assess what the irrigation requirements would be if the area would be brought under irrigation, ETLook will also be applied for a future scenario assuming that areas would be covered by irrigated crops. The evapotranspiration deficits under this scenario can be considered as the net irrigation water requirements. Irrigation water requirements combined with the irrigation efficiency will provide total water requirements for irrigation. Estimates of irrigation efficiencies are difficult to assess as these vary substantially by region and by irrigation practice. Moreover, the irrigation efficiency varies

32

even more between different fields, command areas and irrigation system, as reuse of “non-efficient” water should be taken into account. Water infiltrating beyond the root zones of the crops can be become available for plant transpiration at a downstream location. As such this water cannot be considered as a loss and is available for downstream users. Based on published irrigation efficiencies and the combination of soil and terrain analysis a map of expected irrigation efficiencies at field and command area level will be generated. This map will be multiplied with the net irrigation water requirements resulting in a map showing the total water requirements. During Phase 2 of the project emphasis will be put on the agricultural aspects of the potential irrigated areas. During Phase 1 a yield-gap analysis based on ETLook will be one of the components for selecting the potential irrigated areas, while during phase 2 an expanded yield-gap analysis will be undertaken (details in Chapter 5). The yield-gap is defined as the difference between the actual yield and the maximum obtainable yield. The analysis followed in this study will expand this approach in two main areas. First of all, an economic component will be included where yield-gap will not only be expressed as the difference between maximum yield and actual yield, but also as potential revenue in dollars. Based on a simplified economic analysis these values will be obtained, taking into account issues like depreciation, return on investment, and risks. The second improvement will be that the yield-gap will be defined at two levels. Normally, Yield-Gap analysis is based on the maximum theoretical obtainable yield assuming no growth restrictions (e.g. water, nutrients, pests, diseases, farmer practice). In the current study this will be referred to as the Theoretical Yield-Gap. However, in practice this theoretical maximum yield can hardly be obtained, considering the conditions in the study area. We will therefore use also the Achievable Yield-Gap, which is based on the maximum yield obtained in the area under optimal, but realistic, field situations. This means that it will be based on actual yields and net benefits obtained by experimental farms, or very high input farms in the region. The results of these activities will be documented in a Excel database on potentially suitable area for irrigation, which will include information on: (i) suitability for the main crops, (ii) Theoretical and Achievable Yield-Gap (in kg/ha en $/ha), (iii) main restrictions per crop for each area. During Phase 2 details regarding the focal areas (4-5 per country) will be given. At this phase substantial use of locally obtained data, information, reports and statistics will be made. The Consultants network will be combined with local representatives’ network to collect this information. It is important that once the potential irrigated areas will be determined the irrigation practice will be as efficient as technically feasible. A detailed irrigation performance analysis is beyond the scope of this study, but clear guidelines will be provided on how productive irrigation systems can be developed most effectively. Based on reports, guidelines and publications on irrigation performance an extended summary will be developed and tailored towards the situation in the NELSAP countries. More details on the SEBAL/ETLook approach can be found in the Appendix.

33

5 Methodology

5.1 Overview

The Work Program is divided into two distinct phases (Figure 8). During the first phase focus will be on all the physical components of the assessment and will be undertaken at an intermediate level of detail. At the first phase also some first assessment of the crop yield assessments will be undertaken to support the selection of the focus-areas. Also some dissemination activities are required during the first phase. The RATP Project should communicate the interim results to the countries and project stakeholders. During the second phase we will zoom in on a more detailed scale and focus besides physical analysis on environmental considerations, institutional and legal frameworks, and dissemination activities. The transition from the first to the second phase will be marked by a two days’ workshop with representatives from the Client, experts from the region and the Consultant. Such a two-phase approach ensures the best use of the limited resources and available time and also guarantees that the Phase 2 component can be executed with sufficient level of detail. Phase 2 will focus on an in-depth analysis of a maximum of 4 to 5 potential irrigation schemes (referred to as focal areas) for each of the seven countries. Selection of these 4 to 5 focal areas will be done jointly with all relevant parties.

Figure 8: Two phases approach in the project and as sociated Work Packages.

34

5.2 Work Packages

The overall Work Program has been divided into a consistent set of ten Work Packages:

• 1 Collection and review of existing information • 2 Land suitability assessment

• 3 Water resources assessment • 4 Assessment of irrigation water requirements • 5 Potential crop yield assessment • 6 Environmental, socio-economics • 7 Institutional and legal framework • 8 Integration • 9 Dissemination

• 10 Management and meetings The objectives and the output of each Work Package are presented in the following table, while details will be presented in the subsequent sections: WP 1: Collection and review of existing information on irrigation potential The objective of WP 1 is:

To collect and review the relevant existing information from local and public domain databases and documents.

The outputs of WP 1 are:

• Organised set of data and documents on assessing irrigation potential. • Local complementary information to support the assessment.

Task 2: Land suitability assessment The objective of WP 2 is:

To produce a land suitability assessment of the Nile Equatorial Lake countries focusing on irrigation suitability.


• Identification of potential cropping patterns using an evaluation of terrain, soil and vegetation productivity characteristics.

Work Package 3: Water resources assessment The objective of WP 3 is:

To assess the available water resources at sub-basin level. The outputs of WP 3 are:

• Analysis of water resources availability for irrigation development at different spatial scales.

• Maps with available runoff and base flow for areas with high irrigation potential identified during the land suitability analysis.

Work Package 4: Assessment of irrigation water requirements The objective of WP 4 is:

To assess the irrigation water requirements of potential suitable crops

35

The outputs of WP 4 are: • Maps with net irrigation requirements at field and command area level. • Maps with irrigation efficiencies at field and command area level

Work Package 5: Potential crop yield assessment The objective of WP 5 is:

To assess the crop yield for selected crop commodities. The outputs of WP 5 are:

• Quantification of the potential economic returns for an irrigation system. • Guidelines to develop effective irrigation systems.

Work Package 6: Environmental and socio-economic considerations The objective of WP 6 is:

To assess economic, political, social and environmental aspects of the irrigation developments


• An overall socio-economic and environmental assessment of the impact of developing the proposed irrigation schemes

Work Package 7: Institutional and legal framework The objective of WP 7 is:

To provide a first assessment of institutional and legal framework of the project. The outputs of WP 7 are:

• Guidelines for developing an appropriate institutional and legal framework. Work Package 8 Integration The objective of WP 8 is:

To integrate the analysis into final maps at the intermediate and the detailed scale level. The outputs of WP 8 are:

• Potential irrigated areas • Detailed description of 28-35 sites

Work Package 9: Dissemination The objective of WP 9 is:

To disseminate the study outputs during stakeholder consultations and promote in-country use.

The outputs of WP 9 are: • Reports

• Website • Donor project acquisition package

36

Figure 9. Time planning. (T) indicates travel to th e region (note that one travel can serve more than one Work Package, details in section 6.3) .

5.3 Phase 1

5.3.1 Collection and review of existing information

Review of existing global information

This task will start immediately at the beginning of the project and will focus on collecting all relevant data for this study. The following information will be reviewed:

- Existing studies on irrigation potential - Water sector development programs - Agricultural and/or irrigation sector reviews - Project appraisal documents (planned dams, other infrastructure) - Statistics (FAO)

For the quantitative component of the assessment, the following data will be gathered. Local data will be used, which is in many cases already quality controlled and checked and available from various sources:

- Digital Elevation Models (ASTER, SRTM) - Soil (Harmonized World Soil Database, FAO) - Landuse (AfriCover, Global Irrigated Area Map, SPAM) - Rainfall and other climate variables (FEWS-NET, Persiann, GSOD database) - Streamflow (UNH, GRDC, River Discharge Database)

PHASE I PHASE II

Work Package and Activities 2-11

3-11

4-11

5-11

6-11

7-11

8-11

9-11

10-1

1

11-1

1

12-1

1

1-12

2-12

3-12

4-12

5-12

0 Inception T1 Collection and review of existing information T

1.1 Review of existing global information1.2 Local information / stakeholder consultation

2 Land suitability assessment T2.1 Terrain suitability evaluation2.2 Soil suitability assessment2.2 Land productivity2.3 Potential Cropping patterns

3 Water resources assessment T T3.1 Sub basin delineation3.1 Precipitation3.2 Water balances

4 Assessment of irrigation water requirements4.1 Net irrigation requirements4.2 Total Irrigation Water Requirements

5 Potential crop yield assessment T5.1 Crop and Yield-Gap Analysis5.2 Irrigation performance improvement

6 Environmental, socio-economics T7 Institutional and legal framework T8 Integration T T

8.1 Potential Irrigated Areas8.2 Detailed plans

9 Dissemination T T9.1 Reporting9.2 Website9.3 Donor targeted dissemination

10 Management and meetings T T T T T T

37

- MODIS remote sensing images During Phase 1 globally and regionally available datasets, reports, publications and other relevant publications will to a large extent be used giving the project a quick start and ensuring that at the intermediate level of detail a consistent methodology is applied for all NEL countries.

Local information and stakeholder consultation

During Phase 1 the NBI-NLOs (National Liaison Officers) have been designated to facilitate the local data collection. Contracts have been signed between FutureWater and the seven NLO to act as Facilitators. A copy of the contract is included in the Appendix 0. To streamline the process of the NLOs a data form was prepared by the Consultant that will be used by the NLOs. A copy of this data form is included in the Appendix. It was agreed and confirmed during the Inception Workshop and the contacts between the NLOs and the Consultant, that a very strict scheduling should be applied to ensure that the project will be completed timely. The following two deadlines have to be followed exactly:

• 10-Apr-2011: The NLOs will inform the consultant on the availability of the requested information. For each dataset the NLOs will indicate whether:

o data are directly available, o data are available under specified conditions, o data are not available within one month.

• 30-Apr-2011: Deadline for submitting the data It was agreed during the Inception Workshop that any local data not made available after 30-Apr-2011 cannot be included in the analysis. Any missing information and data will be collected from the global dataset at the risk of lower detail and delays in project progress. It should be however emphasized that most global data sources are based on locally obtained information and/or locally ground-truthing.

5.3.2 Land suitability assessment

The following sub-tasks are included in the Land Suitability Assessment:

• Terrain suitability evaluation • Soil suitability assessment • Land productivity • Potential cropping patterns

Terrain suitability evaluation

Firstly the topography of the NEL countries will be analyzed as slopes are a crucial factor for the possibility of irrigation and the type of irrigation that can be applied. Based on the existing FAO irrigation slope classifications, a map will be generated including this standard legend defining the type of irrigation possible. Obviously, the map will indicate also those areas where irrigation is not possible due to slope restrictions. However, for some areas in Rwanda and Burundi specific interest exist in promoting and developing hill-side irrigation. A specific class will be

38

included where hill-side irrigation might be possible if other defining factors see Section 5.3.8 Integration are favourable. For Phase 1 the public domain available SRTM DEM (90m) will form the base in combination with standardized ARC-GIS procedures.

Soil suitability assessment

In this part of the analysis the soil qualities will be assessed in terms of potential productivity. This soil quality assessment will start with a standardized soil suitability assessment based on soil characteristics such as organic matter, soil water holding capacity, soil carbon and the drainage condition. These soil characteristics will be converted to suitability classes ranging from “no restrictions” for irrigation to “unsuitable”. For Phase 1 the soil assessment will be based on global data sets such as SOTER and FAO soil map.

Land productivity

An integrated land productivity index will be generated based on Remote Sensing analysis. For this the Normalized Difference Vegetation Index (NDVI) will be used. Obviously, the NDVI is also influenced by water availability, but high NDVI values during the wet season, when water is not limiting, indicate productive soils. The MODIS satellite (resolution 250 x 250 m) will be used during Phase 1. In both phases the monthly variation in vegetation conditions will be included in the productivity analysis.

Potential Cropping patterns

Based on the terrain suitability, elevation, the soil suitability and the land productivity a first assessment of potential cropping patterns will be derived. For Phase 1 this will be defined for the major crop classes. The assessment will make use of published crop requirements for soil and land (e.g. FAO, 1975; FAO, 1997a; FAO, 2005; FAO 2008).

5.3.3 Water resources assessment

The following sub-tasks are included in the Water Resources Assessment:

• Sub basin delineation • Quantification of Precipitation • Water balances assessment

Sub basin delineation

Water resources planning will be carried out on the basis of sub basins as described in the ToR. This study will follow the same basin boundaries as adopted by the NBI. However, for this specific study a more detailed sub-basin division might be required, and therefore the same Digital Elevation Model (DEM) the USGS-HydroSHEDS (resolution 3 arc-second (approx. 90 meters at the equator) will be used to delineate these more detailed sub-basins.

39

Precipitation

Precipitation data will be used from local station data. This local station data will be interpolated using the GPCC approach. Moreover, given the sparse coverage of weather stations, especially in the higher mountain areas where rainfall amounts can be significant, satellite rainfall product to obtain reliable estimates of the spatial distribution of rainfall amounts (FEWS). Analysis will be done at a monthly time-scale. Such a monthly based analysis is essential since most water availability – demand issues have a strong monthly variation, often much stronger than the year-to-year variation. Developments in satellite based data products are very fast. Starting with using only satellite data (TRMM), currently most products use local data to adjust the satellite information. Typical examples include the FEWS-NET and the PERSIANN products. These products have been proven to be very accurate nowadays, especially given their high spatial resolution (Dai, 2007). Daily estimates of precipitation for the African continent are prepared operationally at the Climate Prediction Center (CPC) for the United States Agency for International Development (USAID) as a part of the Famine Early Warning System Network (FEWS NET). The satellite rainfall product is available from October 2000 until present with a spatial resolution of 0.1 degree (~10 km) which is more than sufficient for this study. The earlier described new product, PERSIANN, will be considered as well.

Water balances

Profitable irrigation requires excess rainfall beyond evapotranspiration and sufficient runoff that can be channeled or stream flow that can be stored for later use by crops. Given the project’s spatial scale it is proposed to use a hydrological model that is able to generate relevant information in a rather short time frame, but still includes the most relevant processes for this particular project. From the various available hydrological models the PCRaster model is particular useful given the enormous size of the study area. A detailed discussion regarding the PCRaster model and its setup have been provided in Appendix 12.

5.3.4 Assessment of irrigation water requirements

The following sub-tasks are included in the Irrigation Water Requirements Assessment:

• Net irrigation requirements • Irrigation efficiency

Net irrigation requirements

The interaction between crop water needs, rainfall during the cropping season, and the water balance throughout the year determines the water shortage and thus irrigation water requirements. The SEBAL based ETLook remote sensing technique will be used to access this irrigation water requirement.

40

ETLook is a remote sensing algorithm specifically developed to compute the (i) reference evapotranspiration (ET), (ii) potential ET, (iii) actual ET and (iv) ET deficit. Since ETLook is based on radar satellite information it is applicable under all weather conditions. Its main driving force is soil moisture, derived from passive microwave sensors. ETLook provides 8-day estimates of actual evapotranspiration, evapotranspiration deficit and biomass production and these will be accumulated to monthly values. This will provide a first estimate of the water requirements of the area. ETLook analysis indicates water shortage under current conditions. To assess what the irrigation requirements would be if the area would be brought under irrigation, ETLook will also be applied for a future scenario assuming that areas would be covered by irrigated crops. The evapotranspiration deficits under this scenario can be considered as the net irrigation water requirements.

Irrigation efficiency

Irrigation water requirements combined with the irrigation efficiency will provide total water requirements for irrigation. Estimates of irrigation efficiencies are difficult to assess as these vary substantially by region and by irrigation practice. Moreover, the irrigation efficiency varies even more between different fields, command areas and irrigation system, as reuse of “non-efficient” water should be taken into account. Water infiltrating beyond the root zones of the crops can be become available for plant transpiration at a downstream location. Based on published irrigation efficiencies and the combination of soil and terrain analysis from previous Work Packages a map of expected irrigation efficiencies at field and command area level will be generated. This map will be multiplied with the net irrigation water requirements resulting in a map showing the total water requirements.

5.3.5 Potential crop yield assessment

An overall crop and yield potential analysis, based on existing information, will be generated during Phase 1. The approach followed will be based on a yield-gap analysis. The yield-gap is defined as the difference between the actual yield and the maximum obtainable yield. The results of these activities will lead to a map showing this yield-gap. It is clear that areas with a high yield-gap gains of implementing irrigation are much higher than for areas where this yield-gap is smaller.

5.3.6 Environmental, socio-economics

Economic, political, social and environmental aspects will be included in the analysis, although a detailed quantitative analysis, such as a complete Environmental Impact Assessment (EIA) or a full Poverty and Social Impact Analysis (PSIA), is beyond the scope of the current study. During Phase 1 emphasis will be put on country specific policies and regulations, based on existing and relevant studies and information such as Poverty Reduction Strategy Papers

41

(PRSP), Millennium Development Goals Monitoring Project, FAOstat and World Banks Statistical Division. Provisions to start local data collection will be made during phase 1. Two important socio-economic factors will be quantified for the seven countries to be included in the analysis of suitable irrigated areas: (i) distance to nearest roads, and (ii) distance to markets.

5.3.7 Institutional and legal framework

The development of irrigated areas, especially if introduced at large scales, will require the implementation of structural and institutional reforms. This Work Package will assess possible bottlenecks and adaptations needed for a successful development of the NEL region. A full institutional and legal framework assessment is beyond the scope of the current study, but general description at basin level as well as local level will be provided. At the basin level emphasis will be given on national water treaty agreements as far as it concerns water allocations and the impact of development of irrigation schemes. At country level land ownership rights are a key issue, especially if large-scale irrigation is considered. The results of this Work Package will be a map with institutional and/or legal restrictions in developing irrigated agriculture.

5.3.8 Integration

At the end of Phase 1 of the study all information is available to create the overall map of potential irrigated areas. The final map will go beyond the classical yes-no approach, but will include a legend ranging from “very suitable” to “not suitable”. Moreover, the map will indicate as well what the restrictions are and how severe these restrictions are. By combing the information on the terrain and soil suitability assessment and the water resources availability, it will be possible to evaluate the irrigation potential of the different identified schemes. The areas will be classified according to their irrigation potential, based on soil quality and topography, and the proximity to one of the tributaries where sufficient streamflow can be expected or runoff that can be channeled. The final result of this integration will be a map of potential irrigated areas distributed per class. Based on this map basin, sub-basin and country scale data will be extracted and compared to other relevant comparable studies. Deviations will be justified and explained. In summary, the following quantitative maps will be combined to one map indicating suitability classes for irrigation:

• Slope (WP2) • Soil (WP2) • Land productivity (WP2)

• Water availability (WP3) • Distance to stream (WP3)

42

• Water requirements (WP4) • Potential crop yield (WP5) • Population density (WP 6) • Distance to markets (WP 6) • Distance to roads (WP 6) • Institutional/legal limitations (WP 7)

It is important that the decision on which of the potential irrigated areas will be studied in detail, will be taken jointly by all relevant parties (Client, countries, donor, and Consultant). An important issue will be whether focal areas outside the Nile Basin will be considered as well. Information and data outside the Nile Basin will be, most likely, less detailed and therefore additional resources might be required to study focal areas outside the Nile Basin. During the Inception Workshop it was decided to postpone this discussion to the start of Phase 2.

5.4 Phase 2

5.4.1 Collection and review of existing information

During Phase 2 four to five suitable areas (focal areas) per country will be selected for detailed analysis. Local consultants will be used if required to collect all necessary data and information for areas that have the potential for irrigation as defined during Phase 1 of the study. Local available datasets, reports, publications and other relevant information will be collected leading to a balanced and objective feasibility analysis for the specific focal areas based on existing knowledge and information.

5.4.2 Land suitability assessment

The following sub-tasks are included in the Land Suitability Assessment for the focal areas: • Terrain suitability evaluation

• Soil suitability assessment • Land productivity • Potential cropping patterns

Terrain suitability evaluation

A slope analysis will be carried out for the focal areas as this determines potential limitations but also the type of irrigation that can be applied. For Phase 2 this analysis will be done as much as possible on existing locally available DEMs, topographic maps and the ASTER GDEM. Special emphasis will be put on selected focal areas which might be suitable for developing hill-side irrigation like in Rwanda and Burundi.

43

Soil suitability assessment

Soil suitability assessment will be based on existing soil maps. As much as possible local soil maps will be used in combination with the recently updated Harmonized Worlds Soil Database. Soil characteristics such as organic matter, soil water holding capacity, soil carbon and the drainage condition will determine the suitability for specific crops.

Land productivity

An integrated land productivity index will be generated based on Remote Sensing analysis. For this the Normalized Difference Vegetation Index (NDVI) will be used. Obviously, the NDVI is also influenced by water availability, but high NDVI values during the wet season, when water is not limiting, indicate productive soils. Landsat images (resolution 30 x 30 m) will be investigated during Phase 2.

Potential Cropping patterns

Based on the terrain suitability,, the soil suitability and the land productivity a first assessment of potential cropping patterns will be derived. For Phase 2 suitability per crop will be provided. The assessment will make use of published crop requirements for soil and land (FAO, 1975; FAO, 1997a; FAO, 2005; FAO 2008). Especially during this Phase 2 this analysis will take into accountant local practice in terms of cropping patterns.

5.4.3 Water resources assessment

A detailed water resources assessment has already been made during Phase 1 resulting in a clear picture on the availability of water and its source (surface water, groundwater). For the four to five focal areas per country a location specific water resources assessment will be made. The result will be for each focal area a description on the required infrastructure to irrigate the area.

5.4.4 Assessment of irrigation water requirements

During Phase 1 net irrigation requirements were assess using the ETLook approach. In Phase 2 this will be enhanced by using CropWat to analyse crop specific irrigation requirements. For the dominant potential crops in the focal areas a CropWat model will be used to obtain the net irrigation water requirements. Irrigation efficiencies will be determined based on the following factors:

• Soils • Slopes • Irrigation technique

• Farmer’s proficiencies • Local practice

The total water requirements for a focal area will be obtained by combining the potential area, the net irrigation requirements and the irrigation efficiencies.

44

5.4.5 Potential crop yield assessment

The following sub-tasks are included in the Potential Crop Yield Assessment:

• Crop and yield-gap analysis. • Irrigation performance improvement.

Crop and Yield-Gap Analysis

During Phase 2 of the project emphasis will be put on the agricultural aspects of the focal areas. For this an expanded yield-gap analysis will be used defined as the difference between the actual yield and the maximum obtainable yield. The analysis followed in this study will expand this approach in two main areas. First of all, an economic component will be included where yield-gap will not only be expressed as the difference between maximum yield and actual yield, but also as potential revenue in dollars. Based on a simplified economic analysis these values will be obtained, taking into account issues like depreciation, return on investment, and risks. The second improvement will be that the yield-gap will be defined at two levels. Normally, Yield-Gap analysis is based on the maximum theoretical obtainable yield assuming no growth restrictions (e.g. water, nutrients, pests, diseases, farmer practice). In the current study this will be referred to as the Theoretical Yield-Gap. However, in practice this theoretical maximum yield can hardly be obtained, considering the conditions in the study area. We will therefore use also the Achievable Yield-Gap, which is based on the maximum yield obtained in the area under optimal, but realistic, field situations. This means that it will be based on actual yields and net benefits obtained by experimental farms, or very high input farms in the region. The results of these activities will result in information on: (i) suitability for the main crops, (ii) Theoretical and Achievable Yield-Gap (in kg/ha en $/ha), (iii) main restrictions per crop for each area.

Irrigation performance improvement

It is important that once the potential irrigated areas will be determined the irrigation practice will be as efficient as technically feasible. A detailed irrigation performance analysis is beyond the scope of this study, but clear guidelines will be provided on how productive irrigation systems can be developed most effectively. Based on reports, guidelines and publications on irrigation performance an extended summary will be developed and tailored towards the situation in the NEL countries.

5.4.6 Environmental, socio-economics

For the focal areas the local economic situation and socio and environmental issues are of paramount importance. Although a full quantitative analysis (EIA or PSIA) is beyond the scope of the current study, a detailed socio-economic constraints assessment will be made. Local consultant will be supporting in getting the sometimes sensitive issues on transferring rainfed land into irrigated areas.

45

The impact that irrigation schemes may have on downstream water availability will be evaluated more in detail based on the quantitative analysis of previous Work Packages. It should be emphasized that these analysis are very relevant given the transboundary nature of the study area and on-going discussions on the impact of irrigation development on downstream users and downstream countries.

5.4.7 Institutional and legal framework

The intuitional and legal issues of the focal areas will be assessed using qualitative information obtained from local reports and especially local consultants. The development of irrigated areas, especially if introduced at large scales, will require the implementation of structural and institutional reforms. This Work Package will assess possible bottlenecks and adaptations needed for a successful development of the focal areas. The following issues will be described for each focal area: (i) land ownership, (ii) tribal setting, (iii) gender aspects, (iv) water rights, and (v) decision making processes.

46

6 Staffing and Management

6.1 Consultant Staff

A team of international recognized specialists, in combination with regional staff will be working on the project. Key staff are:

• Dr. Peter Droogers (Project Leader / Water Resources Specialist) • MSc. Vincent Kabalisa (Assistant Team Leader, Regional Water Resources Expert) • Prof. Dr. Wim Bastiaansen (Senior Irrigation Specialist) • Dr. Walter Immerzeel (Senior Water Modeler)

• MSc. Wilco Terink (Data Analyst, Hydrologist) • Dr. Wouter Meijninger (Remote Sensing Specialist) • Dr. Petra Hellegers (Water Economist) • MSc. Simon Chevalking (Environmental Expert) • Dr. Frank Steenbergen (Social Geographer)

Dr. Peter Droogers will be project leader and will have overall responsibility including contacts with the Client. Peter will also be the key person for reporting and the consultation and dissemination activities. He will be assisted by Vincent Kabalisa, a senior water resources expert, with many years of experience with projects in the region, and who worked for the NBI before. Dr. Walter Immerzeel performs the modeling activities and the GIS analysis, while MSc. Wilco Terink is responsible for all data management and analysis. Dr. Wouter Meijninger will carry out the remote sensing activities, under supervision of Prof. Dr. Wim Bastiaanssen. Dr. Petra Helleger is responsible for the economic part of the project. Simon Chevalking, and Dr. Frank Steenbergen will ensure that the environmental and the institutional and legal framework components will be covered. Dr. Peter Droogers (FutureWater) is an expert on integrated water resources management at different spatial scales with emphasize on water for food issues, climate change, decision support systems, simulation modeling in combination with data mining and remote sensing. Peter has over 15 years of experience working in The Netherlands and overseas (as a resident in Sri Lanka and Turkey). Non-resident assignments included Cambodia, Ethiopia, France, Gambia, India, Iran, Kenya, Laos, Niger, Pakistan, South Africa, Spain, Tanzania, Thailand and USA. Research was conducted at various institutions including Wageningen University, International Water Management Institute, and FutureWater. Peter is part-time lecturer at several universities and has written over 100 publications of which 50 appeared in peer-reviewed journals. He is reviewer for a number of journals and is one of the associate reviewers of the Journal of Hydrology. MSc. Vincent Kabalisa is a hydrologist-water resources expert with 15 years experiences in the Nile Basin region. Vincent is a graduate from ITC-Enschede, The Netherlands, with majors in water resources, GIS and climate change. He started his career working for the Ministry of Agriculture and Livestock in Rwanda on soil and water conservation planning. Most of his working experiences were obtained working for international organizations such as the Nile Basin Initiative (NBI), FAO and UNDP. Prof. Wim Bastiaanssen (WaterWatch) is a graduate of Larenstein International College for Land and Water Management (Velp, the Netherlands). Wim (1960) worked as a junior/senior

47

hydrologist at the Department of Water Management in Arid Zones of the DLO - Staring Centre in Wageningen (currently Alterra) for 10 years. Wim obtained his Ph.D. related to the development of the SEBAL model from Wageningen University in 1995. Thereafter, from 1997 to 2000, he was a Visiting Scientist and Research Fellow of the CGIAR International Water Management Institute (IWMI) in Colombo. With WaterWatch (2001), Wim became independent, and was able to develop a series of water management information products based on satellite measurements. Since 2007 Wim is also visiting professor at Technical University of Delft, Netherlands, in Water Resources Management and Remote Sensing. Dr. Petra Hellegers is a senior water economist with a proven track record of academic and professional accomplishments and with more than 15 years of experience as a researcher delivering value through both microeconomic as macroeconomic analysis in water management. She has been project leader of projects in various countries of the MENA region (Egypt, Morocco, Oman, Yemen) focusing on ground water extraction and water re-allocation and competing claims on water. She was the main organizer of the International Conference on linkages between energy and water management for agriculture in developing countries that took place in Hyderabad , India, 2007. She supervises (inter)national MSc students of Wageningen University and is fellow of the Mansholt Graduate School of Social Sciences. Besides, she has a large list of peer-reviewed publications in journals related with water and agricultural policy and economics. Dr. Walter Immerzeel (FutureWater) has over ten years experiences in geo-informatics, water resource management and climate change with a special focus on the interface between GIS and simulation models. He also has extensive experience in the application of Remote Sensing in mountain areas for systematically assessing and monitoring droughts and food security. He has worked in the Netherlands as well as in a number of developing countries (Bangladesh, India, Nepal, Philippines, and Tibet). In the Netherlands he has worked on several projects for Dutch waterboards focusing on climate change, water storage and flooding. From December 2002 until June 2004 he was attached to the International Centre for Integrated Mountain Development (ICIMOD) in Nepal as expert on GIS and natural resource management. Dr. Wouter Meijninger (WaterWatch) graduated at Wageningen University in Soil-Water-Atmosphere (with a specialization in Meteorology), and started a PhD-project at the Meteorology and Air Quality Group (Wageningen University) in 1997. His PhD-research was focused on measuring surface fluxes (including ET) over natural (heterogeneous) landscapes using scintillometry, eddy-covariance techniques and remote sensing techniques (such as SEBAL). He conducted field experiments in Turkey (in collaboration with the International Water and Management Institute), Italy, China, Germany and The Netherlands. After finishing his dissertation in 2003, Wouter continued within the same group on a Post-doctoral research project on microwave scintillometry for measuring evaporative fluxes at kilometer scales. In July 2006, Wouter joined WaterWatch as an agro-meteorologist and remote sensing specialist. Wilco Terink (MSc.) is a hydrologist with experience in catchment hydrology, rainfall-runoff modeling and climate change impact assessments. His experience was gained at Wageningen University, where he completed his Masters study in Hydrology and Quantitative Water Management, and where he worked as a researcher in hydrology for three years. During this period, he studied the impact of land use changes and climate change on streamflow dynamics of the river Rhine. He is especially skilled in the use of spatially distributed hydrological models

48

with input from Global Circulation Models (GCMs) and Regional Climate Models (RCMs). Wilco is co-author and first author of several scientific peer-reviewed publications. Johannes Hunink (M.Sc.) is a water management specialist with experience in flood hazard assessments, urban and agricultural drainage studies, and planning of surface - and groundwater resources. His particular interest is in the use of remote sensing data with GIS for water resource management and hydrological modeling. He has experience with a wide range of commercial and public domain hydrological and flood routing models. However, his major skills are in the development of computational tools for water resources planning and forecasting, databases and GIS technology. Johannes has worked for a broad range of organizations and has lived in various countries including Ecuador, The Netherlands and Spain. Ir. Simon Chevalking has over the past 5 years worked in water and natural resources management - covering groundwater modelling, management and governance, water and natural resources assessments, agriculture and drinking water supply and service provision. He has been involved in on the ground project implementation as well as coordinating and preparation of projects (proposals, workshops and networks). His skills include; integrated field surveys, sector assessments, partnership building and policy studies. He has worked in a variety of international geo-physical context with local and international stakeholders combining both professional knowledge and interpersonal skills relating (local) data and knowledge in development, assessment or research projects. Dr. Frank van Steenbergen has worked over the past twenty years in water resource management and local development – covering drinking water, groundwater resource management, small and large scale irrigation, spate irrigation, water harvesting and drainage. He has been involved in both in on the ground project implementation and the preparation of policy documents such as Framework for Action (Second World Water Forum), Dialogue on Water Governance (Third World Water Forum) and the Review of Mainstreaming Water and Environment (DGIS). His skills concerns preparation of business plans and supporting user management, program and project management, capacity building, local planning partnership building, institutional change processes, policy studies, social assessment, strategic planning and financial management. He has worked in a large number of groundwater management projects, prepared policy documents on the topic as well as the training package on “Participatory Groundwater Management”. Various country specific facilitators will be involved in the study to ensure detailed information will be used. First of all, knowledge and information from the following RATP-NLO (National Liaison Officers) staff will be used:

Country Name

Burundi Jean-Marie Bukuru

DR Congo Okitolembo Otshusi Henri

Kenya Hosea Wendot

Rwanda Ngabonziza Prime

Uganda Ssozi Frederick Ivo

Tanzania Amandus David Lwena

Sudan Mary Banjamin

49

These NBI-NLOs has been agreed upon facilitating the country specific data collection. An agreement to this end was signed on 29-Mar-2011 and a copy can be found in the Appendix. The Consultant has contacts and lists with various experts and institutions in all countries who worked on the completed Italian supported FAO project “Information Products for Nile Basin Water Resources Management”. Between 10 and 20 experts per country are known to be available to support the study and will be approached if specific information and/or analysis are required. When the final selection of focal areas has been made, these local consultants will be approached. The strategy to get the required information includes the following steps:

• Approach potential local experts using the inventory as shown in “15 Appendix: Local expert selection”

• Selection and contracting experts • Make detailed list of required information and discuss with local experts • Obtain requested information.

6.2 Management and meetings

The management of the project will be focused on a strict output oriented approach. Planning and delivery will be closely monitored and frequent interactions between Consultant’s staff will be enforced. Essential component of the management is close interaction with Client throughout the entire project. Client has nominated a contact person for this specific project. Many informal meetings are foreseen and two more formal meetings are proposed. At the end of Phase 1, month 6, a small symposium will be organized to discuss results and to mark the start of the detailed level analysis Phase 2. Operational costs for this meeting will be covered by the Consultant. It is proposed to organize one final symposium at the end of the project (costs covered by Client). The Client will cover the organizational cost of the symposium. The Consultant’s operational costs (travel, hotel) will be covered by the Consultant. A final symposium will be organized at the end of the study. The Consultant will facilitate the symposium and cover the operation cost of its Consultants, while the Client will organize the symposium and cover participants’ operational costs.

6.3 Travel

A large component of the work consistent out of using advanced quantitative tools and analysis: modeling, remote sensing, GIS and data assimilation. These activities do require a location with fast computers, and reliable and high-speed internet connection. However to ensure that these analysis will be discussed with the client intensive travel to the region is scheduled. Moreover, location specific data and information require travel to the region as well.

50

The following travel schedule is foreseen. Each travel will vary from a few days up to two weeks, depending on the needs. Note that flexibility in these dates is foreseen and details will be discussed with the client. However, practical conditions (visa, tickets, scheduling) make last minute changes often difficult. Venue of these trips will depend on the need and will be discussed with the client. Finally, it is important that a local representative will be continuously in the region and additional local staff will be used as well (see section on Staffing).

Month Staff Objectives Mar-2011 Peter Droogers Discuss Inception Report May-2011 Wilco Terink Country specific analysis Sep-2011 Peter Droogers Discuss and present Draft Report Phase 1, workshop Sep-2011 Wilco Terink Priority schemes selection, workshop Sep-2011 Wim Bastiaanssen Detailed remote sensing, workshop Dec-2011 Petra Hellegers Economic analysis Dec-2011 Simon Chevalking Environment, socio analysis Feb-2012 Wilco Terink Priority schemes analysis Apr-2012 Peter Droogers Discuss and present Draft Report Phase 2 Apr-2012 Frank Steenbergen Donor targeted mission May-2012 Peter Droogers Discuss and present Final Report

51

7 Deliverables

7.1 Overall

The tangible output of the study will include four main (set of) reports. The date of the reports are revised from the contract reflecting the delayed start of the project (5-feb-2011) and total project duration (16 months). Type of the Report Inception Report Deadline 31-Mar-2011 Content Detailed Mission work plan including: methodology, overall

implementation calendar. Schedule and itinerary of visits and consultations, structure of the reports, team composition and assigned tasks.

Type of the report Draft Report Phase 1 Deadline Sep-2011 Content One report with sub-basins and countries as chapters containing: (i)

Status of data collection and stakeholder feed-back; (ii)The overall irrigation situation in the countries; (iii) Irrigation potential in the NEL watersheds and sub-basins; (iv) List of irrigation schemes in order of priority for more in-depth treatment per country.

Type of report Draft Report Phase 2 Deadline Apr-2012 Content One report per country containing: (i) Findings, conclusions and

project opportunities in the country and in a regional perspective; (ii) Detailed description of selected schemes (4 to 5 schemes per country); (iii) Draft/templates of information materials, maps and other visual outputs for future dissemination.

Type of report Final report Deadline May-2012 Content One report for each country (in English for English speaking countries

and in French for French speaking countries) and one report for the overall Nile Basin, with chapters for each sub-basin presenting: (i) The general irrigation potential; (ii) The most promising irrigation schemes and project opportunities as well as; (iii) The complete set of information and dissemination materials ready for use.

7.2 Draft Report Phase 1

The Draft Report Phase 1 will be completed at the end of July 2011. The main objective of this report is to present the findings of the overall investigation on the potential for irrigated agriculture in the seven countries. This will be achieved by summarizing the findings from previous studies and new knowledge obtained during this study. The outline of the report will be:

52

1. Introduction 2. Review previous studies 3. Methodology and tools 4. Country results

4.1. Burundi 4.1.1.Overview 4.1.2.Biophysical conditions 4.1.3.Socio-economic conditions 4.1.4.Priority areas

4.2. Eastern DR Congo 4.2.1.etc…

4.3. Kenya 4.4. Rwanda 4.5. South Sudan 4.6. Tanzania 4.7. Uganda

5. Sub basin results 5.1. Kagera 5.2. Lake Victoria 5.3. Victoria Nile 5.4. Lake Albert 5.5. Albert Nile 5.6. Sudd 5.7. Bahr el Ghazzal

6. Conclusions

7.3 Draft Report Phase 2

The Draft Report Phase 2 is scheduled to be completed in April 2012. The main objective of this report is to present per country detailed information for a set of selected potential irrigation schemes. In fact the Draft Report Phase 2 consists out of seven reports describing results per country. These seven reports will have the same structure: 1. Introduction 2. Methodology and tools 3. Scheme A

3.1. Overall description 3.2. Biophysical conditions

3.2.1.Water 3.2.2.Climate 3.2.3.Soils 3.2.4.Crop options

3.3. Socio-economic conditions 3.3.1.Economic mapping 3.3.2.Distance to markets

3.4. Summary

53

4. Scheme B 4.1. Etc.

5. Scheme C 5.1. Etc.

6. Conclusions

7.4 Final Report

The final report is due one month after the Draft Report Phase 2. The Final Report will therefore include similar results and analysis, including additional feedback on the Draft Report Phase 1 and 2. The Final Report exists out of two sets of reports: one single report describing the overall irrigation potential in the NEL countries including chapters for each sub-basin. The second set of reports will be for the seven countries separately and will follow the same structure as the Draft Report Phase 2. These reports will have overview investment memos, as appendices, for each potential high irrigation systems. These investment memos will provide a brief overview and can be used as a first document to communicate with potential investors.

54

8 References

Agriculture Sector Development Strategy and Investment Plan: 2010/11- 2014-15. Agriculture for Food and Income Security. 2010. Ministry of Agriculture, Animal Industry & Fisheries. Republic of Uganda.

Alejandro, N., M. Johnson, E. Magalhaes, X. Diao, L. You, and J. Chamberin. 2009. Priorities for realizing the potential to increase agricultural productivity and growth in Western and Central Africa. IFPRI discussion paper 00876.

Anonymous, 2008. Nile Basin Initiative. Best practices for water harvesting and irrigation. EWUAP project, Nile basin Initiative

Baligira R. 2008. Rapid Baseline Assessment. Final Report. Nile Basin Initiative. Efficient Water Use For Agricultural Production (EWUAP) Project.

Bierkens, M.F.P. and L.P.H. van Beek. 2009. Seasonal predictability of European Discharge: NAO and Hydrological Response Time. J. Hydrometeor, 10, 953–968.

CIA. 1999. World Factbook. https://www.cia.gov/library/publications/the-world-factbook/

Dai, A., Xin Lin, Kuo-Lin Hsu. 2007. The frequency, intensity, and diurnal cycle of precipitation in surface and satellite observations over low- and mid-latitudes. Clim Dyn (2007) 29:727–744.

Dawelbeit M.I. 2008. Best Practices for Water Harvesting and Irrigation in Sudan. Nile Basin Initiative. Efficient Water Use For Agricultural Production (EWUAP) Project.

Droogers, P., D. Seckler and I. Makin. 2001. Estimating the potential of rainfed agriculture. IWMI Working Paper 20.

FAO. 1993. Guidelines for land-use planning. FAO Development Series 1.

FAO, 1985. Guidelines: Land evaluation for irrigated agriculture, FAO Soils Bulletin 55

FAO. 1975. Doorenbos, J. and Pruitt, W. O. Guidelines for predicting crop water requirements, Irrigation and Drainage Paper 24, Food and Agriculture Organization of the United Nations, Rome, 179 p.

FAO. 1997a. Irrigation potential in Africa: a basin approach. FAO Land and Water Bulletin 4.

FAO. 1997b. Land quality indicators and their use in sustainable agriculture and rural development. FAO Land and Water Bulletin 5, 212 p.

FAO. 1998. Crop evapotranspiration - Guidelines for computing crop water requirements - FAO Irrigation and drainage paper 56.

FAO. 2005. Aquastat survey: Irrigation in Africa in figures. FAO Water Reports, 29

FAO. 2008. Water and the Rural Poor. Edited by: Jean-Marc Faurès and Guido Santini FAO Land and Water Division.

FAO/IIASA/ISRIC/ISSCAS/JRC, 2009. Harmonized World Soil Database (version 1.1). FAO, Rome, Italy and IIASA, Laxenburg, Austria.

Iessime N. 2007. Best Practices for Water Harvesting and Irrigation in DR. Congo. Nile Basin Initiative. Efficient Water Use For Agricultural Production (EWUAP) Project.

IFPRI, FAO Trade and Food Security Database (2005).

Immerzeel W.W., P. Droogers, W. Terink, J. Hoogeveen, P. Hellegers, M. Bierkens. 2011. Middle-East and Northern Africa Water Outlook. World Bank.

55

Immerzeel, W.W., Beek, L.P.H., Bierkens, M.F.P., 2010, Climate Change Will Affect the Asian Water Towers. Science 328: 1382-1385.

Isaya V.S. 2007. Rapid Baseline Assessment of Agricultural Water in Kenya. Nile Basin Initiative. Efficient Water Use For Agricultural Production (EWUAP) Project.

Iwadra M. 2007. Best Practices for Water Harvesting and Irrigation in Uganda. Nile Basin Initiative. Efficient Water Use For Agricultural Production (EWUAP) Project.

Jarvis A., H.I. Reuter, A. Nelson, E. Guevara. 2008. Hole-filled seamless SRTM data V4. International Centre for Tropical Agriculture (CIAT). Available from http://srtm.csi.cgiar.org.

Loos, S., H. Middelkoop, M. van der Perk and R. van Beek (2009). Large scale nutrient modelling using globally available datasets: A test for the Rhine basin. Journal of Hydrology 369: 403-415.

Loveland, T.R., Zhu, Z., Ohlen, D.O., Brown, J.F., Reed, B.C., and Yang, L., 1999. An Analysis of the IGBP Global Land-Cover Characterization Process. Photogrammetric Engineering and Remote Sensing, v. 65, no. 9, p. 1021-1032.

Mayaux, P., E. Bartholomé, S. Fritz, A. Belward. 2004. A new land-cover map of Africa for the year 2000. Journal of Biogeography (J. Biogeogr.) (2004) 31, 861–877

Mburu D. 2008. Best Practices for Water Harvesting and Irrigation in Kenya. Nile Basin Initiative. Efficient Water Use For Agricultural Production (EWUAP) Project.

Mitchell T.D., T.R. Carter, P.D. Jones, M. Hulme, M. New. 2004. A comprehensive set of high resolution grids of monthly climate for Europe and the globe: the observed records(1901 - 2000) and 16 scenarios (2001-20016), Climate Research Unit, School of Environmental Sciences, University of East Anglia, UK

Niyongabo H. 2007. Best Practices for Water Harvesting and Irrigation in Burundi. Nile Basin Initiative. Efficient Water Use For Agricultural Production (EWUAP) Project.

Ntamavukiro, A. 2007. Rapid baseline assessment of agricultural water in Burundi. EWUAP project, Nile basin Initiative. (French).

Ormsbee L.E. and Khan, A.Q., 1989. A parametric model for steeply sloping forested watersheds. Water Resources Research 20: 1815–1822.

Petrescu, A. M. R., L. P. H. van Beek, J. van Huissteden, C. Prigent, T. Sachs, C. A. R. Corradi, F. J. W. Parmentier, and A. J. Dolman (2010), Modeling regional to global CH4 emissions of boreal and arctic wetlands, Global Biogeochem. Cycles, 24, GB4009.

Reuter H.I, A. Nelson, A. Jarvis. 2007. An evaluation of void filling interpolation methods for SRTM data. International Journal of Geographic Information Science. 21(9). pp. 983-1008.

Rwehumbiza P. 2007. Best Practices for Water Harvesting and Irrigation in Tanzania. Nile Basin Initiative. Efficient Water Use For Agricultural Production (EWUAP) Project.

Salih A.A. 2007. Rapid Baseline Assessment of Agricultural Water in Sudan. Nile Basin Initiative. Efficient Water Use For Agricultural Production (EWUAP) Project.

Siebert, S., Hoogeveen, J., Frenken, K. 2006. Irrigation in Africa, Europe and Latin America. Update of the Digital Global Map of Irrigation Areas to Version 4. Frankfurt Hydrology Paper 05, University of Frankfurt (Main), Germany and FAO, Rome, Italy.

Sijali, I.V. 2007. Rapid baseline assessment of agricultural water in Kenya. EWUAP project, Nile basin Initiative

56

Sisila S. 2007. Rapid Baseline Assessment of Agricultural Water in Tanzania. Nile Basin Initiative. Efficient Water Use For Agricultural Production (EWUAP) Project.

Sloan PG, Moore ID, Coltharp GB and Eigel JD. 1983. Modeling Surface and Subsurface Stormflow on Steeply-Sloping Forested Watersheds. Water Resources Institute Report 142, University of Kentucky, Lexington, KY.

Sloan, P. and I. Moore, 1984. Modeling subsurface stormflow on steeply sloping forested watersheds. Water Resources Research: 20(12), 1815-1822.

Sperna Weiland, F.C., L. P. H. van Beek, J. C. J. Kwadijk, and M. F. P. Bierkens (2010): The ability of a GCM-forced hydrological model to reproduce global discharge variability. Hydrol. Earth Syst. Sci., 14, 1595-1621.

Sutcliffe, J.V. 2009. The Hydrology of the Nile Basin. Monographiae Biologicae. Volume 89, IV, 335-364, DOI: 10.1007/978-1-4020-9726-3_17

Sutcliffe, J.V.; Parks, Y.P. 1999. The hydrology of the Nile. IAHS Special Publication No 5.. Wallingford. UK.

Tapley, B.D., Bettadpur, S., Ries, J.C., Thompson, P.F. and Watkins, M.M., 2004. GRACE measurements of mass variability in the earth system, Science, pp. 503-505.

Todini, E., 1996. The ARNO rainfall-runoff model. Journal of Hydrology 175: 339-382.

UBOS. 2005. 2002 Population and Housing Census. Main Report. Kampala: UBOS.

UBOS. 2008. The Uganda Demographic and Health Survey, 2006. Kampala: UBOS.

UNEP. 2000. Water Sharing in the Nile River Valley. Diana Rizzolio Karyabwite UNEP/DEWA/GRID –Geneva.

Van Beek, L.P.H. and M.F.P. Bierkens. 2009. The Global Hydrological Model PCR-GLOBWB: Conceptualization, Parameterization and Verification, Report Department of Physical Geography, Utrecht University, Utrecht, The Netherlands, http://vanbeek.geo.uu.nl/suppinfo/vanbeekbierkens2009.pdf

Van Engelen, V.W.P., T.T. Wen. 1995. Global and National Soils and Terrain Digital Databases (SOTER). Procedures Manual. International Soil Reference and Information Centre, Wageningen, Netherlands.

Verdoodt, A. 2003 Elaboration and Application of an Adjusted Agricultural Land Evaluation Model for Rwanda. PhD thesis University of Gent, Belgium.

Wada, Y., L. P. H. van Beek, C. M. van Kempen, J. W. T. M. Reckman, S. Vasak, and M. F. P. Bierkens (2010), Global depletion of groundwater resources, Geophys. Res. Lett., 37, L20402

Wesseling, C.G., Karssenberg, D., Van Deursen, W.P.A., Burrough P.A. 1996. Integrating dynamic environmental models in GIS: the development of a Dynamic Modelling language. Transactions in GIS 1:40-48.

You, L., S.Crespo, Z. Guo, J. Koo, W. Ojo, K. Sebastian, M.T. Tenorio, S. Wood, U. Wood-Sichra. 2006. Spatial Produciton Allocation Model (SPAM) 2000 Version 3 Release 2.

Zhao, R. J., 1992. The Xinganjiang Model Applied in China. Journal of Hydrology, 135: pp. 371-381

57

9 APPENDIX: Literature Review

9.1 Introduction

An enormous amount of studies focusing on the Nile has been undertaken over the last centuries. In the scientific literature alone, over half a million publications can be found related to the Nile (Google Scholar, 14-Mar-2011). Beside these scientific studies hundreds to thousands other non-scientific literature and publications are made related to the River Nile. Using the search term River Nile in Google resulted in over five million pages found (Google, 14-Mar-2011). This section of the report is not meant to provide an inclusive summary of all publications related to the Nile. The focus of this section is to give an overview of the most relevant publications and studies in the context of the current project. A distinction is made between more general irrigated related studies, other relevant studies and country specific studies.

9.2 Irrigation related studies

Döll, P. and S. Siebert. (2002) Global modeling of irrigation water requirements. Water Resources Research, Vol. 38, No. 4: Worldwide almost 90% of the water consumption is used for irrigation purpose. With a rapidly increasing population it can be questioned whether enough water will be available to increase the food production accordingly. This study aims to give a global view, at a relatively small scale (0.5º by 0.5º), of the irrigation water requirements. For this reason the current distribution of irrigated land is modelled first. As there is not sufficient information available on what crops are grown under irrigated conditions where and when, the cropping patterns and the growing seasons are also simulated by the model, based on soil suitability and climate. Furthermore, a distinction is made between only two crop types, rice and no rice.

FAO 1997 – irrigation potential in Africa, a basin approach: There is a growing concern about the food security in sub Saharan Africa as the import of cereals is projected to triple from 1990 to 2020. Africa is (apart from Australia) the driest continent in the world, with a highly unstable rainfall regime. Droughts are frequent, which put more people at risk each year. Agricultural productivity has not been able to keep up with the population growth. As the cultivated land can hardly be increased the solution should be to increase the yields. The irrigated area of 8.5% of the cultivated area is far beneath the world average of 17%. In the areas where irrigation is most needed the water is getting scarce due to population growth, urbanization, and industrialization. This study concentrates on the quantitative assessment based on physical criteria.

Definition of irrigation potential

58

This study refers to irrigation as the process by which water is diverted from a river or pumped from a well and used for the purpose of agricultural production. Areas under irrigation thus include areas equipped for full and partial control irrigation, spate irrigation areas, equipped wetland and inland valley bottoms, irrespective of their size or management type. It does not consider techniques related to on-farm water conservation like water harvesting.

Figure 10: Assessment of irrigation potential.

Methodology This study is carried out per river basin. This summary will be focused on the Nile river basin. Criteria are defined to determine the physical resources (Figure 10). The type of irrigation is set on surface irrigation. Annual renewable water resources are calculated per country mainly based on surface water. Non-renewable water resources are not taken into account. Assessment of the irrigation potential, based on soil and water resources, can only be done by simultaneously assessing the irrigation water requirements, which in turn depend on the cropping pattern and climate. For this reason, irrigation cropping pattern zones were defined for current and potential scenarios and (net and gross) water requirements were computed. Although the physical resources are the main concern of this study, it is acknowledged that economic, political, social and environmental issues are essential for a holistic view. This study will highlight the most important environmental issues related to irrigation.

Soil and terrain suitability for surface irrigation Two land use types have been considered, the upland crops and rice under irrigation. In case the soil is suitable for both; priority will be given to rice. The following characteristics have been used to assess the soil quality: topography, drainage, texture, surface and subsurface

59

stoniness, depth, calcium carbonate level, gypsum status, salinity and alkalinity conditions (Table 5). Table 5: Soil and terrain suitability for surface i rrigation by country. 1

Area in ha Country (1)

Total area of the country (2)

Soil suitable for irrigation of rice

(3)*

Soil suitable for irrigation of

upland crops (4)*

Total area of soils suitable for

surface irrigation (5)

As % of total area of

country(5/2)* 100 (6)

BURUNDI 2 783 400 302 100 286 700 588 800 21

CONGO 34 200 000 9 257 600 45 600 9 303 200 27

EGYPT 100 145 000 6 477 400 655 900 7 133 300 7

KENYA 58 037 000 11 405 600 5 979 100 17 384 700 30

RWANDA 2 634 000 220 600 80 300 300 900 11

SUDAN 250 581 000 66 955 100 1 814 100 68 769 200 27

TANZANIA 94 509 000 23 344 700 908 700 24 253 400 26

UGANDA 23 588 000 7 652 000 23 700 7 675 700 33

Total for Africa 3 029 020 800 511 998 900 84 961 100 596 960 000 20

Figure 11: Internal renewable water resources by co untry (in km 3).

1 Note that figures might vary slightly from source to source and here the original numbers are presented.

60

Water resources The water resources can only be assessed on basin level, although the exchange of water through rivers is very important for some countries (Figure 11). The available information comes from a multitude of sources so no reference period has been set. The internal renewable water resources and global renewable water resources have been calculated. If no information was available, estimation was made by multiplying the precipitation by the runoff coefficient. Evaporation from open waters does have a significant influence on the water balance. This has been considered as much as possible. The distribution of the water resources have not been specified further than country level. Irrigation water requirements (IWR) By dividing the available water by the gross irrigation water requirement the maximum irrigated area can be calculated. Because of the scale, assumptions had to be made on the definition of areas to be considered homogeneous in terms of rainfall, potential evapotranspiration, cropping pattern, cropping intensity and irrigation efficiency. First the major irrigation cropping patters where delineated. Second the climatic zones are defined, based on climate stations. The combination of the cropping zones with the climate zones resulted in 1437 areas, homogeneous in irrigation cropping characteristics and climate. The model to calculate the Nett IWR was run for three scenarios and divided by the efficiency to calculate the Gross IWR. The influence of selecting cropping pattern zones and the estimations used for cropping intensity and irrigation efficiencies are of prime importance for the final results. The potential efficiency and the net and gross irrigation water requirement per area have been listed in a table. Results Nile basin A review has been given per river basin. This review describes the hydrological situation, the water resources, and the irrigation potential. Table 6 gives a quick insight. The complete review is available at the following link: http://www.fao.org/docrep/w4347e/w4347e0k.htm#the nile basin. It is evident that the figures for some countries of this table are not accurate. It is why the use of remote sensing data is relevant for this study.

Table 6: Nile basin, irrigation potential, water re quirements, water availability and areas under irrigation

Environmental and socio economic considerations Irrigation has contributed to poverty alleviation and food security, but the sustainability of irrigated agriculture is questioned, both economically and environmentally. To ensure a sustainable project, funds for maintenance should be available, and the project should be

61

environmental and social embedded. Large scale irrigation project can change the hydrological situation, which may cause groundwater drop, reduced downstream water supply, pollution, erosion, waterlogging, salinization and increased nutrient levels. Water-related diseases which are commonly associated with the introduction of irrigation should be considered as well. The construction of an irrigation scheme could have numerous of social impacts, which have to be considered in terms of equity, ownership and poverty to develop a sustainable area. Climate fluctuations may influence the possibilities for irrigation development. In this study this is not taken into account. In regions where the irrigation is most important for agriculture, between 60% and 100% of the potential is already irrigated. Most of the potential is located in humid areas. It is estimated that over 50% from the current irrigation schemes need rehabilitation if they are to be managed to the maximum of their potential. FAO Aquastat survey (2005) Irrigation in Africa in figures: This comprehensive report presents the most recent information available, up to 2005, on water availability and its use on the African continent, with an emphasis on agricultural water use and management. It analyses the changes that have occurred since the first survey in 1995. Many terms related to water and irrigation have been defined. Sutcliffe, J. V., Y. P. Parks. (1999) The hydrology of the Nile. IAHS Special Publication no. 5: Compared to the size of the Nile basin the total flow is relatively small. Higher precipitation is associated with mountainous areas. The furthest tributary to the Nile is the Kagera, which drains the mountain areas of Burundi and Rwanda, as well as for Uganda and Tanzania, into Lake Victoria. A number of tributaries drain the forested escarpment to the northeast of the lake. Other less productive water courses drain the plains of the Serengeti to the southeast of the lake and the swamps of Uganda to the northwest. From Lake Victoria the flow continues towards the north, and reaches Lake Kyoga. This lake is essentially a grass-filled valley. Trough swamps the Kyoga Nile flows towards the west into Lake Albert. The lake also receives the inflow of the river Semliki, draining Lake Edward, Ruwenzori and other mountains. The Albert Nile leaves the lake at its northern end and flows towards Juba and Mongalla. In the reach between Lake Albert and Mongalla the river receives seasonal runoff from a number of streams known as the torrents; these provide the high flows of the river following the single rainfall season. Within the Sudd the higher flows spill from the main channel into swamps and seasonally flooded areas. Evaporation from the flooded areas greatly exceeds rainfall. The effect of this spilling is that the outflow from the swamp is only about half the inflow and has little seasonal variation. At Lake No1 the Bahr el Jebel turns east and becomes the White Nile, and the Bahr el Ghazal flows into the lake from the west. The Bahr el Ghazal basin is relatively large and has the highest rainfall of any basin within the Sudan. However, the flows of the various tributaries of the Bahr el Ghazal are spilled into seasonal and permanent swamps, and virtually no flow reaches the White Nile. This research describes the hydrological situation for every river section, lake or contributory. Possibilities to increase river flow are discussed. Examples include the Jonglei canal, and measures to reduce evaporation from the swamps.

1 This is no spelling error: Lake No is located at the confluence of the Bahr al Jabal and Bahr el Ghazal

62

Figure 12: Schematic balance of Lake Victoria, Kyog o, and Albert (km 3/year) (Source: Sutcliffe and Parks, 1999). You, L. et al (2010) – What is the irrigation poten tial for Africa? A Combined Biophysical and Socioeconomic Approach. IFPRI discussion paper 00993: Although irrigation in Africa has the potential to boost agricultural productivities by at least 50%, food production on the continent is almost entirely rainfed. The area equipped for irrigation, currently slightly more than 13 million hectares, makes up just 6% of the total cultivated area. Eighty-five percent of Africa’s poor live in rural areas and mostly depend on agriculture for their livelihoods. As a result, agricultural development is a key to ending poverty on the continent. Many development organizations have recently proposed to significantly increase investments in irrigation in the region. However, the potential for irrigation investments in Africa is highly dependent upon geographic, hydrologic, agronomic, and economic factors that need to be taken into account when assessing the long-term viability and sustainability of planned projects. This paper analyses large, dam-based and small-scale irrigation investment needs in Africa based on agronomic, hydrologic, and economic factors. This type of analysis can guide country- and local-level assessment of irrigation potential, which will be important to agricultural and economic development in Africa. Food production in the NEL basin is almost entirely rain fed. Although the water resources are ample the variability is high, and the water is spread over a wide range of agro-ecologic zones. Due to irrigation the yield can easily double compared to rainfed agriculture. For this reason irrigation is considered a main cornerstone for agricultural development and rural poverty reduction. About three percent of the cultivated area is equipped for irrigation, which is about 11% of the irrigation potential for the NEL countries.

This study has been carried out in five steps:

63

1. The assessment of the production geography, existing and potential performance of irrigated agriculture is done with the SPAM model.

2. Calculation of the potential runoff that could be used for small-scale irrigation. Attention has been paid to the interaction between crop water needs, rainfall during the cropping season, and excess rainfall throughout the year. These factors determine the potential for yield increases. Calculated with a hydraulic model.

3. Identification of the potentially irrigable areas and associated water delivery costs. All dam and potential dams are mapped, as they assume 30% of dam storage available for large scale irrigation purpose. Rehabilitation of existing dams could play an important role for irrigation. The identification of irrigable areas is based on geographical issues, rather than physical aspects. Assumed is that small scale irrigation does not have any delivery costs1. The cost for large scale irrigation, combined with water storage is calculated.

4. The annual net revenue due to irrigation expansion is maximized across potential areas and crops. The experience with irrigation is taken into account together with the investment potential for each country.

5. The internal rates of return (IRRs) to irrigation are calculated. These results show that the IRR is quite high (7%) in Kenya and probably Sudan, for the other NEL countries this number is much lower at about 2%.

This mainly economic report tries to give a better understanding of the conditions under which irrigation investments will yield their full potential. According to You et al. it is important to ensure that planned investments do not surpass a country’s financial capacity and that investments are proportional to other agricultural expenditures and value added. The investments can be based on pure economic considerations, such as maximizing yields and profits. Another approach could be to secure food to all countries, or to limit the area for instance by targeting the poorer regions. Investment decisions seldom depend on physical or economic criteria alone. Other non-irrigation related factors, like policies, drinking water, energy, rural development or donor suggestion may play an important role. Furthermore, irrigation is one of more productivity improving measures. Other measures include fertilizer use, advanced seed delivery systems, postharvest processing facilities, and access to markets.

9.3 Other relevant studies

Allen, Richard G. et al. Crop Evapotranspiration (g uidelines for computing crop water requirements) FAO Irrigation and Drainage Paper No. 56: This study provides a methodology to calculate the reference evapotranspiration in a more accurate manner as has been done since the publication of FAO Irrigation and Drainage Paper No. 24 in 1977.

Beyene, T., D. P. Lettenmaier and P. Kabat. (2007) Hydrologic Impacts of Climate Change on the Nile River Basin: Implications of the 2007 I PCC Climate Scenarios:

1 There are questions whether this reported statement is correct.

64

A multi-model ensemble method is used to asses climate change inducted changes in hydrology, for the IPCC’s A2 and B1 scenarios. Precipitation is expected to increase up to 117% till 2040 compared to 1950-1999, and from 2040-2100 the average will be below 100% of the reference period.

Camberlin, P. (1996) Rainfall Anomalies in the Sour ce Region of the Nile and Their Connection with the Indian Summer Monsoon. Journal of Climate Volume 10: The author examines both the inter-annual and intra-seasonal variability of the July–September rains and compares them to the Indian summer monsoon. Analysis shows that a direct statistical link exists between monsoon variations in these two regions, independent of the Southern Oscillation.

Camberlin, P. (2009) Nile Basin Climates: The climate is characterized by a gradual transition between the dry north of Sudan and the increased monsoon precipitation south in the Nile basin. The inter-annual climate variation is strong, but is only indirect influenced by El-Nino. Furthermore, the NEL region can be characterized by the occasional very wet years. (e.g. 1961, 1997).

Conway, D. and M. Hulme (1993) Recent fluctuations in precipitation and runoff over the Nile Sub-basins and their impact on main Nile disch arge. Climatic Change 25 Substantial fluctuations in precipitation and runoff have occurred over the Nile Basin in recent decades. Ten-year mean flows of the Blue Nile (Khartoum gauge) during the 20th century have ranged from 42.2 to 56.7 km3 and for the White Nile (Malakal gauge) from 25.5 to 36.9 km3. These fluctuations have been responsible for changes in decade-mean Main Nile discharge of up to ± 20% which have had important consequences for water resource management in both Egypt and Sudan. FAO (2003) Review of world water resources by count ry. Water Reports 23: This review, based on climate and hydrological data sets, of the renewable resources per country presents an overview of the physical internal and external water resources in the current situation. An attempt is made to estimate the exploitable water resources per country.

Inocencio, A. et al. (2005) Costs and Performance o f Irrigation Projects: A Comparison of Sub-Saharan Africa and Other Developing Regions. IW MI research report 109: This study aims to establish systematically whether costs of irrigation projects in SSA are truly high, determine the factors influencing costs and performance, and recommend cost-reducing and performance-enhancing options. Among other recommendation special attention should be paid to the size of the irrigation schemes, the type of crops grown, the farmer’s involvements and the integration of irrigation projects. The high failure rate of irrigation projects in SSA contributes to the fact that irrigation projects in SSA are more expensive than those in other developing regions.

Kay, M. (2001). Smallholder irrigation technology: Prospects for Sub-Saharan Africa:

65

Experience in sub-Saharan Africa has shown that successful smallholders generally use simple technologies and have secure water supplies over which they have full control. The most successful technologies are those that improve existing farming systems rather than those that introduce radically new ideas. Speeding up development does not necessarily mean building irrigation schemes faster but building many more of them. An important lesson learned over the past 20 years is that smallholder schemes develop through a slow incremental process of improvement, usually in response to farmer demand. Unfortunately this is at odds with the way in which most donor and government agencies work to specific time schedules.

Mohamed, Y. A., B. J. J. M. van den Hurk, H. H. G. Savenije, and W. G. M. Bastiaanssen. (2005) Hydroclimatology of the Nile: results from a regional climate model. Hydrol. Earth Syst. Sci. Discuss: 2: A regional climate model is applied in order to reproduce the regional water cycle as close as possible. Observations on runoff, precipitation, evaporation and radiation have been used to evaluate the model results.

Probst, J.L. and Y. Tardy. (1987) Long range strea mflow and world continental runoff fluctuations since the beginning of this century. J ournal of Hydrology, 94 Fifty major rivers, distributed all around the world, have been selected and since the beginning of this century their mean annual discharge fluctuations have been studied by filtering methods. The global runoff has been fluctuating but as an average has only increased about 3% during the last 65 years (1910–1975). The humid years seem to be centered around 1915, 1927, 1950, 1960 and 1972. On the contrary, the dry periods seem to be located around 1920, 1940, 1955 and 1965. Rosegrant, M.W. and N. D. Perez. (1997) Water resou rce development in Africa: a review and synthesis of issues, potentials, and strategies for the future. EDTP discussion paper no.28: This literature review examines how water resources development and water policy reform can be deployed to address the twin problems of food insecurity and water scarcity in Africa. Agricultural water use accounts for approximately 85% of the water withdrawals municipal for 14% and industrial for 3%. The total makes up about 2.5% of the internal water resources in eastern Africa region. Several policy reforms can stimulate and contribute to efficient water (re)use.

Svendsen M., M. Ewing and S. Msangi. (2009) Measuri ng Irrigation Performance in Africa. IFPRI Discussion Paper 00894: This research for Sub-Saharan Africa looks at six indicator categories —institutional framework, water resource use, irrigation area, irrigation technology, agricultural productivity, and poverty and food security — to assess the potential for improving performance in the agricultural food security sector through increasing irrigation sector investments. With these indicators a baseline is set to assess the improvements in the irrigation performance with extra investments. Average groundwater utilization in Sub-Saharan Africa is less than 20 percent of renewable supplies.

66

Groundwater is a resource particularly well suited for small-scale irrigation and for multiple-use systems.

Tate, E., J. SUTCLIFFE, D. CONWAY, F. FARQUHARSON. (2004). Water balance of Lake Victoria: update to 2000 and climate change modelli ng to 2100. Hydrological Sciences 49: Change in precipitation and to a lesser extent temperature over the Nile basin, could have serious consequences on regional water resources throughout the basin. To understand runoff the processes of precipitation and evapotranspiration should be understood first. Taye, M. T., V. Ntegeka, N. P. Ogiramoi, and P. Wi llems (2011) Assessment of climate change impact on hydrological extremes in two sourc e regions of the Nile River Basin. Hydrol. Earth Syst. Sci., 15. The potential impact of climate change was investigated on the hydrological extremes of Nyando River and Lake Tana catchments, which are located in two source regions of the Nile River basin. The results reveal increasing mean runoff and extreme peak flows for Nyando catchment for the 2050s while unclear trend is observed for Lake Tana catchment for mean volumes and high/low flows. The unclear impact result for Lake Tana catchment implies that the GCM uncertainty is more important for explaining the unclear trend than the hydrological models uncertainty. Yates, D. N., and K. M. Strzepek. (1998) Modeling the Nile basin under climate change. Journal of hydrologic engineering: A monthly water balance model is used to assess the potential climate change impacts on Nile runoff. Almost all models give a significant increased discharge for the NEL region. You, L. et al. (2007) Generating Plausible crop dis tribution and performance maps for Sub-Saharan Africa using a spatially disaggregated data fusion and optimization approach. IFPRI discussion paper 00725. Agricultural production statistics reported at country or sub-national geopolitical scales are used in a wide range of economic analyses, and spatially explicit (geo-referenced) production data are increasingly needed to support improved approaches to the planning and implementation of agricultural development. However, it is extremely challenging to compile and maintain collections of sub-national crop production data, particularly for poorer regions of the world. Using the modified spatial allocation model, a 5-minute (approximately 10-km) resolution grid maps for 20 major crops across Sub-Saharan Africa was generated The approach provides plausible results but also highlights the need for much more reliable input data for the region, especially with regard to sub-national production statistics.

67

9.4 Country specific studies

In the Appendix more details regarding the seven countries included in this study are described. In this section an overview of the relevant country studies is included.

9.4.1 Burundi

Ntamavukiro, A. (2007) Rapid baseline assessment of agricultural water in Burundi. EWUAP project, Nile basin Initiative. (French). Niyongabo, H. (2007) Best Practices in Water Harves ting and Irrigation in Burundi. EWUAP project, Nile basin Initiative. In this EWUAP report Burundi is divided in 5 agro ecological zones. 1.57% of the total irrigable area is currently under irrigation, of which most is situated in the IMBO plain. Irrigated area has grown from 3000ha in 1960 till around 17000ha in 2008. More than 80% of the irrigation consists of rice irrigation and the other 20% is made up by food crops like tomatoes, onions, corn and potatoes. There is hardly any diversification in irrigation techniques; nearly all irrigation is done by gravity, mostly by flooding. Energy deficit doesn’t allow for mechanization and modern techniques. Multi-criteria analysis is carried out to assess the used technologies and the sites with a high potential for irrigation. Water harvesting for irrigation purpose is not commonly used in Burundi, although new initiatives are emerging. Four sites are known were runoff is stored in artificial ponds to irrigate in the dry season. Water harvesting for domestic use is more common. The terrain is a limiting factor for irrigation, as irrigation in mountains is almost non-existent. However, irrigation in marshes is increasing. Points of attention are: stagnant water to cause diseases; cattle destroy canals in search for food left after harvesting or soil compression by cattle. Competition for water is a serious issue is some places as well as erosion. A major challenge is to make irrigation systems sustainable in all aspects, such as management, farmer involvement and design as deficient maintenance and vandalism are reported. A list of actors is included.

9.4.2 Eastern DRC

Lessime, N. Best Practices in Water Harvesting and Irrigation in DR. Congo EWUAP project, Nile basin Initiative. (French)

9.4.3 Kenya

Blank, H.G. et al. (2002). The changing face of irr igation in Kenya: Opportunities for anticipating changes in eastern en southern Africa. IWMI. Kenyan government (2010) Agricultural sector develo pment strategy 2010-2020. Kenyan government, ministry of water and irrigation (2009) Irrigation and drainage master plan.

68

Mburu, D. (2008) Best Practices in Water Harvesting and Irrigation. EWUAP project, Nile basin Initiative: The FAO decided on seven agro-climatic zones for Kenya. Combined with the six main agro-ecological zones these results in a large number (71) of sub-agro-ecological zones. A diversity of water harvesting techniques are discussed and classified as in which agro-climatic zone they are most commonly practiced. Furthermore a criterion based ranking of the different water harvesting techniques gives a clear overview of the best techniques per agro-climatic zone. For zone 1-3 bench terraces are most suitable and for zone 4-7 the sand dams are ranked best. Over 80% of the irrigation methods are made up of furrow or basin irrigation. Half is fed by gravity, the other half is pumped. In the list of best practiced irrigation method the gravity fed sprinkler is ranked first, followed by drip irrigation. The best practices sites are ranked as well. The greatest challenge in Community Managed Irrigation (CMI) is lack of stable and organized market. Gravity fed systems proved to be more sustainable than pumped systems. Ngigi,. S.N. Review of Irrigation Development in Ke nya. University of Nairobi. Sijali, I.V. (2007) Rapid baseline assessment of ag ricultural water in Kenya. EWUAP project, Nile basin Initiative: The EWUAP project is mandated to bring together the regional and national stakeholders in the riparian countries to develop a shared vision on common issues related to the increase of the availability of water and its efficient use for agricultural production. Development of the irrigation sector in Kenya is still very low as indicated by the small percentage of the developed potential. In addition, a substantial proportion of the developed schemes are performing poorly due to poor system operation and maintenance and weak farmers’ organizations. Irrigation development in the Lake Victoria basin is low with only 5% of the potential exploited. In order to accelerate, policies are being reformed. Water use efficiency could be improved significantly till over 60% by scaling up new technologies like drip and sprinkler irrigation. Land and water degradation is an issue, and should be dealt with to avoid water scarcity. Thurlow, J., J. Kiringai, M. Gautam. (2007) Rural I nvestments to Accelerate Growth and Poverty Reduction in Kenya. IFPRI Discussion Paper 00723: In order to meet the Milenium Development goal (MDG) to half poverty by 2015, it is necessary to accelerate agricultural growth. The increase of agricultural spending to 10 percent of total spending is insufficient to meet this MDG. Achieving this target requires nonagricultural investments, such as in roads and market development. UN (2005) Kenya National Water Development Report. World water assessment program.

9.4.4 Rwanda

Baligira, R. (2008) Rapid Baseline Assessment. EWUA P project, Nile Basin Initiative: Agriculture in Rwanda is executed on all land types, including on land of marginal quality and on moderate to steep sloping hillsides. In large parts, soils are originally fertile, and the bimodal rainfall makes two crops a year possible, with a third crop grown in the bottom valley and drained marshlands. Due to the slopes soil erosion conservation measures are needed.

69

Rwanda’s National Agricultural Commission estimated that half the country’s farmland suffers from moderate to severe erosion. Demographic pressure is driving soil degradation in Rwanda. The country's main exports remains tea and coffee and, to a lesser extent, pyrethrum extract. Main constraints for agriculture are: the very steep slopes (50% of fields have a slope gradient above 35%), uncontrolled deforestation which leads to erosion, and the population growth which causes over-farming. In large areas the top soil layer is relatively thin, which under present land use, can be result is a total disappearance of the fertile arable layer in less than 30 years. This study includes social aspect related to agriculture. Current fertilizer use stands on less than 2 kg/farmed ha, compared to the 150 kg recommended. All types of water resources are briefly discussed. Anonymous, (2008) Nile Basin Initiative. Best pract ices for water harvesting and irrigation. EWUAP project, Nile basin Initiative A description of all 12 agro-ecological zones is given. Criteria to assess the best practices are discussed, and irrigation sites mentioned. Some gaps are noted, as a recommendation is given to carry out a study to assess the water resources per watershed. In addition the soil fertility should by monitor in order to note changes and adopt cropping pattern. Another advice is that the government thinks about a policy to stimulate private irrigation initiatives. Government of the Republic of Rwanda, Ministry of A griculture and Animal Resources (2007) Action Plan for Implementation of Agricultur al Rainwater Harvesting Interventions in Rwanda. Government of the Republic of Rwanda, Ministry of f inance and economic planning (2000) Rwanda vision 2020.

9.4.5 Sudan

Dawelbeit, M. I. (2008) Best practices for water ha rvesting and irrigation. EWUAP project, Nile basin Initiative: Sudan is divided in six agro-ecological zones of the Sudan according to Harrison and Jackson (1958). A criterion based ranking of the best practices water harvesting and irrigation areas is carries out. Salih, A. A. Rapid baseline assessment of agricultu ral water in Sudan. EWUAP project, Nile basin Initiative. UNESCO (2008) Case study volume: facing the challen ges. World water development report 3.

9.4.6 Tanzania

Droogers, P., W. Bastiaanssen (2008) Irrigation Pot ential Lake Victoria, Tanzania:

70

This study evaluates and ranks five potential irrigation schemes in the Tanzania part of the Lake Victoria Basin. Schemes included in the analysis are Bugwema, Manonga, Isanga, Nkona and Mara Valley. Rwehumbiza, P. (2007) Best practices for water harv esting and irrigation. EWUAP project, Nile basin Initiative: The agricultural sector is very much affected by inadequacy, seasonality, and unreliability of rainfall as well as periodic droughts. The three highly ranked RWH practices are bunded field plots (jaluba), spate irrigation, and ndiva in first, second and third position respectively. The second and third ranked practices are well adopted in the best site. Sisila, S. Rapid baseline assessment of agricultura l water in Tanzania. EWUAP project, Nile basin Initiative: This study addresses some general constraints for agriculture and for community managed irrigation schemes. General constraints include: poor rural infrastructure and market access, low investments in irrigation and the decline of the use of improved seeds, fertilizer and agrochemicals. Constraints for the community managed irrigation schemes mainly focus on management issues like: leadership, management formation and irrigation knowledge. Strategies to address these constraints focus on a proper project planning and documentation, to contribute to a good interaction between science and practice and to stimulate good management. Tanzanian government, ministry of water and irrigat ion. (2009). The national irrigation policy. Tanzanian government (2005) Agricultural sector dev elopment program (ASDP).

9.4.7 Uganda

Iwadra, M. (2007) Best practices for water harvesti ng and irrigation. EWUAP project, Nile basin Initiative: Due to the fact that farmers can produce at least one crop or two per year using rain fed agriculture, irrigation development is rather low in Uganda although the need for irrigation is becoming increasingly serious due to unreliable rainfall and the effect of global warming. Harvest could be increased by 50% or 100% or more if supplemental irrigation was used. The best roof water harvesting is Ferro Cement Tank (FCT) and valley tank for on stream surface runoff harvesting. These technologies are relatively cheaper to install and manage than similar category technologies. The cost of construction can be recovered in one year. PELUM (2010) A Review and Analysis of Agricultural Related Policies that Support Sustainable Agriculture. Rugumayo, A. I., N. Kiiza and J. Shima. (2003) Rain fall reliability for crop production: A case study in Uganda. Diffuse Pollution Conference Dublin 2003.

71

Ugandan government, Ministry of Agriculture, Animal Industry & Fisheries (2010) Agriculture Sector Development Strategy and Investm ent Plan: 2010/11- 2014-15. Ugandan government (2010) National development plan (2010/11 – 2014/15) UN (2005) Uganda National Water Development Report. World wat er assessment program: An extensive study concerning the water resources and all it sustainable uses. Key challenges per sector are given and the actions taken are mentioned.

72

10 Appendix: Summary Countries

10.1 Burundi 1

10.1.1 General

Burundi (Figure 13) is situated in the Great Lakes region, Central Africa. Burundi has a total area of 27,834 km2, of which 25,200 km2 consists of land, and 2,634 km2 is covered with lakes (Niyongabo, 2007). The natural and planted forests are of major importance in maintaining the ecological and hydrological balances, covering an area of almost 2,000 km2. This area, however, tends to decrease as a result of population growth. The mountainous terrain of Burundi gives it a tropical altitude climate, which is hot and humid on low altitudes, and temperate and wet on the mountains. The country’s river system is divided into two major watersheds: the Nile and the Congo basins. Statistical projections, based on the census of people in 1979 and 1990, indicate a current population estimated around eight million. The average population density would be 317 inhabitants per km2. In densely populated areas, like e.g. Buyenzi, Kirimiro, and Mimirwa, this would peak to 400-500 inhabitants per km2.

10.1.2 Socio-economy

In Burundi the majority (more than 90%) of the population depends on extensive agriculture. In 2003, agriculture was providing 95% of the total food supply, and contributed to 49% of the Gross Domestic Product (GDP) (95 USD per person per year), and 90% of foreign exchange earnings (FAO, 2005). According to socio-economic indicators, Burundi belongs to the five poorest countries in the world. The assessment of the irrigation potential project comes at the right time. This will certainly improve the living circumstances of the local population by increasing the agricultural productivity.

10.1.3 Relief, climate, and hydrography

Burundi is a mountainous area with some plains in the Imbo, Buragane, Mosso and Bugesera natural regions. The climate is tropical and tempered by altitude. Average temperature ranges between 15 and 24 degrees Celsius. However, extreme high temperatures of 33 degrees Celsius during the day are not an exception. Despite the climate challenges currently observed within the Eastern Africa region, Burundi detains an important potential for irrigation. Unfortunately this is currently underused at the moment. The average annual rainfall in Burundi is sufficient, ranging from 700 to 2000 mm per year. It is partly for this reason that rainfed agriculture is by far more dominant than irrigated agriculture. Agricultural activity is marked by two rainy seasons: the first season from February to May, which provides 60% of the total

1 This section is based on Ntamavukiro, 2007 and Niyongabo, 2007.

73

precipitation, and the second season from September to December, delivering 40% of the total precipitation. Burundi is divided into eleven natural regions and five agro-ecological zones. The plain of Imbo: lowlands (774-1000 m) with a warm tropical climate (23°C average temperature), a low amount of rainfall (annual 800-1000 mm), and a dry season of 5-6 months. The west slope of the Congo-Nile ridge: a mountainous area with elevations ranging from 1000 to 2000 m, Annual rainfall ranges from 1100 to 1800 mm and temperatures vary between 23 and 17°C. The Congo-Nile ridge: elevations range from 2000 to 2670 m, and the annual rainfall varies between 1500 and 2000 mm, and mean annual temperatures ranging between 12 and 16°C. The central plateau: elevation varies between 1500 and 2000 m, while the average annual rainfall varies between 1150 and 1500 mm, and temperatures between16 and 18°C. The East and Northeast depressions: a ltitude varying between 1320 and 1500 m, rainfall between 600 and 1100 mm, and temperature around 20oC.

10.1.4 Main crops and land use

The agricultural sector is the dominant activity in Burundi with arable land under permanent crops occupying 12,000 km2, which is 43% of the total area of Burundi. The agricultural products are mainly food crops (46% of GDP), fish products, and oilseed crops (7% of GDP and 98% of exports). The climatic conditions prevailing in the country is encouraging for a variety of food crops of which the most important in volume are:

• Bananas • Tubers (sweet potatoes, potatoes, cassava) • Legumes (beans) • Cereals (sorghum, rice) • Vegetables • Fruits

Oil crops which are essentially made of peanuts, palm oil and cotton produce about 19,000 tons of oil per year. The industrial crop production (coffee, tea, cotton, palm oil, sugar cane, tobacco, rice, and cinchona) is organized into the agro-industrial sectors. This agricultural sector provides the main export products of the country and is the main source of foreign currency. That is why it has benefited from a preferential treatment while allocating financial resources for agricultural development. Burundi has a total land area of 25,200 km2, of which 23,500 km2 is considered as potentially agricultural. Currently, the cultivated area covers about 14,000 km2, which is split up in peasant mountain farms, culture in marshes and industrial crops.

10.1.5 Agriculture

It is estimated that Burundi has about one million farms, with an average size of 0.8 ha. In these farms they practice mixed crops (mainly food crops) incorporating more or less breeding and afforestation. In the more densely populated regions (Buyenzi, Kirimiro, and Mumirwa center), the average size of farms is 0.5 ha. The largest holdings (2 to 5 ha) are located in the plains of Imbo and Moso where population densities are lower. The three seasons of agricultural

74

production allows the small producer, through cropping intensity, to develop a cultivated area multiplied by 1.5 to 2 of the real size of the holding. An issue, however, is that without the input of organic matter and without refund of minerals, the soil fertility deteriorates, the production declines, and the small farm is insufficient to sustain the family. The agricultural development in Burundi, mainly characterized by small private farms, has a lack of mechanization. This factor in combination with the mountainous terrain and the energy deficit are a major obstacle to practice large scale irrigation. Therefore irrigation is hardly developed in Burundi. Most of the irrigated fields (99%) are located in the plains of Imbo, Moso, Bugesera, and in the marshes. Mountain irrigation represents only a tiny fraction, less than 1% of the total irrigated area. It is known that the total irrigated area in Burundi represents only 0.65% of the total surface landmass, and only 1.57% of the total irrigable area. Table 7 gives an overview of the current irrigable and irrigated areas in Burundi. Table 7: Overview of irrigable and irrigated areas in Burundi (Niyongabo, 2007).

BOX: FAOstat definitions Agricultural area This category is the sum of areas under a) arable land; (b) permanent crops - land cultivated with long-term crops which do not have to be replanted for several years (such as cocoa and coffee); land under trees and shrubs producing flowers, such as roses and jasmine; and nurseries (except those for forest trees, which should be classified under "forest"); and (c) permanent meadows and pastures - land used permanently (five years or more) to grow herbaceous forage crops, either cultivated or growing wild (wild prairie or grazing land). Arable land This category include land under temporary agricultural crops (multiple-cropped areas are counted only once), temporary meadows for mowing or pasture, land under market and kitchen gardens and land temporarily fallow (less than five years). The abandoned land resulting from shifting cultivation is not included in this category. Data for “Arable land” are not meant to indicate the amount of land that is potentially cultivable. Total area equipped for irrigation Area equipped to provide water (via irrigation) to the crops. It includes areas equipped for full and partial control irrigation, equipped lowland areas, pastures, and areas equipped for spate irrigation.

75

Int$ International US Dollar. MT Metric Ton, a unit of weight equivalent to 1000 kilograms. Source: FAO Statistics Division.

76

Figure 13: Map of Burundi with the Nile basin.

Bujumbura

Kivoga

Kabezi

Musenyi

Rumonge

Karonda

Muramvya

Kayongozi

Nyanza-Lac

±0 20 40 60 8010

Kilometers

Burundi

Legend

Water

Nile

Roads

Nile basin

Borders

major town

Capital

Height in m

High : 2400

Low : 600

77

10.1.6 Summarized facts

The following Figures and Table give a quick overview of some key facts regarding agriculture

and irrigation in the country. All data are based on FAOstat and AquaStat.

Figure 14: Agricultural area and arable land in Bur undi.

Figure 15: Agricultural production in Burundi. Table 8: Area equipped for irrigation in Burundi. Burundi ha

1965 14,000

1975 14,000

1985 14,000

1995 18,000

2005 23,000

Burundi

0

500

1000

1500

2000

2500

1961

1964

1967

1970

1973

1976

1979

1982

1985

1988

1991

1994

1997

2000

2003

2006

Years

Are

a in

thou

sand

s of

ha

Agricultural area Arable land

Burundi

0

50

100

150

200

250

300

350

400

Ban

anas

Bea

ns, d

ry

Sw

eet

pota

toes

Mai

ze

Sor

ghum

Cas

sav

a

Pe

as, d

ry

Fru

it F

resh

Nes

Veg

etab

les

fres

h ne

s

Ric

e,pa

ddy

Commodity

Are

a ha

rves

ted

in 1

000

ha

78

10.2 Eastern DRC 1

10.2.1 General

The Democratic Republic of Congo (Figure 16) is the third largest country in Africa after Sudan and Algeria. The country lies on the southwestern fringes of the Nile Basin, and the Nile portion constitutes less than 2% of the national land area. It contributes to flow into the equatorial lakes region, lying along the border with Uganda. DRC has a vast central basin in a low-lying plateau with mountains in the east. In order to distinguish it from the neighboring Republic of Congo to the west, the Democratic Republic of Congo is often referred to as DRC. The DRC borders the Central African Republic and Sudan to the north, Uganda, Rwanda, and Burundi in the east, Zambia and Angola to the south, and the Atlantic Ocean to the west, and it is separated from Tanzania and Burundi by Lake Tanganyika in the east. The Democratic Republic of Congo has a total area of 2,344,585 km2, of which 2,267,048 km2 is occupied by land, and 77,810 km2 is occupied by water. Current estimates of the country’s population are 71.7 million.


Agriculture occupies an important place in the economy of the DRC. It contributed to 58% of GDP in 2002 against 35% in 1985 and less than 10% in the 1970s. But since then, exports of cash crops have plummeted. The agricultural sector provides a living for 70% of the population since the year 2002 against 63% in 2000. The largest economic activity still occurs in the informal sector. Renewed activity in the mining sector, the source of the most export income, boosted a GDP growth from 2006-2008. The government’s review of mining contracts that began in 2006, however, combined with a fall in world market prices for the DRC’s key mineral exports temporally weakened output in 2009, leading to a balance of payment crisis. The recovery of mineral prices, beginning in mid-2009, boosted mineral exports, and emergency funds from the IMF boosted foreign reserves. The country faces several issues: water pollution, deforestation, soil erosion, wildlife poaching, and mining of minerals, which cause significant environmental damage.

10.2.3 Climate

The climate in the DRC is hot and humid in the equatorial river basin, and cooler and drier in the southern highlands. North of the Equator we have a wet season (April-October) and a dry season (December-February). South of the Equator we also have a wet season (November-March), and a dry season (April-October). According to Iessime (2007), the country is divided into three agro-climatic zones: equatorial, tropical wet, and mountain climate. A summary of these agro-climatic zones is represented in Table 9.

1 This section is based on Iessime (2007).

79

Table 9: Agro-climatic zones in DRC (Iessime, 2007) .

10.2.4 Agriculture and main crops

In DRC, food crops and vegetables are currently the bulk of agricultural productions through small farm families grouped in community associations located in rural, urban and sub-urban areas. Yields and crop production are low and the country depends on imports of food products to cover domestic needs. The low development rate of irrigable land resources, whose potential reaches over 7 million hectares, is one of the major factors in the persistence of low levels of agricultural production in the DRC. The lack of mastering the techniques of water harvesting and irrigation during the rainy season declines the yields in agricultural production. Initiatives to improve water retention of rainfall on rainfed upland agriculture are lacking in the DRC. Currently the percentage of arable land in DRC is 2.9%. Permanent crops occupy an area of 0.5%. The main crops which are grown in DRC are: coffee, sugar, palm oil, rubber, tea, quinine, cassava (tapioca), palm oil, bananas, root crops, corn, fruits, and wood products.

80

Figure 16: Map of Eastern DRC with the Nile basin.

ByumbaRuhengeri

Nyagatare

Butiti

Gayaza

Ibanda

KanoniKasese

Kiboga

Kitoba

Kitoma

Masaka

Bushenyi

Kakumiro

Ntungamo

Hoima

Kabale

Masindi

Mbarara

Kabatoro

Kikagati

Kikungiri

Nyakibale

Fort Portal

±0 30 60 90 12015

Kilometers

Eastern DR Congo

Legend

Water

Nile

Roads

Nile basin

Borders

Height in m

High : 2400

Low : 600

Major town

Other town

81

10.2.5 Summarized Facts

The following Figures and Table give a quick overview of some key facts regarding agriculture and irrigation in the country. All data are based on FAOstat and AquaStat.

Figure 17: Agricultural area and arable land in DR Congo.

Figure 18: Agricultural production in DR Congo. Table 10: Area equipped for irrigation in DR Congo. DR Congo ha

1965 N/A

1975 N/A

1985 9,000

1995 11,000

2005 11,000

DR Congo

0

5000

10000

15000

20000

25000

196

1

196

3

196

5

196

7

196

9

197

1

197

3

197

5

197

7

197

9

198

1

198

3

198

5

198

7

198

9

199

1

199

3

199

5

199

7

199

9

200

1

200

3

200

5

200

7

Years

Are

a in

th

ousa

nds

of

ha


DR Congo

0

200

400

600

800

1000

1200

1400

1600

1800

2000

Cas

sava

Mai

ze

Gro

undn

uts,

with

she

ll

Ric

e, p

addy

Pla

ntai

ns

Bea

ns,

dry

Oil

palm

fru

it

Cow

pea

s,dr

y

Mel

onse

ed

Ban

anas

Commodity

Are

a ha

rves

ted

in 1

000

ha

82

10.3 Kenya 1

10.3.1 General

Kenya (Figure 19) covers an area of 582,000 km2 and has wide variations in climate, land forms, geology, soils, and land use. Elevations range from sea level at the Indian Ocean to the top of Mt. Kenya with snow at 5,200 MASL. The Nile basin in Kenya represents only 8.5% of the total area of the country. This area, however, contains over 50% of the national freshwater sources with four major rivers (Nzoia, Yala, Nyando and Sondu Miriu) draining directly into Lake Victoria. The Mara River also drains into this lake, but runs through Tanzania. Kenya is an agricultural country and depends entirely on agricultural production for subsistence and socio-economic development. About two thirds of the land area in Kenya is in the arid and semi-arid lands. The pressure exerted on the fragile ecosystems that characterize the arid and semi-arid lands lead to severe land degradation. The agricultural sector faces the challenge of producing food for a rapidly growing population. Most of the agricultural activities in Kenya are rainfed and therefore the rainfall amount and distribution are vital components of agricultural production systems. Agricultural activities contribute significantly to the economic growth and GDP of Kenya. Compared to the other sectors of development, agriculture is the main consumer of water. Due to increasing competition for water amongst other sectors, agriculture is therefore expected to produce more crop per given volume of water if agricultural production is to be sustained as a viable economic activity. There is therefore a dire need to improve water use efficiency in irrigated agriculture.


Agriculture in Kenya contributes directly to 26% of the GDP, and indirectly a further 27% of the GDP through linkages with manufacturing, distribution, and other service-related sectors. The sector produces the bulk of the country's food requirements in years of favorable weather. The agricultural sector accounts for 80% of rural employment with women providing 75% of the labor force. Agriculture contributes 60% of export earnings, 45% of annual Government revenue and produces almost all the raw materials for agro-industries. With this important contribution, development of the sector should have the greatest impact on the livelihood of the people. Kenya is, however, largely arid and semi-arid (83%) with only 17% considered as medium and high potential. Thus this limits the production potential and often leads to chronic deficits in maize, wheat, rice, sugar and edible oils.

10.3.3 Climate and hydrography

The average annual rainfall in Kenya ranges from 250 to 2500 mm, while the average potential evaporation ranges from less than 1200 to 2500 mm. The average annual temperature ranges from 10 to 30°C. From the total land area of 582,000 km2, only 16% is considered to be of high potential for agriculture (Mburu, 2008). This high potential area receives over 1000 mm of annual rainfall and accounts for less than 20% of the agricultural land. More than 50% of the

1 This section is based on Mburu ( 2008)and Sijali (2007)

83

country’s population lives in this area. The medium potential area receives between 750 and 1000 mm of rainfall per annum. This area occupies 35% of the agricultural land and carries 30% of the total population. The remaining part of Kenya (80%) is classified as arid and semi-arid land with mean annual rainfall of less than 750 mm, carrying 20% of the total population. These numbers show that the country is poorly endowed with potential for rain-fed agriculture. The future growth and development of the agricultural sector will rely on integrated water resources management that encompasses water harvesting and irrigation. The land potential in Kenya can be based on agro-climatic zones or agro-ecological zones. Agro-climatic zoning is based on rainfall amount and distribution and temperature. The main agro-climatic zones are based on their probability of meeting the temperature and water requirements of the main leading crops. There are many different rainfall distribution types in Kenya which make it difficult to produce a detailed agro-climatic zone classification to cater for all variations in rainfall and temperature. There are seven main agro-climatic zones in Kenya according to Mburu (2008), based on the average monthly rainfall and potential evapotranspiration.


The humid, sub-humid and semi-humid areas are mainly above 1,500 MASL and are characterized by intensive farming for cash and subsistence. Large farms and estates with tractor mechanization coexist with small holdings using oxen or hand labor. Major crops include tea, coffee, maize, wheat, cut flowers, vegetables, fruits, sugarcane, beans and bananas. High grade dairy cattle are common in these areas but are often stall fed due to shortage of land for grazing. Improved breeds of sheep, pigs and poultry are also found in these high potential areas. The main forest areas, both indigenous and planted, are found above 1,500 MASL but occupy less than 3% of Kenya’s land area. The semi-arid areas are characterized by mixed crop and livestock farming whereas the arid and very arid areas are associated with pastoralism and wildlife. Crops grown in the semi-arid areas include maize, sorghum, millet, beans, cow peas, pigeon peas and irrigated vegetables. Cotton and sisal are sometimes grown. The arid and semi-arid lands support 35% of Kenya’s cattle, 67% of sheep and goats and all camels. Irrigation is practiced on a relatively small but increasing scale depending on water availability. The agriculture in Kenya is characterized into smallholdings, medium holdings and large holdings (Table 11). The high and medium potential areas continue to be devoted to intensive crop and milk production systems. Small-scale farming is mainly practiced in the high and medium potential areas and accounts for 75% of the total agricultural output and 70% of the marketed agricultural produce. Small-scale farmers produce over 70% of maize, 65% of coffee, 50% of tea, 65% of sugar-cane, 80% of milk, 70% of beef and related products, and almost 100% of the other food crops (millet, sorghum, pulses, vegetables, roots and tubers) (Isaya, 2007). Smallholdings, defined as agricultural land between 0.2 ha and 10 ha in size, occupy 3.2 million ha (46% of the total agricultural land) and accommodate 3.5 million households (98% of the total farm households). The average size of smallholdings is 0.9 ha. Kenya’s large-scale farming is practiced on farms averaging 50 hectares. The large scale sub sector accounts for 30% of marketed produce and is mainly involved in growing crops such as tea, coffee, horticultural produce, maize and wheat.

84

Table 11: Agricultural land sizes in Kenya (Isaya, 2007).

85

Figure 19: Map of Kenya with the Nile basin.

Lira

Jinja

Mbale

Kitgum

Soroti

Tororo

Voi

Molo

Gazi

EmbuNyeri

Kisii

Thika

Nakuru

Magadi

Kitale

Kisumu

Kilifi

Gilgil

Nanyuki

Nairobi

Mombasa

Malindi

Kericho

Kajiado

EldoretBungoma

Naivasha

Machakos

Kakamega

Elburgon

Matondoni

Athi River

Mwaluvanga

±0 70 140 210 28035

Kilometers

Kenya

Legend

Water

Nile

Roads

Nile basin

Borders

Major town

Other town

Height in m

High : 5882

Low : -13

86



Figure 20: Agricultural area and arable land in Ken ya.

Figure 21: Agricultural production in Kenya.

Table 12: Area equipped for irrigation in Kenya. Kenya ha

1965 14,000

1975 40,000

1985 42,000

1995 70,000

2005 103,000

Kenya

0

5000

10000

15000

20000

25000

30000

196

1

196

3

196

5

196

7

196

9

197

1

197

3

197

5

197

7

197

9

198

1

198

3

198

5

198

7

198

9

199

1

199

3

199

5

199

7

199

9

200

1

200

3

200

5

200

7

Years

Are

a in

th

ous

and

s o

f ha


Kenya

0

200

400

600

800

1000

1200

1400

1600

1800

Mai

ze

Bea

ns,

dry

Pig

eon

peas Tea

Cof

fee,

gree

n

Cow

pea

s,dr

y

Whe

at

Pot

atoe

s

Sor

ghum

See

d co

tton

Commodity

Are

a ha

rves

ted

in 1

000

ha

87

10.4 Rwanda 1

10.4.1 General

Rwanda (Figure 22) is located in the Southern West of the Victoria Basin and belongs to the Upper Nile River States. The country shares its borders with the Democratic Republic of Congo in the west, Uganda in the north, Tanzania in the east, and Burundi in the south. Rwanda has a total area of 26,338 km2 and is divided into two main basins: the Congo basin representing 17% of the area, and the Nile basin representing 83% of the area (Baligira, 2008). Its relief comprises succession of relatively large hills and valleys. More than 40% of the country is located on an altitude of between 1,500 MASL and 1,800 MASL. 90% of the national water resources are drained through the southern and eastern part by the main rivers Nyabarongo, Akanyaru and Akagera. The surface occupied by lakes, rivers and marsh is 2,125 km2, hence approximately 8% of the national territory. Lakes have an area of 1,282 km2, whereby the Kivu Lake alone accounts 1,028 km2. Permanent rivers cover 73 km2 whereas the marshes and bottom valleys add up 1700 km2. Rwanda belongs to one of the highest populated countries in Africa with 321 persons per km2, and 90% of the population lives from food subsistence agriculture.


Farming is the principal economic activity of the Rwandan people, carried out on more than 1.4 million farm households. According to the general population census in 2002, a ratio of 8 people per 10 is used in agriculture whereby most of them are women. The agricultural sector produces 45% of the GDP in the past decade (1995-2004) and generates nearly 75% of the foreign exchange earnings. In some areas, particularly in the Nile Basin catchment, the population pressure has reduced the area of arable land available per household to about 0.75 ha per household. Table 13: Agro-climatic zones in Rwanda (AQUASTAT, 2005).

1 This sections is based on Baligira( 2008) and Anonymous (2008).

88


Annual rainfall ranges from 800 mm to above 1,600 mm, divided between two rainy seasons (March-May and September-December). The amounts of rainfall are good in most parts of the country, but there is a persistent risk of drought in most areas. The temperature regime is specified as “moderate highland equatorial” with average temperatures between 16° and 23°C. Based on elevation, available rainfall and soil conditions, the country has been divided into eight different agriculture regions (Baligira, 2008). Those regions include the Volcanoes Highlands, Buberuka North ridges, Buberuka foot ridges, Gikongoro, Lakes Kivu shores, Central plateau, Eastern lowlands and Kibungo. From the study made by Aquastat in 2005, Rwanda has been divided into 3 main regions. These regions with their characteristics are shown in Table 13. According to Verdoodt (2003), twelve agro-ecological zones (AEZs) are currently recognized in Rwanda. These zones have been determined based on climate, soil suitability, geology and geomorphology. A detailed description of these zones can be found in Verdoodt (2003).

10.4.4 Agriculture, land use, and main crops

Rwanda is facing a serious problem of low availability of the cultivable lands. The arable land is 13,850 km2, corresponding to 52% of the total surface of the country. 39% of the arable land has a high erosion risk. A consequence of farming more intensively, and farming on steep slopes is the high incidence of soil loss due to erosion, and, along with it, declining soil fertility. Rwanda’s National Agricultural Commission estimated that half the country’s farmland suffers from moderate to severe erosion. Demographic pressure is driving soil degradation in Rwanda. The cultivated area is 8,520 km2, i.e. 61.5% of the arable land and 31% of the total surface of the country. The size of cultivable land per family is 0.6 ha. Each farm in Rwanda comprises 5 to 6 members, half of them below 15 years of age. The crop for food consumption occupies 92%. The export of agricultural goods is dominated by coffee and tea. Rwanda does not satisfy the food needs for its population with its own agricultural production. In terms of potential agriculture lands, Rwanda has 1,649 km2 of swamps of which 1,119 km2

belong to the lower hydrographic systems and 531 km2 to the primary system. The total surface area under use is estimated at 938 km2, equivalent to 57% of the total area of marshlands in the country. Those marshlands are regularly flooded during the rainy season and prevent any agriculture activity. These marshlands, however, reduce the maximal flow rates during the rainy seasons and maintain a relatively high flow rate during dry seasons. Only 130 km2 of swamp is currently managed with moderate irrigation structures. Hillside irrigation is not known yet in the country. Different techniques, however, aiming at water control and soil conservation in the steep terrains are used. The agricultural survey carried out in 1984, showed that of the 1.34 million ha available for agricultural sector, only 1.1 million ha were effectively used for food production, reforestation and pastures. In regard of the Marshland Master Plan framework, Rwanda defined the priority of products to be cultivated in order to reduce the costs intended for food importation.

89

Figure 22: Map of Rwanda with the Nile basin.

MusenyiMuramvya

Butare

Byumba

Kigali

Kibuye

Gisenyi

Cyangugu

Gitarama

Ruhengeri

Nyanza

Kabuga

Ruhango

Rwamagana

Gikongoro

Nyagatare

±0 20 40 60 8010

Kilometers

Rwanda

Legend

Water

Nile

Roads

Nile basin

Borders

Height in m

High : 2400

Low : 600

Major town

Other town

90



Figure 23: Agricultural area and arable land in Rwa nda.

Figure 24: Agricultural production in Rwanda.

Table 14: Area equipped for irrigation in Rwanda. Rwanda ha

1965 4,000

1975 4,000

1985 4,000

1995 5,000

2005 9,000

Rwanda

0

500

1000

1500

2000

2500

1961

1963

1965

1967

1969

1971

1973

1975

1977

1979

1981

1983

1985

1987

1989

1991

1993

1995

1997

1999

2001

2003

2005

2007

years

Are

a in

thou

sand

s of

ha


Rwanda

050

100150200250300350400450

Bea

ns,

dry

Pla

ntai

ns

Mai

ze

Cas

sava

Sw

eet

pota

toes

Pot

atoe

s

Sor

ghum

Pum

pkin

s, s

quas

h an

dgo

urds

Cof

fee,

gre

en

Pea

s, d

ry

Commodity

Are

a ha

rves

ted

in 1

000

ha

91

10.5 Sudan 1

10.5.1 General

Sudan (Figure 25) is the largest country in Africa with an area of 250 million hectares and a population of around 35 million, living on 15% of the land mainly around the Nile and its tributaries. Sudan holds 60% of the Nile basin within its borders. The sum of the internal and external water resources available to Sudan is about 30 BCM. This shows that Sudan is already below the water stress margin of 1000 m3 per capita. The actual per capita consumption is much less than 1000 m3 as the water used fluctuates between 14 and 18 BCM out of the 30 BCM. This is mainly because of the varying nature of rain and flow of the Nile and the non-Nile streams coupled with limited available storage capacities, which are continuously decreasing by silt accumulation. The annual projections of water consumption in Sudan are shown in Table 15. Table 15: Annual projection of water consumption (i n BCM) in Sudan in 2003 (Salih, 2007).

In Sudan, many irrigation schemes have the entire infrastructure, but their cropping intensity is very low because of scarcity of water during the long dry season. Heightening of Roseires Dam, which is currently slowly executed from the country's financial resources, is supposed to avail such water. The current situation is that irrigated agriculture consumes about 94% of the water, 5% goes to human and animal consumption and 1% to industrial and other uses. Watershed degradation and sedimentation is a current issue in Sudan. Watershed degradation mainly resulted from the clearance of vast areas of forested lands for cultivation, fuel wood, brick making, and over grazing. Silt deposition in the Blue Nile and Atbara Rivers has interfered with their flow regimes. Bank erosion along the rivers has contributed to increased sedimentation elsewhere. Another issue is that Sudan has experienced many devastating floods and droughts during the last two decades.


Although production and export of oil are growing significantly in importance, agriculture still remains the major source of income for most of the country’s population, whereas 70% of them live in rural areas. This makes millions of people in the country directly dependent on natural resources for their livelihood and employment. In the period 2002-2006, agriculture contributed between 39% and 46% to GDP, employed 57% of the total economically active population, and

1 This section is based on Dawelbeit (2008) and Salih (2008).

92

contributed about 90% of the non-oil export earnings. Within the agricultural sector, crop production accounts for 53% of agricultural output, livestock 38% and forestry and fisheries 9%. The most salient features of agricultural production in Sudan are low productivity, low value of crops, high fluctuation in areas and low water use efficiency.


Sudan is characterized by its high climatic and ecological diversity, ranging from no rain desert in the north to high rainfall humid areas in the south. The country is a gently sloping plain with the exception of Jebel Marra in the West, the Red Sea Hills in the East, Nuba Mountains in the center and Imatong Hills in the South. Its main features are the alluvial clay deposits in the central and eastern, the stabilized sand dunes in the western and northern part and the red ironstone soils in the south. Annual rainfall ranges from less than 50 mm in the north, 350–800 mm in the central clay plains and savannah belt to more than 1500 mm in West Equatorial region in the south. The main rainy or monsoon season is from June to September but the duration will vary with latitude. In the south, there are two rainy seasons. There are six agro-climatic zones in Sudan (Dawelbeit, 2008). These include: desert, semi-desert, low rain savanna, high rain savanna, flood, and mountain zones. Details of these zones can be found in Dawelbeit (2008).


Land is by far the most important resources for over 80% of the population who live in rural areas for farming and herding. The arable land constitutes one third of the country area. Pasture and forest accounts 40% of the total land. Nearly 84 million hectares are cultivable. Only about 10% of this is currently utilized for agriculture. Because a large portion of these cultivated lands depend on rainfall, the amount actually cultivated in any particular year can greatly vary due to fluctuations in rainfall. Only 21% of the total arable land is under cultivation. Both annual and perennial crops are grown over a wide range of climatic conditions. The occupied areas of arable land by farming sector are shown in Table 16. The major crops grown in Sudan are sorghum, cotton, wheat, sugarcane, and gum Arabic. Table 16: Total arable land by farming sector in Su dan (Dawelbeit, 2008).

93

Figure 25: Map of Sudan with the Nile basin.

Juba

Uwayl

Torit

Rumbek

Yambio

Malakal

±0 90 180 270 36045

Kilometers

Southern Sudan

Legend

Water

Nile

Roads

Nile basin

Borders

Major town

Other town

Height in m

High : 5882

Low : -13

94


The following Figures and Table give a quick overview of some key facts regarding agriculture

and irrigation in the country. All data are based on FAOstat and AquaStat.

Figure 26: Agricultural area and arable land in Sud an.

Figure 27: Agricultural production in Sudan. Table 17: Area equipped for irrigation in Sudan. Sudan ha

1965 1,550,000

1975 1,700,000

1985 1,763,000

1995 1,946,000

2005 1,863,000

Sudan

0

20000

40000

60000

80000

100000

120000

140000

160000

196

1

196

3

196

5

196

7

196

9

197

1

197

3

197

5

197

7

197

9

198

1

198

3

198

5

198

7

198

9

199

1

199

3

199

5

199

7

199

9

200

1

200

3

200

5

200

7

Years

Are

a in

tho

usa

nds

of

ha


Sudan

0

1000

2000

3000

4000

5000

6000

7000

Sor

ghum

Mill

et

Ses

ame

seed

Gro

undn

uts,

with

she

ll

Whe

at

Veg

etab

les

fres

h ne

s

Tea

Nes

Cow

pea

s,dr

y

fora

geP

rodu

cts

Sun

flow

erse

ed

Commodity

Are

a ha

rves

ted

in 1

000

ha

95

10.6 Tanzania 1

10.6.1 General

Tanzania (Figure 28) is located in central East Africa and shares its borders with Kenya and Uganda in the north, Rwanda, Burundi and the Democratic Republic of Congo to the west, and Zambia, Malawi and Mozambique to the south. The Indian Ocean is located at the eastern border of Tanzania. Tanzania covers an area of 945,203 km2, of which 6.2% is water. The 2009 population estimates is 43.7 million.


Agricultural Sector is still the leading sector of the economy of Tanzania, despite the rapid growth of the mining sector, and accounts for over half of the GDP and export earnings. The agricultural GDP has grown at 3.3% per year since 1985, the main food crops at 3.5%, and export crops at 5.4% per year. Over 80% of the poor people live in rural areas and their livelihood depends on agriculture. Moreover, about 80% of the population live and earn their living in rural areas with agriculture as the mainstay of their living. It has linkages with the non-farm sector through forward linkages to agro-processing, consumption and export. These linkages provide raw materials to industries and a market for manufactured goods. The agricultural sector has maintained a steady growth rate of over 3% per annum over the last decade. Although this is greater than the growth rate of the population, this rate is considered to be unsatisfactory, because it has failed to improve the livelihood of the rural population whose major occupation is agriculture. This has often resulted in localized food insecurity and hunger, which has been intensified by the lack of access to external resources for households.


Tanzania is mountainous in the northeast, where Mt. Kilimanjaro, Africa’s highest peak, is located. To the north and west are the Great Lakes of Lake Victoria and Lake Tanganyika. Central Tanzania comprises a large plateau, with plains and arable land. The eastern shore is hot and humid, with the island of Zanzibar lying just offshore. Tanzania has a tropical climate with temperatures in the highlands ranging between 10 and 20°C during cold and hot seasons, respectively. The remaining part of the country has temperatures rarely falling below 20°C. The hottest period extends between November and February (25-31°C), while the coldest period occurs between May and August (15-20°C). The annual temperature is 32°C and the climate is cool in the high mountain regions. Tanzania has two rainfall regions, with one being modal (December-April), and the other being bimodal (October-December and March-May). The former is experienced in southern, south-west, central and western parts of the country, and the latter is found to the north and northern coast. One third of Tanzania receives less than 800 mm of rainfall, which is characterized as the arid or semi-arid area. Only one-third of the rest of the country has precipitation of above 1,000 mm. According to Rwehumbiza (2007), seven main agro-ecological zones are present in Tanzania. 1 This section is based on Rwehumbiza (2007) and Sisila (2007)

96


Agricultural production in Tanzania is dominated by smallholder farms (peasants), cultivating on average farm sizes of between 0.9 hectares and 3.0 hectares. About 70% of Tanzania’s crop area is cultivated by hand, 20% by ox plough, and 10% by tractor. It is mainly rainfed agriculture. Food crop production dominates the agriculture economy, where 5.1 million ha is cultivated annually, of which 85% is under food crops. Women constitute the main part of the agricultural labor force. The major constraint facing the agriculture sector is the falling labor and land productivity due to poor technologies, and dependence on unreliable and irregular weather conditions. Both crops and livestock are adversely affected by periodical droughts. Irrigation holds the key to stabilizing agricultural production in Tanzania to improve food security, increase farmers’ productivity and incomes, and also to produce higher value crops such as vegetables and even flowers. Due to variations in climatic and agro-ecological conditions, different crops are grown under different farming systems. The major staples include: maize, rice, wheat, sorghum, millet, pulses (mainly beans), cassava, potatoes, bananas and plantains. The important export crops are: coffee, cotton, cashew nut, tobacco, sisal, pyrethrum, tea, cloves, horticultural crops, oil seeds, spices and flowers.

97

Figure 28: Map of Tanzania with the Nile basin.

Same

Kyela

Lindi

Mbeya

Moshi

Ujiji

Nzega

Vwawa

Tanga

Kilosa

Kondoa

Mtwara

Muleba

Mwanza

Njombe

Newala

Songea

Bukoba

Dodoma

Iringa

Kigoma

MasasiMbinga

Musoma

Tabora

Tukuyu

Chunya

Kibaha

Arusha

Kasulu Kahama

Rujewa

Tunduma

Singida

Kibondo

Korogwe

Manyoni

Pangani

Tunduru

Kiomboi

Mahenge

Ifakara

Handeni

Ruangwa

Morogoro

Zanzibar

Bagamoyo

Makambako

Mikindani

Sengerema

Shinyanga

Biharamulo

Nachingwea

Sumbawanga

Kilwa Kivinje

Dar es Salaam

±0 90 180 270 36045

Kilometers

Tanzania

Legend

Water

Nile

Roads

Nile basin

Borders

Major town

Other town

Height in m

High : 5882

Low : -13

98



Figure 29: Agricultural area and arable land in Tan zania.

Figure 30: Agricultural production in Tanzania. Table 18: Area equipped for irrigation in Tanzania. Tanzania ha

1965 28,000

1975 52,000

1985 127,000

1995 150,000

2005 184,000

Tanzania

0

5000

10000

15000

20000

25000

30000

35000

40000

196

1

196

3

196

5

196

7

196

9

197

1

197

3

197

5

197

7

197

9

198

1

198

3

198

5

198

7

198

9

199

1

199

3

199

5

199

7

199

9

200

1

200

3

200

5

200

7

Years

Are

a in

th

ous

and

s o

f ha


Tanzania

0

500

1000

1500

2000

2500

3000

3500

Mai

ze

Bea

ns,

dry

Sor

ghum

Ric

e, p

addy

Cas

sava

Sw

eet

pota

toes

Ban

anas

Gro

undn

uts,

with

she

ll

See

d co

tton

Coc

onut

s

Commodity

Are

a ha

rves

ted

in 1

000

ha

99

10.7 Uganda 1

10.7.1 General

Uganda (Figure 31) is located in East Africa and occupies an area of 236,040 km2, of which 15.4% consists of water. Based on the 2009 population estimate, a total of 32.4 million people live in Uganda. Uganda shares its borders in the east with Kenya, in the north with Sudan, in the west with the Democratic Republic of Congo, in the southwest with Rwanda, and in the south with Tanzania. The southern part of the country includes a substantial portion of Lake Victoria. Uganda lies almost completely within the Nile basin. The Victoria Nile drains from the lake into Lake Kyoga, and thence into Lake Albert on the Congolese border. From there it runs northwards into Sudan. One small area on the eastern edge of Uganda is drained by the Turkwel River, part of the internal drainage basin of Lake Turkana.


Agriculture is arguably the most important sector of the Ugandan economy. It contributes up to nearly 20% of the GDP, accounts for 48% of exports (UBOS, 2008), and provides a large proportion of the raw materials for industry. Food processing alone accounts for 40% of total manufacturing. The sector employs 73% of the population aged 10 years and older (UBOS, 2005). Agriculture will be the key determinant in the country’s efforts to reduce poverty in the immediate years ahead.


Uganda is located on the East African plateau, and it averages about 1,100 MASL. The country slopes very steadily downwards to the Sudanese Plain to the north. The southern part of the country, however, is poorly drained. The center of Uganda is dominated by Lake Kyoga, which is surrounded by marshy areas. Although generally equatorial, the climate is not uniform as the altitude modifies the climate. The southern part of Uganda is wetter with rain generally spread throughout the year. At Entebbe on the northern shore of Lake Victoria, most rain falls from March to June and during November-December. Further to the north a dry season gradually emerges. At Gule, which is about 120 km from the Sudanese border, November-February is much drier than the rest of the year. The northeastern Karamoja region has the driest climate and is prone to droughts in some years. Rwenzori in the southwest on the border with DRC receives heavy rain all year round. Lake Victoria influences large parts of the south of the country.

Temperatures in Uganda show little variation throughout the year, with maxima ranging between 25-31°C for most areas. The rainfall distribution has generally been categorized as:

• High: over 1700 mm per annum – 4% of the land area • Moderate: 1000-1750 mm per annum – 70% of the land area • Low: under 1000 mm per annum – 26% of the land area

1 This section is based on Iwadra (2007).

100

Rainfall distribution in southern Uganda is bimodal, allowing two crops annually, and adequate grazing for livestock throughout the year. Around Lake Victoria the annual rainfall averages 1200-1500 mm, and is well distributed. To the north, the two rainy seasons gradually merge into one. Dry periods at the end of the year become longer, with annual rainfall ranging between 900-1300 mm, this restricts the range of crops that can be grown. These conditions are not suitable for bananas, but favor extensive livestock production. The influence of soils, topography and climate on the farming systems in Uganda has led to the dividing of the country into seven broad agro-ecological zones. These zones are:

• The banana-coffee system • The banana-millet-cotton system • The montane system • The teso system • The northern system • The West Nile system • The pastoral system


The agricultural system in Uganda is based on the agro-ecological zones as mentioned previously. Between 1999/2000 and 2005/2006, the production trends of the major crops are inconsistent. Positive increases were recorded for cereals (maize, millet, rice and sorghum), beans and simsim, while significant declines were noted for root crops (cassava, Irish and sweet potatoes) and export crops (cotton and coffee). Due to the fact that farmers can produce at least one crop or two per year using rain fed agriculture, irrigation development is rather low in Uganda although the need for irrigation is becoming increasingly serious due to unreliable rainfall and the effect of global warming.

101

Figure 31: Map of Uganda with the Nile basin.

Arua

Gulu

Lira

Hoima

Jinja

Mbale

Kabale

Kitgum

Soroti

Tororo

Entebbe

Kampala

Masindi

Mbarara

Kabatoro

Kikagati

Kikungiri

Nyakibale

Fort Portal

±0 40 80 120 16020

Kilometers

Uganda

Height in m

High : 3500

Low : 500

Legend

Water

Nile

Roads

Nile basin

Borders

Major town

Other town

102



Figure 32: Agricultural area and arable land in Uga nda.

Figure 33: Agricultural production in Uganda. Table 19: Area equipped for irrigation in Uganda. Uganda ha

1965 3,000

1975 4,000

1985 9,000

1995 9,000

2005 9,000

Uganda

0

2000

4000

6000

8000

10000

12000

14000

196

1

196

3

196

5

196

7

196

9

197

1

197

3

197

5

197

7

197

9

198

1

198

3

198

5

198

7

198

9

199

1

199

3

199

5

199

7

199

9

200

1

200

3

200

5

200

7

Years

Are

a in

tho

usa

nd

s o

f ha


0

200

400

600

800

1000

1200

1400

1600

1800

Pla

nta

ins

Be

an

s, d

ry

Ma

ize

Sw

ee

t p

ota

toe

s

Mil

let

Ca

ssa

va

Co

ffe

e,

gre

en

So

rgh

um

Se

sam

e s

ee

d

Gro

un

dn

uts

, w

ith

she

ll

Are

a h

arv

est

ed

in

10

00

ha

Commodity

Uganda

103

11 Appendix: PCRaster-NELmod

11.1 Hydrological modeling

A huge number of hydrological models exits, applications are growing rapidly and a relevant question for hydrological model studies is therefore related to appropriate model selection. An important issue to consider here is the continuum between physical detail and spatial scale. In general it can be stated that the larger the spatial scale the less physical detail is included (Figure 34). A field scale model that aims at simulation crop growth, water transport through he unsaturated zone, percolation to the groundwater and atmosphere land surface interaction requires a lot of data and is computational intensive and can therefore only be applied at the field scale. If one wants to study for example the impact of climate change at the continental scale, then different algorithms are used which are less data intensive. If we consider irrigation schemes, then we are looking at the spatial scale of a system (Figure 34), which has more detail than modeling at the basin scale, but less detail than modeling at the field scale.

Figure 34: Relation between spatial scale and physi cal detail. The green ellipses show the position of different models. Besides these important considerations there are a number of other factors influencing the choice of the model such as the availability of source code, documentation, support, user friendliness, and inclusion of crucial processes relevant to a particular study (see reference list). Hydrological models use input data that have, by definition, inaccuracies. These input data or parameters must be estimated for a given catchment and for each computational segment of the model. They must be estimated either by some relationship with physical characteristics or by tuning the parameters so that model response approximates observed response, a process known as calibration. The process of model calibration is quite complex because of limitations of the models and especially of data. An example of a limitation is the mathematical description that can be imperfect and/or the understanding of the phenomenon may not be complete. Another example of model limitations is that the mathematical parameters used in models to represent real processes are often uncertain because these parameters are empirically determined or represent multiple processes. Also the initial conditions and boundary conditions in a model may not be known. These model limitations, together with the input and output data

104

limitations, and ability to express quantitatively our preferences for how best to fit the models to the data. As a result of these limitations, it is even not clear that a unique set of values exists for the model parameters for a given watershed. The use of remote sensing in hydrological modeling is a growing field and proves to be highly relevant, especially in areas where data are scarce, unreliable or unusable. This situation is regularly encountered in many areas across the world in developing countries. Obviously, groundtruthing is an important aspect of quality and will in general improve accuracy of the data. Remote sensing provides objective and continuous information on relevant variables and could provide a solution for this issue. As far as the link with models is concerned a distinction should be made in applications aimed at model parameterization and in applications aiming at model calibration. Remotely sensed parameterization is more common, and could for example include land cover classification, inclusion of digital elevation model in catchment delineation, use of vegetation indices to derive surface roughness and the use of precipitation radar data as input to a model. In the current study we will use a combination of remote sensing data, global data and local data to calibrate our model. For the assessment of the irrigation potential in in Burundi, Eastern DRC, Kenya, Rwanda, Southern Sudan, Tanzania and Uganda a hydrological model will be used and a number of criteria are important in the selection of the model and these include:

• The model should include a reasonable level of physical detail and include all major hydrological processes at the basin and sub-basin scale at a level of data availability.

• The model should be applied on a very large scale (~ 2.4 million km2). • The model should be run on a daily timescale. • The model should be fully distributed and raster based to provide as much detail as

possible. • There needs to be a clear link with continental and global public domain data sources

(climate forcing, land use and soil).

• There needs to be a clear link with remote sensing datasets to calibrate and parameterize the model.

• The model needs to be user-friendly and in the public domain. We will first describe the PCRaster model and previous applications and then provide a justification for the use of the PC-Raster based approach.

11.2 PC-Raster

11.2.1 Introduction

The PCRaster Environmental Modeling Language1 (Wesseling et al., 1996) is a computer language for the construction of iterative spatial-temporal environmental models. The PCRaster Environmental Modeling Language is developed at the department of physical geography of Utrecht University in the Netherlands. A huge advantage of PCRaster is that it is open source software, and therefore enables it’s user to easily change or extend the model code to satisfy 1 http://pcraster.geo.uu.nl/

105

the user’s wishes as will be done for the current study. One hydrological model which is successful applied in the Middle East and North African (MENA) countries, and written in the PCRaster language, is PCR-GLOBWB. This name stands for PCRaster Global Water Balance. This model is developed at the department of physical geography of Utrecht University in the Netherlands with the explicit aim to simulate terrestrial hydrology at macro-scales, under various land use and climate conditions, with a temporal resolution of one to several days (Van Beek, 2009). FutureWater successfully applied this model in the MENA region to assess the water availability under climate change (Immerzeel et al., 2011). An example of the model output is given in Figure 35. Immerzeel et al. (2010) also successfully applied this model in Asia with the aim to assess future water availability in large Asian river basins in relation to food security.

Figure 35: Example of output of PCR-GLOBWB hydrolog ical model output, in this case the internal renewable water resources based on the 2000-2009 climatology in the MENA region (Immerzeel et al., 2011). For the water resources assessment in the MENA study, PCR-GLOBWB was set-up at a spatial resolution of 10 km. This resolution of 10 km was considered by World Bank as very high given that previous studies focused often on basin, country or sub-basin level only. Moreover, this high resolution over such a large area was only possible given that the normal restriction of data was partly overcome by using remotely sensed data. For the current study model result of NELmod will be at a resolution of 250m. This resolution can be obtained by running the NELmod model at a spatial resolution of 1 km and resample the results using the 250m DEM to the final output resolution of 250 m.

11.2.2 Previous applications

In addition to the above he PC-Raster based hydrological model has been applied successfully in a number of cases under varying conditions and the results have been published in top scientific journals:

• Bierkens and van Beek (2009) have applied the model in Europe and they have developed a seasonal prediction systems for river discharge based on the North Atlantic Oscillation (NAO) and anomalies in sea surface temperature.

• Loos et al. (2009) use a PC Raster based hydrological model to assess nutrient and sediment loads for the Rhine river basin and they show that this can be simulated with a relaive high degree of accuracy.

106

• Sperna Weiland et al. (2010) test the usefulness of GCM data for hydrological studies, with focus on discharge variability and extremes using bias-corrected daily climate data from a selection of twelve GCMs as input to the global hydrological model.

• Petrescu et al. (2010) use the hydrological model in upscaling methane emission of boreal and arctic wetlands.

• Wada et al. (2010) map global groundwater depletion and assess how much this contributes to global sea level rise.

11.2.3 Discretization

The optimal model resolution is a tradeoff between the detail of the available input data, the desired output resolution, the physical detail of the model, and the calculation time. In general, resolution of data availability is the limiting factor. Given these constraints and previous experiences for World Bank, ADB and EU, results will be presented at a resolution of 250 m. The NELmod model will run on a resolution of 1 km and results will be resampled to 250 m using the DEM.

11.2.4 Model concept

The model concept as will be used in our hydrological model is represented in Figure 36. In the remainder of this study we will refer to this model as NELmod. NELmod simulates the most direct pathways of water that reaches the earth surface back to the open water (streams, ponds, and lakes) or atmosphere; within each cell precipitation in the form of rain or snow either falls on soil or in open water surface. Additional specific cell features can be added if necessary. If there is a pumping station in a specific area for example, then this can be implemented into that grid cell. The left side of Figure 36 shows the soil compartment, which is divided in the two upper two soil (root zone and sub-soil layer) stores and the third groundwater store, and their corresponding drainage components: direct runoff (QDR), subsurface flow (QSf) (drainage) and base flow (QBf). In the center of the figure, the resulting discharge along the channel (QChannel) with lateral inflow is depicted. Any precipitation that falls on the soil surface can be intercepted by vegetation and in part or in whole evaporated. A part of the liquid precipitation is transformed in direct or surface runoff, whereas the remainder infiltrates into the soil. The resulting soil moisture is subject to soil evaporation when the surface is bare and to transpiration when vegetated. A certain amount of moisture in the root zone will contribute to subsurface flow, also known as drainage. The remaining part will percolate to the sub-soil layer. The sub-soil moisture content can recharge the groundwater layer, or it can be used for capillary rise to the root zone. Water used for recharge of the groundwater layer will eventually exit the layer as baseflow.

107

Figure 36: Hydrological model concept as will be us ed in this study.

Runoff

Liquid water passed on to the soil surface will infiltrate if sufficient storage capacity is available, else it will drain over the surface as direct runoff. Following the concept developed by Zhao (1977) and Todini (1996), the partitioning into infiltration and direct runoff is dependent on the degree of saturation and the distribution of available storage in the soil. In other words, if the root water volume exceeds the saturation volume, then the part that exceeds the saturation volume becomes runoff. This is shown in the following equation: �� , 0� Where: Runoff = Runoff on a specific day [mm];

RootWater = Moisture in root zone on a specific day [mm]; RootSat = Saturated root water volume [mm];

Actual evapotranspiration ETact The amount of water which evaporates from a grid cell can be either bare soil evaporation, or transpiration from a crop. Water from bare soil or open water will evaporate at the potential rate. For vegetated areas, however, the situation will be different. Each type of crop will have a different rate of potential evapotranspiration (ETpot), depending on the crop factor and ETref (reference evapotranspiration). If the soil becomes too wet, then the roots cannot breathe and as a result there will be a reduction in potential evapotranspiration, known as the actual transpiration. The same is true for too dry conditions. If the soil is too dry, then the crop will reduce its transpiration because there is a stress situation. Therefore the model incorporates an evapotranspiration reduction for too wet and too dry conditions. These are shown in the following two equations: �� 0 �� 1�

108

Where: ETreductionWet = Reduction for wet conditions [-]; RWater = Moisture in root zone on a specific day [mm]; RSat = Saturated root water volume [mm];

�� 1 � �� /�� 2 � �� 1,1�,0� Where: ETreductionDry1 = Reduction for dry conditions [-]; RWater = Moisture in root zone on a specific day [mm]; RDry = Permanent wilting point [mm]; RWilt = Wilting point [mm]; ETreductionDry2 = Final reduction for dry conditions [-];

Then the actual evapotranspiration will be calculated as follows: �� #�� ∗ �� ∗ �� 2 Where: ETact = Actual evapotranspiration [mm] on a specific day; ETpot = Potential evapotranspiration [mm] on a specific day;

Infiltration

As mentioned before the amount of precipitation is added to the root zone. A part of that will leave the grid cell as runoff, and another part evaporates into the air. The remaining part (RootWater – Runoff – ETact) stays in the root zone and can be seen as the updated root water moisture content. This can be seen as the amount of water which has infiltrated into the root zone. Not all the infiltrated water in the root zone will stay in the root zone. A certain amount of this water will leave the grid cell as subsurface flow, also known as drainage, and another amount of this water will percolate to the sub-soil layer. Drainage Drainage will only be significant in areas with soils having high hydraulic conductivities and significant slopes. Drainage in the NELmod follows the concept of a kinematic storage model for subsurface flow developed by Sloan et al. (1983) and summarized by Sloan and Moore (1984). This model simulates subsurface flow in a two-dimensional cross-section along a flow path down a steep hillslope. This model is based on the mass continuity equation, with the entire hillslope used as the control volume. The excess from the root zone is considered whenever the root zone water content exceeds the root zone’s field capacity: �� %� ��&�# if RWater > RFieldCap �� 0 if RWater <= RFieldCap Where: RWexcess = Drainable volume of water in the root zone [mm]; RWater = Moisture in root zone on a specific day [mm]; RFieldCap = Field capacity of root zone [mm];

Then the lateral volume at the hillslope outlet is given by: '�� () ∗ *+,-

109

Where: Qlat = Net drainage at hillslope outlet [mm]; H0 = Saturated thickness normal to the hillsope at the outlet

expressed as a fraction of (RootSat – FieldCap) [-]; vlat = Velocity of flow at the outlet [mm/d];

The velocity of flow at the outlet is defined as: *+,- � ./,- ∗ ��# Where: vlat = Velocity of flow at the outlet [mm/d]; Ksat = Saturated hydraulic conductivity [mm/d] of root zone; slp = Slope as the increase in elevation per unit distance [-];

In large sub-basins with a time of concentration greater than one day, only a portion of drainage will reach the main channel on the day it is generated. Therefore a drainage flow lag is incorporated in the NELmod. So once the lateral volume is calculated, the amount of drainage released to the main channel is calculated as:

��0 1 � 231 � # 4 56788-99:; ∗ '��<= 4 # 4 56

788-99: ∗ ��0 156: Where: Drainagei = Drainage [mm] on day i; RootTT = Lateral flow travel time [d]; Qlat = Lateral volume generated on day I; Drainagei-1 = Drainage [mm] on day i-1;

The travel time of lateral flow is calculated as:

�� 7>,-57?1@+AB,CD/,-

Where: RootTT = Travel time of lateral flow [d]; RSat = Saturated root water volume [mm]; RFieldCap = Field capacity of root zone [mm];

Ksat = Saturated hydraulic conductivity [mm/d] of root zone;

Percolation Percolation occurs from the root zone to the sub-soil (second layer), and from the sub-soil into the groundwater store. Percolation from the root zone to the second soil layer is only allowed if the water content in the root zone exceeds the field capacity of the root zone and the sub-soil layer does not have a seasonal high water table. The equation for root water excess is already shown earlier on page 109. Then the amount of percolation from the root zone to the sub-soil layer is:

�E �� 2�� F G�%� ��&�# = 0.5 ∗ �� %� ��&�#�J�� 0 �� ∗31 � # 4 56

788-99:;<

Where: Rperc = Water percolating to the sub-soil layer [mm]; SWater = Water content of sub-soil layer [mm]; SFieldCap = Field capacity of sub-soil layer [mm]; SSat = Saturated sub-soil water volume [mm];

110

RWexcess = Drainable volume of water in the root zone [mm]; RootTT = Travel time of flow in root zone [d];

The equation for RootTT was shown before on page 110. Percolation from the sub-soil to the groundwater layer is only allowed if the groundwater store water content is lower than the saturated content of the groundwater store. Then percolation is calculated as:

�# �� KL�� F L�� 0 �� 2�� ∗ 31 � # 4 56>MN99:;<O

Where: Sperc = Water percolating to the groundwater layer [mm]; GWater = Water content of the groundwater layer [mm]; GSat = Saturated groundwater store volume [mm]; SubTT = Travel time of flow in sub-soil layer [d];

The travel time of flow in the sub-soil layer is calculated as:

��P�� >>,-5>?1@+AB,CD/,-

Where: SubTT = Travel time of flow in sub-soil layer [d]; SSat = Saturated sub-soil water volume [mm]; SFieldCap = Field capacity of sub-soil layer [mm];

Groundwater

The third store of the soil compartment represents the deeper part of the soil that is exempt from any direct influence of vegetation and constitutes a groundwater reservoir fed by active recharge. The water balance of the groundwater store is as follows: L�� 1 � L�� 156=L�QRSQT �'N �L�Q@U,C � L�CMVC

Where: GWateri = groundwater storage on day i [mm] GWateri-1 = groundwater storage on day i-1 [mm] GWrchrg = groundwater recharge on day i [mm] Qb = baseflow on day i [mm] GWrevap = water moving to sub-soil due to deficiencies [mm] GWpump = water extracted from groundwater storage [mm]

The groundwater recharge depends on the recharge entering on the previous day and the percolation exiting the sub-soil on the current day according to:

L�QRSQT,1 � G1 � #W�1/XTYZJ ∗ �# �� = # [� 6\]^_ ∗ L�QRSQT,156

Where: δgw = delay time over overlaying formations [days] Sperc = percolation exiting from sub-soil [mm] GWrchrg,i-1 = recharge entering groundwater store on day i-1 [mm]

Baseflow from the groundwater store is related to the recharge to this groundwater store. Finally baseflow contributes to the total discharge from a grid cell. Baseflow is calculated as follows:

111

'N,1 � 'N,156 ∗ #W�`TY ∗ ∆�Z= L�QRSQT,1 ∗ G1 � #W�`TY ∗ ∆�ZJ Where: Qb,I = baseflow on day i [mm] Qb,i-1 = baseflow on day i-1 [mm] αgw = baseflow recession coefficient [d-1] ∆t = time step [days] GWrchrg,I = recharge to groundwater store on day i The parameters in the equations above will determined during the calibration process.

Vegetation

NELmod is simulating plant growth and water uptake based on the well-known FAO-56 approach. For each land cover monthly Kc factors and rooting depths are defined and will be used to calculate actual evapotranspiration and water shortage.

11.2.5 Model data

For every grid cell of the model data on the elevation, land use, soils and irrigation practices are required and the model is driven by daily fields of precipitation, air temperature and reference evapotranspiration. It is crucial that the forcing and parameterizations are derived in a consistent and reproducible way for all countries and that it is similar for all NELSAP countries. The delineation of the basins and sub-basins is described below.

Basin delineation

HYDRO1K1 is a geographic database developed to provide comprehensive and consistent global coverage of topographically derived data sets, including streams, drainage basins and ancillary layers derived from the USGS 30 arc-second (~1 km) digital elevation model of the world. HYDRO1K provides a suite of geo-referenced data sets, both raster and vector, which will be of value for all users who need to organize, evaluate, or process hydrologic information on sub-basin scale. The HYDRO1K dataset provides hydrological correct DEMs along with ancillary data sets for use in continental and regional scale modeling and analyses. Based on HYDRO1K, the basins and sub-basins in the 7 NEL countries within the Nile basin will be delineated.

Model validation

The performance of NELmod will be evaluated by comparing the simulated streamflow with the measured streamflow. Local and global data sources of measured streamflow will be used for this. NELmod will be calibrated in order to achieve the most reliable model results.

1 http://eros.usgs.gov/#/Find_Data/Products_and_Data_Available/gtopo30/hydro

112

11.3 Conclusions

In the original proposal SWAT was proposed as hydrological model. However during the inception phase the use the PC-Raster based hydrological model (NELmod) appeared to be much more favourable for the following key reasons:

1. SWAT is a semi-distributed model and irregular hydrological response units are the smallest unit of calculation, whereas NELmod is fully distributed at a high level of spatial detail (1 km). All results of NELmod will be downscaled using DEM to a resolution of 250 m.

2. NELmod is grid based making interactions with Remote Sensing efficient. Moreover, many other analysis for the study (soils, slopes, elevation, NDVI, precipitation, population density) are also based on grids.

3. A SWAT model has a maximum of about 1000 and 2000 HRUs. This means that for the current study area (~2.470 million km2) the level of detail will be only about 1235 km2. It was therefore proposed by the Consultant to use PCRaster..

4. NELmod is highly efficient, public domain and all results are available in GIS format that can easily be shared and distributed.

5. Over the last year three major studies for World Bank, ADB and EU has been executed where clients asked specifically the use of PCRaster.

113

12 Appendix: Facilitator Contract

115

13 Appendix: Data Form

118

14 Appendix: Minutes Validation Workshop

127

15 Appendix: Local expert selection

irrigation potential kenya

Documents

water resources assessment

irrigation potential

land suitability assessment

main crops

irrigation related studies

draft report phase

summarized facts

general information