Initial TAGMI Project

The Challenge Programme for Water and Food Basin Development Challenges

Since its 2002 launch, the CGIAR Challenge Program on Water and Food (CPWF) has become a comprehensive global research effort on water and food. More than 10 years of CPWF research includes 100+ research-for-development projects and the involvement of 400+ partners in international river basins. More than 1.5 billion people live in CPWF basins, amongst whom half of the poorest people on Earth. In Phase 1 (2002-2007), projects focussed on discovering successful water-related strategies for improving food security and reducing poverty. CPWF Phase 2 builds on prior work, and is organised into integrated research programmes involving 5+ projects and partners to address Basin-specific Development Challenges (BDC):

Volta BDC

Strengthen integrated management of rainwater and small reservoirs so that they can be used equitably and for multiple purposes (

V1 Targeting and scaling out

V2 Management of rainwater for crop livestock and agroecosystems

V3 Management of small reservoirs

V4 Governance of rainwater and small reservoirs/p>

V5 Coordination and enabling change

Limpopo BDC

To improve smallholder productivity and livelihoods and reduce livelihood risk through integrated water resource management (

L1 Targeting and scaling out

L2 Small-scale infrastructure

L3 Farm systems and risk management

L4 Water governance/p>

L5 Learning for innovation and adaptive management

V1 and L1 Targeting and scaling out

Numerous pilot studies and case studies in the Volta Basin have evaluated practices, methods, and tools that could prove beneficial to others, both within the basin and outside of it. However, the question whether an intervention successfully applied in one location has a reasonable chance of success at any other location remains extremely difficult to answer. A consistent finding in pilot studies is that detailed characteristics of the study location – economic, biophysical, institutional, and cultural –all have essential roles in the eventual success or failure of achieving a successful land and water management outcomes. For out-scaling of initiatives it is impractical to collect detailed information at every potential site where an agricultural land and water management (AWM) intervention might be introduced. This project starts with the premise that, while certainty is unobtainable, degrees of certainty are both obtainable, using available information in a systematic way, and useful.

Despite hosting some of the most developed sub Saharan countries, a majority of rural smallholder farmers in the Limpopo basin still live in poverty. The challenge of low and highly variable rainfall together with inadequate technology transfers, inadequate policy and investment context all act to disable successful transitions out of poverty. The CPWF Phase I identified several opportunities to manage rainfall in more efficient and productive manners at field to basin scales. The challenge of successful targeting and scaling out is still a key research and development area to contribute towards the Limpopo development challenges with opportunities to enable transformations of rural livelihoods at a greater scale.

The CPWF Project V1 and L1 projects have thus developed this web-based "decision support" targeting and scaling out tool to map the likelihood that a given AWM intervention will be successful in given locations. The tool is aimed at non-expert users who could use it to identify likely sites where introducing AWM interventions for smallholder farming systems are likely to succeed.

Consultation process

As the convening institution, the Stockholm Environment Institute, with partners, led the development of TAGMI through a series of national and local consultations in Ghana, Burkina Faso, and with representatives of all four basin countries attending meetings in Zimbabwe and South Africa in 2011 and 2012. The first consultation in 2011 was aimed at identifying successful AWM cases in the country and region and identifying factors and contexts that contributed and/or hindered outscaling of AWM technologies based on known examples of successful and unsuccessful outscaling in the respective basin. In the second consultation, which took place in 2012, the aim was to get feedback on the model. Local partners and stakeholders discussed the relations between factors and indicators, and the importance to the likelihood of success of an AWM intervention. The Bayesian Network Model, which informs the web-tool’s map display output, is a combination of local, national and regional datasets with the expert input of our Basin project partners. The final engagements in the region as part of this project took place in July and August 2013.

Success and failure of AWM interventions

Experts identified a range of technologies from rainfed to full irrigation. However, there was no obvious pattern of success or failure of interventions across this AWM technology gradient. Both rainfed and full irrigation interventions were referred to as successful or failed depending on the context.

An analysis of the criteria used to identify the success or failure of an intervention showed that in all four countries the success of AWM cases was especially due to factors not related to natural, physical, human, social, or financial aspects. More important for the success or failure of an intervention were factors related to project implementation, process and participation.

Detailed narratives of successful AWM interventions

Consultation process

In addition to these three rounds of consultation, the project partners Waternet, Witwatersrand University, INERA and SARI collected more detailed information about successful AWM interventions in their respective countries. They used a methodology that combines maps and participation in focus groups (Participatory GIS) to collect a narrative of project implementation and enabling factors that had resulted in a successful AWM intervention. Stakeholders at district level participated in these events and their stories were complemented by community level narratives.

The Project Team

Dr Jennie Barron is the Theme Leader for Sustainable Agricultural Water Management at the International Water Management Institute and affiliated to Stockholm Resilience Centre. Her research is focussed on agriculture, water management and ecosystem services in the field to meso-scale landscapes for food production, livelihood improvements and agro-ecosystems sustainability. During the initial TAGMI project, Dr. Barron was the theme leader of Managing Environmental Systems at the Stockholm Environment Institute (SEI). SEI was the institutional coordinator for the V1 and L1 projects. The project was carried out in close collaboration with the other CPWF BDC projects (V2-V5 and L2-L5) with the support of the respective BDC Leaders : Olufunke Cofie (Volta) and Amy Sullivan (Limpopo).

Volta Basin Research Partners

Institut National de l’Environnement et de Recherches Agricoles (INERA); Civil Engineering Dept. of the Kwame Nkrumah University of Science and Technology (KNUST); Savanna Agricultural Research Institute of the Council for Scientific and Industrial Research, Ghana (CSIR-SARI); Département de Géographie de l’Université de Ouagadougou

Limpopo Basin Research Partners

WaterNet; University of Witwatersrand; International Water Management Institute-South Africa

CPWF Upper East Region, Ghana

Enhancing uptake and socio-economic benefits of shallow groundwater irrigation in the White Volta Basin (January 01, 2012 - April 30, 2014)

Project Leader: Pamela Katic

The project continued the research carried out in the CPWF Phase 1 project 65 ('Shallow groundwater irrigation in the White Volta Basin') in the Atankuidi catchment of Upper East Region, Ghana. This project explored the potential for greater use of groundwater irrigation in the White Volta Basin. The specific objectives of the project were: to identify and, where possible introduce affordable technologies adapted to local conditions, to improve the understanding of the economics of current and potential crop enterprises, to explore interventions that help address the marketing challenges faced by shallow groundwater irrigation practitioners, and investigate ways of fostering and accelerating uptake and adoption of groundwater irrigation in other areas of the White Volta Basin. To achieve these objectives, the project implemented several research-into-use activities, which were grouped into three interrelated and mutually reinforcing work packages:

  • Evaluation of existing irrigation methods and alternative technologies to improve the efficiency of shallow groundwater utilization;
  • Improving profitability and market access: A value-chain analysis; and
  • Improving uptake and out-scaling of shallow groundwater irrigation technologies and management practices.

The project was led by the International Water Management Institute (IWMI) and funded by the CGIAR CPWF program and IWMI.

Location: Upper East Region, Ghana

Technologies: Small Petrol Pumps

Project partners: International Development Enterprises, University of Development Studies, Stockholm Environment Institute.

WLE White Volta Basin

Targeting investments in Groundwater irrigation in the White Volta Basin (May 01, 2014 - June 30, 2015)

Project Leader: Pamela Katic ([email protected])

The project served to test and expand the applicability of the Bayesian-based decision tool TAGMI (developed by SEI as part of CPWF-V1). TAGMI provided recommendations for scaling-up the adoption of a specific irrigation technology. The project carried out specific ground-truthing and expansion of the TAGMI model adapted to petrol pumps for shallow groundwater irrigation (CPWF-RiU project) by:

  • verifying the results of the model with field observations and expert opinions.
  • re-assessing the model structure (Are the lines and relationships accurate? What should change?) and data sources (Are there others to include? What organizations have the data?)
  • exploring avenues for model refinement through additional data sources and spatial disaggregation
  • exploring the potential of the model to other pumping and groundwater application technologies
  • exploring the relevance and data requirements of expanding the model to other areas in the wider Volta/Niger Basin.

The project was led by the International Water Management Institute (IWMI) and funded by the CGIAR Water, Land and Ecosystems program.

Location: Districts of Burkina Faso and Ghana that fall within the White Volta basin

Technologies: Drip irrigation (both), small petrol/ diesel pumps (Burkina Faso) and small electric pumps (Ghana)

Key team members: Pamela Katic, Joanne Morris, Mathieu Chaix-Bar, Kalifa Dembele, Euloge Kabore, Anna Minkah

Project partners: International Development Enterprises (Ghana), International Development Enterprises (Burkina Faso), Stockholm Environment Institute.

WLE Niger in the Niger Basin

Managing water and food systems in the Volta - Niger basins (January 01, 2014 - December 31, 2016)

Project Leader: O. Cofie, ([email protected])

The improvement of TAGMI and development of a TAGMI tool for smallscale irrigation in Niger was one component of the larger project. Extensive groundtruthing of landuse around small reservoirs was carried out in December 2014, to support the development of an improved map of smallscale irrigation in Burkina Faso, to serve as a validation method for the original small-scale irrigation TAGMI tool developed under the CPWF V1 project.

The larger project sought to build and expand on earlier work carried out in the CPWF Volta from 2010-2013. It aimed at improving the management of water for sustainable food production at local and sub-basin levels. The project would produce outputs of: 1) Innovation and learning, including publication of special journal edition and policy briefs. 2) Consolidated knowledge of the historical evolution and resilience of small reservoirs in West Africa documented as a book 3) Recommendations on optimizing the use of small reservoirs for multiple purposes. 4) Recommendations for improving water governance options at watershed level. 5) Improved decision support tool for better targeting investments in agricultural water management interventions (TAGMI). The project was led by the International Water Management Institute (IWMI), and funded by EC/IFAD and the CGIAR Water, Land and Ecosystems program.

Location of new TAGMI application(s): Districts of Niger that fall within the Niger basin

Technologies: Small Scale Irrigation (initial draft)

Key team members (in the TAGMI development): Joanne Morris, Mathieu Chaix-Bar

Project partners: CIRAD-UPR GREEN, Institut de recherche pour le développement (IRD), International Livestock Research Institute (ILRI), Kwame Nkrumah University of Science and Technology (KNUST), Stockholm Environment Institute, University of Utrecht, Volta Basin Development Foundation (VBDF)

Bayesian Network

The Bayesian Model calculates a desired outcome, 'Success', which is the likelihood that an AWM technology introduced in a target community will still be in use 2 years after the intervention project has ended. Based on participants' discussions, and using the DFID Sustainable Livelihood Framework (DFID, 1999), 'Success' is conditional on adequate levels of 5 capitals: Human, Social, Financial, Physical and Natural. Water resources are included as a separate 6th capital given its centrality to AWM. Each capital comprises 2-4 key factors (e.g. Human capital is a combination of Labour availability, Skills, and Health). Each factor is described by 1-3 data variables, which are the foundation of the model (e.g. Labour availability is indicated by the relative size of the working age population and the gender ratio in the population).

The linking arrows convey the conditional probabilities of how each node in the network influences the presence of the next node. The model calculates the probability that the factor is present given knowledge about the state of its data variable (high, medium or low), then the probability that the capital is present given the calculated state of its factors, then the probability that Success is present given the calculated state of all capitals. A similar application of Bayesian network modelling to analyse the likelihood of water poverty is explained in detail in Kemp-Benedict et al. (2009).

The Volta Basin

The Volta Basin has a surface area of 407,000 km2, covering parts of Burkina Faso, Ghana, Benin, Cote d'Ivoire, Mali and Togo. The TAGMI for the Volta Basin covers Northern Ghana and Burkina Faso, which collectively represent 85% of the basin. The basin's farmers, some of the poorest in the world, generally rely on rainfed agriculture. Annual average rainfall distribution ranges from 2000mm in the south to 500mm in the north. Potential evaporation rates range from 1500 mm in the south to over 2500 mm in the north, meaning that less than 10% of rainfall is available for river flow (GLOWA 2013). Even when rainfall is adequate for cropping, its uneven distribution leads to a high risk of crop loss; moreover, the lack of incentives means a situation where farmers are reluctant to invest in agriculture and water management. Climate change is making already variable rainfall less reliable. Farmers must have access to reliable water supplies. Small reservoirs, locally maintained and requiring no recurrent energy input are a sustainable supply option (CPWF 2013). The most significant hydrological features is Lake Volta, created post-construction of the Akosombo Dam for hydropower, which is the second largest man-made lake in the world. The dam also generates hydroelectricity, ~80% of Ghana's power production, and hosts an important small-scale fishery industry (AgWater Solutions 2012).

basin image

basin image

The Limpopo Basin

The Limpopo Basin has a surface area 408 000 km2, covering parts of Botswana, South Africa, Zimbabwe and Mozambique*. The basin supports more than 14 million people in areas of Botswana, Mozambique, South Africa and Zimbabwe. Rainfall is highly variable with little run-off available in many parts of the basin to produce crops and livestock (CPWF 2013). Annual average rainfall distribution ranges from 200 mm to 1200 mm, with an average of 530 mm per annum. Potential evaporation ranges from 800 mm to 2400 mm per annum. The Limpopo basin is a relatively dry basin with most of the water in the highly productive areas already used for irrigation. Irrigated agricultural covers 244 000 ha, with the potential for a further 122 000 ha in selected sub-catchments. Dryland or rain-fed agriculture includes crop production (234 000 ha), pastures (1 780 000 ha) and forestry (455 000 ha) (ARC 2013). Prolonged dry spells are common, making rainfed farming in many areas difficult, while occasional high rainfall events result in uncontrolled flooding (CPWF 2013).

* While TAGMI for the Limpopo Basin covers all countries, it is primarily driven by available data and learning events undertaken in Zimbabwe and South Africa. Data for Mozambique and Botswana are derived from publicly available global data sets.

CA image

Conservation Agriculture

Conservation Agriculture (CA) is an approach to managing agro-ecological systems. It is designed to enable sustainable and profitable agriculture while improving farmers' livelihoods. It is governed by three CA principles: 1) continuous and minimal soil disturbance, 2) permanent (organic) soil cover and 3) crop diversification and rotation. CA is scalable, its principles universally applicable to all agircultural agricultural landscapes but its adoption is arguably most urgently required by smallholder farmers. Introducing CA requires that farmers accept a change in their existing crop management system, employ technologies that help manage cover crops, and adopt a new way of thinking with respect to weed management and livestock/cropping interactions. More info from the FAO

Small scale irrigation

These are irrigation systems that are developed, owned or managed by individual farmers. They buy or rent irrigation equipment and draw water from nearby sources without depending on public agencies or Water Users' Associations, for example. In many African countries the smallholder private irrigation sector is more important than public irrigation in terms of the number of farmers involved and the value of the production. More info from IWMI

tapping-groundwater image

SA image

Small Reservoirs

In most of Sub-Saharan Africa, small reservoirs are earthen or cement dams that are less than 7.5 meters high. They provide significant opportunities for soil and water conservation, drought proofing and for developing small-scale, community-based irrigation schemes. A well-designed reservoir can support multiple water uses from livestock watering, fisheries, domestic and small business water use to brick making and handicraft activities. More info from IWMI

This is open source software, released under the Apache License v. 2.0.

The data used in the tool are all sourced from the public domain. Please refer to the metadata for details of the relevant copyright notices.