Targeting AGwater Management Interventions

AWM banner

The Targeting AGwater Management Interventions (TAGMI) is a decision support tool that facilitates targeting and scaling-out of three different Agricultural Water Management (AWM) technologies in the Limpopo and the Volta River Basins. This online tool displays the output of a Bayesian network model that assesses the influence of social and bio-physical factors on the likelihood of success of implementing different AWM technologies. The Bayesian network model was developed iteratively, in collaboration with local researchers and experts, and merges knowledge pools from technical experts to local agriculture extension agents (read more about the consultation process).

TAGMI displays spatially explicit model results at the district scale, based on available data, to determine which districts may be better suited than others for a particular technological intervention in Volta and Limpopo Basin countries. TAGMI helps to answer the question: will an intervention successfully applied in one location have a reasonable chance of success at other locations? The answer, provided with a measurable degree of certainty, suggests a way forward for scaling-out AWM interventions.

TAGMI Assesses the Likelihood of Success. The tool models the relationship between social and bio-physical factors and successful implementation and long-term adoption of agricultural water management technologies. It is intended for non-technological expert users who want to know which parts of the river basins have conditions suitable for successful implementation of a planned AWM intervention. For more detailed information about using the tool see the User Manual.

It is Science Based. Taking social and human resources into account reflects the fact that there are further enabling conditions required beyond the purely bio-physical conditions that dictate whether or not a technology is appropriate for introduction. The conceptual framework for the Bayes model is informed by the Sustainable Livelihoods Framework (DFID 1999). For more detailed information about the Bayes model behind the tool see the Model Technical Documentation

It is Evidence Based. The Bayesian network model makes use of available data on key characteristics in a systematic way to suggest the likelihood of success of an intervention. It estimates how different contextual factors interact to influence success. This model and tool are based on the premise that, while absolute certainty is unobtainable, degrees of certainty are both obtainable and useful when using the available information in a systematic way.

This tool was developed as part of the 3-year CGIAR Challenge Program on Water and Food's Volta and Limpopo Basin Development Challenges (2011-2013) (About), for:

  • Volta Basin: Soil and Water Conservation, Small reservoirs, Smallscale irrigation
  • Limpopo Basin: Conservation agriculture, Small reservoirs, Smallscale irrigation
Further developments to the tool:
  • Upper East Region, Ghana: Small petrol pumps - as part of the project "Enhancing uptake and socio-economic benefits of shallow groundwater irrigation in the White Volta Basin" (Research into Use), funded by CPWF, Challenge Program of Water and Food (2013)
  • White Volta Basin: Drip irrigation, Small electric pumps (GH), Small petrol/ diesel pumps (BF) – as part of the project "Targeting investments in Groundwater irrigation in the White Volta Basin", funded by the CGIAR Water, Land and Ecosystems program (2014-2015)
  • Niger Basin: Smallscale irrigation – part of the project "Managing water and food systems in the Volta - Niger basins", within the project "WLE in Africa", funded EC/IFAD and the CGIAR Water, Land and Ecosystems program (2014-2016)