JP Venot

Remote sensing for irrigation in sub-Saharan Africa: Promises and pitfalls

Compelling discussion, commentary, stories on agriculture within thriving ecosystems.

There is growing demand for sustainable development policy that is data driven, but our unprecedented and increasing capacity to generate large volumes of quantitative data also requires a much greater consideration of the ways these data are generated and interpreted. This, after all, determines if and how newly generated data are harnessed by policy and with what effects.

The growing use of remote sensing to detect the extent and location of irrigation in sub-Saharan Africa is a case in point. Since the 2007/08 food price crisis there has been a renewed focus on investment in sub-Saharan Africa's agricultural sector. Irrigation has either returned to or newly entered the portfolios of many development agencies, promoted as a key component of Africa's agricultural development. There is, however, widespread debate regarding what is 'the right irrigation' development pathway for the continent, and notably the place smallholders have in shaping rural transitions.

The growing appeal of remote sensing

The widespread exploitation of water by smallholders to intensify their agricultural production and supply sub-Saharan Africa's rapidly urbanizing food markets has meant that mapping and characterizing the diffuse and complex phenomenon of farmer-led irrigation development has become a key issue for researchers and decision makers alike.

In parallel, over the last two decades, remote sensing has witnessed tremendous refinement in processing techniques. This has created growing expectations for the generation of remote sensing data to inform policy, not least because of systematic methodological approaches and the attractive maps generated at basin, national, continental and global levels.

Since early efforts to map irrigated areas using remote sensing techniques were spearheaded by the International Water Management Institute (IWMI) in the early 2000s, increased availability of open access, high-resolution satellite imagery and refinement of methodologies have led to the improved reliability and accuracy of irrigated area assessments.


IWMI data portal (accessed May 13, 2021)

IWMI

A new study, 'Below the Radar: Data, Narratives and the Politics of Irrigation in Sub-Saharan Africa', illustrates the opportunities that radar-based remote sensing techniques offer to better detect instances of farmer-led irrigation development. Specifically, the paper shows that radar analysis can illuminate the widespread irrigation of rice in low-lying areas, as shown in the picture below, which may be well known to national policy makers but often remains 'below the radar' of irrigation statistics.

A critical look at remote sensing

Since these new tools and methods make it possible to locate irrigation at spatial and temporal resolutions that many other datasets cannot, and provide important data that could inform policy, it is essential to look at the underbelly of the 'attractive beast' that remote sensing has become. It is critical that we investigate the ways in which remote sensing data are generated and the ideologies that underpin them -- as these will impact if and how they are used in decision-making contexts.

The paper, published in Water Alternatives, stresses that remote sensing analysis of irrigated areas hinges on the identification of 'typical signals' -- spectral light or microwave radiation reflected from the earth's surface which provides a measure of a plant's greenness. The signals serve to differentiate different classes of irrigation from other non-irrigated landscape categories.

This means that resulting maps reflect first and foremost the analyst's understanding of the boundary between those categories rather than farmers' dynamic agricultural practices which tend to display extreme variability and heterogeneity and rarely conform to ideal type categories. For instance, a crop that is only irrigated at specific times to cope with sudden dry spells, a technique known as supplementary irrigation, may be classified as rainfed or irrigated depending on the time window used for the analysis but also on what the analyst considers to be irrigation.

Reinforcing colonial legacies

The paper argues that in sub-Saharan Africa these boundaries are inadequately defined -- and driven by a conventional view of irrigation that relies heavily on colonial legacies of engineering design and agricultural modernization. It further stresses that farmer-led irrigation development challenges this 'modernist' understanding of irrigation and raises questions about the relevance of conventional irrigation categories used in existing global databases that continue to inform remote sensing analysis.

Dependence on bounded categories risks reinforcing an established view that reduces irrigation to concrete irrigation schemes. Although it can be detected in remote sensing analysis, farmer-led irrigation development often remains classified as 'other' in such exercises. The global irrigated area map displayed on the IWMI data portal is a case in point. It introduces two categories entitled 'water managed non irrigated' because spectral light signals from these areas do not conform to either of the ideal signatures that analysts associate with 'rainfed' and 'irrigated' land use.

Without the legitimacy provided by the term 'irrigation', and in contexts in which water allocation policy is dominated by a narrative of water scarcity and smallholders are routinely dismissed as inherently inefficient, there are risks that new remote sensing data revealing increased water use by smallholders may lead to regulation that further marginalizes them.


Rice irrigated agriculture in Moshi rural district, Tanzania.

JP Venot

The paper concludes that newly generated data alone, however accurate, may fall short of a much needed re-framing of irrigation dynamics in sub-Saharan Africa. Since farmer-led irrigation development is now on the radar of national governments and development agencies, there is a need to move away from a debate on irrigation data and monitoring towards a more holistic discussion that reconsiders what can be officially defined as 'irrigation'.

It is important to work collectively towards a grounded definition of irrigation (categories) that takes farmers' practices -- rather than infrastructure -- as its entry point. This means challenging entrenched vested interests that profit from an orientation towards infrastructure development, changing development practices, and generating different data that link farmers' innovation and entrepreneurial initiatives with key policy objectives such as food security and enhanced rural incomes.

Selected references

Harrison, E. 2018. Engineering change? The idea of the 'scheme' in African irrigation. World Development 111: 246-255.

Harrison, E. and Mdee A. 2017. Successful small-scale irrigation or environmental destruction? The political ecology of competing claims on water in the Uluguru Mountains, Tanzania. Journal of Political Ecology 24: 406-424.

Lankford, B. 2009. Viewpoint- The right irrigation? Policy directions for agricultural water management in sub‐Saharan Africa. Water Alternatives 2(3):476 :480.

Veldwisch G J.; Venot, J-P.; Woodhouse, P.; Komakech, H. and Brockington, D. 2019. Re-introducing politics in African farmer-led irrigation development: Introduction to a Special Issue. Water Alternatives 121: 1-12.

Venot, J-P.; Bowers, S; Brockington, D.; Komakech, H.; Casey, R.; Veldwisch G J. and Woodhouse, P. 2021. Below the Radar: Data, Narratives and the Politics of Irrigation in Sub-Saharan Africa. Water Alternatives 14(2): 546-572.

Woodhouse, P.; Veldwisch, G.J.; Venot, J.P.; Brockington, D.; Komakech, H. and Manjichi, A. 2017. African farmer-led irrigation development: reframing agricultural policy and investment? Journal of Peasant Studies 441: 213-133.

---

Thrive blog is a space for independent thought and aims to stimulate discussion among sustainable agriculture researchers and the public. Blogs are facilitated by the CGIAR Research Program on Water, Land and Ecosystems (WLE) but reflect the opinions and information of the authors only and not necessarily those of WLE and its donors or partners.

WLE and partners are supported by CGIAR Trust Fund Contributors, including: ACIARDGISFCDOSDC, Sida and others.