Review maps CGIAR’s 12-year push on extreme rainfall and floods

The analysis, published in Agriculture & Food Security, synthesizes twelve years of publications from the Consultative Group on International Agricultural Research, better known as CGIAR, covering the period from January 2012 to December 2023. Out of 2,846 papers initially identified from CGIAR’s climate research repositories, 73 studies met the review’s strict inclusion criteria and were examined in depth, spanning research conducted across Asia, Africa and Latin America.

The stakes are considerable. According to figures cited in the review, floods now account for 19 percent of total crop damage worldwide, and extreme rainfall events explain 18 to 43 percent of observed anomalies in global crop yields. Some modeling studies indicate a global weighted mean rice yield loss of 9.0 percent by 2080 under a high-emissions scenario. In sub-Saharan Africa, one source cited in the review reports a 33 percent reduction in crop yields over the past 60 years due to climate-related weather extremes, including floods, while other work points to a potential 13 percent reduction for major cereal crops in a worst-case scenario for West and Central Africa. The review also notes an estimate that in Africa, about 12 percent of people who experienced food insecurity between 2009 and 2020 had their food security status affected by flooding.

The physical science underpinning these concerns is well established. The IPCC’s assessments, referenced extensively in the review, conclude that human-induced greenhouse gas emissions have increased the frequency and intensity of weather and climate extremes, with global surface temperature reaching 1.1 degrees Celsius above 1850 to 1900 levels during 2011 to 2020. Heavy precipitation is expected to intensify at roughly 7 percent per 1 degree of global warming, as a warmer atmosphere holds more moisture. For farmers, this translates into escalating risks: waterlogged soils, drowned crops, salinity intrusion in coastal fields, soil erosion, nutrient leaching and heightened outbreaks of pests and diseases.

The review’s methodology itself reflects modern evidence synthesis practice. The team searched the CGSpace database, compiled from nine CGIAR institutional repositories, using topic modelling, an unsupervised machine-learning technique, to generate 20 climate-relevant themes, three of which concerned extreme rainfall and floods. After screening 2,850 articles with artificial intelligence assistance and double-screening titles and abstracts in the Rayyan software, 107 articles proceeded to full content review, and 34 were excluded for treating floods only in passing. The remaining 73 articles were then queried with OpenAI’s GPT-4o-mini model, with every answer manually cross-checked by the reviewers before synthesis in Excel.

Geographically, the dataset is heavily skewed. Asia dominates with 45 articles, or 61.6 percent of the total, followed by Africa with 24 articles at 32.9 percent, and Latin America with just 4 articles at 5.5 percent. Bangladesh leads the country count with 10 articles, followed by India with 9 and Nepal with 8; in Africa, Ethiopia tops the list with 6, then Kenya with 4. The authors attribute this concentration to the fact that these South Asian countries frequently experience extreme rainfall and floods, but they also flag it as a limitation, noting that flood-prone African nations such as Mozambique and Madagascar are conspicuously underrepresented.

The reviewed papers split almost evenly across three analytical clusters: 37 articles, or 50.7 percent, addressed impacts; 35, or 48 percent, addressed methods; and 34, or 46.6 percent, addressed adaptation solutions. Most studies, 56 articles or 76.7 percent, relied on observed real-world data rather than projections, which the authors interpret as grounding the research closely in farmers’ lived realities. Among impacts, the most prominent themes were physical crop losses from coastal inundation, waterlogging and salinity intrusion. Eight papers directly addressed coastal flooding and salinity, showing that crop losses increase exponentially when flooding and salinity intrusion exceed average levels, particularly in Bangladesh and Nepal.

On the methods side, CGIAR researchers deployed a diverse toolkit. Fifteen articles used modelling and forecasting techniques, with most projecting that average annual and seasonal precipitation in Asian and African study sites will increase by at least 25 percent under various scenarios by 2050, suggesting flood risk will worsen. At least eight papers applied the Mann-Kendall trend detection test, including a 68-year analysis of Malaysia’s Sarawak tropical peatland covering 1948 to 2016, which found significantly increased extreme rainfall expected during the Southwest monsoon. Fifteen articles used historical time-series analysis, fifteen combined reanalysis techniques with observational datasets, and seven used remote sensing, satellite imagery and GIS to decode flood inundation patterns and identify flooding hotspots. Cross-sectional surveys of farmer perceptions formed the largest methods subset at 17 papers.

The adaptation solutions documented by CGIAR research form perhaps the most actionable part of the review. Climate-smart agriculture features prominently, with conservation agriculture-based wheat systems in Haryana, India, showing yield advantages over conventional tillage that doubled during bad weather years, at 16 percent versus 8 percent. Flood-tolerant crop varieties, salinity barrier infrastructure, agroforestry and early warning systems all appear repeatedly. In Vietnam, farms with trees showed shorter recovery times after most types of natural disasters than farms without trees, and in Cameroon, mangrove reforestation emerged as a key adaptation pathway. In Sri Lanka, the International Water Management Institute established a platform delivering early warnings for floods, landslides and droughts using space-based information.

Yet the review is candid about the barriers. Lack of finance and knowledge consistently limit smallholder farmers’ ability to adopt technological innovations, prompting the authors to call for microcredit, low-interest loans and climate risk insurance schemes tailored to smallholders, alongside robust capacity-building programs. The identified research gaps are equally instructive: CGIAR’s portfolio underrepresents roots and tubers, livestock, fish, and peri-urban agriculture; only one paper quantified economic losses from floods; and attribution studies linking specific extreme rainfall events to climate change are entirely absent from the dataset. The authors recommend expanding geographic coverage, broadening thematic focus, and integrating attribution science to inform targeted interventions, noting that every dollar spent on CGIAR and similar agricultural research yields roughly 10 dollars in benefits for farming communities through higher yields, better nutrition and reduced crop losses.

For policymakers grappling with how to scale adaptation solutions, the review offers a consolidated evidence base and a clear agenda. The CGIAR launched its dedicated Research Program on Climate Change, Agriculture and Food Security in 2009, and the publications captured here represent the maturing of that investment. The authors caution that their findings cannot be generalized wholesale, given the context-specific nature of the underlying studies and the dataset’s skew toward Southeast Asia and staple crops like rice, wheat and maize. They also note the review excluded non-English publications and grey literature.

Subject of Research: A systematic review of CGIAR research on climate-induced extreme rainfall, floods and agricultural adaptation from 2012 to 2023

Article Title: A review and synthesis of the Consultative Group on International Agricultural Research’s work on climate-induced extreme rainfall and floods, 2012–2023

Article References: Chikozho, C., Mukherji, A., Bosire, C., Kujinga, K., & Managa, R. (2026). A review and synthesis of the Consultative Group on International Agricultural Research’s work on climate-induced extreme rainfall and floods, 2012–2023. Agriculture & Food Security, 15(1), Article 54. https://doi.org/10.1186/s40066-026-00640-w

Image Credits: AI Generated

DOI: 10.1186/s40066-026-00640-w

Keywords: CGIAR, extreme rainfall, floods, agri-food systems, climate adaptation, smallholder farmers, climate-smart agriculture, food security, early warning systems, salinity intrusion, waterlogging, systematic review