Five newsletters landed in November 2026 from the Carbon Trust, CGIAR Climate Action, CGIAR Scaling for Impact, CGIAR Policy Innovations and the WFP Innovation Accelerator. Read together, they tell one story that matters to every African agri value chain: warnings are now precise, but a warning only protects a farmer if it becomes advice, money and verified action before the loss arrives. This article follows that chain through Kenya’s 2026 short rains, where an unusually strong El Niño is testing it in real time, and draws in evidence from IFPRI, FAO, ACRE Africa and the Kenya Meteorological Service Authority.
Read this first. We read all five newsletters in full. We then checked their claims against the institutions’ own pages and press coverage, and we say where we read only a summary. We did not attend any event. Research stops at 11 November 2026, and the El Niño season is still unfolding, so figures may change. The Carbon Trust’s newsletter points to an agriculture vision page asking how to feed 8.3 billion people; we have not read that page and use it only as framing.
Six findings
- The warning is loud and early. Kenya’s seasonal outlook expects above average October, December rain across about 80% of the country.
- The cost is mostly weather, not fuel. IFPRI’s model puts Kenya’s 2026 GDP loss at 1.1% to 1.5%, with fuel prices adding almost nothing.
- The short rains cut both ways. Good rains could shrink the loss, while flooding or poor rains could widen it.
- Forecasts reach farmers only through channels. A new Kenya pilot will test turning official forecasts into advice through extension, media, SMS and digital platforms.
- Early action pays. FAO measured returns of $3.50 to $9 for each dollar spent before the 2017 drought in Kenya and Ethiopia.
- Finance must carry the risk. ACRE Africa’s AI risk tool aims to put loans bundled with insurance in the hands of 50,000 Kenyan smallholders by 2028.
The warning is loud
On 26 August the Kenya Meteorological Service Authority forecast above average rainfall over about 80% of the country for October, December, with prolonged wet spells and isolated storms. Some areas carry an 85% to 90% likelihood of above normal rain, according to Big3 Africa’s report. The Greater Horn of Africa outlook forum made a similar call for the equatorial and southern parts of the region. Probabilities for the El Niño itself differ by source: one agricultural outlet cites NOAA at a 65% chance of a strong or very strong event between October 2026 and February 2027, while IFPRI cites NOAA at above 90% for a very strong one. Both agree on direction.
Kenya Meteorological Service Authority outlook, 26 August 2026, as reported by K24 and Capital FM. The 20% remainder is our arithmetic
The surprise is that the long rains had already failed in places. IFPRI notes that drought hit much of western Kenya in 2026, that many areas had their driest or second driest June, August since at least 1981, and that the long rains normally supply about 70% of crop output. El Niño does not act alone either: the Kenya Meteorological Department has stressed that the Indian Ocean Dipole can strengthen or soften its effects.
What the shocks cost
IFPRI and KIPPRA used Kenya’s RIAPA-AI model to test three production scenarios. National GDP loses 1.1% to 1.5% in 2026, and agricultural GDP loses 4.5% to 6.1%. Maize productivity falls 13%, 15% or 18% by scenario and livestock productivity 7%, 8% or 9%. In the moderate case about 756,000 more people fall into poverty and 4.9 million more become undernourished, on top of a baseline in which 35.3% of Kenyans, about 20.8 million, already fall below the hunger threshold. These are model scenarios, not forecasts.
Share of GDP lost, by scenario. IFPRI and KIPPRA, RIAPA-AI model, October 2026
Assumed fall in national productivity, per cent. IFPRI and KIPPRA, October 2026
The Iran war matters less than expected for 2026. Brent crude is assumed at $94 a barrel against $67 without the conflict, yet national GDP moves only by about 0.01%, because most fertiliser was bought before April and a subsidy programme cushions prices. IFPRI warns that 2027 input costs could feel more of the spike, and a fast switch to La Niña could raise drought risk in March, June. For farmers the lesson is that the weather pushes the loss, and fertiliser prices add a second risk for the next season.
The short rains cut both ways
IFPRI tested the short rains as the swing factor. If output runs 15% above average, the national GDP loss falls by 0.4 percentage points and 1.7 million fewer people become undernourished. If it runs 15% below, the loss rises to 1.7%. Above average rain can help crops and still flood them, which is why timing and drainage advice matter as much as the headline forecast.
National GDP loss in 2026, per cent. Average case 1.3%, below average case 1.7%; the 0.9% figure is our arithmetic (1.3 minus the 0.4 point offset). IFPRI
From national forecast to the farmer
This is where the chain usually breaks. The Alliance of Bioversity International and CIAT, the Kenya Meteorological Service Authority and AIM for Scale have launched a Forecast to Farmer pilot, set out in the CGIAR Scaling newsletter. A shared, non binding charter commits partners to turn authorised forecasts into location specific advice via extension services, mass media, SMS and digital platforms. The October, December 2026 short rains are the proof of concept, with wider dissemination in the March, May 2027 long rains. Workshop lessons from Ethiopia and India showed that even accurate forecasts have limited value when they are not communicated in ways that support decisions. Kenya has had a National Framework for Climate Services since 2024, and Farmer Field Schools and county level scenario planning are the channels the partners point to.
Based on the Forecast to Farmer pilot (CGIAR, Alliance of Bioversity and CIAT). The last step is our addition
The scale ambition is wider. AIM for Scale’s weather package, with WMO as technical adviser, works in Kenya, Ethiopia, Nigeria, Bangladesh and Chile, and plans to reach 30 countries by 2027. Outside Africa, CGIAR’s climate newsletter reports that IWMI’s forecast based advice helped Sri Lankan rice farmers harvest 226 kg more per acre than those outside the programme, an IWMI figure we did not check.
Releasing money before the loss
CGIAR’s climate director argues that response systems must release funding before losses mount. African evidence supports acting early. In 2017 FAO measured benefits of about $3.50 for each dollar spent on livestock in northern Kenya, rising to $9 with avoided food aid and restocking costs, and about $7 in Ethiopia’s Somali region. These are FAO’s own results from a drought year pilot, not a general rule.
Benefit per $1 spent, US dollars. FAO, 2018; Ethiopia figure reported as $6.9 in a 2017 account
For farmers who need credit rather than relief, ACRE Africa’s Tool for Agricultural Risk Assessment (TARA) combines climate, economic, satellite and smartphone image data into lending decisions, insurance recommendations and advice. IFPRI reports that ACRE has bundled index insurance with credit and inputs for close to 200,000 farmers across Kenya, Rwanda and Tanzania. A June workshop agreed to reach 50,000 Kenyan smallholders with TARA enabled loans bundled with insurance by 2028, with 50% women, 25% youth and 5% marginalised people. An earlier CGIAR case study counted 334 Kenyan farmers approved for funding from 569 trained, unlocking US$520,800 in credit.
Inclusion targets applied to 50,000 farmers (our arithmetic; groups may overlap). IFPRI and Alliance workshop report, 2026
Finance works best when it travels with skills and markets. The WFP newsletter reports that all 30 members of a Rwandan group, Twitezimbere Bweramana, now run businesses, up from four, and that their pepper enterprise grew from half a hectare to four. WFP’s SheCan programme combined finance, training and market access. The figures are WFP’s own.
Policy tools that answer in minutes
The CGIAR Policy Innovations newsletter describes Policy AI, a platform that lets ministers ask scenario questions in plain language, such as what a 20% fertiliser price rise would do to food security in Ethiopia. RIAPA AI is piloted in Ethiopia, Kenya, Malawi and Nigeria; the other modules are still at concept stage. The Kenya analysis above shows the idea at work, since RIAPA-AI produced its numbers. The event was an invitation only planning workshop, not a public launch, so Policy AI is promising but unproven at scale.
Where verified field data fits
Triggers, insurance and early finance all depend on knowing what happened on each plot. AgriKonnekt is built to check farm plots with sensors and satellites so cooperatives can show results to funders, according to its own description at AgriKonnekt. Whether it can feed forecast triggered advisories or insurance is our proposal, not a tested feature. The reason to raise it is that the same data could support farmers’ claims, loan decisions and the impact evidence that funders now ask for.
What to do next
- Check your county’s October, December advisory and plan drainage, harvest timing and storage before heavy rain.
- Ask extension officers and local radio for the Forecast to Farmer advisories during the short rains, and give feedback so the March, May 2027 rollout improves.
- Plan for 2027 input costs now, since IFPRI warns the fertiliser price spike may reach farmers next year.
- Cooperatives should ask lenders and insurers about loans bundled with insurance and keep plot level records that can verify losses.
- Funders should release early action money against agreed triggers rather than after losses show.
- Governments should publish scenario analyses early and keep climate services funded beyond pilots.
What we can and cannot claim
| Statement | Evidence status |
|---|---|
| Above average short rains are forecast across about 80% of Kenya | KMSA outlook as reported by K24 and Capital FM; we did not read the outlook document |
| National GDP loss of 1.1% to 1.5% in 2026 | IFPRI and KIPPRA model scenarios, read in full; not a forecast |
| Chance of a very strong El Niño | Sources differ: 65% for strong or very strong, and above 90% for very strong |
| Forecast to Farmer pilot runs this short rains | CGIAR newsletter and Alliance pages; no results yet |
| Returns of $3.50 to $9 per dollar in early action | FAO’s own 2017 pilot results; not independently audited |
| ACRE reached close to 200,000 farmers; target of 50,000 by 2028 | IFPRI reporting; targets are ambitions, not results |
| Verified plot data could support forecast based finance | Our argument; not tested |
Kenya’s 2026 season is a stress test for a chain Africa has been building for years: a forecast that is trusted, advice that is local, money that moves early and evidence that proves what happened. The science is ahead of the delivery. The evidence from FAO, ACRE Africa and the new pilot suggests the weakest links are institutions and finance, not meteorology. Farmers, cooperatives and lenders that connect those links this season will be better protected in March, when the long rains and a possible La Niña arrive.
Sources & notes
SOURCES: Carbon Trust, CGIAR Climate Action, CGIAR Scaling for Impact (The Scaling Signal), CGIAR Policy Innovations and WFP Innovation Accelerator newsletters, November 2026; IFPRI, Double shock: El Niño and rising global commodity prices in Kenya; IFPRI on Malawi; K24 on the KMSA short rains outlook; Capital FM on the outlook; Big3 Africa on Kenya Met; Farmbiz Africa on El Niño; ICPAC and FSNWG El Niño special report, August 2026; Big3 Africa on the Indian Ocean Dipole; Alliance of Bioversity and CIAT, Forecast to Farmer pilot; Alliance on farm level decisions; WMO, AIM for Scale technical advisory; IFPRI Africa on ACRE Africa and TARA; CGIAR case study on bundled credit; FAO, Acting early to prevent humanitarian emergencies; Down To Earth on early action in Africa; AgriKonnekt.
Differences between sources: NOAA odds are cited at 65% (strong or very strong) and above 90% (very strong); Ethiopia’s early action return is $7 per FAO and $6.9 in a 2017 account.
Our own arithmetic: the 20% remainder of Kenya; the 0.9% GDP loss under above average short rains; the 25,000, 12,500 and 2,500 farmer figures.
Not read: the Carbon Trust agriculture page, the KMSA outlook itself, the Policy AI platform and the ACRE Africa data. Programme results (WFP, IWMI, FAO, ACRE) are as stated by those organisations.