At a glance · 3-minute read
This week's central question is whether public institutions can act on earlier climate information. On 3 September, the World Meteorological Organization (WMO) said El Niño was firmly established and likely to become very strong, with a near 100 per cent likelihood of persisting through February 2027. For Kenya, that global signal coincides with regional and national outlooks for wetter-than-normal October–December conditions, including a 90 per cent probability of enhanced rainfall over north-eastern Kenya. A more confident forecast does not by itself protect people or services.
The week's most consequential institutional development was the Addis Ababa Call to Action on Early Warnings in Africa. It treats warning systems as national infrastructure and calls for pre-arranged finance linked to forecasts and triggers. Kenya's immediate question is practical: can national forecasts activate county decisions before heavy rain becomes a compound emergency?
The other developments test how institutions use new evidence. Kenya has commissioned work to connect climate exposure with health investment. African finance is being mobilised against fertiliser and energy shocks, although it may reduce structural dependence or merely postpone it. Air-pollution evidence now bears on health and climate policy, while data centres and artificial intelligence (AI) in aviation create new demands on energy, water and regulation. In each case, the practical question is whether information leads to authorised, funded and measurable action.
Which public authority must act, with what funds, when an official forecast gives earlier warning?
- Regional seasonal outlook from the 74th Greater Horn of Africa Climate Outlook Forum
- Kenya Meteorological Department outlook
- WMO El Niño update
- Addis Call to Action
- Expected onset, north-east
Seasonal timing is indicative. Shorter-range forecasts should govern operational decisions.
How the week's developments connect
How the week’s evidence affects people, services and public decisions
| Pressure or opportunity | Systems interacting | How the effect travels | Who or what is affected first | Further consequence |
|---|---|---|---|---|
| Very strong El Niño; wetter October–December outlook | Weather, water, food, health, transport and finance | Forecasts alter planting, drainage, surveillance and contingency decisions | Arid and semi-arid lands, lake basins, the coast and informal settlements | Rain may restore water and pasture while increasing flooding and disease |
| Early-warning reform | Meteorology, telecoms, local government and social protection | Warnings require trusted channels, authority and pre-arranged finance | Remote, displaced and low-connectivity communities | Warnings without feasible action can erode public trust |
| Climate-health assessment | Climate data, epidemiology, hospitals and county planning | Exposure evidence can redirect investment before caseloads rise | Heat-exposed, flood-prone and pastoral populations | Surveillance choices shape future health capital |
| Energy and fertiliser shock | Trade, farming, inflation and fiscal policy | Import costs reach inputs, food prices, services and budgets | Farmers, households and import-dependent states | Short-term relief may entrench structural dependency |
| AI and digital infrastructure | Electricity, water, land, aviation and regulation | Computing consumes resources; prediction may reduce warming | Data-hub cities, grid users and air corridors | A climate solution in one system may create scarcity in another |
Principal developments
What changed during the week.
El Niño becomes an unequivocal planning condition
The WMO update published on 3 September is unusually definite: El Niño is established, is expected to intensify to very strong levels and is almost certain to persist through February 2027. WMO also cautions that intensity does not determine impacts uniformly, because the Indian and Atlantic oceans and local conditions can modify the result.
For Kenya, that distinction is essential. The Intergovernmental Authority on Development (IGAD) Climate Prediction and Applications Centre (ICPAC) assigns a 90 per cent chance of enhanced October–December rainfall to north-eastern Kenya, while the Kenya Meteorological Department's October–December outlook anticipates above-average rain in most counties, with heavy events possible. Marsabit, Mandera, Wajir, Garissa and Isiolo are expected to receive onset in the second or third week of October and cessation in the first or second week of December.
This is not a flood prediction. It is a narrowing planning window. Counties should inspect drainage, water-retaining structures, health supplies, livestock vaccination, road access and food storage against shorter-range forecasts. Aquifer recharge, pasture recovery and higher output remain possible if water, animal-health and post-harvest systems are ready.
Africa defines early warning as infrastructure, not messaging
The Africa Early Warnings Forum met in Addis Ababa from 2 to 4 September, and its Call to Action was published on 7 September. Forty-one African countries now report on early-warning systems, and WMO says their comprehensiveness has more than doubled since 2022. The remaining weaknesses are revealing: observing networks, last-mile communication, institutional mandates, domestic finance and the ability to act.
The Call's most important move is fiscal. It asks governments to place warning systems in annual budgets and establish pre-arranged funding linked to forecasts and triggers. Performance would be judged by whether warnings reach people, prompt action and reduce losses.
For Kenya, cell broadcast, radio, extension services and local leaders are delivery channels, not the whole system. A complete system determines who authorises action, which budget becomes available, what counties deliver and how results are audited. In border areas, pastoral landscapes and informal settlements, a correct warning may still be impossible to follow.
Kenya begins building an evidence base for climate-health investment
On 4 September, Kenya's Ministry of Health published terms of reference for a national Climate Change and Health Vulnerability and Adaptation Assessment. The proposed work covers selected counties across climatic zones, combines climate, health and socioeconomic data, includes field inquiry and modelling, and is intended to support a health adaptation plan and investment case. It seeks projections for communicable and non-communicable disease risks to 2050.
The assessment aims to turn climate-health risk into evidence that officials can use. Floods can damage sanitation, heat can reduce productivity, and ecological change can alter vector and livestock disease. The assessment should not end as a national average. Counties need thresholds for cholera supplies, vector surveillance, heat protocols and continuity of care. County selection, model transparency and the budget response are the points to watch.
Wildfire smoke becomes a public-health planning issue in Africa
The WMO Air Quality and Climate Bulletin No. 6, published on 7 September, reports that 2025 fine-particle concentrations were above long-term averages in western-central Africa because of heightened fire activity. It also stresses that heat, wildfire smoke and ground-level ozone can offset gains from reducing transport and industrial emissions. Research cited by WMO associates the global tripling of extreme fire-smoke events since the 1990s with nearly 100,000 additional deaths a year from 2010 to 2018, but that estimate should not be mechanically transferred to Kenya.
Air basins cross administrative boundaries. Health surveillance, fire management, land policy, meteorology and urban pollution control cannot remain separate. Kenya's monitors are more useful when readings inform hospitals and advice for children, older people, those with respiratory disease and outdoor workers. Smoke and heat should enter the new climate-health assessment alongside rainfall-sensitive disease.
A US$5.1 billion response raises a longer-term question about import dependence
The African Development Bank announced on 7 September that its Board had approved, on 1 September, a Global Energy and Fertilizer Crisis Response Framework of up to US$5.1 billion. The one-year, demand-driven framework combines US$4.1 billion in additional African Development Bank lending with up to US$960 million from the African Development Fund. It responds to higher energy, food, fertiliser and logistics costs associated with conflict and disrupted trade routes.
Expensive fertiliser can reduce yields and income, raise food prices and encourage land conversion. Expensive fuel raises transport, irrigation and service costs. Indiscriminate subsidy or emergency lending can lock governments into the imported inputs and carbon-intensive systems that created the exposure.
Kenya should distinguish immediate supply protection from structural investment. Beyond targeted support before planting, longer-term value would come from lower-emission regional fertiliser production, efficient nutrient use, soil testing and stronger logistics. Transparent loan terms and protection of health and social spending will determine whether the framework absorbs a shock or moves it onto public balance sheets.
South Africa's data-centre boom faces a human-rights inquiry
On 4 September, the Associated Press reported that the South African Human Rights Commission had received more than 250 submissions after opening an inquiry into the expansion of data centres. The Commission's formal notice asks about electricity, tariffs, water and cooling, land, climate justice, e-waste, privacy, consultation and regulatory gaps. Civil-society organisations have sought a pause on new construction pending investigation, while industry representatives contest claims that data centres materially drive scarcity and point to renewable energy and lower-water cooling.
Artificial intelligence infrastructure is a land, water, electricity, industrial and rights question. Kenya's renewable advantages may attract similar investment. It should require facility-level disclosure of energy source, hourly demand, water consumption, backup generation, land impacts, employment and public-system costs. The same rules should apply to domestic and foreign operators.
AI-assisted contrail avoidance moves into live long-haul operations
On 7 September, Cathay Pacific and Google announced an expanded Asia-Pacific contrail-avoidance trial. Their system combines predictive models, weather information and satellite detection to identify conditions in which persistent warming contrails may form, allowing dispatchers and pilots to consider small altitude changes. More than 80 flights followed avoidance routes in an initial trial, and Google estimates a roughly 40 per cent reduction in contrail warming on those flights.
The new element is operational testing on ultra-long-haul routes in an under-researched region. The result comes from the partners, applies to selected flights and does not establish system-wide benefit after fuel penalties and air-traffic constraints. For Kenya Airways, the civil aviation regulator and African air-navigation institutions, this is a standards question before procurement. African airspace and weather data should enter validation; methods and fuel trade-offs should be auditable.
Differentiated implications for Kenya
The same outlook creates different risks and opportunities across Kenya
In Garissa, Wajir, Mandera, Isiolo and Marsabit, better rain can restore water points, pasture and livestock value after recurrent drought. The same rainfall can sever roads, contaminate shallow sources, displace households and create conditions favourable to Rift Valley fever and other vector-borne disease. County action should therefore pair water harvesting and rangeland recovery with animal surveillance, safe-water, sanitation and hygiene measures, and access planning. Pastoral mobility and cross-border information are central, not peripheral.
In the Lake Victoria basin and western highlands, rain raises crop and hydropower opportunity alongside flood, landslide and post-harvest risks. In central highlands, soil conservation and storage will decide whether moisture becomes output or erosion. On the coast, warmth and heavy rain implicate vector control, transport, tourism, fisheries and drainage. In major towns, low-income residents carry disproportionate flood and heat exposure.
The macroeconomic channels converge. Better harvests could ease food inflation, but fertiliser and fuel costs may suppress that benefit. Flood damage can enlarge supplementary budgets and disrupt trade. Health outbreaks move costs from prevention to treatment. Data-centre investment can add capital, skills and demand for renewable power, but opaque water and grid obligations may shift costs to households and municipalities.
Responsibility is divided between national and county institutions under Kenya's devolved system. National institutions provide authoritative forecasts, national financing, standards, major infrastructure and regulatory frameworks. Counties deliver much of the practical response through local health, agriculture, water and sanitation, county roads, planning and emergency services. The operational requirement is a common trigger, a costed county action, a named authorising officer, available money and a public record of what happened.
Africa and the wider international system
Climate exposure differs by ecology, infrastructure, trade dependence and institutional capacity
The week's evidence shows that exposure differs across Africa according to institutional and ecological conditions, not simply national borders. Island states and fragile countries face thin observing networks and financing. Import-dependent economies are exposed to energy and fertiliser routes. South Africa and emerging data hubs confront resource allocation questions sooner than countries with less digital capacity. The Greater Horn faces an immediate warning-to-action test, while fire-prone parts of western and central Africa need integrated smoke, land and health systems.
The African Union Commission has been asked to put into practice the Addis Ababa Call. Regional bodies must make forecasts, disease surveillance and warning protocols interoperable across borders. Development-finance institutions should publish whether crisis lending reduces import dependence and protects essential spending. Scientific networks should negotiate access to model inputs, validation data and intellectual property.
Africa's bargaining position improves when it can specify the public value demanded from technology and finance: open methods, local validation, fair infrastructure pricing, domestic skills, transparent environmental performance and institutions capable of saying no. That is more consequential than the language of partnership alone.
Public finance and preparedness
The warning-to-budget gap.
A warning reduces harm only when each of six steps works: observation, forecast, an authorised warning, predefined action, finance and accountable delivery. Investment often stops after the first three. A credible forecast cannot protect people if the responsible institution cannot release funds or households cannot act.
Kenya has many components: seasonal and shorter-range forecasts, National Drought Management Authority practice, county sector departments, emergency instruments, scalable social protection and livestock insurance. The Disaster Risk Financing Strategy 2026–2030 seeks a shift from reactive reallocations to layered, pre-arranged finance and notes uneven coverage for floods and other sudden hazards.
What remains missing is a routine rule connecting a probabilistic warning to a limited, reversible set of funded actions. A forecast should not automatically declare a disaster or suspend normal accountability. It should authorise preparedness within agreed bounds: clearing specific drains, protecting water points, pre-positioning tests and medicines, vaccinating animals, moving learning materials, inspecting dams or expanding temporary cash support. The action must be costed before the warning, matched to a responsible agency and assessed after the season.
Performance should be judged on whether a trigger was reached, finance moved on time, action reached the intended place and avoidable loss fell. Forecasting institutions would need dated thresholds, stated uncertainty and reviews of false alarms. Public value comes from acting early while keeping uncertainty visible and decisions reviewable.
Developments to watch
Six developments that could change the assessment
| Item | Why and when | What would alter the assessment |
|---|---|---|
| El Niño and October–December forecasts | Weekly to monthly updates through December will sharpen location and timing. | A material change in Pacific intensity, Indian Ocean Dipole conditions or Kenya Meteorological Department county guidance. |
| Addis Ababa implementation | The African Union Commission and partners must translate the Call into mandates, budgets and dates during 2026–27. | Published responsibilities, domestic allocations and loss-reduction indicators. |
| Kenya climate-health assessment | Procurement, county selection and data design should emerge in the coming months. | Transparent methods, representative counties and a funded adaptation pathway. |
| African Development Bank crisis framework | Country operations and terms should appear within its one-year window. | Evidence that lending protects essential spending and builds lower-dependence supply. |
| Data-centre rights review | The South African inquiry may produce a continental regulatory reference. | Facility-level disclosures and enforceable resource rules rather than voluntary claims. |
| Contrail trial phase two | Results should develop as routes and operational conditions expand. | Independent methods, fuel-penalty accounting and regulator engagement. |
A proposal for Kenya
A rule for releasing county preparedness funds when forecast thresholds are met
Kenya should create a public-finance protocol that makes a small, predefined portion of existing national and county budgets available when authoritative hazard thresholds are crossed. It is a rule for moving money before impact, not a new fund, platform or coordinating body.
How the rule would work
The case for the rule rests on a very strong El Niño outlook, an African call for forecast-linked finance and Kenya's assessment that disaster finance remains too reactive. Existing components include Kenya Meteorological Department forecasts, National Drought Management Authority systems, county emergency arrangements, the Hunger Safety Net Programme, livestock insurance, the Financing Locally Led Climate Action programme and the national disaster-risk financing strategy. The proposal would add a common legal standard linking a dated forecast, a place-specific action plan and a capped release of funds.
Ownership should sit with the National Treasury and county treasuries, with thresholds certified by the Kenya Meteorological Department and the relevant hazard authority. Sector agencies would maintain pre-costed actions. The Council of Governors should agree a model protocol; county assemblies would appropriate eligible lines; the Controller of Budget and Auditor-General would retain oversight. Public-finance regulations, county legislation and existing disaster-risk policy could provide the legal base.
Money, data and safeguards
Different sources of finance would cover different scales of action. Frequent actions would use preparedness lines and county emergency appropriations. Larger events could activate national contingencies, scalable social protection, insurance or partner matching. Release would be capped and time-limited. Budgets should disclose the amount exposed to triggers and the actions purchased.
The proposal needs only two linked public records: one for forecasts and one for actions. The forecast record would preserve the issuing agency, model cycle, probability, geographic unit and revision history. A linked action register would show the threshold, responsible office, budget code, supplier or delivery channel and completion evidence. Non-personal records should be open. Household data should remain within protected government registries, with role-based access, data minimisation and independent audit. AI may assist anomaly detection or local downscaling, but it should neither authorise public expenditure nor override accountable officials. Counties and Kenyan institutions should own locally generated operational data; vendors should receive limited licences, not control.
100 days, three years, 2035
Select three actions each in Garissa, Kisumu and Mombasa; define thresholds linked to Kenya Meteorological Department forecasts; map the lawful budget route; run a tabletop activation; publish the result.
Fifteen counties operate audited protocols, common budget codes and independent evaluations. Communities and universities test whether action is equitable and locally useful.
Kenya routinely releases and audits preparedness funding against forecast thresholds, under national rules and through county delivery.
The risks are automatic spending based on an inaccurate forecast, gamed thresholds, procurement shortcuts, exclusion through weak registries and intrusion into county functions. Caps, multiple evidence thresholds, manual confirmation, sunset clauses and audits can contain them.
This is a policy proposal, not a forecast. It would allow government to act earlier on authoritative warnings while retaining legal limits, human approval and public audit.
Sources and evidence notes
- World Meteorological Organization, El Niño set to become very strong, raising risks of extreme weather into 2027, 3 September 2026.
- Climate Prediction and Applications Centre of the Intergovernmental Authority on Development, Greater Horn expected to experience a wetter-than-normal October–December 2026 season, 18 August 2026; read alongside the 3 September World Meteorological Organization update, which strengthened the global El Niño signal.
- Kenya Meteorological Department, October–December 2026 Seasonal Outlook, issued 31 August 2026.
- World Meteorological Organization, Addis Ababa Call to Action on Early Warnings in Africa, 7 September 2026; forum held 2–4 September 2026.
- Ministry of Health, Kenya, Terms of Reference: Climate Change and Health Vulnerability and Adaptation Assessment, published 4 September 2026.
- World Meteorological Organization, Air Quality and Climate Bulletin No. 6, 7 September 2026.
- African Development Bank, Global Energy and Fertilizer Crisis Response Framework, announced 7 September 2026; approved 1 September 2026.
- South African Human Rights Commission, Call for public submissions on data centres and human rights, 22 May 2026; Associated Press reported on 4 September 2026 that the Commission had received more than 250 submissions.
- Cathay Pacific, AI-powered contrail avoidance trial with Google, 7 September 2026.
- National Treasury, Kenya, Disaster Risk Financing Strategy 2026–2030, 2026.
Evidence labels show how firmly each cited claim is established, not how important the issue is. Forecast probabilities are not impact probabilities; institutional announcements are not proof of implementation; commercial trial results require independent validation.