When Weather Meets the Artery War

Weekend edition. Evidence to 24 October 2026.

The rain that may fall on Garissa in November and the diesel sold in Nairobi this month are set by two different instruments, on two different clocks. One is a seasonal forecast signed off on 31 August. The other is a monthly price cap that took effect on 15 October. Both now shape the choices of households whose budgets have to absorb them together.

Calling them one crisis would be lazy. The Pacific and the shipping lanes of the Gulf are separate physical systems, with separate physics and separate institutions. But a household budget is a single ledger. Compound risk is what happens when two of its lines tighten at once, and the IPCC's Africa assessment says plainly that exposure and vulnerability on the continent are multi-dimensional, with socioeconomic, political and environmental factors intersecting (IPCC AR6 Working Group II, Chapter 9). Rain is a hazard. What it becomes depends on where it lands.

What the ocean is doing, and what that does not tell you

On 8 October NOAA's Climate Prediction Center kept its El Niño Advisory in place and reported the event still strengthening. Sea surface temperature anomalies exceeded +4.0 °C in portions of the eastern equatorial Pacific; the Niño-3.4 index stood at +2.1 °C. Using its relative index, the CPC put the odds of a three-month value at or above +2.5 °C — a level that would exceed every El Niño in its record back to 1950 — at 83 per cent for October–December, and said a strong-to-very strong event was likely to persist through January–March 2027 (NOAA Climate Prediction Center, 8 October 2026).

Read that 83 twice. It is a probability about an index of ocean temperature. It is not a probability that a river will break its banks, that a road will close, or that a child will fall ill. NOAA says so in the same document: with an event of this magnitude, the chances of experiencing impacts consistent with El Niño are larger "though not guaranteed".

In September the World Meteorological Organization called the event firmly established and put near-100 per cent odds on its persistence through February 2027 — the first time one of its updates had been so unequivocal — while warning that the strength of an event "does not determine how severe its impacts will be in any individual country or region", because the Indian and Atlantic oceans can reinforce, weaken or redirect the effect (WMO, 3 September 2026). Its October Global Seasonal Climate Update forecasts a seasonal average Niño-3.4 anomaly near +3.7 °C, peaking around December, above the +2.6 °C record set in 2015–16, with a positive Indian Ocean Dipole in play. It lists a wetter-than-normal signal across the Greater Horn of Africa, then adds the sentence that should sit above every climate headline: these signals "are not a substitute for national and local forecasts" (WMO, 8 October 2026).

What the national forecast actually says

Kenya's Meteorological Service issued its own answer on 31 August. Above-average rainfall is expected in a list of counties that includes Garissa, Wajir, Mandera, Tana River, Kilifi, Nairobi, Kisumu, Nakuru and much of West Pokot, Samburu and Marsabit; near-average to above-average conditions are likely in Turkana. Onset windows run from the second week of October in the highlands to the third and fourth weeks in the arid north-east, with cessation there in the first or second week of December. The season supplies more than 30 per cent of annual rainfall in several counties and can exceed 700 mm in places. Drought probabilities are low, except for a 15 to 20 per cent chance of a mild drought in Turkana (Kenya Meteorological Service, OND 2026 outlook, 31 August 2026).

Regionally, ICPAC's outlook forum in August put a 90 per cent chance on enhanced rainfall over southern Ethiopia, central-to-southern Somalia and north-eastern Kenya, and called seasonal totals above 400 mm likely in parts of central Kenya, the Lake Victoria basin, Burundi, western Rwanda and western Tanzania. It named 1997 and 2023 as analogue years (ICPAC, GHACOF 74, 18 August 2026).

Both are forecasts, and the ground they land on is not neutral. Kenya Met's own review of June to August found most of the country below average, with Lodwar, Garissa and Mandera recording no rain at all. FEWS NET goes further: June–August was the driest on record across the North Rift and western unimodal areas; the long-rains maize crop came in an estimated 22 to 28 per cent below the five-year average nationally and about 43 per cent below in the North Rift breadbasket counties; the government planned to facilitate imports of 25 million 90-kilogram bags; and its August analysis assumed at least a 50 per cent likelihood that short-rains rainfall would rank in the wettest quintile of recorded seasons (FEWS NET Kenya, August 2026). Its September update retained the outlook for above-average short rains and said expected flooding threatened recovery in eastern Kenya (FEWS NET Kenya, September 2026 update).

That deficit changes what a wet season means. Compacted, crusted or poorly drained ground can generate rapid runoff during heavy rain; FEWS NET warns about early-season flooding as conditions change. It also notes that the North Rift maize harvest will coincide with the peak of the forecast rains, raising the risk of post-harvest losses, rotting and aflatoxin contamination — a wet season that ruins the crop it was meant to rescue. Kenya Met's sectoral annex makes the same point in another register, listing waterlogging, nutrient leaching, pests, spoiled produce, and outbreaks of cholera, Rift Valley fever, chikungunya and malaria.

Hazard is not exposure

The same regional outlook can mean heightened flood risk in the Tana basin and potential pasture relief elsewhere, depending on local rainfall and exposure. Somalia's latest IPC analysis projects 5.4 million people in Crisis or worse between October and December, 1.5 million of them in Emergency, with 1.92 million children needing treatment for acute malnutrition in the year to June 2027; the country's disaster agency estimates 2.5 to 3.5 million people could be exposed to flooding and a million displaced (IPC Somalia, published by UNICEF, 23 September 2026). FEWS NET's September update projects Dadaab worsening to Emergency between October and January, while other pastoral areas improve to Stressed; it expects Crisis to persist in Garissa, Mandera, Tana River, Wajir and Turkana (FEWS NET Kenya, September 2026 update). Whether rain is a disaster depends on drainage, roads, storage, herd sizes and prices — on exposure, in the IPCC's language, and on vulnerability measured in who can move and who cannot.

The IPCC's Africa chapter explains why the channel is narrow: 55 to 62 per cent of the sub-Saharan workforce is employed in agriculture and 95 per cent of cropland is rainfed. It also finds increasing evidence linking higher temperatures and drought to conflict risk in Africa, while cautioning that climate is one of many interacting factors and may explain only a small share of the variation in conflict incidence. That caution matters, because the same ocean is delivering the opposite sign next door. On 12 October the IPC warned of a growing famine risk across parts of Greater Darfur and Greater Kordofan, where, in its words, "an El Niño-induced drought, affecting roughly half of the country, is compounding the impacts of conflict and displacement" (WHO note for media on the IPC Sudan alert, 12 October 2026). One signal, two outcomes, mediated by politics, infrastructure and money.

The artery bill arrives first

While the rain was still a probability, the regulated fuel price was a fact. EPRA's maximum retail price schedule for 15 October to 14 November set Nairobi diesel at 227.76 shillings and kerosene at 219.87; those caps are published by the regulator itself (EPRA pump prices). News reports of the October adjustment agree that diesel rose by 9.90 shillings a litre and kerosene by 28.49 after the subsidy was withdrawn, but differ on whether super petrol rose by 12.33 or 12.93 (Daily Nation, 19 October 2026; The Eastleigh Voice, 19 October 2026).

Illustrative arithmetic, with its assumptions stated: a household burning eight litres of kerosene a month pays roughly 228 shillings more than in the previous cycle (8 × 28.49), at unchanged volumes; a matatu burning 40 litres of diesel a day faces about 396 shillings a day more (40 × 9.90), before any fare adjustment. These are multiplications of announced per-litre changes. They are not forecasts of what anyone will spend.

In the days after that cycle began, EPRA's director-general told a parliamentary committee that two tankers carrying petroleum products destined for Kenya had been immobilised by drone strikes in the Strait of Hormuz, that transferring their cargo would take about 60 days, that importers were turning to costlier sources including Singapore, and that supply was being secured through the government-to-government arrangement with Saudi Arabia (Daily Nation, 19 October 2026). That is a supply risk, not yet a shortage: the cargo was reported safe, and rising prices are not the same thing as empty pumps. The distinction is worth keeping past the weekend.

Fuel does not stay in the tank. FAO's East Africa price round-up on 13 October attributes higher year-on-year coarse grain prices in most countries partly to elevated transport costs "reflecting higher fuel prices due to ongoing supply chain disruptions associated with the Middle East crisis". In Juba, diesel prices more than doubled between February and September; in Sudan, petrol and diesel in August stood 68 and 133 per cent above their year-earlier values (FAO GIEWS/FPMA East Africa round-up, 13 October 2026).

FAO's own index of internationally traded food commodities averaged 136.0 points in September, 1.5 per cent above August and 5.8 per cent above a year earlier — a measure of export quotations, not of Nairobi's market stalls. Its chief economist attributed the build-up to disruptions in the Strait of Hormuz and the Black Sea combining with climate shocks, and warned that if sustained, these pressures "will soon pass through to consumer food prices, especially in food and energy import dependent countries" (FAO, 2 October 2026). In Kenya, transport prices rose 15.6 per cent in the year to September against headline inflation of 6.8 per cent (Kenya National Bureau of Statistics, September 2026 CPI).

Where the two bills meet

Kenya Met's outlook is not only a rainfall document. Its macroeconomy section anticipates "increased fiscal pressure due to additional government spending" — repairs, search and rescue, evacuations — alongside lower revenue collection, and it lists food and freight subsidies, temporary tax deferrals and emergency credit lines among the responses. A seasonal rain forecast is already a budget forecast.

That budget has little slack. September inflation sat near the top of the central bank's target band, with transport its sharpest large component, and the Monetary Policy Committee left the Central Bank Rate unchanged at 8.75 per cent on 7 October (Central Bank of Kenya, 7 October 2026). The reduced 8 per cent VAT rate on petroleum products — down from the standard 16 per cent — was extended for another 90 days effective 15 October; without that, January's price cycle would start from a worse place (KBC, 24 October 2026).

The health system is carrying its own load. WHO estimates that more than 37.8 million people were experiencing, or projected to experience, acute food insecurity across six Greater Horn countries, driven by "the convergence of acute food insecurity, conflict, climate shocks, population displacement, recurrent disease outbreaks, economic pressures and fragile health systems"; it asked for about 25.4 million dollars to sustain priority work for six months (WHO, Greater Horn of Africa situation report No. 1, 7 July 2026). Kenya's own weekly emergency reporting in June listed cholera in two counties, dengue in Garissa, measles in several and mpox across nineteen — before the forecast October–December rains (WHO Kenya health emergencies, week 26 of 2026). The country has also recorded one imported case of Bundibugyo virus disease, with 28 contacts listed and more than 652,000 travellers screened by early October (WHO Regional Office for Africa, October 2026).

What would change the picture — and what would not

Some of what sits in this article is evidence: price changes, caps, index values, case counts. Some is forecast, carrying dates and probabilities. Some is scenario — the harvest lost to a flood, the camp that deteriorates. And some is belief: that a coordinated war on trade routes is being fought; that this El Niño will be catastrophic for Kenya; that the one proves the other. Beliefs may be argued. They cannot be laundered into measurement.

The checks are scheduled. NOAA's next ENSO discussion is due on 12 November. Kenya Met issues monthly and weekly updates and instructs users to read the seasonal outlook with them rather than instead of them. The rains will either start inside the window the outlook drew, or they will not.

What the evidence supports is narrower and more useful: a wet season arriving into a deficit, while transport prices were already 15.6 per cent higher year on year in September. The preparation that pays either way does not require knowing which. Clear the drains and inspect the culverts. Move medicines, fuel and food above the floor. Agree the alternative route for a referral and for a harvest. Keep the storage dry. And do not treat a single supply corridor as a guarantee.

The forecast is probabilistic. The cost of being wrong is not symmetric. That asymmetry is the reason to read the rain and the freight bill in the same sitting.

Floodwater on the C12 Narok–Sekenani road in Kenya, photographed 29 December 2019. This file photograph shows a transport disruption, not current road conditions. Credit: Mikenjau. Image source · CC BY-SA 4.0 International