El Niño in East Africa: Heavy Short Rains, Floods and a Fragile Recovery

For most regions in this series, El Niño's arrival reads as a threat. East Africa is more complicated: the same event that parches Indonesia and Australia historically delivers rain to the Horn of Africa — in the October–December "short rains" season that can refill reservoirs and pastures, or drown them, depending on how hard the pattern hits. With a strong El Niño declared in June 2026 and the short rains four months away, the region faces both faces of the phenomenon at once — while still climbing out of its worst drought in decades.

How El Niño changes East African weather

East Africa's equatorial belt gets two rainy seasons: the "long rains" of March–May and the short rains of October–December. It is the short rains that ENSO grips. During El Niño, the warm Indian Ocean and the reorganized Walker circulation feed moisture and rising air over the region precisely in those months — one of the cleaner teleconnections in the tropics.

The signal rarely travels alone. El Niño years frequently coincide with a positive Indian Ocean Dipole — anomalously warm seas off East Africa — which independently pumps up short-rains totals. When the two align, as in 1997 and 2019 (an IOD-dominant year), the region's rainfall can run to multiples of normal for weeks at a time.

The flip side of the seesaw: the northern parts of the region tied to June–September rains — much of Ethiopia's highlands and Sudan — historically lean drier in El Niño years, the same broad mechanism that pressures the Indian monsoon. One event, two opposite regional stories.

The historical record

1997-98 remains the benchmark. The short rains of late 1997, supercharged by the century's strongest event and a strong positive IOD, brought catastrophic flooding across Kenya, Somalia and southern Ethiopia — hundreds of thousands displaced, roads and rail cut for months, and a massive Rift Valley fever epidemic through livestock and people. Humanitarian logistics were rewritten by that season.

2015-16 showed the split screen. The record event's short rains ran generously wet in the equatorial belt, while Ethiopia's June–September season failed badly — one of its worst droughts in decades, with tens of millions needing assistance. Same El Niño, opposite emergencies a few hundred kilometers apart.

2023's short rains — under the moderate El Niño and positive IOD of that year — again brought serious flooding to Kenya and Somalia, straight on the heels of the 2020–23 multi-year La Niña drought that had pushed the Horn to the edge of famine. That whiplash — years of drought, then floods on baked, degraded land — is the region's defining climate-risk pattern.

What the 2026–27 forecast implies

The base case from history: a wetter-than-normal October–December across Kenya, Somalia, southern Ethiopia and Tanzania, with flood risk concentrated in river basins (the Tana, Juba and Shabelle have form) and in cities whose drainage has not kept pace with their growth. If a positive Indian Ocean Dipole develops alongside — the historical companion worth watching — the high-end scenarios firm up.

The drought-recovery context cuts both ways. Pastures and water tables devastated by the 2020–23 drought genuinely need a generous season, and a well-behaved wet year would be a humanitarian dividend. But drought-hardened ground sheds torrential rain rather than absorbing it, and livestock herds rebuilt near rivers sit in the flood path. The difference between relief and disaster is mostly rainfall intensity — which no seasonal forecast can promise.

Health systems get actionable lead time: Rift Valley fever risk maps key directly off El Niño short-rains forecasts, and vaccination and surveillance campaigns launched before the rains have historically paid for themselves many times over. The same logic applies to cholera pre-positioning and malaria-control planning in highland fringe zones.

The economics of acting early are unusually well documented for this region. Anticipatory-action programs — cash transfers, destocking support and supply pre-positioning triggered by forecasts rather than disasters — were pioneered in the Horn precisely because its El Niño and La Niña signals are strong enough to act on. The 2026 declaration, arriving four months before the short rains, is the textbook activation scenario: the cost of preparing for a wet season that behaves is a rounding error against the cost of an unprepared 1997-style one.

Urban exposure has grown faster than any other variable since the benchmark years. Nairobi, Mombasa, Dar es Salaam and Mogadishu have all added millions of residents — many in informal settlements on floodplains and drainage paths — since 1997. The same rainfall anomaly now meets far more people and pavement, which is why recent wet seasons produced urban flood emergencies from rainfall totals the region's history would rate as merely generous.

For the northern, boreal-summer-rain zones, 2026's June–September season carries the historical dry lean — a reminder that Ethiopia and Sudan planning cannot simply mirror Nairobi's.

Who should prepare, and how

Governments and humanitarian agencies: the 1997 and 2023 playbooks — pre-positioned supplies in cut-off-prone areas, early-warning dissemination down to river-basin level, livestock vaccination ahead of the rains — exist because this exact setup recurs. River-adjacent and urban flood-plain communities: the season to clear drainage and plan evacuation routes is before October. Insurers and lenders exposed to the region's agriculture: parametric structures triggered on rainfall indexes were built for seasons like this (insurance and risk guide).

What to watch

• The Indian Ocean Dipole index through August–September — the amplifier that separates wet from catastrophic.
• Official short-rains outlooks from the regional climate centers (ICPAC) as October approaches.
• River-level monitoring on the Tana, Juba and Shabelle once the rains begin.
• The event's strength on the live dashboard — the stronger the Pacific signal, the firmer the regional tilt.