ECMWF and Copernicus Models Point to a Record December Peak — With an Important Caveat

The latest round of seasonal forecasts coordinated by Europe's Copernicus Climate Change Service delivers the most striking numbers yet for the 2026 El Niño: a multi-model mean Niño 3.4 anomaly of roughly 3.5°C for December, with about 91% of ensemble members exceeding the strongest event ever observed. But the forecasters supplying the numbers are also supplying a caveat that should temper how "record" is read.

The projections come from the C3S multi-system ensemble, which combines nine of the world's leading seasonal prediction models, initialised with data from around the start of August. The spread of the ensemble is as notable as its centre: the middle 80% of members sits entirely at or above the all-time record, and even the low end of the plume lands near 2.8°C — above anything in the observational books.

How far past the record?

For context, the strongest El Niño events in the instrumental record peaked at an Oceanic Niño Index of +2.4°C in 1997–98 and +2.8°C in 2015–16. A December mean near 3.5°C would not merely edge past those benchmarks; it would clear them by close to a degree, a margin without precedent in seasonal forecasting of ENSO.

The models do not fully agree on timing. The American CFSv2 system favours a November peak, while ECMWF's own system points to December. That kind of one-month disagreement is normal at this range and matters little for impacts planning; the consensus on a late-2026 peak of exceptional amplitude is the operative signal, and it is consistent with what NOAA laid out in NOAA's July diagnostic discussion.

Confidence has firmed markedly since spring. In an ECMWF science blog published in April, forecast researcher Tim Stockdale had cautioned that "it is too early to assign high confidence" in the more dramatic scenarios. Subsequent forecast cycles tightened the ensemble upward rather than back toward the mean — the pattern that typically precedes a major event.

The relative-index caveat

Here is where ECMWF urges care. In a companion blog on measuring El Niño strength, the centre makes a point that complicates any "strongest ever" headline: the background ocean is at record warmth, and absolute SST anomalies are computed against historical baselines. Part of a 3.5°C anomaly in 2026 therefore reflects long-term warming of the entire tropical ocean, not the El Niño pattern itself.

The remedy is what ECMWF calls relative Niño indices — versions of the standard indices that subtract the tropical-mean warming, isolating how much the central Pacific stands out from its surroundings. On a relative basis, the 2026 event still measures as very large, but the comparison with 1997–98 and 2015–16 becomes less extreme than the raw numbers suggest. Australia's Bureau of Meteorology has adopted the same philosophy operationally, running its ENSO monitoring on a relative Niño3.4 index for exactly this reason.

The distinction is not academic. Many El Niño impacts depend on where atmospheric convection sets up, which responds to temperature contrasts across the tropics rather than absolute values — a point developed further in the science of El Niño. An event that looks unprecedented in absolute terms may drive an atmospheric response closer to that of past strong events.

What the ensemble can and cannot say

Seasonal ensembles quantify the range of plausible outcomes, not a single destiny. What this one says is that an exceptionally strong peak is now the central expectation, that even the cautious edge of the model spread implies a top-tier event, and that the peak arrives in the November–December window. What it cannot say is how the impacts — the rainfall shifts, the marine heat, the regional droughts — will scale against a baseline ocean already warmer than any previous event encountered.

C3S refreshes its multi-model forecasts monthly, with releases around the 10th to 13th. The September cycle, initialised as the event approaches its projected peak, will be the next major checkpoint for whether the 3.5°C December figure holds.