What Is the IRI ENSO Forecast?
The IRI ENSO forecast comes from the International Research Institute for Climate and Society, a climate research center at Columbia University. Every month, IRI — in partnership with NOAA's Climate Prediction Center — publishes a probabilistic outlook for the state of the El Niño-Southern Oscillation (ENSO) over the next nine months.
The output has two parts that get combined into one chart. The first is the plume: a set of thin lines, each one a different model's forecast for the Niño 3.4 sea surface temperature anomaly. The second is the probability bar underneath, which converts that ensemble into plain numbers — the official chance of El Niño, neutral, or La Niña conditions in each upcoming three-month season.
IRI doesn't run one giant model. It aggregates roughly 20-25 models — a mix of full-physics dynamical models and simpler statistical models — which is exactly why the chart shows a spread rather than a single line. That spread is not noise; it's the honest representation of how uncertain the forecast actually is.
How to Read the IRI Plume
If you're staring at the plume for the first time, start with three elements:
The horizontal threshold lines. The red dashed line at +0.5°C is the El Niño threshold; the blue dashed line at −0.5°C is the La Niña threshold. The index has to hold at or above +0.5°C for five consecutive overlapping three-month seasons to qualify as an official El Niño, but for reading the plume, the simple question is: are the lines above red, below blue, or squeezed between them?
The ensemble mean. The thick line running through the middle is the average of all the models. It smooths out the outliers and gives you the best single-number answer. When the thick line sits well above +0.5°C through the winter months, that's the models collectively saying a sustained event is likely.
The spread. Watch how the individual lines fan out as the forecast extends into the future. A tight cluster in the first three months means high confidence. A wide fan by month six or seven means the models genuinely disagree — and you should treat any single prediction that far out with a grain of salt.
One trap to avoid: reading a single low line as "the forecast is wrong." The plume is an ensemble on purpose. One statistical model dipping negative while every dynamical model stays positive is not a contradiction — it's a reminder that ENSO is a noisy system and no model is perfect.
What the 2026 Plume Shows
The 2026 event is already underway, so the plume is less about whether El Niño arrives and more about how long it lasts and how strong it peaks. NOAA declared El Niño conditions on June 11, 2026, and the Oceanic Niño Index reached +1.39°C in the May-July period — a fast climb from the −0.21°C of January-March, and one of the sharpest La Niña-to-El Niño flips on record.
The IRI/CPC ensemble has been pointing in one direction for months: continued El Niño through the 2026-27 winter. NOAA's Climate Prediction Center currently gives a 63% chance this event reaches Super El Niño strength (an ONI of +2.0°C or higher) by mid-winter. That would put it in the same category as 1997-98, 2015-16, and 2023-24.
Because the plume is updated monthly — and each month's data reshapes the spread — the exact probabilities shift. For the current, month-by-month numbers, the authoritative source is the live plume on iri.columbia.edu. The dashboard on this site tracks the underlying ONI and Niño region indices that feed that forecast.
IRI vs. NOAA CPC and NMME
It's easy to conflate the IRI forecast with NOAA's, because the two institutions co-publish the official outlook. Here's the practical distinction:
IRI plume is the visualization and the model aggregation — the ensemble of ~20-25 models and the resulting probability bars. NOAA CPC is the operational authority that issues the official ENSO status and the textual forecast discussion that accompanies those probabilities. NMME (the North American Multi-Model Ensemble) is a subset of dynamical models — NOAA's CFSv2, NASA's GEOS, NCAR's CCSM4 and others — that IRI and CPC both draw on for the plume.
In practice, they agree more often than they disagree, because they're built on overlapping data and models. When they do diverge — say, the plume's statistical models lean neutral while the dynamical NMME members lean strongly El Niño — the dynamical models have historically carried more weight for forecasts beyond three to four months, as detailed in our forecast methods guide.
Why the IRI Plume Matters for You
You don't need to be a climate scientist to get value from the plume. It's the difference between "there might be an El Niño" and "there's a 63% chance of a strong one by winter." For a homeowner in California deciding whether to buy flood insurance, a farmer in Australia weighing crop choice, or a ski resort owner planning staffing, that number — and how it shifts month to month — is actionable.
Watch the trend more than any single month. If the ensemble mean keeps climbing update after update, confidence is building. If the probability bars swing back toward neutral, the event may be fading. For a plain-English read on where things stand right now, start with our El Niño 2026 hub, then check the real-time dashboard for the underlying numbers.