A Just Have a Think video pointed me at Zeke Hausfather’s piece in The Climate Brink. The headline: seasonal models give the 2026 El Niño about a 90 percent chance of setting a record, with the central Pacific forecast to peak near +3.6 °C late in the year, past the 2015 to 2016 event.
I wanted to understand how a number like that is measured and forecast. The map above is what I built to find out. It redraws from NOAA satellite data every morning, so we can watch the event unfold through December. Here is what its pieces mean.
Temperature is the wrong number
Raw ocean temperature is dominated by the seasons and by latitude, neither of which has anything to do with El Niño. So climate science plots the anomaly instead: how far the temperature is from normal for that spot on that day of the year.
The second term is the long-term average temperature for that spot on that day of the year, taken here over 1971 to 2000. This baseline is the climatological mean. Subtract it and the seasonal cycle drops out, leaving the departure from normal. The red tongue on the map is that anomaly, not the raw temperature, and it is the view that brings El Niño into focus.
The ocean lives on a grid
The data is NOAA’s OISST, version 2.1: a daily estimate of sea surface temperature built by blending satellites with ships and buoys. It is stored on a grid of quarter-degree cells, about 25 km each, one anomaly value per cell, every day back to 1981.
Zoom into the map and the smooth field breaks into those cells. This matters because everything below, every index and every forecast, is just arithmetic on this grid.
The index is a box
El Niño is measured inside named boxes on the equator. The map draws four; the standard one is Niño 3.4, from 170°W to 120°W. The index is nothing more than the average anomaly of the cells inside that box. Toggle Grid on the map to see the definition directly: a few thousand cells collapsing into one number. The official ONI is just a three-month average of that value. It crosses +0.5 °C for an El Niño and +2.0 °C for a very strong one.
The scorecard tracks all four boxes at once. The warmth leans east, with Niño 3 and Niño 1+2 running hotter than Niño 4, which is the shape of a strong event.
The forecast is a crowd
The record claim is about the peak, months from now, so it comes from seasonal forecast models: physics simulations started from July’s observed ocean and run to the end of the year, drawn from two collections, the NMME and Copernicus C3S. Each model runs an ensemble of slightly different copies to map out the uncertainty.
The article distilled 14 of them to a single peak Niño 3.4 value each. They range from +2.2 °C to +5.3 °C, with a median of +3.6 °C. The map plots those as dots at each model’s own peak month. The observed line climbs from the left toward that cluster, and the gap between the two is the whole story.
The warming underneath: RONI
Here is the catch. The entire tropical ocean has warmed over the past half century, so measured against a 1971 to 2000 baseline, part of every warm reading is that background, not El Niño. The fix is RONI, the relative index, which subtracts off what the rest of the tropics is doing at the same moment.
Toggle RONI on the map and the field cools: the uniform warming is stripped away, leaving the El Niño pattern behind. The Niño 3.4 number drops with it, from about +2.2 °C to +1.4 °C in mid-July. Same ocean, two honest numbers, measuring two different things. This is why the article leans on RONI to compare 2026 fairly against older events, back when the background was cooler. (The model dots stay in raw terms, since no per-model RONI is published, so that half of the comparison is not yet like-for-like.)
Watching it play out
A small job fetches the latest NOAA field every morning, recomputes the anomalies and box averages, and keeps a rolling four-month window. As of mid-July the observed Niño 3.4 was climbing steadily but still sat well short of the model peaks out in November and December.
So the map is really a scoreboard. Come back through the autumn and watch whether the real ocean reaches up into that crowd of forecasts, or stalls below it. The data will settle it, one grid cell at a time.
Sources and data
- Video: Just Have a Think, “El Niño is here and it has SMASHED all records”.
- Article: Zeke Hausfather, The Strongest El Niño Ever, The Climate Brink, July 2026. The 14-model forecast figure is drawn from this piece.
- Sea surface temperature: NOAA OISST v2.1, accessed daily through NOAA CoastWatch ERDDAP.
- Forecast models: the North American Multi-Model Ensemble and Copernicus C3S seasonal forecasts.
- Niño regions, ONI, and RONI definitions: NOAA Climate Prediction Center.
- Map rendering: Deck.gl.