№ 170 · earth science

El Niño is one ocean sloshing, egged on by the wind

The trade winds normally push the sun-warmed surface water of the tropical Pacific west, toward Indonesia, where it piles up. When those winds slacken, some of that warm water spreads back east and the cold water that usually wells up off South America stops reaching the surface. The eastern Pacific warms. The warming weakens the winds further, and the ocean and the air keep pushing each other until the warm water drains away. That is El Niño. Warm and cold swings of this kind come every two to seven years.

A tilted bathtub

Peruvian fishermen named the warm current that showed up off Peru and Ecuador around December and January "El Niño", perhaps centuries ago. Today the name means something much bigger than a coastal current, and Michelle L'Heureux of NOAA's Climate Prediction Center explains why. It starts with the average state of the Pacific.

Along the equator the trade winds blow from east to west. They drag the surface water with them, and it warms in the sun on the way. On the far side it has nowhere to go, so it piles up around Indonesia. In the east, with the surface water pushed away, colder water from below rises to replace it. So the ocean has a warm layer that is thick in the west and thin in the east, sitting on cold water. The boundary between the two, the thermocline, slopes up from west to east like the water in a tilted bathtub.

The air answers the ocean. Over the warm west, air rises, builds thunderstorms and rains. Over the cool east it sinks and dries out. The surface winds blow from where air sinks toward where it rises, which is east to west: the trade winds again. The tilt keeps the winds blowing, and the winds keep the tilt.

Predict first

In a strong El Niño the surface of the eastern Pacific can run several degrees warmer than usual. Where does most of that warmth come from?

Check your answer

The third. L'Heureux describes the warming as weaker trade winds cutting the upwelling of cold water in the east and moving less water to the west, so the east warms. The heat was stored in the warm layer before the event began.

The wind and the water egg each other on

The loop that grows an El Niño is called the Bjerknes feedback, after Jacob Bjerknes, who saw in the 1960s that the swings in surface pressure and the swings in Pacific temperature were one coupled system. Suppose the eastern Pacific warms a little. The temperature difference across the ocean shrinks, so the rising air and rain spread east and the trade winds weaken. Weaker winds mean less cold water welling up in the east and less warm water being pushed west, so the east warms more. Sooman Han, Alexey Fedorov and Jérôme Vialard, in a 2025 paper, add the ocean's side: the eastern thermocline sinks, and warm water from the west spreads east. The same loop run backwards, with stronger winds and a cooler east, grows a La Niña.

What starts the loop is often unclear. L'Heureux notes that forecasters frequently cannot tell whether the ocean or the wind moved first. A strong westerly wind burst, a spell of wind blowing from the west against the trades, is one common trigger. Timing matters too: the Pacific is most ready to couple in the northern summer and autumn, and events usually peak in the northern winter.

Interactive A recharge oscillator, the two-number model of El Niño, using the parameters Han, Fedorov and Vialard fitted in 2025. One number is how warm the eastern equatorial Pacific is; the other is how much warm water is stored in the west. Top: a cartoon slice along the equator at the month you pick. Bottom: forty years of the model. Set how strongly the wind and the ocean feed each other and how hard random wind bursts kick, then scrub through time or press Play. Every control works with Tab, Enter and the arrow keys.

wind–ocean feedback R0.03 /month
random wind bursts1.0×
month shownyear 0.0
west: Indonesia east: South America yellow: warm water above, cold below +3 K 0 −3 K year 0 year 40 solid: east temperature · dashed: west warm-water store (±27 m at the ±3 K lines)
The model's average state: strong trade winds, warm water piled up in the west, cold water near the surface in the east.

Why it ends: the slower slosh

If the feedback only amplified, an El Niño would never end. L'Heureux describes how it sows its own end, through what is called the recharge–discharge cycle. During neutral years or a La Niña, warm water builds up below the surface along the equator, first in the western Pacific, over months or even years. That is the recharge. Once the store is big enough, a trigger can tip it into El Niño. During the event, the higher sea level in the east and the westerly wind anomalies drive currents that carry warm upper water away from the equator, north on one side and south on the other. That is the discharge. With the warm water drained, colder water sits closer to the surface on the equator, and sometimes, not always, the Pacific swings into La Niña. Some of the heat also escapes into the atmosphere and raises global air temperatures. L'Heureux adds that a second explanation, the delayed oscillator, tells the story through ocean waves, called Kelvin and Rossby waves, that move heat east and west along the equator. The two are not rivals, and both can act at once.

Han and colleagues boil this down to two numbers: how warm the eastern Pacific is, and how much heat is stored in the western Pacific. Each pushes the other. A full store warms the east; a warm east drains the store. That makes the pair swing like a pendulum, with the store running ahead of the temperature, and that is what the figure runs. With their fitted values a swing left alone fades out on a timescale of about 20 months, shorter than its natural rhythm of about four years. So the Pacific behaves as a strongly damped oscillator. Random wind bursts keep restarting it, which is why the gaps between events are irregular. They also tested versions that keep swinging on their own, and found their swings too regular to match the statistics of the real record.

The real cycle is also lopsided. In Han and colleagues' look at the record since 1970, moderate-to-strong El Niño events peaked at about 1.9 K above normal in their eastern Pacific box and took about 7 months to fade from the peak. La Niña events peaked at about 1.3 K below normal and took about 14 months. One year after such an El Niño, the Pacific was in La Niña half the time, and in none of the cases was it in another El Niño.

The stored warm water is a forecasting tool, but its lead has changed. In a 2026 paper, Zhongtian Wu, Li Yan, Jianjun Xu and Shaojun Zheng found that before 2000 the equatorial warm-water volume predicted ENSO best about 8 months ahead. Since 2000 its best lead for La Niña has shrunk to about 3 months, while for El Niño it still works 8 months or more ahead. They link the change to the multi-year La Niñas that have become more common since 2000.

How NOAA calls one now

Since 1 February 2026, NOAA's Climate Prediction Center has used a new yardstick, the Relative Oceanic Niño Index. It starts from the sea surface temperature anomaly in the Niño 3.4 box (5° N to 5° S, 120° W to 170° W) and subtracts the average anomaly of the whole tropical belt (20° N to 20° S). The result is then rescaled to the size of the old index. The reason is warming: as the whole tropical ocean warms, a fixed 30-year average falls behind, and part of a box's warmth is then the general trend rather than El Niño. El Niño conditions mean this relative index, averaged over three months, is +0.5 °C or more, and La Niña means −0.5 °C or less. To count a past event, the index has to stay past that line for five overlapping three-month seasons in a row.

The one under way in 2026

As of NOAA's discussion of 8 October 2026, a strong El Niño was still growing. The Niño 3.4 anomaly reached +2.1 °C in September, and the eastern thermocline was deeper than normal. Westerly wind anomalies stretched from the western to the east-central Pacific, and rain was heavier than normal from the central to the eastern Pacific and lighter over Indonesia: the Bjerknes feedback in action. In a post the day before, Rebecca Lindsey gave the July–September value of the relative index as 1.7. NOAA put the chance of a strong-to-very-strong event lasting through January–March 2027 above 83%, and the chance of beating every El Niño since 1950 (a three-month value of +2.5 °C or more) at 83% for October–December. These are forecasts from one month. NOAA updates them every month, so check the latest discussion before relying on them.

In short

The trade winds keep warm water piled in the western Pacific and cold water near the surface in the east. When they weaken, warm water spreads east, the cold upwelling fades, and the warming weakens the winds further: that loop grows an El Niño. Meanwhile the event drains warm water away from the equator, which ends it and can set up a La Niña. Then the trade winds slowly rebuild the store. The pendulum is heavily damped and kept going by random wind bursts, so events come every two to seven years, not on a schedule.

Where this comes from

  1. Realistic ENSO Dynamics Requires a Damped Nonlinear Recharge Oscillator reuse permitted with attribution
    Sooman Han, Alexey Fedorov, and Jérôme Vialard · arXiv:2506.11206 · 2025
    arxiv.org/abs/2506.11206
  2. The Rise of El Niño and La Niña (NOAA ENSO Blog) linked only, not reproduced
    Michelle L'Heureux · 2020
    www.climate.us/news-features/blogs/enso/rise-el-nino-and-la-nina
  3. The life and death of El Niño (NOAA ENSO Blog) linked only, not reproduced
    Michelle L'Heureux · 2016
    www.climate.us/news-features/blogs/enso/life-and-death-el-ni%C3%B1o
  4. Asymmetric decadal changes in the relationship between warm water volume and ENSO around 2000 (Frontiers in Marine Science) linked only, not reproduced
    Zhongtian Wu, Li Yan, Jianjun Xu and Shaojun Zheng · 2026
    www.frontiersin.org/journals/marine-science/articles/10.3389/fmars.2026.1813774/full
  5. Public Information Statement 26-05: Implementing a Relative Oceanic Niño Index effective February 1, 2026 linked only, not reproduced
    NOAA National Weather Service and Climate Prediction Center · 2026
    www.weather.gov/media/notification/pdf_2026/pns26-05_Relative_ONI.pdf
  6. El Niño/Southern Oscillation (ENSO) Diagnostic Discussion, 8 October 2026 linked only, not reproduced
    NOAA Climate Prediction Center · 2026
    www.cpc.ncep.noaa.gov/products/analysis_monitoring/enso_disc_oct2026/ensodisc.shtml
  7. Highlights from NOAA's October 2026 El Niño forecast (ENSO Blog) linked only, not reproduced
    Rebecca Lindsey · 2026
    www.climate.us/news-features/blogs/enso/highlights-noaas-october-2026-el-nino-forecast