The tropical Pacific Ocean lives in two rhythms. The first is fast: a wave of thunderstorms born over the Indian Ocean races eastward and dies out off the coast of the Americas within a month or two. The second is slow: every few years, the ocean surface warms enough to alter weather across the entire planet. That is El Niño. Between these rhythms lies a bridge, and building it took climatologists decades.

Earlier models either captured the fast rhythm but missed the slow one, or vice versa. The new model combines both, adding a third, slowest rhythm: decadal changes in the strength of the trade winds. It turns out that this third rhythm determines whether El Niño will be weak or extreme.

The model reproduced what is observed in the ocean: warm patches appear sometimes off the coast of South America, sometimes in the central ocean, and occasionally persist for two consecutive years. Event frequencies matched the real ones over 71 years of observations. Even the rare super El Niños, which occur every 15–20 years, were correctly reproduced.

The model is lightweight: it runs on a regular laptop what would take supercomputers weeks. This opens the door to long simulations, thousands of years instead of hundreds. And long records are needed to study rare catastrophes and train neural networks to predict climate.

The bridge is built. Now we can cross it in both directions: from fast wind bursts to slow currents and back. And each time, we find something new about how the ocean breathes.