In 2021, two tools designed to forecast weather a year ahead showed different results. One, built on connections between temperatures at different points in the Pacific Ocean, was silent: no El Niño in 2022. The other, which measures how chaotically temperature changes in the central ocean, was instead raising the alarm: a strong event is expected, one that would warm the sea surface nearly two degrees above normal.
Both approaches rely on the physics of the climate system, not on statistical guessing. The first assembles something like a telephone network from the ocean: fourteen points in the equatorial zone and almost two hundred outside it. When temperatures at these points start to change in sync, the network "comes alive": connections between nodes strengthen, and a year before El Niño this signal becomes noticeable. In 2021, no such strengthening occurred.
The second method looks at the same ocean but differently. It takes twenty-two series of temperature anomalies in the region where El Niño typically originates and calculates how complex their behavior is. Imagine recording the temperature at one point every month. If it changes predictably, the record is easy to compress: lots of repetition. But if the data show unexpected diversity, compression becomes harder. High complexity in 2021, according to calculations, pointed to a future strong El Niño.
History knows six cases where these two tools disagreed just as they do now. In four of them, the network method turned out to be right, meaning El Niño did not occur. Therefore, the authors lean toward the view that in 2022 an event most likely will not happen. But if it does start, it will be large: the entropy method rarely errs in estimating strength.
And so 2022 remained suspended between two forecasts. One tool advised preparing for droughts and floods, the other advised expecting nothing unusual. The situation itself resembles two barometers on the same wall: one shows clear, the other shows a thunderstorm. Which one would prove more accurate became clear only at the end of the year, when the ocean gave its final answer.
In 2012, when most forecasting centers expected El Niño with 75% probability, the climate network method had already predicted its absence in late 2011—and turned out to be right. This was one of the first public successes of the network approach.