The Pacific Ocean breathes. Every few years, its warm western "lung" exhales eastward, and from Peru to Indonesia, rains, droughts, and hurricanes shift. This breath is called El Niño–Southern Oscillation, or ENSO. The question that keeps climate scientists up at night: Does this oscillation damp out on its own, like a pendulum gradually slowing down, needing constant nudges from random wind gusts? Or does it have an internal motor that keeps it going?

Testing this is not as straightforward as it seems. Scientists build simplified models in which the ocean and atmosphere exchange heat and momentum. One of the most popular is the recharge oscillator model. It reduces the whole system to two numbers: sea surface temperature and the thickness of the warm layer. From these two numbers, you can compute whether the anomaly grows or decays. And here's the strange part: almost always, the model says ENSO is damped. The growth rate comes out negative.

The authors of the new study decided to check whether the method itself might be fooling us. They created a computer simulation in which they knew exactly whether ENSO was damped or not. Then they applied the standard parameter estimation method to this simulation, the same one used on real observations. The result was sobering: even when the simulated ENSO was self-sustaining, the method almost always said it was damped. The error was systematic, not random.

Where does this blindness come from? The recharge oscillator is a very coarse description of reality. It ignores many details that influence ocean and atmosphere behavior. When we fit such a simplified model to real data, some of those ignored details get "absorbed" into the model parameters, distorting them. In particular, the parameter governing the Bjerknes feedback is estimated with a large bias. It's like trying to determine how strong the spring in a pendulum is by watching its motion through frosted glass: you'll always see damping, even if the spring is actually giving it a push.

What does this mean for our understanding of climate? If the real ENSO growth rate is near zero or even positive, then the system is in a more fragile equilibrium than previously thought. A tiny nudge, such as a change in greenhouse gas concentrations or melting ice, could shift it into a different regime. Perhaps ENSO is not just passively reacting to random disturbances but actively participates in shaping climate anomalies. This also changes forecasts: if the system is self-sustaining, its behavior could be more unpredictable over long timescales.

So, the pendulum we thought was damping might simply be hard to see. We were looking at it through the frosted glass of a simplified model. The new study wiped that glass clean, and it turns out the damping might be an illusion. The ocean breathes, and its breath may be driven by its own rhythm, not just by gusts of external wind.