Every few years, the water off the coast of Peru becomes unusually warm. Fish disappear, rains flood the desert, and Australia suffers drought. This phenomenon is called El Niño. At the same time, over the warm ocean, huge cloud waves are born every month or so, slowly crawling eastward. For a long time, scientists could not understand how these two rhythms are connected.

The researchers built a simple model: they divided the ocean into three basins and the atmosphere into three traveling waves. Between them, they laid two wires. But the model was still too 'well-behaved.' So they added randomness that depends on the situation: when the water is already warm, random gusts more often heat it up even more. This is like a snowball rolling down a slope: the bigger the ball, the faster it grows.

Now the model can simulate thousands of years of climate in minutes and capture rare extreme events. It can be used as a training tool for forecasters and as a data generator for artificial intelligence that learns to predict weather a season ahead.