A new study finds that the most intense and destructive rainstorms in Portugal, particularly those fueled by atmospheric rivers, are not the most chaotic, but among the most predictable. These events form within large, well-organized atmospheric systems that strengthen winds and channel moisture efficiently, producing significantly heavier rainfall while also creating clearer, more coherent signals in the atmosphere. As a result, the very storms that pose the greatest risk to infrastructure and public safety may also offer the best opportunity for earlier and more reliable forecasts.

In December 2022, relentless rains battered western Portugal, swelling rivers, flooding streets and leaving a trail of damage that lingered long after the skies cleared. To many, the storm felt chaotic — another reminder of a changing climate’s volatility. But new research suggests that some of the most intense rainfall events may be far less unpredictable than they appear.

A study led by Mr. Ehud Bartfeld and Dr. Assaf Hochman of the Hebrew University of Jerusalem, together with Dr. Alexandre M. Ramos of the Karlsruhe Institute of Technology, finds that the fiercest downpours in Portugal often follow atmospheric patterns that are not only identifiable, but intrinsically more predictable.

Published in Weather and Climate Extremes, the research focuses on Heavy Precipitation Events (HPE) in the western Iberian Peninsula, storms that pose growing risks to infrastructure, water systems and public safety. At the heart of many of these events are Atmospheric Rivers, long, narrow corridors of moisture that transport vast amounts of water vapor across oceans and into coastal regions.

The team discovered that when these atmospheric rivers are involved, rainfall becomes significantly more intense, on average 36% stronger than storms without them. Surprisingly, the added intensity does not stem from more moisture in the air overall, but from stronger low-level winds that funnel moisture more efficiently into the region.

“It’s not just how much water the atmosphere holds,” said the researchers. “It’s how effectively the system delivers that water to the ground.”

But the study goes further, asking a question that has long challenged meteorologists: When are extreme events actually predictable?

Using a novel dynamical systems approach, the researchers analyzed how atmospheric conditions evolve before and during storms, examining patterns in both lower and upper layers of the atmosphere. What emerged was a striking divide.

The most predictable extreme rainfall events were tied to well-organized, deep extra-tropical cyclones forming over the North Atlantic, near 50°N and 15°W. These systems exhibited pressure anomalies roughly twice as strong as their less predictable counterparts, along with clearer jet stream interactions and more coherent large-scale wave patterns in the atmosphere.

In practical terms, the difference is dramatic: these highly predictable events produced rainfall intensities about 80% greater than less predictable storms.

“The irony is that the most dangerous events are often the ones the atmosphere signals most clearly,” the researchers said. “When the large-scale structure is strong and organized, the system becomes more ‘readable’.”

The devastating December 2022 storm in Portugal provided a vivid case study. The researchers showed how the alignment of an atmospheric river with a powerful cyclone and structured jet stream created both extreme rainfall and a relatively high level of forecast confidence, a combination that, if better understood, could improve early warnings.

The findings suggest that integrating atmospheric river detection with dynamical systems analysis could sharpen forecasting tools, not only in Portugal but in vulnerable coastal regions around the world.

As climate change continues to amplify rainfall extremes, the distinction between chaos and predictability may prove critical. The atmosphere, it turns out, does not always hide its intentions. Sometimes, it signals them, strongly, coherently, and in ways scientists are only now beginning to decode.