Over the warm waters of the Atlantic, air begins to swirl like water going down a drain. This is a future hurricane, and its fate depends on what happens thousands of kilometers away in the Pacific Ocean. When the El Niño current heats up there, Atlantic storms become fewer. Models that predict the hurricane season know about this link. But they convey it like a translator who translates every word too literally and loses the meaning.

Six European models were compared with real observations over 22 seasons. The overall conclusion is correct: in El Niño years, there are fewer hurricanes. But in the eastern Caribbean, the models see a strong link, while real data show none. The reason lies in wind shear, that is, the difference in wind speed at different altitudes. El Niño strengthens the shear, and the models blow this link out of proportion.

[caption]The error is that the models incorrectly shift the boundary of the large-scale air circulation between the tropics and subtropics[/caption]

This error comes at a high cost. Even if a model correctly predicts the total number of hurricanes in a season, the forecast for the Caribbean basin may be wrong. Model developers will need to improve the physics of atmospheric circulation, and operational centers should use such forecasts cautiously for individual regions.