The dynamics that govern a hurricane's path and intensity are incredibly complex, and one of the least understood is eyewall replacement. In this process the wind speed drops initially when the first cloud walls collapse. But the new walls that move in to replace them re-intensify the wind as they shrink inward – a similar result of angular momentum conservation that makes ice skaters spin faster as they fold their arms.

Precisely when eyewall replacement will occur, and to what extent it will re-intensify a storm, is of utmost importance to meteorologists. For example, in 1999 Hurricane Floyd's eyewall replacement reduced a potential catastrophe to a rainstorm as it reached land. But in 1992 Hurricane Andrew – the US's second most destructive tropical storm – an eyewall replacement led to fierce wind intensification at the worst conceivable moment.

Now, Shuyi Chen from the University of Miami and colleagues from other US research institutions say they have a method of forecasting eyewall replacement. The physicists took data from their own Hurricane Rainband and Intensity Change Experiment (RAINEX), a collaboration that took radar data from Hurricanes Rita, Katrina and Ophelia, which caused massive devastation to the Gulf Coast of the US in 2005. They then used a computer model to interpret the data, which had a resolution of 1.6 km – more than three times as detailed as previous models.

Chen's team discovered that, prior to replacement, a "moat" of dry air forms that separates the existing eyewall from a new wall farther out. This moat gradually merges with the inner wall until it finally collapses, allowing the outer wall to take its place.

Chen told Physics Web that although their research is available now, it may take some time for it to be implemented in hurricane forecasting. "As with all new techniques, it needs to be improved and more research is needed for a transition to operations."