Researchers at the National Institute of Standards and Technology developed a new method of simulating hurricanes.

Using artificial intelligence and 100 years of hurricane data, the simulations can accurately predict the trajectory and wind speeds of actual storms, according to a March 29 news release. The researchers suggest simulating numerous hurricanes with this approach can help to develop improved guidelines for building design in hurricane-prone regions.

"Imagine you had a second Earth, or a thousand Earths, where you could observe hurricanes for 100 years and see where they hit on the coast, how intense they are,"  NIST mathematical statistician Adam Pintar, a co-author of the study, said in the release. "Those simulated storms, if they behave like real hurricanes, can be used to create the data in the maps almost directly."

Building codes, which regulate building design and construction, point designers to standardized maps that show the wind levels their structure must handle based on its location and importance, the release reported. These maps are based on computer models of hypothetical hurricanes, which are, in turn, based on real-life hurricane records. 

With the new AI approach to hurricane simulation, engineers can determine the right level of winds to design for in a more accurate and effective way, according to the release.

Being able to withstand hurricane-force winds is crucial for the longevity of buildings on the eastern seaboard and gulf coast of the U.S., the release reported. The new AI approach to hurricane simulation may offer a simple solution for determining the right level of winds to design for. 

By simulating numerous realistic hurricanes, researchers can develop improved guidelines for building design in hurricane-prone regions, which could lead to buildings that are better equipped to withstand extreme weather conditions, according to the release.

Researchers split data on more than 1,500 storms into sets for training and testing of their model, the news release said. 

“It performs very well," Pintar said in the release. "Depending on where you're looking at along the coast, it would be quite difficult to identify a simulated hurricane from a real one, honestly.”