Conjuring hurricanes with computer models

Written by
Catherine Zandonella
Catherine Zandonella, Office of the Dean for Research
March 24, 2021

Using computer bits and bytes, Ning Lin, an associate professor of civil and environmental engineering, has become an expert in predicting the probability of hurricanes.

“The formation of hurricanes is actually quite a complex physical process that is not well understood,” Lin said.

But by looking at the myriad ingredients of hurricanes, such as warm surface temperatures, high humidity and high vorticity, and putting these into a computer, Lin has been developing computer models that generate data on the likelihood and intensity of hurricanes.

Lin used these data back in 2012 to predict the probability that New York City would get hit by an intense storm. Her models gained validity in October of that year when Hurricane Sandy made landfall in the city.

Recently, Lin and her colleagues have been using a type of artificial intelligence known as machine learning to streamline hurricane predictability. They have also used these models to estimate the risk of storm surge along the East and Gulf coasts.

With the prevalence and intensity of hurricanes likely to change because of a warming planet, models like Lin’s will help researchers make better projections of future storms.

Environment Tags