NOAA researchers are launching a massive high-tech effort to track tropical cyclones this hurricane season. The agency is combining advanced artificial intelligence with a fleet of uncrewed ocean and air drones to gather data from the most dangerous parts of these powerful storms.
Scientists will deploy a mix of robots that can go where human crews cannot safely travel. This includes 10 Saildrone Explorer vehicles alongside 16 other remote-controlled surface craft, like the Oshen C-stars and Seasats Lightfish. Powered by wind and solar energy, these machines measure wind speed, wave height, air temperature, and ocean salinity. Below the surface, ocean gliders, drifters, and floats will track underwater temperatures to see if the ocean will strengthen or weaken a storm.
Up in the air, NOAA is using the Black Swift uncrewed aircraft. This is a tiny 2.75-pound drone with a 54.6-inch wingspan. For the first time, data collected by this small drone inside the storm’s eyewall will feed directly into NOAA’s main forecast model, the Hurricane Analysis and Forecast System, in near-real time.
“This is an important milestone in advancing NOAA research to operations,” said Joe Cione, lead meteorologist for emerging technologies at NOAA’s Atlantic Oceanographic and Meteorological Laboratory. “We’ve been testing the Black Swift uncrewed aircraft for several years and we’ve now shown by analysis that the data it collects can improve hurricane intensity forecasts by 10 percent.”
Forecasters will also use artificial intelligence to make their predictions smarter and faster. At NOAA’s Geophysical Fluid Dynamics Lab, scientists are blending traditional computer models with AI tools to look at ocean conditions months in advance.
READ: AI Storm Chasers: How The National Hurricane Center Is Training Computers To Track Killer Storms
“Instead of replacing traditional models, AI is helping them to become smarter, faster and more effective,” said Hiro Murakami, a scientist at the lab. “Early results show this approach can improve forecasts of how active a hurricane season will be.”
AI is also being used to clean up radar data collected by human crews flying aboard NOAA Hurricane Hunter aircraft.
“Our current quality control method removes most non-meteorological information but also discards a significant amount of meteorological data,” said Paul Reasor, a meteorologist at the laboratory. “With the support of machine learning, we recover much of the previously discarded meteorological data resulting in a 25% increase in usable data while maintaining the high quality of operational products.”
Human crews will still fly directly into the storms using two WP-3D Orion planes and a Gulfstream IV-SP jet. This season, they will test the benefits of flying the jet over the storms more frequently to collect extra data from the inner core, which has been shown to improve track and intensity forecasts.
Additionally, regional research teams are studying the long-term impacts of these storms. This includes using AI to predict storm-triggered landslides in the Caribbean, examining how stalling storms increase flood risks, and tracking how fallen trees from past storms, like Hurricane Helene, increase wildfire risks in the Carolinas.
Please make a small donation to the Tampa Free Press to help sustain independent journalism. Your contribution enables us to continue delivering high-quality, local, and national news coverage.
Sign up: Subscribe to our free newsletter for a curated selection of top stories delivered straight to your inbox
