Buoys have it tough.
Built to endure some of the harshest conditions on the planet, the instruments are thrashed by ocean waves, buffeted by high winds, corroded by sea salt, and scorched by the sun’s ultraviolet rays. Their measurements on everything from solar radiation to seawater salinity, barometric pressure, and the still-alarmingly-warm water temperatures in the Gulf of Mexico (by any other name) provide crucial information for the experts making forecasts for weather patterns like the El Niño and the Southern Oscillation variations, which have impacts felt around the world. The buoys also provide life-or-death data used to make informed forecasts for the 60 million Americans living in the Atlantic and Gulf regions — i.e. those most vulnerable to hurricanes.
The job of maintaining the government’s more than 200 moored buoys across the Pacific and Atlantic basins falls to the National Oceanic and Atmospheric Administration’s National Data Buoy Center, based out of southern Mississippi. Like many teams at NOAA, the NDBC consists of a small group of oceanographers, computer scientists, engineers, and meteorologists that play an outsized role in shaping our understanding of what’s happening in the ocean. Also like many teams at NOAA, it has been hit hard by the Trump administration’s sweeping layoffs and buyouts. Of its 34 full-time employees, the NDBC had already lost three as of March 1, while the fate of another 120 contract employees — who help keep the buoys maintained and operational — is in limbo. “Hopefully it won’t get to the point where [the system] kind of falls apart,” one engineer who retired this year worried to The Columbian.
Against this bleak backdrop, independent forecasters have begun to release their predictions for the 2025 hurricane season. Groups like Colorado State University’s Department of Atmospheric Sciences and the media company AccuWeather, which publish highly regarded outlooks every April, rely almost entirely on data from NOAA’s buoys, satellites, and weather stations.
“NOAA is critical,” Levi Silvers, a research scientist and a co-author of CSU’s 2025 outlook, told me. “If you look back 20 or 30 years ago, we didn’t have nearly as many buoys out there. That meant forecasters “couldn’t really tell how deep the warm or cool layers of the Pacific went,” which led to more unpleasant surprises, he said. “We can see that now because of the buoys from NOAA.”
This year, government-provided data informed CSU’s forecast of 17 named storms in 2025, as well as AccuWeather’s prediction of 13 to 18 named storms. Both groups’ forecasts are slightly lower than their 2024 predictions, although Silvers stressed that the dip shouldn’t be the emphasis. “It’s still above normal,” he said, noting that the average number of named storms between 1991 and 2020 was fewer than 15. “I hope that people don’t get the impression that it’s a below-average season because it’s less than last year.”
Paul Pastelok, AccuWeather’s head of long-range forecasting, likewise told me that while water temperatures aren’t as warm in the Atlantic basin’s main development area as last year, they’re still pretty close. Hurricanes primarily draw their power from heat at the sea’s surface, so early season temperature readings can tip off forecasters to increased storm activity. There is no reason to write off the possibility of another storm as powerful as Hurricane Milton making U.S. landfall this year. (AccuWeather predicts three to five Category 3-strength or higher storms this year, while CSU expects four.) The potential for rapid intensification — where a storm significantly increases in wind speed over a period of 24 hours or less, as we saw with last year’s Hurricane Beryl — also remains.
Pastelok was particularly alarmed by the numbers NOAA has reported in the Gulf, which could mean a lot of “homegrown activity.” “In the past, we’ve seen these long-track systems coming off the African coast that can produce some big storms — Category 4 or 5 — but we have time to react and see where they’re going to go,” Pastelok said. “It’s the ones that develop closer to the county that could catch people off guard.”
Pastelok added that AccuWeather hasn’t had any issues receiving NOAA data, and he isn’t worried yet about continuing to obtain quality data to tweak their predictions, including potentially accounting for a late-season La Niña, a pattern typically conducive to more hurricanes. Silvers sounded less sure: “I don’t think people realize how much work it takes to get information from a satellite or a buoy to make a picture in your computer,” he said. “It has to be collected, and there’s a huge process of quality control, where we have to make sure the data is good.” NOAA has — or at least had — many employees doing the “grueling, tedious, computer-science-type work” to provide good data to hurricane forecasters.
NOAA’s National Hurricane Center also provides its own forecast of named storms, which is usually released just before the June 1 start of the season. In response to my emailed questions about how the administration’s layoffs may affect NHC’s forecasting capabilities, a communications officer reiterated the agency’s policy of not discussing internal personnel matters or engaging in speculative interviews. She added, however, that NOAA “remains dedicated to its mission, providing timely information, research, and resources that serve the American public and ensure our nation’s environmental and economic resilience.”
Other branches of NOAA responsible for observations and communications related to hurricanes also appear to be in trouble. Mission-critical flight directors for the Hurricane Hunters, who measure the intensity of developing storms by flying through the eyewall, have been among those laid off by the agency, reducing NOAA’s aviation capacity by 25%. The National Weather Service has also indefinitely suspended its extreme weather alerts in languages other than English — including Spanish, which is spoken at home by 20% of Floridians and nearly 30% of Texans. And while NOAA noted to me that its Weather Prediction Center, National Water Center, and National Weather Service offices around the country issued a “rare coordinated NOAA news release” ahead of last year’s devastating Hurricane Helene, that kind of inter-department cooperation and messaging gets harder as the contact information of former point-people goes dark and one-time colleagues are no longer around to answer a call.
For now, at least, the 2025 hurricane predictions remain high quality and trustworthy; Pastelok sounded confident of the range AccuWeather had landed on, and Silvers also sounded assured in the numbers CSU put out. But buoys break — the NDBC’s annual “maintenance mission” alone lasts eight months, not to mention its constant backlog of as-needed repairs — and other ocean monitoring programs are also at risk of losing their funding.
At the end of the day, a forecast is only as good as the data fed into it. Hurricanes are highly complicated systems, and every degree of water temperature, shift in wind shear, and variation in tropical waves can change the character of a storm. If NOAA’s data and quality control degrades in the coming months or even years, it’s not an exaggeration to say that the fallout could be catastrophic.
And as hurricane forecasters like to say: All it takes is one storm.
