Article first published in The New York Times.
As darkness descended on the Gulf of Mexico in October, a 1970s-era U.S. government turboprop plane neared the eye of the newly formed Hurricane Milton. When the plane’s first radar scan arrived by satellite communications, I pounced and took to the airwaves, describing to viewers what I saw inside the storm: a dreaded vortex alignment signaling the early stages of rapid intensification. On social media I put it more plainly: “Katy bar the door, this one’s about to put on a show.”
And Milton did just that, strengthening at a breathtaking rate over the next 24 hours to a 180-mile-per-hour Category 5 monster, the strongest Gulf hurricane in almost 20 years. But there was no October surprise on the Florida coast because we’d had ample warning from the National Oceanic and Atmospheric Administration’s hurricane hunters — enough time for people in the highest-risk areas to safely evacuate and businesses to prepare for the worst.
But as we head into what NOAA forecasts will be another active Atlantic hurricane season, the Trump administration and the so-called Department of Government Efficiency are downsizing the agency, which houses the National Weather Service, the hurricane hunters and many other programs crucial to hurricane forecasters. Without the arsenal of tools from NOAA and its 6.3 billion observations sourced each day, the routinely detected hurricanes of today could become the deadly surprise hurricanes of tomorrow.
The National Weather Service costs the average American $4 per year in today’s inflated dollars — about the same as a gallon of milk — and offers an 8,000 percent annual return on investment, according to 2024 estimates. It’s a farce for the administration to pretend that gutting an agency that protects our coastlines from a rising tide of disasters is in the best interests of our economy or national security. If the private sector could have done it better and cheaper, it would have, and it hasn’t.
Losing the hurricane hunters would be catastrophic, but that would be only the forerunner wave in a brutal, DOGE-directed tsunami to weather forecasting. In just three months DOGE has dealt the National Weather Service, which operates 122 local forecast offices around the country, the equivalent of over a decade of loss to its work force. Some offices have hemorrhaged 60 percent of their staff members, including entire management teams.
National Weather Service forecast offices — typically staffed 24 hours a day, seven days a week, 365 days a year — are the source of all weather warnings received by Americans by phone, TV and radio. Without these warnings and data, local weather broadcasts and private weather apps couldn’t operate.
With dozens of local forecast offices struggling to maintain 24/7 operations, NOAA put out a mayday on May 13 asking remaining staff members to temporarily vacate their posts to salvage what was left of the nation’s critical warning network. Nearly half of local forecast offices are critically understaffed, with a vacancy rate of 20 percent or higher, and several are going dark for part of the day, increasing the risk of weather going undetected and people going unprotected and unwarned.
The staff reshuffling is just the latest move from an agency fighting for survival. Weather balloons, a mainstay of data collection for more than 60 years, usually launch twice a day from 100 sites around North America, the Caribbean and the Pacific. But recently some of their flights were reduced or suspended so that limited staffs can tend to other priorities.
Even in the satellite era, weather balloons have been shown to markedly improve forecast accuracy, so much so that twice-daily launches are commonly supplemented with up to four launches a day ahead of major hurricane threats. The extra balloons increase forecast confidence and allow time-sensitive decisions like evacuation orders to be made sooner. On the precipice of a new hurricane season, balloon launches are down 15 to 20 percent nationwide, throwing the nation into a risky experiment that nobody wanted to run.
Within NOAA, research and forecasting are inextricably linked. In budget documents released last month, the White House proposed eliminating NOAA’s research wing, the Office of Oceanic and Atmospheric Research, which lends mission-critical support to the hurricane hunters. Taking a sledgehammer to OAR would shatter decades of progress in hurricane forecasting, one of the roaring success stories of predictive sciences. The fate of the agency’s research arm is now in the hands of Congress.
Thirty years ago, forecasters couldn’t detect a hurricane until it formed, and once it formed, we were lucky to give two or three days’ notice that it might strike land. Today, our forecast models — developed, maintained and improved by NOAA scientists and their supercomputers — routinely and reliably predict hurricanes sometimes a week or more before the first puff of clouds. At two or three days out, we’re able to whittle hurricane forecasts to within a county or two.
Rapidly intensifying hurricanes like Milton remain difficult to forecast, eluding conventional satellites and outpacing weather prediction models. In recent years, however, NOAA has developed powerful high-resolution hurricane models that both see the small details and skillfully forecast episodes of rapid intensification. Slashing funding for NOAA research and development would effectively unplug these world-class models and eviscerate the institute that supplies the only rapid-intensification prediction tools available to forecasters. Without them, forecasters like me are flying a plane in the clouds with no navigation system. It’s a recipe for disaster.
I’ve spent over two decades working to reduce the loss of life from hurricanes, from revolutionizing the way we forecast and warn for storm surge at the National Hurricane Center to overhauling FEMA’s hurricane response and recovery plans to guiding viewers on air at The Weather Channel and at stations in hurricane-prone South Florida through dozens of storms. Hurricanes aren’t an afterthought to the more than 60 million Americans living in the hurricane zone. They’re our highest priority during the Atlantic hurricane season, which runs from June 1 through November 30. I’ve not seen a bigger threat to weather and climate science than what we’re witnessing now.
The irreparable harm the Trump administration is doing will imperil the nation’s longstanding weather warning network for hundreds of millions of Americans in the decades ahead. It’s only a matter of time before the next Milton is at our doorstep — but with our weather intelligence severely compromised, will we know it?