Hurricane Katrina struck the Gulf Coast on Aug. 29, 2005, with wind gusts exceeding 130 miles per hour and a surge of water taller than a two-story building. The water left the coast in ruins and swelled through the broken levees of New Orleans as a wave that left 80 percent of the city flooded for days. Nearly 1,400 people died.
Two days earlier, Max Mayfield, then the director of the National Hurricane Center, had made a call to the city’s mayor. Mr. Mayfield said he recalled that after they had hung up he turned to a person next to him and expressed concern: “He doesn’t get it.”
Days before the storm hit the Gulf Coast, forecasters at the Hurricane Center knew where Katrina would go and how strong it was likely to be. They knew it would follow nearly the exact same path that Hurricane Camille — still one of the most intense hurricanes on record — had in 1969.
But Mr. Mayfield said he and others had trouble persuading people in the storm’s path to take it seriously. “If you don’t communicate that forecast to the right people effectively, it’s no good,” he said. “If people don’t respond, it is wasted effort.”
Michael Brennan, the current director of the Hurricane Center, said that since Katrina, meteorologists and other emergency officials had learned crucial lessons not just about how to forecast hurricanes but about communicating the dangers they pose.
Katrina wasn’t the first hurricane to hit the city, and won’t be the last. Here’s what’s different now.
Forecasting got a major funding increase after Katrina.
Since Katrina, “everything’s changed,” Dr. Brennan said.
After that season, Congress created the what’s now known as the Hurricane Forecast Improvement Program, which expanded the number of staff members at the Hurricane Center in Miami.
The program also made significant investments into hurricane-specific computer modeling. Technological advances have increased the quality and the amount of data that goes into a model in the early stages of a storm, which gives forecasters a leg up on knowing where it’s likely to go next.
Both Dr. Brennen and Mr. Mayfield said that given the tools available in 2005, the forecast for Katrina had been unusually good, though it wasn’t perfect.
The satellites that tracked Katrina were at the cutting edge then, able to capture an image of a storm every five minutes. But they had a critical flaw: Their solar-powered batteries ran low when they passed through the Earth’s shadow. The day before landfall, the satellites went dark for two hours right as the storm was intensifying.
Newer satellites, in operation since 2016, can provide high-definition images every minute, without the lengthy downtime.
Katrina’s forecast track initially indicated that it would move into the Florida Panhandle, but by Aug. 27, the track had shifted to Louisiana. At that point, Mr. Mayfield thought the storm might intensify, but forecast models then were notorious for not accurately predicting the rapid intensification that occurred with Katrina the next day.
Today, errors in forecast tracking have dramatically decreased, and the specialized forecast models, while still not perfect, give a much better indication of whether a storm is likely to intensify, Dr. Brennen said.
But meteorologists can still be surprised by a storm, he warned. He recalled Hurricane Otis in 2023, which formed as a tropical storm on Oct. 24 and made landfall a day later near Acapulco, Mexico, as a Category 5 hurricane. The models had failed to predict what Otis would do and how quickly it would do it, and more than 50 people were killed.
Climate change may not be leading to more hurricanes, researchers have found, but it is most likely causing some storms, similarly to Katrina, to intensify more quickly.
There’s a better understanding of storm surge now.
One of the deadliest parts of Hurricane Katrina was the storm surge, the wall of water pushed inland by the storm. It was the surge that broke the levees and flooded New Orleans, and immediately after the storm, it was the subject that most baffled some experts. They struggled to identify how high the surge had even been because the gauges that would have measured it had been washed away.
Hermann M. Fritz, a professor of civil engineering at the Georgia Institute of Technology, is an expert in coastal hazards like storm surge. His first stop after the storm was the barrier islands in the Gulf of Mexico, where he saw that the high water had stripped the bark off the trees, which helped him and other specialists estimate not only how high the storm surge had been but how powerful the winds had been as Katrina moved toward shore.
“Most people don’t die from wind,” Dr. Fritz said. “Even with 200-mile-per-hour winds, you’re not going to get thousands of fatalities. You’re only going to get that with storm surge flooding and inundation.”
Mr. Mayfield said the Hurricane Center had known it was possible that the levees would be affected. He had forecast models that showed it. But compared with today’s technology, he said, those forecast computer models were rudimentary. Today’s models show not only the surge but also potential inundation in low-lying areas, giving officials a better understanding of whether they need to evacuate certain neighborhoods.
Storm surge has historically caused about half of hurricane fatalities, but since 2013, the number of deaths attributed to it have declined significantly, to about 11 percent, according to research from Dr. Brennan and others.
Warnings can come sooner now, with more detail.
Forecasters now know more about a storm in its earliest days, even before it forms. In 2005, it was possible to track a weak cluster of storms before it grew organized enough to become a named tropical storm, but the ability to visualize it with graphics was limited. An official written forecast, telling people that a storm might exist soon and where it might go, was issued only up to two days in advance of the formation of a storm.
The Hurricane Center can now forecast the possible formation of a storm up to seven days in advance and can issue watches, warnings and potential paths sooner.
But then as now, there’s a risk with being too certain too early: Initial forecasts for Katrina predicted it would turn more abruptly toward the Florida Panhandle, but by Saturday, it had shifted toward Louisiana. In 2022, Hurricane Ian shifted south at the last minute and made landfall on what had been the edge of the forecast zone, which caught some local officials and residents off guard.
Where does forecasting go now?
The Trump administration has made sweeping cuts to many of the agencies involved in natural disasters, including the Federal Emergency Management Agency, the National Weather Service and the National Oceanic and Atmospheric Administration.
James Franklin, a retired meteorologist who was once the head of the hurricane team at the Hurricane Center, called the investment after Katrina the “greatest success that I can think of from the budget perspective.”
He added, “That led to some really remarkable improvements in forecast accuracy.”
But with the cuts this year, he and others fear future progress may be stymied.
“The next generation of improvements are less likely to happen,” he said.
Dr. Brennan hopes to see forecasts for inland flooding improve next. During Hurricane Helene last year, the Hurricane Center issued what he called an “unprecedented” news release about the potential for flooding in North Carolina, to “basically ring the alarm bell about the inland flooding risk in the Appalachians.” Still, the extent of the storm’s impact that far inland caught many people off guard.
“We learn something from every event,” Dr. Brennan said. “We learn lessons from Katrina that help to improve forecasts and products and services and how we communicate going forward. And we learn from Ian and we learn from Helene.”
Judson Jones is a meteorologist and reporter for The Times who forecasts and covers extreme weather.
The post 3 Ways Hurricane Forecasts Have Changed Since Katrina appeared first on New York Times.
