On July 8, New Mexico’s Rio Ruidoso unbound from its banks for the second year in a row and swelled to 20 times its typical knee-high depth. The cascade of water roared like a train, Kathy Papasan, a longtime resident on the river’s edge, told me, and dark waves battered her porch. She and her husband had to flee uphill to a neighbor’s house. Floods have always funneled down the canyons that surround their village, which is also named Ruidoso, but these were the worst ever recorded, at least in terms of damage (400 homes and buildings affected and 293 destroyed, Ruidoso’s spokesperson, Kerry Gladden, said). The July flood also killed three people, including two children visiting an RV park with their family.The Ruidoso flood was one of many across America this summer—a more active flood season than those in the previous couple of years, Kelly Mahoney, a research meteorologist for the National Oceanic and Atmospheric Administration, told me. At least four people died when rain swamped the suburbs of Chattanooga, Tennessee, this month. In Wisconsin, Milwaukee flooded after days of rain, during which multiple rivers hit record levels. Tropical Storm Chantal parked itself over North Carolina in July and caused six deaths. And in Texas, Kerrville was inundated by a slow-moving storm, causing one-in-1,000-year rains and a flood that killed at least 135 people.Mahoney told me that it was too soon to conclude if this season is abnormal or the product of climate change. Some years simply have more, or worse, floods. Climate change just makes the chance of a bad year higher. The quickly warming atmosphere holds more precipitation, which can unbuckle more often and in heavier bursts. Landscapes scarred by previous disasters—drought, wildfire—are at greater risk of flooding too.Still, the Trump administration is lowering the odds that the country will be prepared for bad years. The government has moved to freeze NOAA’s budget and cut staffing and has halted agency programs, delaying scientists from analyzing rainfall data that could help predict future storms. The National Weather Service, perhaps the best-known office within NOAA, lost almost 600 staff members as a result of cuts by the Department of Government Efficiency, and then received permission to hire as many as 450 earlier this month. NOAA’s research arm, Oceanic and Atmospheric Research, could be disbanded, according to the Trump administration’s proposed 2026 budget. (Congress, however, has moved to fund NOAA just below the 2025 budget.)If this administration does hobble NOAA, the U.S. will limit its ability to forecast extreme precipitation at a time when scientists know that these events are becoming more severe, Rick Spinrad, who led the agency under the Biden administration, told me: “It’s a double whammy—here is Mother Nature adding fuel to the fire, but it’s more rain to the storm, at the same time we’re reducing our capabilities.” (In an email, NOAA’s communications director, Kim Doster, said the administration’s budget request would allow NOAA “to deliver the world’s best science and services” while eliminating “ideological initiatives and inefficient green policies.”)Before this, NOAA spent well more than a decade improving the nation’s capacity to predict and prepare for extreme rainfall. Experts prioritized precipitation because so many types—thunderstorms, tropical cyclones, hurricanes, atmospheric rivers—affect the U.S., and understanding where and how much rain will fall is one of the most challenging parts of atmospheric science, Thomas Graziano, who had been the director of NOAA’s Office of Water Prediction before he retired in April (a few months earlier than planned), told me. The U.S. has developed the best researchers on these subjects, Spinrad said, so “there’s no backup team we can turn to.”Graziano spent his own career mapping the effects of flooding. Before leaving NOAA, he helped launch flood-inundation maps that can predict where and how much each part of the U.S. would flood under a certain weather forecast. In 2023, that inundation mapping covered an area inhabited by just 10 percent of the U.S. population, but by September 2026, it should cover nearly the entire U.S., except for remote areas of Alaska, Graziano said.One of the biggest threats to research like this would be cutting funding to Oceanic and Atmospheric Research. Dismantling that office could limit precipitation predictions used for the inundation maps, hurting their accuracy, Graziano told me. Shrinking NOAA’s staff would also leave fewer people to deliver the services that the maps provide. National Weather Service staff, spread out in more than 100 offices across the U.S., establish trust with emergency managers and city officials who make critical decisions regarding where to close roads or set up emergency shelters during floods. Without them, communities could be less prepared for disasters.The country’s aging infrastructure simply isn’t equipped to handle influxes of water as dramatic as those brought by the floods it now faces, Dave Easterling, a former division chief at the National Centers for Environmental Information, NOAA’s archive for climate and weather data, told me. (Easterling said he retired a year and a half early, in April, to avoid more DOGE cuts in his department.) The nation’s roads, bridges, buildings, roofs, and drainage systems are under-designed for both the rainfall rates of today and the increased rates of the future. And to make sure that bridges are high enough and sturdy enough that they won’t be washed away in a one-in-1,000-year flood, for instance, civil engineers and city planners need precipitation data. Right now, they’re planning to rely on a massive data set, called Atlas 15, that’s being compiled by NOAA. The second part of the two-volume project was canceled earlier this year; officials un-canceled it in July, right after the Texas floods. That could be a coincidence, Easterling said, but to him, it looks a lot like “somebody somewhere was looking at what just happened to Texas and what happened in New Mexico days later, and then what happened in central North Carolina.”Another NOAA project could help civil engineers plan for the worst-case scenarios—the biggest storms the atmosphere could create in the current world, and in a hotter one. The lab where Mahoney works, in Boulder, Colorado, studies the most precipitation that could ever fall in a given place for a given amount of time. These most extreme rainfall scenarios can inform design thresholds for high-hazard infrastructure, such as dams and nuclear facilities. But the model—the probability-maximum precipitation model, or PMP—hasn’t been updated since the 1950s, Mahoney said. Publication of a modernized version requires funding through 2030, but funding has been appropriated only through 2026.“The methods that went into developing PMP in the 1950s were based on limited observations and scientific understanding, and we have a much greater understanding of extreme precipitation now,” Mahoney said. “But it’s all happening on the backdrop of a changing atmosphere.” New Mexico, for example, has generally gotten drier: Annual precipitation has decreased over a 30-year climate record, Scott Overpeck, the warning-coordination meteorologist for the National Weather Service office in Albuquerque, told me. But the past several years have had more single heavy-rain events than previous years did, Overpeck said.“It might be that you get precipitation events fewer and farther apart, but when you do get it, you’re getting a lot more than what we might’ve observed in a single event many years ago,” Overpeck said. He also noted that future forecast models would rely on NOAA research, including Atlas 15 and the flood-inundation maps.That information is becoming more important to towns such as Ruidoso as they brace for more severe and frequent floods. This spring, the town spent millions of dollars on flood-preparedness efforts, including dredging Rio Ruidoso to make it wider and deeper after a combination of wildfires and floods had wrecked many homes the summer before. Much of those mitigation efforts were ripped out by this year’s storm, Jason Kean, a research hydrologist for the United States Geological Survey, told me. People will continue living in places such as Ruidoso and Chattanooga and Kerrville. But with the breadth of these floods likely to continue to grow, investing in scientific research is one of the clearest ways to help communities prepare for both the rains that are already here and the rains that are coming.