Cuts to NOAA increase the risk of deadly weather tragedies

The deadliest U.S. flash flood from a thunderstorm complex in nearly a half-century swept through the Hill Country of Texas early morning on July 4, killing at least 89 people, including 27 from Camp Mystic, a Christian summer camp for elementary school girls with nearly a century of operations along and near the Guadalupe River.

The disaster has raised hard questions about whether short staffing from DOGE-directed cuts and compelled early resignations at the National Weather Service, or NWS, contributed to the disaster. It’s true that the weather service office at New Braunfels responsible for the warnings had lost 22% of its staff. However, the New Braunfels office had “surge staffing” for the event, with five forecasters instead of the usual two working, according to the Associated Press. And as Bob Henson and I wrote in our previous post, the weather service performed admirably, issuing the proper flood watches and flood warnings.

But we’re pushing our luck if we think the cuts to NOAA, which oversees the weather service, won’t cause a breakdown in our ability to get people out of harm’s way in the future. In particular, the loss this year of many of the weather service’s most experienced leaders – people with decades of experience in the particular weather vulnerabilities in local areas – poses a significant danger to the mission of protecting people and property.

And the dismissal of all probationary NOAA employees earlier this year means that we have eaten our seed corn: There is no incoming fresh blood to replace the old guard. Even if the weather service does start hiring again, it may have trouble attracting talented people. Who would want to work for a government where science is not valued, where you could be suddenly dismissed for no reason with no warning?

Short staffing increases the risk of bad outcomes

As Bob Henson and I wrote at the beginning of hurricane season, massive staffing losses at NOAA from firings of probationary employees, DOGE buyouts, and early retirements have left at least eight of the 122 National Weather Service offices unable to operate around the clock. Because of the loss of staff, regular twice-per-day upper air balloon soundings, which are typically the most important ingredient in making reliable model weather forecasts, have been lost from about 18% of the nation’s upper air stations. Some locations have been reduced to once-per-day launches, and a number are doing no launches at all.

The Washington Post reported that for the month ending on May 26, 17% of all U.S. balloon launches that should have occurred have not, mostly because of NOAA staffing losses. Though it appears that lack of balloon data did not have a significant detrimental impact on the Texas flood forecasts, it is a certainty that this level of data loss will cause significant degradation for some forecasts of extreme weather events – potentially including hurricanes making landfall along the U.S. Gulf and Atlantic coasts.

New NOAA budget would eliminate the lab that developed a key flash flood forecasting tool

The Texas tragedy has eerie similarities to the flash flood that tore through Colorado’s Big Thompson River on July 31, 1986, killing 144 people. Like the July 4 flood, the Big Thompson flood occurred on a holiday — the Saturday night of Colorado’s centennial weekend — so even more people than usual were camping and recreating in the narrow canyon. While more than 12 inches of rain was falling upstream, forecasters in Denver had no visual access to radar data because of technical difficulties.

The Big Thompson disaster led to awareness tools such as the “Climb to Safety” signs that are common in Colorado canyons. And it stimulated research that led to many of the flash flood warning practices that are now commonplace, including coordination between the National Weather Service and state and local stakeholders that can save lives when a disaster hits.

NOAA’s National Severe Storms Laboratory in Norman, Oklahoma, has developed and is researching one of the key tools used to make flash flood forecasts: the Flooded Locations And Simulated Hydrographs Project (FLASH), which improves the accuracy, timing, and specificity of flash flood warnings using high-resolution, accurate rainfall observations from the Multi-Radar Multi-Sensor project. This program was made operational in 2016 after researchers showed it doubled the skill over the legacy weather service flash flood guidance system.

But NOAA’s just-released budget plan for 2026 would close all NOAA labs, including the National Severe Storms Laboratory (which was founded in 1964), and others with similar long histories of innovation and accomplishment. This includes the two labs most instrumental in improving hurricane forecasts, the AOML and GFDL.

The justification for doing so comes from Project 2025, which describes NOAA’s primary research branch, which operates all of these labs – the Office of Oceanic and Atmospheric Research – as “the source of much of NOAA’s climate alarmism.”

As meteorologist Michael Lowry wrote today, “One of the primary tools we use to predict flash floods like the ones in Central Texas come from the Multi-Radar/Multi-Sensor System, a project of the National Severe Storms Laboratory or NSSL in Norman, Oklahoma. I’ve zero doubt weather service forecasters were leveraging that tool that evening to issue flash flood warnings. The National Severe Storms Laboratory and associated projects like this one are slated for elimination in NOAA’s proposed 2026 budget, which would be detrimental to our ability to forecast these types of deadly floods in the future.”

NOAA’s new budget would devastate flash flood research

Homeland Security Secretary Kristi Noem said Saturday, “We know that everyone wants more warning time, and that’s why we’re working to upgrade the technology that’s been neglected for far too long to make sure families have as much advance notice as possible.”

And in comments Monday, she doubled down on this theme, saying that the weather service “has been working to put in new technology and a new system because it has been neglected for years. It’s an ancient system that needed to be upgraded and so President Trump recognized that right away and got to work on it when he came into office in January but that installation is not complete and that technology isn’t fully installed.”

It’s unclear which system Noem may have been referring to. But in reality, the proposed shuttering of the National Severe Storms Laboratory would significantly degrade our ability to improve flash flood forecasting by eliminating the three most promising research programs to do so. These include the FLASH program mentioned above; the Forecasting a Continuum of Environmental Threats (FACETs), a next-generation weather forecast and warning framework for all weather hazards, which will be modern, flexible, and designed to communicate clear and simple hazardous weather information to better serve the public; and Warn-on-Forecast, which aims to increase lead time for tornado, severe thunderstorm, and flash flood warnings.

In his Balanced Weather Substack feed, Alan Gerard, the former director of the analysis and understanding branch at the National Severe Storms Laboratory, wrote on Sunday, “Not only is the administration not doing what Secretary Noem claims, they are clearly doing the opposite, not only with NOAA but other federal science agencies, destroying the research apparatus that would actually upgrade technology and provide more advance notice of horrific events such as this flash flood.”

Wholesale destruction of weather research infrastructure

Why are we cutting NOAA and weather service capability to make lifesaving forecasts at a time when climate change is making extreme rainfall events more frequent and severe? And why are we destroying our ability to do the research needed for the weather service to better forecast extreme rainfall events?

If approved by Congress as it now stands, the latest budget proposal for NOAA will close the government research labs needed to create those tools. And these are not simple cuts, but a destruction of the basic infrastructure needed to do weather research – including hurricane forecasting improvements. This will take years or decades to recover from, even if funding is fully restored next year. It’s like blowing up a dam and trying to rebuild it by gluing the pieces back together. It won’t work. You have to start from scratch.

The rationale the administration has given is that it wants to privatize the National Weather Service. This sentiment is expressed in Project 2025, which claims, “Studies have found that the forecasts and warnings provided by the private companies are more reliable than those provided by the NWS.” To support this notion, Project 2025 cites a 2020 AccuWeather press release. In reality, most private forecasting firms and broadcast meteorologists rely heavily on the weather modeling carried out by the National Weather Service, and the insights from NOAA’s online forecast discussions and other products provide value to the entire weather enterprise as well as interested citizens.

But if we do go ahead to privatize the weather service, there needs to be a plan for an orderly transition to a new forecasting enterprise phased in over a period of years. Otherwise, we risk higher death tolls from extreme weather events because of a lack of warning from hasty, unwise cuts to NOAA and the weather service.

There is no such plan.