“This is the most dangerous river valley in the United States. . . We do not have a warning system.” — Judge Rob Kelly, Kerr County, Texas, 4 July 2025
As I write this, the death toll in the Texas flash floods now exceeds 70, with 12 people still missing, including 11 girls and one camp counselor. It is a heartbreaking and horrific tragedy.
Many have been quick to politicize the tragedy in an effort to support whatever agenda that they were promoting before the disaster — climate change, DOGE budget cuts, operations of the National Weather Service, the Biden Administration. The one political implication of the disaster that I’m ready to call for is to reassert the importance of establishing a U.S. Disaster Review Board, a case made here at THB by Mike Smith last March.
Today, I share some data and context on the event for those wanting to go beyond seeking to use tragic deaths in hopes of scoring online partisan points. Shameful.
Before getting to relevant data and research, my view — This tragedy occurred in a location that has among the greatest risks in the nation of flash flooding, where kids in summer camps have previously been swept away to their deaths, and where warning systems are (apparently and incredibly) not in place. This tragedy never should have happened and it should never happen again.
Just a bit more background — early in my career I studied the use of weather forecasts and warnings at the National Center for Atmospheric Research, including flood warnings. Overall, the U.S. has seen tremendous progress in forecasts, warnings, evacuations, with a long-term drop in death rates from flooding. However, this week’s tragedy shows that we still have much work to do.
Where did this flood occur?
The flood took place in a region of Texas that has long been called, “flash flood alley,” and is pictured in the image above from a 2022 article by Accuweather. That article explained:
[F]airly regular flash floods have led officials to nickname a part of the state ‘Flash Flood Alley,’ a geographic region that tracks through many of Texas’ major metropolitan areas, including San Antonio, Dallas, Austin and Waco.
The Balcones Escarpment, which roughly parallels Interstate 35, marks the location of flash flood alley. The inactive fault zone formed a rise in the topography in the area, which enhances storm systems that pass over it, causing them to dump more rain there than they might elsewhere.
“We’re going from the coastal plains right into the hill country. There’s a rise of at least about 500 feet in elevation,” Pete Rose, a meteorologist with the Lower Colorado River Authority, told AccuWeather National Reporter Bill Wadell.
“Along with that, you have a lot of your hills and valleys that go along with that type of topography, and these hills don’t contain a lot of soil; they have very thin soil. So when rain does hit them, not much of it gets absorbed,” Rose said, noting that water will rush down the valleys and pile into creeks and streams.
Warm, moist air from the Gulf of Mexico helps fuel storms as well, giving them ample moisture to dump lots of rain in a short amount of time across the dusty Texas soil.
If there is anywhere that should be prepared for flash flooding, it is “flash flood alley.”
Was the flooding unprecedented or unusual?
The flooding was certainly extreme but it should not have been historically unexpected. The documented record of extreme flooding in “flash flood alley” goes back several centuries, with paleoclimatology records extending that record thousands of years into the past.
Consider the figure above, from a classic 1940 historical text on U.S. floods, which shows that the same region of Texas that experienced this week’s floods has long been known to be a bullseye for flash flooding. In fact, almost a century before Hoyt and Langbein, Texas experienced one of the greatest losses of life in U.S. history related to extreme weather.
In 1846, in the months after Texas became a U.S. state, massive flooding compounded the many problems facing thousands of recent immigrants from Germany who had been settled in New Braunfels, Texas, which was significantly impacted by this week’s floods.
According to a contemporaneous 1846 account, cited in a fantastic 2006 PhD dissertation on flooding in Texas by William Keith Guthrie, at the University of Kansas:
The Guadalupe [River] would often rise fifteen feet above its normal stand after these heavy rains, carrying with it in its swift torrent a number of large trees, uprooted farther up the hills. Smaller brooks, ordinarily not containing flowing water, became raging torrents which could be crossed only by swimming.
Newsweek this week recalled an eerily similar event involving summer campers from 1987:
The disaster echoes aspects of a 1987 flood that killed 10 campers at a nearby Christian camp, local meteorologist Cary Burgess told Newsweek on Sunday via email. . .
She noted that, while flooding like this “doesn’t happen often,” large scale ones “have occurred at least four times in the last 50 years.”
“Kerrville actually had a higher crest of 37.4″ in a July 1987 event where 10 campers were lost downstream,” Burgess told Newsweek. The crest refers to the highest level a river reaches before it goes back down.
She added: “This flood resembled all of those previous events, although we have a much higher population now compared to 40 or 50 years ago.”
During the 1987 flood, about 11 inches of rain fell on the area, sending buses of teenagers fleeing from the Pot O’ Gold Christian Camp near Comfort, Texas, into the water. Search-and-rescue crews saved dozens, but 10 teenagers died.
“Each flood event of this magnitude has occurred in the month of June, July or August, and usually has some kind of tropical characteristic with it,” Burgess noted, adding that “remnants of former Tropical Storm Barry that made landfall in Mexico last week and that circulation is still spinning across Texas today” . . .
Burgess also noted to CNN that the Guadalupe River is bedded with limestone rather than mud, “so it acts like concrete or a street.” The river, which is approximately 250 miles, typically has a fast stream as it flows over limestone.
This week’s event should have surprised no one.
Has extreme precipitation or flooding become more common in “Flash Flood Alley”?
The figure above shows for 1981 to 2022 two measures (ERA5 top and CHIRPS bottom) of one metric of trends in of extreme precipitation for the continental U.S. — R95p, “annual precipitation on the days above the 95th percentile of total precipitation.” Blue coloring indicates an increasing trend and red indicates a decreasing trend. Hatching indicates that the trend is statistically significant.
The data show considerable variation across the U.S. but also no indication of an increase in this metric for “flash flood alley” over this 42-year analysis. In fact, with the exception of the southwest (decreasing trend) and parts of Appalachia and the Ohio River Valley (increasing trend) there is little overall indication of significant trends in R95p metric of extreme precipitation.
The IPCC AR6 WG1 concluded the following on U.S. river floods:
There is limited evidence and low agreement on observed climate change influences for river floods in North America (Section 11.5). Trends in streamflow indices are mixed and difficult to separate from river engineering influences, with large changes but little spatial coherence across the USA, making it difficult to identify trends with confidence . . .
Based on the peer-reviewed literature and observational records, there is little empirical basis to claim that extreme precipitation has increased in “flash flood alley” (or indeed, most of North America or the world). Similarly, there is little basis for claims that flooding has become more common or severe.
Has preparation and response to flooding improved?
The figure above shows that as the population of Texas increased from ~9.2 million in 1958 to ~28.6 million in 2018, overall flood deaths remained fairly constant, meaning that the fatality rate dropped by about two-thirds.
These data are strongly suggestive that actions to reduce the risks of flooding — through better structural mitigation, improved forecasts and warnings, etc. — have been highly effective. The trends in decreasing mortality rates from flooding in Texas have also been documented globally.
The time series also indicates that in terms of fatalities, the 2025 flood tragedy is exceptional in recent historical context.
I’ll end where I started
This tragedy occurred in a location that has among the greatest risks in the nation of flash flooding, where kids in summer camps have previously been swept away to their deaths, and where warning systems are (apparently and incredibly) not in place. This tragedy never should have happened and it should never happen again.
If there was ever an issue where politicians should come together to take action to ensure that this type of tragedy never happens again — This is it.