[SystemSafety] Texas Floods

[Prof. Dr. Peter Bernard Ladkin]

In the early morning of Friday July 4th, the upper reaches of the Guadelupe River from Kerrville 
west into the Texas Hill Country flooded. Some 80 or so people lost their lives, including a number 
of children at a kids' summer camp, Camp Mystic, on the banks of the South Fork of the Guadelupe.

The Texas Hill Country is known for its flash floods. The Guadelupe River valley is known locally as 
"Flash Flood Valley". They have had a number of serious events there -- let me take a guess at every 
decade or so.

This was different. 10" = ~ 254 mm of rain fell (somewhere) in some 3 hours. The River is reported 
to have risen in places some 24 ft = ~ 8m in less than an hour.

I reported in 2022 to the SSS22 on the German floods of July 2021. The ECMWF had give warning four 
days before of an unusual high-altitude accumulation of moisture that was going to come down 
somewhere, sometime.

Texas is known for occasionally severe weather, as the nickname for the Guadelupe Valley indicates. 
Warm air comes north from the Gulf and meets an atmosphere that is generally unstable in the summer 
months, with frequent severe thunderstorms. Westwards this atmosphere yields afternoon thunderstorms 
in the deserts of New Mexico and Arizona; eastwards thunderstorms often coalesce into supercells and 
generate tornados in "Tornado Alley" in Oklahoma.

In Texas, good weather forecasting is a big deal. It is not just the NWS and the NOAA involved, but 
universities such as Texas Tech. I have a colleague who lives in Austin and knows about safety. I 
told him about the ECMWF warning in July 2021 and asked why nobody new that there was all that 
moisture up there in the atmosphere in Texas waiting to come down. He didn't know. But he said it 
has been raining regularly and strongly for a couple of weeks, generally far more than had been 
forecast.

It turns out, according to this Guardian article, that in fact the NWS did know. 
https://www.theguardian.com/us-news/2025/jul/06/texas-floods-forecast-climate-crisis A weather 
balloon sent up from Del Rio, a town on the Rio Grande (and thus the Mexican border) som ~ 200 km W 
of San Antonio, on Thursday noted record levels of atmospheric moisture. The NWS Weather Prediction 
Center in Maryland held numerous discussions about it on Thursday. There was a warning issued by the 
Austin/San Antonio NWS office on Thursday of "excessive runoff" and "5 to 7 inches of rain". At the 
time of the downpour, the office issued further warnings.

In the Eiffel floods in 2021, the maximal precipitation did not fall in the area of main disaster 
but some 100km NNE in Hagen, on the other side of the Rhine. In Friday's flood, the maximal 
precipitation did not fall in the area of main disaster, but in San Angelo, some ~ 200km NE of 
Kerrville, capital of Kerr county where the main disaster occurred. On Saturday, some 14" = ~ 355 mm 
fell in 5 hours just west of Austin.

As in the 2021 floods, early warnings did not appear to reach the vulnerable. In the German 2021 
floods, they reached the Federal Agency for Disaster Protection and Recovery, which passed them on 
to the States, in this case my state of North-Rhine Westphalia (NRW), and the state of 
Rhineland-Palatinate (in which most of the Ahr river watershed lies). But then there seems to have 
been a disconnect. In NRW, there is a fairly labyrinthine collection of agencies responsible for 
waterways and their monitoring and thus for warning residents of emergencies. It devolves from 
state-level agency to county-level agencies and these county.level agencies are responsible for 
different watersheds. There was no straight line cutting through them all to initiate emergency 
warning. (I illustrated this in my talk, but not in the accompanying paper). This has changed. My 
city of Bielefeld has alone installed upwards of 150 sirens, including some very loud ones on a 
building 100m away from mine, and more are planned. Cell broadcast is used for a country-wide 
emergency notification system. You have to turn on "cell broadcast" in your smartphone settings, but 
then it works effectively. This is all tested once a year (and scares the bejeebers out of the cats 
and dogs). And I have received one warning of smoke development in the Bielefeld city (from a fire). 
The next time some atmospheric lake decides to unload itself on us, we'll all know in real time.

In contrast, warning systems have been mooted in Kerr County for many decades. Here, there is a 
systematic disadvantage. Such warning systems are county responsibility, not State and not Feds. The 
have to be paid for out of the county budget and Kerr County is not by any means wealthy. The annual 
budget is some $60+m per year and the costs of almost any warning system are significant, a siren 
network for example, so they have always been turned down; outside financial support has been 
applied for but not obtained. Cell broadcast is now available, but of course that works only where 
there is cell-phone coverage, which in some parts of the Hill Country there isn't. Maybe minds will 
now have changed on the expense.

There is one factor in the damage in Texas that was not present in any great degree in the 2021 
German incident. My colleague in Austin tells me that most deaths in most floods occur to people 
with SUVs and large trucks who think they can barrel their way through moving floodwaters and find 
out quickly and disastrously that they can't. Such incidents also tie up emergency-service 
resources. This didn't happen this time around because the rains and rising water levels happened at 
night. I went through a NYT profile of some of the missing/dead and a significant proportion of them 
seemed to be tourists in their RVs who'd gone down to the Guadelupe River for a pleasant 
Fourth-of-July weekend. There is a question to be asked locally as to why the Thursday NWS warnings 
of "excessive runoff" did not reach these people and induce them to move away from Flash Flood Alley 
until the situation improved.

PBL

Prof. Dr. Peter Bernard Ladkin
Causalis Limited/Causalis IngenieurGmbH, Bielefeld, Germany
Tel: +49 (0)521 3 29 31 00