“How could this happen?” It’s a question many of us are asking with broken hearts as the death toll continues to mount in the wake of the horrific Texas Hill Country flooding. We want to make sense of this tragedy, and we never want to see it happen again. As a Houston-based meteorologist who was born and raised on the outskirts of the Texas Hill Country and witnessed these types of flash floods first hand, I want to lend my voice and meteorological expertise to help answer some of the questions I’ve been receiving on the meteorology side of this tragedy.
Catastrophic flash floods have always been a part of life in the Texas Hill Country. These naturally-occurring rainstorms that stall out are infrequent but not unprecedented. This is the birthplace of the phrase “Turn around, don’t drown!” with the nickname “Flash Flood Alley.” Due to the hilly terrain, when high amounts of rain fall in a short period of time, the water flows quickly through the creek and river valleys, sometimes with devastating and deadly consequences like we are witnessing on this July 4th holiday weekend. (See also the Blanco River flood in Wimberley on Memorial Day weekend in 2015, the Canyon Lake flood in July 2002, or the Guadalupe River flood in July 1987 for some of the more recent examples.)
Often times these floods strike with little warning in the middle of the night due to the tropical nature of the air and the atmospheric physics at play.
So what happened this time?
Remnant tropical moisture from what was once Tropical Storm Barry is primarily responsible for this flood event, with an assist from upper level moisture peeled off from what was once Hurricane Flossie in the Pacific. A weak circulation in the mid-levels of the atmosphere developed over central Texas within this super-moist airmass to help organize the storms into a low-pressure complex called a “mesoscale convective vortex.” The steering currents were rather slow around this low with a steady feed of moist air called a “l…
How difficult is it to predict these flash floods?
When I was studying meteorology at Texas A&M over 20 years ago, I remember being taught that the most difficult thing to predict was how much rain would fall and where that rain would fall. While we have made tremendous progress in “quantitive precipitation forecasts” over the last two decades, this is largely still true today. No meteorologist could have told you with high confidence more than a few hours in advance that *this* much rain would fall in those *exact* locations.
But thanks to a forecasting technique known as “pattern recognition” (which comes with experience and studying past weather patterns) and the guidance offered up by more sophisticated “convection allowing models,” meteorologists already had a sense flash floods were possible on Thursday afternoon (when the National Weather Service issued a Flood Watch), and by Thursday evening they had a sense that something big was potentially brewing. I have attached the “mesoscale precipitation discussion” issued by NOAA’s Weather Prediction Center issued shortly after 6PM CT Thursday to put an extra focus on the dangerous flood ingredients that were coming together over the Hill Country. It is loaded with meteorological jargon, but it’s there for you to read if you so desire. I have also attached the “quantitative precipitation forecasts” (or QPF) from the high resolution “convection allowing models” (or CAMs) that came out Thursday morning. Some of them do show over 10” of rain falling north and west of San Antonio through 7AM on July 4th.
Yet even on Thursday evening, no meteorologist could have accurately predicted with high confidence the specific rain amounts that fell in the precise locations where they fell, and the atmosphere exceeded even the worst projections of the computer models both in coverage and amounts of heavy rain, as is often the case with tropical rain events over Texas.
Did recent staffing cuts and vacancies within the National Weath…
I have looked back at the timeline of warnings issued by the National Weather Service along with communications in a National Weather Service Slack channel where NWS/NOAA meteorologists, broadcast media, and emergency managers share information during extreme weather events, and I agree with assessments I’ve seen from other meteorologists that the National Weather Service issued timely warnings as the event unfolded in real-time. The first Flash Flood Warning for Kerr County was issued around 1 a.m., and the first Flash Flood Emergency (highest level of flood warning issued in the most extreme, life-threatening situations) was issued around 4AM, which automatically triggers a “Wireless Emergency Alert” to all cell phones within the warned area (unless users have elected to turn off those alerts). I’m also told by houston’s former NWS Warning Coordination Meteorologist Dan Reilly that flash flood warnings issued hours earlier also contained a “considerable tag” that would’ve also triggered cell phone alerts in the middle of the night where cell service was available.
It certainly didn’t help the situation that the Austin/San Antonio NWS weather forecast office is understaffed by 22% and without a Warning Coordination Meteorologist, but I see no evidence yet that it hurt the situation either with what I know.
What I am not privy to are all the communications the NWS had with emergency managers and local officials in the middle of night outside of the Slack Channel we use to communicate and share information. Therefore, I cannot comment on what was done with the information and warnings pushed out by the National Weather Service, nor do I know how many meteorologists were staffed in the office to work this event.
What role did our changing climate have in this flood?
The scientific body of evidence indicates that our warmer atmosphere and oceans increase the likelihood of these extreme precipitation events and the amount of rain that falls when they occur. The physics is pretty simple: Warmer oceans release more moisture into the atmosphere, and warmer air also holds more moisture. This enhances rainfall amounts above and beyond what would happen in a cooler world with cooler oceans.
What about cloud seeding? (Added at 2PM because of all the genuine questions about it.)
Cloud seeding cannot create a storm of this magnitude or size. In fact, cloud seeding cannot even create a single cloud. All it can do is take an existing cloud and enhance the rainfall by up to 20%. Most estimates have the rainfall enhancement in a much lower range. I am unaware of any active cloud seeding operations that occurred on Thursday night, but it is physically impossible for that to have created this weather system. This is a matter of scale. If I blow out a candle with my breath, does that mean I can then go blow out a raging wildfire? It is the same with cloud-seeding. For more, see this excellent report by my ABC colleague Ginger Zee: https://abcnews.go.com/.../ginger-zee-clears-air-cloud...
There are many other questions that remain, like what can be done to protect people from future floods, and what can we do to help now? Hopefully there will be changes made and systems put in place to help protect people in the Hill Country from future flash floods that occur in the middle of the night, and as soon as I have a list compiled of ways to help the flood victims and their families, I will share it here.
For now on this Day of Prayer in Texas, my prayers remain with the missing, the families of the flood victims, the first responders, and those tasked with leading us through the aftermath. 🙏🏻 🙏🏻🙏🏻
this is an excellent discussion of the July 4th Hill Country floods by @travisabc13.bsky.social, ABC13's chief meteorologist and a graduate of our program, @tamuatmo.bsky.social.