BalancedWx Special: Watches and warnings in the Texas flood tragedy
Emergency management's role in the process examined
WGUS64 KEWX 031818
FFAEWX
URGENT - IMMEDIATE BROADCAST REQUESTED
Flood Watch
National Weather Service Austin/San Antonio TX
118 PM CDT Thu Jul 3 2025
TXZ183>187-202>204-041200-
/O.NEW.KEWX.FA.A.0003.250703T1818Z-250704T1200Z/
/00000.0.ER.000000T0000Z.000000T0000Z.000000T0000Z.OO/
Val Verde-Edwards-Real-Kerr-Bandera-Kinney-Uvalde-Medina-
Including the cities of Del Rio, Leakey, Brackettville,
Kerrville, Bandera, Hondo, Uvalde, and Rocksprings
118 PM CDT Thu Jul 3 2025
...FLOOD WATCH IN EFFECT THROUGH FRIDAY MORNING...
* WHAT...Locally heavy rainfall could cause flash flooding across
portions of South Central Texas. Rainfall amounts of 1 to 3 inches
with isolated amounts of 5 to 7 inches are possible.
* WHERE...A portion of south central Texas, including the following
counties, Bandera, Edwards, Kerr, Kinney, Medina, Real, Uvalde and
Val Verde.
* WHEN...Through Friday morning.
* IMPACTS...Excessive runoff may result in flooding of rivers,
creeks, streams, and other low-lying and flood-prone locations.
Creeks and streams may rise out of their banks.
* ADDITIONAL DETAILS...
- A moist tropical airmass combined with a slow moving storm
system will bring rounds of scattered to widespread showers
and storms with heavy rain rates possible.
- http://www.weather.gov/safety/flood
PRECAUTIONARY/PREPAREDNESS ACTIONS...
You should monitor later forecasts and be alert for possible Flood
Warnings. Those living in areas prone to flooding should be prepared
to take action should flooding develop.
&&
$$
Just a little more than a week ago, the National Weather Service (NWS) office in Austin/San Antonio issued the flood watch which was intended to put the western parts of their forecast area - an area that stretches from the Rio Grande to central Texas - on alert for heavy rainfall and potential flash flooding for the upcoming night and into the morning of Independence Day. In their technical forecast discussion that accompanied that flood watch, they noted that the morning precipitable water value at Del Rio - the nearest NWS upper air site - was 2.31”, near a record maximum for the date, then added:
Cold pools expanding under weak deep-layer flow may become a focusing mechanism for concentrating a few clusters of showers and storms as depicted on some of the high-resolution mesoscale guidance, and the heaviest rain totals may be associated with these slow-moving clusters. Models remain in disagreement over
the placement of the heaviest rain totals, though the greatest rainfall potential this afternoon into tonight is over the Highway 90 corridor, southern Edwards Plateau and Hill Country. Rainfall totals of 1 to 3 inches appear likely, but locally higher amounts upwards of 5 to 7 inches could materialize if slow-moving storms begin to cluster.What this is saying in lay terms is that the storms would be slow moving and that small scale influences from the thunderstorms themselves would be the primary mechanism for focusing maxima in rainfall - and that as is typically the case, high resolution models differed on the location and amounts of those higher rainfall totals.
As this article published by Scott Dance in the Washington Post yesterday noted, as the evening wore on, updated runs of the high resolution models continued to increase the rainfall amounts, with some runs showing extreme totals of 10-20” over the next day or so. The article correctly points out that the NWS did not increase the rainfall totals mentioned in their flood watch product on evening updates - and it is fair in hindsight to question that and if any increase in the rainfall totals forecast would have further raised awareness with the public.
However, as my colleague Matt Lanza alluded to in a quote in the same article, in situations such as this with slow moving thunderstorm complexes the high resolution models often forecast totals like this that do not pan out, and it certainly still would not have been possible to narrow down the location of those potential heavier rainfall totals. As shown above, the area in the headwaters of the Guadalupe River where the flash flood occurred ended up having - by far - the heaviest rainfall in the flood watch area overnight, and those totals were 6-10”.
Furthermore, the critical step that the NWS has to raise awareness had already been taken, namely the issuance of said flood watch. As I have alluded to in earlier posts, these sort of small scale thunderstorm driven flash flood events really need to be viewed through a similar prism as a tornado situation: meteorologists can tell you conditions are favorable, but (at least with today’s science) they cannot tell you the precise timing, magnitude and location of the impacts until the storms have developed on radar.
When the NWS issued their initial flash flood warning for life-threatening flash flooding at 1:14 am, storms had only been ongoing in the warning area for about 60 to 90 minutes, and river levels had yet to start rising significantly. The NWS would update the warning four additional times over the next nearly 3 hours, culminating with the issuance of a flash flood emergency - their highest level that warns of catastrophic flash flooding - at 4:03 am.
While obviously we as meteorologists would love to have as accurate and precise of a forecast as possible ahead of a hazardous weather impact such as this to maximize awareness and warning efficacy, the reality is that storm scale driven events like tornadoes and flash floods - even ones with catastrophic impacts such as this - are not always apparent ahead of time. Localized flash floods, which form through a complex interaction of rainfall rates and underlying terrain, in particular are less likely to be well anticipated ahead of time; just this week, major urban flash flooding occurred in Chicago due to localized intense rainfall without a preceding flash flood watch.
Hence, the short-fused warning product is the ultimate safety net that emergency managers and meteorologists focus on for protection of life and property. Because of the critical importance of these short-fused warnings, the National Weather Service has a long-standing program called StormReady that encourages local communities and businesses to establish robust procedures to be able to quickly receive and effectively communicate these warnings. To become recognized by the NWS as StormReady, a community must establish a 24-hour warning point and emergency operations center and have multiple ways both to receive severe weather warnings and to disseminate them.
Unfortunately, Kerr County and the city of Kerrville are not StormReady communities. More worrisome than that fact are some of the stories that have arisen in the last couple of days about confusion in the emergency notification process during the flooding. KXAN-TV reported that Kerr County does have CodeRED, a system that governments and organizations can purchase to send out notifications about emergency situations through telephone and cell phone notifications. Emergency services radio recordings from Kerr County during the flood event indicate that volunteer firefighters requested a county dispatcher to activate CodeRED at 4:22 am based on what they were seeing on the ground as far as flooding. This was of course after the flash flood emergency had already been issued, but the dispatcher replied that they could not activate CodeRED without permission of a supervisor. When exactly CodeRED was finally activated is still unclear, but KXAN reported that it does not appear to be any earlier than 5:34 am, more than 4 hours after the initial flash flood warning for life-threatening flooding, and 90 minutes after the issuance of the flash flood emergency.
All of this suggests to me as someone who is experienced in both meteorology and emergency management (EM) - and who has certified numerous communities as StormReady - very poor warning receipt and dissemination procedures. One of the core aspects of StormReady is that on-duty personnel such as dispatchers must have multiple ways to receive warnings and the authority to activate warning dissemination systems without needing to receive authorization from another entity. This is to ensure that warnings are disseminated into the community as rapidly as possible. Whether this was a chronic issue or unique to this event, Kerr County did not seem to have these sort of well organized warning receipt and dissemination procedures on July 4th.
This is particularly concerning given the obvious vulnerability the area along the north and south forks of the Guadalupe River presents. Emergency management principles clearly shows the importance of having a robust plan for warning receipt, notification and action given the large numbers of people, including young children, in an area that has been impacted by major, rapid onset flash floods in the past. Obviously, the local government is not the only entity with responsibility for receipt and dissemination of warnings. Individual campgrounds and resorts along the river had their own responsibility to be able to receive warnings, as of course do individual members of the public. However, aside from all of the discussion about sirens and more elaborate warning systems, the first and most important step of any government entity with life safety responsibility is to have a well organized EM program that can effectively receive and disseminate warnings. For whatever reasons, that does not appear to me to have existed in Kerr County on the morning of July 4th.
This then ties in to some of the broader discussions that have been ongoing about emergency management at the national level, and the concept that has been touted by President Trump and Department of Homeland Security Kristi Noem to push more EM responsibility to the state and local levels while reducing or even eliminating FEMA. As I have pointed out in earlier posts on this topic, emergency management is already organized in this country around the concept that EM is primarily a local function, with state level EM agencies becoming engaged when local capacity is overwhelmed, and similarly FEMA becoming engaged when states are overwhelmed.
However, state and federal EM agencies do also have an important role in providing collaboration, standards and funding for local and county level emergency management programs. Today, Republican Texas state senator Paul Bettencourt told CNN he planned to introduce legislation that would allow the state to actually go in and install emergency sirens in local counties given the flood catastrophe last week.
Pressed on why sirens weren’t in place before, knowing that the area is in a flood zone, Bettencourt said the state government needs to step in and install them — not leave it to individual counties to figure out how to fund projects themselves.
“You’ve got 4,000 governments in Texas. It’s a big place. Not everybody has the resources to get it done,” he said.
I agree with the senator - and I think the same concept obviously applies at the national level. The United States is a big place, with thousands of different government entities; not everybody has the resources to have the emergency preparedness capacity their vulnerabilities demand. The administration’s view seems to be that simply providing the states with federal funds after disasters will be sufficient to meet response and recovery needs. The obvious problem that I see with this concept is that smaller, less populated states do not have the permanent infrastructure to be able to manage and utilize these resources effectively without federal support.
Furthermore, putting more of the responsibility on the states to manage these post-disaster response and recovery efforts will mean even less bandwidth for states to be able to assist places like Kerr County with mitigation and preparedness activities, and ultimately it is those mitigation and preparedness activities - like developing robust warning receipt and action plans - that reduce impacts and save lives. If we continue to push more and more EM responsibilities to local and state levels, I fear we are setting the stage for increasing inequity in capacity and services around the country, and that resource rich areas will fare better than resource poor areas. In my view, we as a nation are only as strong as our weakest link, and the reality is most of us spend time in both resource rich and resource poor areas. It truly benefits all of us to think closely about how we can work together as a nation to identify our most at-risk areas like the Guadalupe River, and then mitigate the vulnerabilities to try to make sure heartbreaking tragedies like July 4th never happen again.