Daniel Swain
There have been claims that NOAA/NWS did not foresee catastrophic TX floods--but that's simply not true. This was undoubtedly an extreme event, but messaging rapidly escalated beginning ~12 hrs prior. Flood Watch mid PM, "heads up" outlook late PM, flash flood warnings ~1am.

RE: https://bsky.app/profile/did:plc:teckhxpypg6v46gj7iysmmqt/post/3lt6heyboa22l
Jul 5 at 3:50pmWeb
Show threadDaniel Swain2d ago
As always, this is not to blame the victims! Quite the opposite; this truly was a sudden & massive event and occurred at worst possible time (middle of the night). But problem, once again, was not a bad weather prediction: it was one of "last mile" forecast/warning dissemination.
Show threadDaniel Swain2d ago
I am not aware of the details surrounding staffing levels at the local NWS offices involved, nor how that might have played into timing/sequence of warnings involved. But I do know that locations that flooded catastrophically had at least 1-2+ hours of direct warning from NWS.
Show threadDaniel Swain2d ago
One thing I do know is that this part of TX Hill Country is (in)famous for sudden and violent flood risk; that's an intrinsic product of being a hilly region with "flashy" watersheds subject to occasional but very extreme precipitation events arising from bathtub-warm Gulf.
Show threadDaniel Swain2d ago
I'm not really clear on why a region so well known for its severe flash flood susceptibly apparently did not have a better warning system in place. That's something I'm sure others with better local knowledge can dissect in greater detail.
Show threadDaniel Swain2d ago
But this does illustrate a few tragic and uncomfortable truths. The first is that even quite good weather forecasts do not automatically translate into life-saving predictions--there's a lot of other work that has to take place to contextualize the forecast and ensure it gets to right people.
Show threadDaniel Swain2d ago
The second is that the NWS historically has done a very good job at that forecast contextualization (outreach to local governments, emergency managers, outdoor recreation facilities, etc.). But that's one of the first things to go away when offices are critically understaffed.
Show threadDaniel Swain2d ago
The third is that this kind of record-shattering rain (caused by slow-moving torrential thunderstorms) event is *precisely* that which is increasing the fastest in warming climate. So it's not a question of whether climate change played a role--it's only a question of how much.
Show threadDaniel Swain2d ago
The fourth is that *exact* location and intensity localized to regional-scale convective storms (i.e., clusters of intense thunderstorms) are something that, in almost all cases, cannot be pinpointed days in advance with extant predictive tools--even in theory.
Show threadDaniel Swain2d ago
That means that while predictions can correctly highlight specific regions at high risk of extreme rainfall/flash floods from such events (as was the case here), it's not possible to predict the exact amount of rain at specific points from t-storm downpours so far in advance.
Show threadDaniel Swain2d ago
Predicting such storms is at the cutting edge of science right now, and the stakes are rising in a warming world in which they are intensifying. Yet this is precisely the kind of research that NOAA/NSF have funded in the U.S. over decades that is at imminent risk of disappearing.
Show threadDaniel Swain2d ago
Additionally, some of NOAA's very high resolution convective-resolving models (designed specifically for this purpose) were the ones that best predicted this incipient disaster. Yet these very same models are on the chopping block this year with the proposed NOAA budget.
Show threadDaniel Swain2d ago
All of this is to say: I think it's simply untrue to say this is a story about how the NWS somehow made a bad prediction or did not issue timely warnings in this case--that's just demonstrably untrue. The more proximal causes of the tragedy are otherwise. But...
Show threadDaniel Swain2d ago
But there are clear intersections between flood disaster in TX & ongoing conversations surrounding federal budget & massive cuts to NWS operations & NOAA/NASA/NSF weather/climate research--precisely at a moment in which we are seeing more events like this due to climate change.
Show threadGLC2d ago

@weatherwest.bsky.social

Tossing in this one (same basic message).

https://bsky.app/profile/nycsouthpaw.bsky.social/post/3lt7r2mjlrc25

southpaw (@nycsouthpaw.bsky.social)

The missing piece of this viral post is that the Texas officials are lying and deflecting blame. The NWS, hobbled as it is, issued an accurate flood watch for Kerr County the evening before and accurate escalating warnings overnight as the flood was developing. [contains quote post or other embedded content]

Bluesky Social
Show threadTotal Sonic Media2d ago
@weatherwest.bsky.social - the Guadalupe near Kerrville raised 26 feet in 45 minutes. This was a historical flood.
Show threadDaniel Swain2d ago
Indeed, and some areas saw an even faster/larger rise. Still, possibility of significant flash flooding in this region was explicitly in NWS forecast; emergency warnings issued 1-3 hours in advance of flood wave's arrival described extreme/exceptional nature of imminent threat.
Show threadTotal Sonic Media2d ago
@weatherwest.bsky.social - if you have ever been out in Kerrville, then you would understand why a 1 to 2 hour warning would not necessarily get to people in a timely way. It's a really small town, surrounded by rural and undeveloped areas, with camp grounds scattered in this, all with spotty cell reception.
Show threadTotal Sonic Media2d ago
@weatherwest.bsky.social - for context, my wife was nearly swept away (multiple times) while driving home from her gig last night (she's a professional fiddle player), and one of our friend lost his home and cars in the floods in Marble Falls and had to be pulled out by helicopter. I'm super thankful they are safe now, super sad so many can not say the same.