• Welcome to TalkWeather!
    We see you lurking around TalkWeather! Take the extra step and join us today to view attachments, see less ads and maybe even join the discussion.
    CLICK TO JOIN TALKWEATHER

Severe Weather Threat May 16-18, 2026

CheeselandSkies said:
Even that still seems a good bit east of where things ought to be from a synoptic perspective based on the 12Z NAM...more like eastern Kansas and Nebraska. I wonder if these "AI" probabilities are based on some sort of ensemble of the physical models, if they run their own models, or what?
Click to expand...
Stormnet is an aggregation of models across a timespan, here it's telling us what are the tornado and lightning probabilities from 1pm CST Monday to 1am Tuesday. The data gets more precise the closer you get to Day 1.
 
Well, if model outputs are anything to go by, Monday definitely has the look of a red-letter day across the core of the Midwest. High run-to-run confidence on a powerful trough punching out towards Nebraska and Iowa will make for a potent, classic Midwest tornado setup, resulting in some pretty ludicrous parameters. Would note that the strongest of the LLJ comes in towards the evening, so I would strongly caution against assuming nothing will happen just because it's 4 PM and no storms are going in the OWS. The presence of a moderate cap will also reinforce this, and add to complacency of people in the core risk area. Even with attending concerns, the high-end thermodynamics and classic synoptics of this setup should tell you enough: the Midwest will need to take Monday very seriously.
1778860865274.pngtrend-gfs-2026051512-f084.500wh.conus.gif1778860921098.png
 
CheeselandSkies said:
Interesting that CSU remains focused a good bit further north and east of where current model trends/SPC forecasts would suggest the greatest threat taking shape.

I rarely look at their outputs unless someone posts it because that site doesn't work for me. The images won't load in any browser on my PC. However I seem to recall their output for 3/31/23 was pretty close, but displaced a little further east (referring specifically to the northern mode). Like, it bullseyed N IL/S WI so if I had targeted based solely on that, I would have missed Keota.
Click to expand...
Try clicking the actual day up top on the header menu, then try the the 00 and 12 UTC INIT. That fixes it for me.
 
slenker said:
Wouldn’t be surprised if member 18 has something seriously wrong with that output, insane outlier of higher values. What could it possibly be doing to even yield values that significant across that wide of an area
Click to expand...
It glitches like that pretty regularly. Just a bugged run. It also causes the mean SCP to go crazy sometimes, so you always gotta check all the members before taking the mean seriously.
 
GXSC said:
CSU-MLP I find has a slight northward bias. Stormnet is showing a severe risk that is further south than predicted, around southern IA/northern MO/northeast KS.View attachment 52838
View attachment 52839
stormnet.ai

StormNet | Severe Weather Forecast + Super-Res Radar

AI-powered severe weather forecast system and super-res radar to help predict the risk of lightning, hail, tornadoes, and damaging winds for the continental United States.
stormnet.ai stormnet.ai
Click to expand...
Holy $hit that tornado probability
 
Clancy said:
Well, if model outputs are anything to go by, Monday definitely has the look of a red-letter day across the core of the Midwest. High run-to-run confidence on a powerful trough punching out towards Nebraska and Iowa will make for a potent, classic Midwest tornado setup, resulting in some pretty ludicrous parameters. Would note that the strongest of the LLJ comes in towards the evening, so I would strongly caution against assuming nothing will happen just because it's 4 PM and no storms are going in the OWS. The presence of a moderate cap will also reinforce this, and add to complacency of people in the core risk area. Even with attending concerns, the high-end thermodynamics and classic synoptics of this setup should tell you enough: the Midwest will need to take Monday very seriously.
View attachment 52841View attachment 52843View attachment 52844
Click to expand...
With any luck we'll just have a line of storms pass through up here and no discrete supercells...
 
Clancy said:
Well, if model outputs are anything to go by, Monday definitely has the look of a red-letter day across the core of the Midwest. High run-to-run confidence on a powerful trough punching out towards Nebraska and Iowa will make for a potent, classic Midwest tornado setup, resulting in some pretty ludicrous parameters. Would note that the strongest of the LLJ comes in towards the evening, so I would strongly caution against assuming nothing will happen just because it's 4 PM and no storms are going in the OWS. The presence of a moderate cap will also reinforce this, and add to complacency of people in the core risk area. Even with attending concerns, the high-end thermodynamics and classic synoptics of this setup should tell you enough: the Midwest will need to take Monday very seriously.
View attachment 52841View attachment 52843View attachment 52844
Click to expand...

EHI plot is valid for 0Z Monday, which is 7 PM CDT Sunday evening. The one valid for 0Z Tuesday is actually surprisingly tame, except for across southern KS and northern OK, where verbatim spreads would be an issue (89/67 on a sounding I cherry-picked from within the >10 EHI bullseye).
 
UpperLevelLOL said:
Should we change the thread title to include today? Pretty much the whole state of IA is under an ENH

Click to expand...

It's for wind and hail. The tornado risk is 2-5% unhatched.
 
I’d probably keep it separate just because today’s setup is tied more to a lead wave, while the 16-18th period still looks like its own more substantial synoptic system with a broader warm sector, stronger moisture return, and overall higher end severe potential. Even though they’re part of the same active pattern, the evolution and forcing mechanisms still seem different enough that it makes more sense to treat them as separate waves.
 
Ozonelayer said:
12z NAM is in range for Monday, at the end though so take it with a grain of salt.

That being said:

2026051512_NAM_084_39.78,-96.06_severe_ml.png


(Taken in NE KS)
Click to expand...
That is a prime Mike Tyson-esque hook on that hodograph. Yikes.
 
CheeselandSkies said:
EHI plot is valid for 0Z Monday, which is 7 PM CDT Sunday evening. The one valid for 0Z Tuesday is actually surprisingly tame, except for across southern KS and northern OK, where verbatim spreads would be an issue (89/67 on a sounding I cherry-picked from within the >10 EHI bullseye).
Click to expand...
Whoops, yeah that's my bad!
 
CheeselandSkies said:
Interesting that CSU remains focused a good bit further north and east of where current model trends/SPC forecasts would suggest the greatest threat taking shape.

I rarely look at their outputs unless someone posts it because that site doesn't work for me. The images won't load in any browser on my PC. However I seem to recall their output for 3/31/23 was pretty close, but displaced a little further east (referring specifically to the northern mode). Like, it bullseyed N IL/S WI so if I had targeted based solely on that, I would have missed Keota.
Click to expand...

Not sure where you're looking. The D4 CSU outlook is in tight agreement with the SPC's 30% D4 risk area. Like, really tight.

IMO, CSU machine learning model does a pretty good job at forecasts. It's usually pretty close to the SPC.
 
You must log in or register to reply here.
Back
Top