Knowledge Base

On April 24th, 2025, there was a small outbreak of very photogenic tornadoes across the Great Plains, from the Texas Panhandle through the Denver, CO area. This tornado outbreak was unique in that it was largely unexpected by most meteorologists until the day-of, yet it produced several long-track tornadoes. This case study is one of the most impressive for StormNet because the signal for a potential tornado event was present 2 weeks ahead of the event, while the SPC didn't denote a tornado threat (in Texas) until the day-of. 

14 days before this tornado outbreak, there was no expectation in the meteorological world for an impending tornado event for Texas. No existing forecast method for tornadoes or severe storms goes out 14 days. 

StormNet's 336hr forecast, however, already had a notable risk area for tornadoes in the southern Plains.

Probabilities between 1% and 10% may seem low at first glance, but these values are significant at this forecast range! Seeing probabilities above 1% at 336hrs out isn't an immediate alarm that a tornado will occur, but rather an indication that a tornado threat may be brewing. At 336hrs out, actual tornado points have an average probability of around 2% while non-tornado points (the baseline) have an average probability of around 0.0001%. 

10 days out, StormNet probabilities started to hone in on the specific region that would be impacted by tornadoes. Interestingly, the overall probabilities/confidence decreased some, but the spatial extend was more specific to where the main tornadoes occurred. 

I suspect that the overall max probabilities came down as a result of StormNet 'realizing' that the higher ceiling outcomes were less likely, thus removing those higher end ensemble members and honing in on a more specific outcome.

Still no mention of possible impending tornadoes by SPC, NWS, or others. I couldn't find any mention of tornadoes or even a favorable severe weather setup on any social media during this period.

At 7 days prior to this tornado outbreak, there was still little expectation in the meteorological community that there would be a tornado outbreak in a week. The SPC, who starts issuing general severe weather outlooks at 8 days out, had no area highlighted in their day 7 outlook for the outbreak day. "They did mention in their outlook: a rather nondescript surface pattern amid weak large-scale forcing will likely preclude widespread severe potential, though localized areas of severe storms could still occur given favorable moisture/instability, particularly along any dryline features."

StormNet's 7 day tornado forecast painted a bit of a different picture: increasing confidence in tornadoes across the high plains from Texas into Colorado:

At 5 days before the tornado outbreak, there was some growing suspicion in the meteorological community that some thunderstorms may be possible somewhere in the southern Plains, but there was still no explicit discussion of tornado potential. The 'big question' at the time was 'where will thunderstorms occur? Might they be severe?". Confidence was increasing in subtle shortwaves moving over the unstable airmass in place.

The SPC maintained a 'predictability too low' outlook, not highlighting any specific regions. They noted in their outlook: "orecast guidance suggests weak shortwave impulses may float through modest southwesterly flow Days 4-5/Wed-Thu, perhaps providing some support for isolated to scattered thunderstorm potential across the warm sector over TX/OK/KS. Given a moist and unstable airmass, at least some severe potential will exist."

StormNet's 5-day tornado forecast was quite a bit more aggressive with a bullseye centered in the Texas panhandle. Probabilities exceed 20% with increasing confidence in some tornadoes.

At 4 days out, the SPC's severe thunderstorm outlook remains 'predictability too low' with no areas highlighted. They note in their outlook: "At least some risk for isolated severe thunderstorms appears possible within the moist/unstable warm sector. However, convection may be ongoing across portions of the region Thursday morning, and given a lack of any stronger upper forcing regime or surface cyclogenesis, confidence in where greater severe potential may develop is low, precluding 15 percent delineation at this time."

StormNet, at 4 days out, continues to increase confidence in a potential tornado threat across the Texas panhandle and into Colorado. 

At 2 days out, the SPC issues their first tornado outlook for the event. The forecast remains a marginal risk for severe storms (level 1 out of 5) with a 2% chance of a tornado (also level 1 out of 5 categorically) centered in western Oklahoma. They state in their outlook: "Isolated severe thunderstorms are possible on Thursday from eastern Colorado into Kansas, and southward across west Texas and Oklahoma. Large hail and strong wind gusts will be possible with this activity."

Interestingly, despite the marginal (2%) risk of tornadoes, they have no mention of tornadoes in their outlook discussion.