Follow along with the video below to see how to install our site as a web app on your home screen.
Note: This feature may not be available in some browsers.


Did somebody say...
pi?
(No, Diaz is still not an EF5 in my book. Came damn close, though!)




Interesting post. My takes are about both the March 2021 risks:
To this day, I think the 45% on March 17th 2021 was one of the weirdest SPC decisions of all time. Even ignoring hindsight, when I look back at the models for March 17th on the HREF archive, I am puzzled that anything higher than 15% was considered for the tornado forecast. This was certainly a case where an event being spotted early and seemingly initially impressive built hype that led to over forecasting at Day 1.
The kinematics were just unimpressive... the trough was not moving overly quickly and the 500mb flow was just 50-60kts over the warm sector during the event. Dixie in early spring heavily relies on powerful LLJs, and across AL the LLJ on the day was in some spots lower than 40kts. The most favourable environment was further West in MS - where sustained discrete convection struggled (and was modelled to do so). The model signal was for copious discrete storms to develop early across AL in the Eastern half of the warm sector (which happened) but as stated the environment was not close to high-risk level and so the strongest tornado was EF2. Had the trough ejection occurred earlier near peak heating the story would be different, imo. But given the lack of CAM signal coinciding with the impressive environment, the 45% let alone 30% seem very out of character for the often conservative SPC.
I think March 25th 2021 was almost the opposite. Admittedly, I have only been tracking setups in detail since about 2019 which is not that long in the grand scheme of weather but that day was and still is the most impressive setup I can recall happening to be honest. The agreement for an extremely high end environment across N MS/AL was literally unlike anything I can remember... literally of the charts.
View attachment 49484
What was even more ridiculous was for literally all CAMS to have a very strong signal for multiple robust, discrete supercells through that environment. Almost every CAM's simref looked basically like 4/27. Even the NAM which notoriously favoured linear modes over anything at that time had a line of discrete cells. I think the SPC was well within their right to go 45% or even higher, and I say this with hindsight. I truly think the only reason the 45% wasn't issued was the failure of the one the previous week to verify. The way the initial round of storms was more significant than expected and managed to save MS/AL from storms utilising the peak parameter space a couple hours later I still think about often. Even then - the event managed to produce multiple intense tornadoes that were destructive. That event was driven by a quick moving shortwave rather than the larger scale troughs you tend to see with major outbreaks (which is what made 3/15 this year look so scary until D1 when timing seemed a bit off).

That means it’s been five years since James Spann’s house got hit by a tornado.Welp I missed the 3/25/21 anniversary, but luckily its biggest tornado occurred after midnight on the 26th, so I can at least not be inaccurate when I say that today marks the 5 year anniversary of the Newnan, Georgia "EF4". It is the most recent violent tornado to not have any clear funnel pics exist.