Severe Weather Threat - November 29th-30th, 2022

[CheeselandSkies]

Washington County, AL tornado rated low-end EF3.

 

[Shakespeare 2016]

Washington County, AL tornado rated low-end EF3.

3 significant tornadoes so far.

 

[Shakespeare 2016]

The Caldwell County, Louisiana tornado from yesterday was also upgraded to a low-end EF3.

 

[tornado examiner]

The Caldwell County, Louisiana tornado from yesterday was also upgraded to a low-end EF3.

that tornado pulled a smithville...digging up the ground and debarking trees immediately after touching down.

 

[atrainguy]

that tornado pulled a smithville...digging up the ground and debarking trees immediately after touching down.

Just looked it up on Twitter after seeing your post, and yep, that thing sure did tear up that field.

 

[tornado examiner]

pretty nuts. it also seemed to have lost most of it's strength very quickly too...never again reaching its former power

 

[Shakespeare 2016]

pretty nuts. it also seemed to have lost most of it's strength very quickly too...never again reaching its former power

It may have been stronger but some people suggest that maybe the tornado was able to scour that area easily and a violent tornado isn't required.

 

[tornado examiner]

eh...

 

[Fred Gossage]

Looks like large swaths of the moderate risk area were devoid of severe reports.
While certainly not a busted event, I definitely expected higher-end supercell behavior and consistent severe weather/tornado production from at least 1 or 2 storms, something like the storm of the day on 12/10/21, storms A and B on 5/3/99, every storm on 4/27/11, etc.

Besides the two EF3s, the EF2 that killed folks north of Montgomery, and the Lowndes County, MS EF2, we also have an EF2 southeast of Bay Springs, MS in Jasper County. That's five strong tornadoes already with surveys still ongoing, and that also doesn't count the weaker tornadoes that add to the total count. I think it's a matter of us needing to readjust our expectations some to what verified but not under-forecasted Moderate Risk days are really supposed to be like. For the most part, they shouldn't quite achieve violent tornadoes like any of the individual storms you mentioned (although that does happen in more than isolated cases for sure). The SPC risk system is completely probabilistically driven now and has been for several years, but back when it was reports-driven, for an area the size of roughly Oklahoma without the panhandle, the criteria for a Moderate Risk was 6 to 19 tornadoes. That isn't how they verify the risks anymore, but the probabilistic risks now are supposed to be smoothly comparable to historic risks from the reports-driven era (a MDT now should = a MDT from the 90s, etc). A lot of people don't realize it, but while the MDT on 12/10/21 was a good call because of the question marks ahead of time, it firmly would've verified a High Risk, both in strong/violent count and in total tornado count. Even the outbreak back on November 4th, for the concentrated area impacted, it would've firmly verified a High Risk as well.

 

[Shakespeare 2016]

Besides the two EF3s, the EF2 that killed folks north of Montgomery, and the Lowndes County, MS EF2, we also have an EF2 southeast of Bay Springs, MS in Jasper County. That's five strong tornadoes already with surveys still ongoing, and that also doesn't count the weaker tornadoes that add to the total count. I think it's a matter of us needing to readjust our expectations some to what verified but not under-forecasted Moderate Risk days are really supposed to be like. For the most part, they shouldn't quite achieve violent tornadoes like any of the individual storms you mentioned (although that does happen in more than isolated cases for sure). The SPC risk system is completely probabilistically driven now and has been for several years, but back when it was reports-driven, for an area the size of roughly Oklahoma without the panhandle, the criteria for a Moderate Risk was 6 to 19 tornadoes. That isn't how they verify the risks anymore, but the probabilistic risks now are supposed to be smoothly comparable to historic risks from the reports-driven era (a MDT now should = a MDT from the 90s, etc). A lot of people don't realize it, but while the MDT on 12/10/21 was a good call because of the question marks ahead of time, it firmly would've verified a High Risk, both in strong/violent count and in total tornado count. Even the outbreak back on November 4th, for the concentrated area impacted, it would've firmly verified a High Risk as well.

A lot of moderate risks have overperformed when it comes to violent tornadoes although. Like March 13, 1990, and December 10th-11th 2021 for example etc.

 

[Fred Gossage]

A lot of moderate risks have overperformed when it comes to violent tornadoes although. Like March 13, 1990, and December 10th-11th 2021 for example etc.

And by very nature of them overperforming, that means that they have done more than what you should expect from such a risk day, as determined by how the whole system was designed to operate. It doesn't mean that's the caliber of event that the Moderate Risk is designed to depict. That means, for whatever reason (justified or otherwise), the event was actually under forecast in comparison to the actual results of the event.

 

[CheeselandSkies]

Thanks, @Fred Gossage . I didn't realize they'd changed the criteria, I still remember the 6-19 tornadoes for a moderate risk, 20+ for a high risk in an area of Oklahoma w/o the Panhandle from reading the SPC site in 2003 or so.

While certainly a destructive and dangerous event, something was clearly keeping a leash on the storms to some extent. Do you think it was the weak low level lapse rates some mentioned ahead of time, or something else?

 

[Sawmaster]

As I tell folks "when you're getting hit by a tornado, you don't know how strong it will be till it's over". Occasional under-performance and over-performance is normal, and what happened here is pretty close to what was forecast. Something stronger was possible but highly unlikely, but a few EF-2 and EF-3 tornadoes were likely.

 

[Fred Gossage]

Thanks, @Fred Gossage . I didn't realize they'd changed the criteria, I still remember the 6-19 tornadoes for a moderate risk, 20+ for a high risk in an area of Oklahoma w/o the Panhandle from reading the SPC site in 2003 or so.

While certainly a destructive and dangerous event, something was clearly keeping a leash on the storms to some extent. Do you think it was the weak low level lapse rates some mentioned ahead of time, or something else?

Probably, and that even the mid-level lapse rates didn't take long to get worked over. There's also the issue that we went almost the full first half of the event with hodographs on the area VWPs a good bit smaller than most model guidance had suggested. There's also the part where the event was primary directly tied with the subtropical jet, and that almost always means an event that doesn't end up being quite as bad as it otherwise could be, even if the forecasts for it are on point in the end.

As far as that criteria with the risks goes, it's not really that they changed it. Outlooks from now are supposed to smoothly be comparable to outlooks from back when that's how verification was driven (similar to how F4 should be cross-comparable to EF4 in a perfect world where there's no personal subjectivity, personal interpretation, and personal or office-level biases or politics at play). It's just that that's not what they have in the front of their minds now when they draw up a risk. From the way I understand it, they draw up the individual probabilities first, and then they match the categories to that. Having said that, I'm sure they often have in mind what categories they want to have on the board and where before they ever begin drawing out the probabilistic lines since it's ultimately the categorical risk areas that are the public face of the outlooks.

 

[Fred Gossage]

Besides the two EF3s, the EF2 that killed folks north of Montgomery, and the Lowndes County, MS EF2, we also have an EF2 southeast of Bay Springs, MS in Jasper County. That's five strong tornadoes already with surveys still ongoing, and that also doesn't count the weaker tornadoes that add to the total count. I think it's a matter of us needing to readjust our expectations some to what verified but not under-forecasted Moderate Risk days are really supposed to be like. For the most part, they shouldn't quite achieve violent tornadoes like any of the individual storms you mentioned (although that does happen in more than isolated cases for sure). The SPC risk system is completely probabilistically driven now and has been for several years, but back when it was reports-driven, for an area the size of roughly Oklahoma without the panhandle, the criteria for a Moderate Risk was 6 to 19 tornadoes. That isn't how they verify the risks anymore, but the probabilistic risks now are supposed to be smoothly comparable to historic risks from the reports-driven era (a MDT now should = a MDT from the 90s, etc). A lot of people don't realize it, but while the MDT on 12/10/21 was a good call because of the question marks ahead of time, it firmly would've verified a High Risk, both in strong/violent count and in total tornado count. Even the outbreak back on November 4th, for the concentrated area impacted, it would've firmly verified a High Risk as well.

We can honestly say the same thing about High Risk days too. A lot of heifers out here on this bird app and on this Facebag think that if the event isn't worse than 4/27/2011 and has a massive fatality count, then a High Risk wasn't needed, even if the event had 40 or 50-something tornadoes and like 7 or 8 of them were strong. Some of these over-performing MDT Risk days have skewed our perception a bit of what a MDT Risk was really designed to cover, just like 4/27/11 skewed a lot of people's perception (even in the degreed met community) of what verifies a High Risk or what constitutes a "tornado outbreak". I don't think any of it is truly intentional. I think it's a byproduct of going through some of the over-performing events.

We shouldn't be walking into a Moderate Risk day thinking another 12/10/21 is the flavor of what we're expecting. We should be walking out of 12/10/21 saying "Yeah, I get why SPC couldn't do it beforehand because of the uncertainty, but holy crap, that would've actually verified a High Risk instead."

 

[Justin Hindman]

I’ve always believed if it’s a moderate risk, I prepare for a high impact day just like a high risk. The weird thing to me with this system is the spacing of storms this time of the year. I don’t waste my time on moderate and high risk bc it really doesn’t matter, it’s going to be a bad day for someone. Just being the first one on the scene for the Caledonia tornado and seeing people crawling out of their houses and walking around is enough for me.

 

[Equus]

Moderate is certainly scary enough, even a 10 hatch enhanced should be enough to get people's attention. It's not super intuitive probably for the public to think about the risk levels probabilistically instead of just assuming higher probs = more/stronger tornadoes so I can see why some hold moderate risks to way higher a threshold/standard than they should. Also definitely agreed events like 4/27 and 12/10 (Easter 2020 really) horribly skew the perception of what constitutes that risk level when in reality most big time red letter days are dramatic overperformers instead of the norm. This one certainly had potential to be so much worse but as the surveys come in it's clear that it was certainly a worthy moderate risk tornado outbreak by the standards we should be assuming risk levels are meant to cover

 

[CheeselandSkies]

Probably, and that even the mid-level lapse rates didn't take long to get worked over. There's also the issue that we went almost the full first half of the event with hodographs on the area VWPs a good bit smaller than most model guidance had suggested. There's also the part where the event was primary directly tied with the subtropical jet, and that almost always means an event that doesn't end up being quite as bad as it otherwise could be, even if the forecasts for it are on point in the end.

Now that is interesting. We've also had some overperforming events where the crucial factor was the hodographs verifying larger than guidance had suggested. Of course models aren't going to be perfect, but that seems like a pretty fundamental thing to get wrong especially at short range, and it seems to happen fairly often.

 

[Equus]

 

[Equus]

Interesting thing about the scouring tornado in Louisiana is that the worst damage occurred the moment it touched down and it didn't come anywhere close to that strong ever again, really wish that had been caught on video because that must have been fascinating (and terrifying)