Severe Weather Threat - November 29th-30th, 2022 (4 Viewers)

Fred Gossage

Member
Meteorologist
PerryW Project Supporter
Messages
593
Reaction score
2,057
Location
Florence, AL
I'm counting six strong/significant tornadoes in the Damage Assessment Toolkit so far:

- EF3 in Caldwell Parish, LA
- EF3 near Tibbie, AL in Washington County
- EF2 near the Flatwoods area in Montgomery/Elmore Counties of AL
- EF2 in southeastern Elmore County, AL separated from the Flatwoods track as a different tornado
- EF2 near Steens, MS in Lowndes County
- EF2 southeast of Bay Springs, MS in Jasper County
 

xJownage

Member
Messages
506
Reaction score
364
Location
Jacksonville, Florida
I've been in discussion with several people on other platforms regarding why the event was not a higher end event. This was most certainly a verified MDT, and anybody suggesting otherwise has very unrealistic expectations, but an environment of this caliber had a much higher ceiling than what actually happened. We had seven discrete supercells over an open warm sector under well vented, curved hodographs with significant low level instability. One of the things I noticed was that certain locations verified better than others. When discussing lapse rates, for example, lapse rates weren't necessarily better in C MS and closer to the MS/AL border, which is where storms were much more likely to produce tornadoes than in LA and western MS. The big question to ask here is why? Synoptically and climatologically speaking, the area right around Yazoo City and just west of there was where the hotspot should've been, but we didn't see that. I've taken a bit of a dive into what happened and want to bring up a few points on potential failure modes for this part of the risk area, and what consequently prevented this event from realizing its ceiling of a high end event.

A couple factors have already been mentioned on this thread by @Fred Gossage, being the LLLRs and the smaller than expected hodographs earlier in the event. I have a couple things to add to each one.

First of all, on the topic of LLLRs, these usually aren't a big deal in dixie. You usually have such deep moisture that LLLRs don't have much of an effect on storms, and even moreso, you'll have so much deep layer moisture that it's not even possible to get the higher lapse rates you see in the plains, especially when your LCLs are expected to be basically ground scraping. To provide an example of this, here is the 18Z BMX sounding from 3/25/21. Notice how poor the lapse rates are? This is normal for low LCL dixie setups. Now, what made the bad LLLRs more impactful was that there was more dry air aloft than what was forecast. Here's the JAN (Jackson) sounding for 11/29. This dry air ended up mixing and slightly raising the LCL. Raised LCLs meant that the weak LLLRs became much more significant than how they were in the forecast. 11/4 had garbage LLLRs but still performed well. In addition, you expect storms to take a while to get going in an environment characterized by poor lapse rates, which is also what we saw on 11/4 (sidenote, I know 11/4 and this event weren't even remotely comparable in terms of synoptic setup, I'm just using it as a lapse rate example) and what was noted in the SPC's mesoscale discussion for the PDS watch.

With regards to the hodographs, model guidance showed the LLJ and 700mb winds intensifying significantly around 22Z-23Z. The intensification of both features was about 2 hours late, which added a litany of problems, particularly in W MS into LA. This could be seen on the VWPs during the event. Here's the VWP at 2347Z in Jackson. Looking at this, our 850mb winds were about 10kts weaker than forecast, and the 700mb winds are about 20kts weaker than forecast. This resulted in much smaller hodographs than forecast for the early portion of the event. Particularly, this was a significant issue for W MS because training cells were leaving large cold pools, and by the time the hodograph was improving, the low level thermals were trash due to cold pooling. With regards to the hodograph, however, 200+ SRH still should result in at least some tornadoes throughout the early part of the event in W MS, given the thermodynamics in place with this event. Instead, what happened was weird: we had several vigorously rotating supercells that just refused to produce tornadoes. The storm that passed near Yazoo City near sundown tracked over a heavily sheared and thermodynamically supported environment for 3+ hours, and even saw improved hodographs during that timespan, but never produced until the very end of its life cycle...while embedded in a rain shield. This storm in particular was a mystery to me and a few others.

Now, there's two other factors that I think were at play here. This is based on my own coverage and talking to a couple other mets who were kind enough to share their personal thoughts with me.

Let's go to the 00Z JAN sounding for 11/29. If you didn't look closely, you may not notice this, but there's just a small inversion right near the surface. I'm not *exactly* sure what caused this, but it was likely the LLJ that was so prevalent for such a large portion of the day causing just enough downward mixing to cause an inversion. Usually, dixie events don't care about this; strong moisture convergence will easily break that inversion. This setup was different however, as we had much weaker moisture convergence than normal. This led to a much more discrete storm mode, but the lacking moisture convergence probably resulted in storms struggling to break that low-level inversion. I don't have any maps to verify, unfortunately, but SFC moisture convergence was in the range of 15-20 g/(kgs) according to a meteorologist I was discussing this with, which is about half that of what we see during typical dixie events. This failure mode is typically more of a plains failure mode rather than a dixie one, but this very unique setup caused a lot of stark differences when compared to normal dixie events.

One of the observations I made during the event was that while storms appeared normal on radar, all the images I saw showed shelf clouds or other outflow-dominant features. Why would storms be outflow dominant? Looking back and the previous 00Z JAN sounding, you have 932 DCAPE. This is a pretty normal value for a plains setup, but this is relatively high for a dixie setup with low cape. Compare this with the value of 532 on the 3/25/21 BMX sounding and you see what I mean. Downdraft Cape, in my experience, is generally balanced by SR Inflow strength. For an example of this, look back to 5/2/22 this year, in which there was a monster discrete supercell in SW/W OK that never produced a tornado until after dark east of OKC (produced the first of the two Seminole, OK EF-2s this may). Chasers reported that it "basically had zero RFD". DCAPE was as high as 800 that day, but when your SR inflow is extremely strong, you won't get strong RFDs that way. SR inflow was much weaker on 11/29, and 900+ DCAPE probably caused storms to be outflow dominant. This is also verified by the chasers I was talking to on the ground, who reported cold, wet RFDs and large cold pools around the storms they chased.

Not only did this cause the original storms to be outflow dominant, it also caused cold pools to be much larger than forecast. By 02-03Z, basically all of N MS was a giant cold pool from the previous storms, and it was plainly visible on mesoanalysis. The OFD issue, whether caused by DCAPE or other factors, further hindered the thermos for any storm passing generally over the path of a previous one.

-----------------------------

As I mentioned in the beginning, it would be silly to call this event a bust. We had at MINIMUM 6 sigtors, posssibly more. However, this was an event with an extremely high ceiling that didn't reach it's potential, and I personally think there's several factors as to why. I'd love to hear some feedback from you guys. Do you think one or more of these issues were more significant than others? Or do you agree with me and believe it was all of these issues that prevented this event from reaching it's ceiling? Let me know what you guys think. I love doing post-event analysis and deep dives like this. It truly is the best way for all of us to learn.
 
Last edited:

UK_EF4

Member
Messages
166
Reaction score
354
Location
NW London
I've been in discussion with several people on other platforms regarding why the event was not a higher end event. This was most certainly a verified MDT, and anybody suggesting otherwise has very unrealistic expectations, but an environment of this caliber had a much higher ceiling than what actually happened. We had seven discrete supercells over an open warm sector under well vented, curved hodographs with significant low level instability. One of the things I noticed was that certain locations verified better than others. When discussing lapse rates, for example, lapse rates weren't necessarily better in C MS and closer to the MS/AL border, which is where storms were much more likely to produce tornadoes than in LA and western MS. The big question to ask here is why? Synoptically and climatologically speaking, the area right around Yazoo City and just west of there was where the hotspot should've been, but we didn't see that. I've taken a bit of a dive into what happened and want to bring up a few points on potential failure modes for this part of the risk area, and what consequently prevented this event from realizing its ceiling of a high end event.

A couple factors have already been mentioned on this thread by @Fred Gossage, being the LLLRs and the smaller than expected hodographs earlier in the event. I have a couple things to add to each one.

First of all, on the topic of LLLRs, these usually aren't a big deal in dixie. You usually have such deep moisture that LLLRs don't have much of an effect on storms, and even moreso, you'll have so much deep layer moisture that it's not even possible to get the higher lapse rates you see in the plains, especially when your LCLs are expected to be basically ground scraping. To provide an example of this, here is the 18Z BMX sounding from 3/25/21. Notice how poor the lapse rates are? This is normal for low LCL dixie setups. Now, what made the bad LLLRs more impactful was that there was more dry air aloft than what was forecast. Here's the JAN (Jackson) sounding for 11/29. This dry air ended up mixing and slightly raising the LCL. Raised LCLs meant that the weak LLLRs became much more significant than how they were in the forecast. 11/4 had garbage LLLRs but still performed well. In addition, you expect storms to take a while to get going in an environment characterized by poor lapse rates, which is also what we saw on 11/4 (sidenote, I know 11/4 and this event weren't even remotely comparable in terms of synoptic setup, I'm just using it as a lapse rate example) and what was noted in the SPC's mesoscale discussion for the PDS watch.

With regards to the hodographs, model guidance showed the LLJ and 700mb winds intensifying significantly around 22Z-23Z. The intensification of both features was about 2 hours late, which added a litany of problems, particularly in W MS into LA. This could be seen on the VWPs during the event. Here's the VWP at 2347Z in Jackson. Looking at this, our 850mb winds were about 10kts weaker than forecast, and the 700mb winds are about 20kts weaker than forecast. This resulted in much smaller hodographs than forecast for the early portion of the event. Particularly, this was a significant issue for W MS because training cells were leaving large cold pools, and by the time the hodograph was improving, the low level thermals were trash due to cold pooling. With regards to the hodograph, however, 200+ SRH still should result in at least some tornadoes throughout the early part of the event in W MS, given the thermodynamics in place with this event. Instead, what happened was weird: we had several vigorously rotating supercells that just refused to produce tornadoes. The storm that passed near Yazoo City near sundown tracked over a heavily sheared and thermodynamically supported environment for 3+ hours, and even saw improved hodographs during that timespan, but never produced until the very end of its life cycle...while embedded in a rain shield. This storm in particular was a mystery to me and a few others.

Now, there's two other factors that I think were at play here. This is based on my own coverage and talking to a couple other mets who were kind enough to share their personal thoughts with me.

Let's go to the 00Z JAN sounding for 11/29. If you didn't look closely, you may not notice this, but there's just a small inversion right near the surface. I'm not *exactly* sure what caused this, but it was likely the LLJ that was so prevalent for such a large portion of the day causing just enough downward mixing to cause an inversion. Usually, dixie events don't care about this; strong moisture convergence will easily break that inversion. This setup was different however, as we had much weaker moisture convergence than normal. This led to a much more discrete storm mode, but the lacking moisture convergence probably resulted in storms struggling to break that low-level inversion. I don't have any maps to verify, unfortunately, but SFC moisture convergence was in the range of 15-20 g/(kgs) according to a meteorologist I was discussing this with, which is about half that of what we see during typical dixie events. This failure mode is typically more of a plains failure mode rather than a dixie one, but this very unique setup caused a lot of stark differences when compared to normal dixie events.

One of the observations I made during the event was that while storms appeared normal on radar, all the images I saw showed shelf clouds or other outflow-dominant features. Why would storms be outflow dominant? Looking back and the previous 00Z JAN sounding, you have 932 DCAPE. This is a pretty normal value for a plains setup, but this is relatively high for a dixie setup with low cape. Compare this with the value of 532 on the 3/25/21 BMX sounding and you see what I mean. Downdraft Cape, in my experience, is generally balanced by SR Inflow strength. For an example of this, look back to 5/2/22 this year, in which there was a monster discrete supercell in SW/W OK that never produced a tornado until after dark east of OKC (produced the first of the two Seminole, OK EF-2s this may). Chasers reported that it "basically had zero RFD". DCAPE was as high as 800 that day, but when your SR inflow is extremely strong, you won't get strong RFDs that way. SR inflow was much weaker on 11/29, and 900+ DCAPE probably caused storms to be outflow dominant. This is also verified by the chasers I was talking to on the ground, who reported cold, wet RFDs and large cold pools around the storms they chased.

Not only did this cause the original storms to be outflow dominant, it also caused cold pools to be much larger than forecast. By 02-03Z, basically all of N MS was a giant cold pool from the previous storms, and it was plainly visible on mesoanalysis. The OFD issue, whether caused by DCAPE or other factors, further hindered the thermos for any storm passing generally over the path of a previous one.

-----------------------------

As I mentioned in the beginning, it would be silly to call this event a bust. We had at MINIMUM 6 sigtors, posssibly more. However, this was an event with an extremely high ceiling that didn't reach it's potential, and I personally think there's several factors as to why. I'd love to hear some feedback from you guys. Do you think one or more of these issues were more significant than others? Or do you agree with me and believe it was all of these issues that prevented this event from reaching it's ceiling? Let me know what you guys think. I love doing post-event analysis and deep dives like this. It truly is the best way for all of us to learn.
Good thread... I think the fact large hail was apparently more prevalent than forecast is quite telling, indicating that in some form low level shear was weaker (based on Cameron Nixon's research). Whether that was by storms being slightly elevated, weaker LLJ and associated smaller hodographs, or a more NE as opposed to E storm motion which would have reduced SRH... or all of them, I don't really know, but certainly interesting and very lucky things didn't reach full potential.
 

xJownage

Member
Messages
506
Reaction score
364
Location
Jacksonville, Florida
Good thread... I think the fact large hail was apparently more prevalent than forecast is quite telling, indicating that in some form low level shear was weaker (based on Cameron Nixon's research). Whether that was by storms being slightly elevated, weaker LLJ and associated smaller hodographs, or a more NE as opposed to E storm motion which would have reduced SRH... or all of them, I don't really know, but certainly interesting and very lucky things didn't reach full potential.
Yeah. That VWP screenshot I took tells its own story. That looks a lot more like a hail hodo than a tor hodo. NE/E storm motion isn't something I caught, but good point, motion was a bit more NE than forecast and not a lot of storms took a more deviant E motion which probably resulted in them ingesting less streamwise vorticity than forecast.
 

Fred Gossage

Member
Meteorologist
PerryW Project Supporter
Messages
593
Reaction score
2,057
Location
Florence, AL
I'm counting six strong/significant tornadoes in the Damage Assessment Toolkit so far:

- EF3 in Caldwell Parish, LA
- EF3 near Tibbie, AL in Washington County
- EF2 near the Flatwoods area in Montgomery/Elmore Counties of AL
- EF2 in southeastern Elmore County, AL separated from the Flatwoods track as a different tornado
- EF2 near Steens, MS in Lowndes County
- EF2 southeast of Bay Springs, MS in Jasper County
We can add another short-tracked EF2 from Marion County, MS. That brings the count to seven strong/significant tornadoes from the event so far.
 
Messages
1,958
Reaction score
2,184
Location
Madison, WI
Thanks for that analysis, @xJownage . That's exactly the sort of thing I was looking for. Sure we verified a MDT, but we didn't have that one storm with off-the-charts VROT and a debris ball for hours when it seemed like the forecast environment was one that could easily have supported such.

This meshes well with a recent PM discussion I was having with @Shakespeare 2016 about how finicky forecasting for these things can be, and how the crucial details that distinguish a truly high-end/"historic" tornado setup from one that is merely "significant" may not be apparent until the event is already underway.
 

Users who are viewing this thread

Top