• A friendly and periodic reminder of the rules we use for fostering high SNR and quality conversation and interaction at Stormtrack: Forum rules

    P.S. - Nothing specific happened to prompt this message! No one is in trouble, there are no flame wars in effect, nor any inappropriate conversation ongoing. This is being posted sitewide as a casual refresher.

La Nina and DJF tornadoes. Early Prelude to 2022 seasonal forecast?

adlyons

EF2
Joined
Feb 16, 2014
Messages
167
Location
Norman, Oklahoma
I found an interesting paper by several well-known seasonal tornado researchers that may be relevant to the coming months. Im sure most remember the headlines hyping 2021 as an active year due to La Nina. I can bet ALL of you remember how that turned out :). What interests me this year is that La Nina is already behaving like La Nina whereas last year's event seemed to lack a more traditional response over the CONUS. I recommend an excellent article by Dr. John Allen on this very topic. La Nina 2021 A quick summary, 2021 La Nina failed to show up in the ways we expect and terminated early which may have contributed to a wonky tornado season. For visualization:

2021 vs a typical La Nina pattern 500mb anomalies.
Screen Shot 2021-12-01 at 11.55.49 AM.png
1638381773182.png

We have regressed back to La Nina for the second time in 2021 and forecasts continue to show a moderate event developing through Winter 21-22. In most respects, this event has already behaved more prototypically with an abundance of cooler and wetter conditions along the West Coast and BC. We have also developed the typical cool anomalies off the coast of Alaska helping to bolster above-average ridging over the Pacific. This ridging is important for westerly jet extensions and eventual trough development. So how does this tie into the paper I linked and possibly have implications for spring 2022?

The paper focused on the role that ENSO state and Gulf of Mexico SST anomalies have on wintertime (DJF) tornadoes in the CONUS. To quickly summarize the findings;
  • EF1+ Tornadoes are more common in La Nina and El Nino than in neutral conditions and significant tornadoes make up a larger fraction of tornadoes in DJF.
  • Gulf of Mexico SST anomalies independently contributes to seasonal variability at least through DJF by modulating surface moisture fluxes and buoyancy.
  • SST anomalies in the eastern Pacific are also contributing to the variability and are more readily influenced by ENSO state.
While njot necessarily surprising, this shows us that ENSO state and SST temps in the GOM and E Pacific are significant drivers of wintertime tornado variability. As ENSO weakens in the spring things get a bit more muddled but we can use this (perhaps) as a proxy to early season tornado potential.
So let's compare these conclusions to some current data.

Starting with ENSO State, we are currently under a La Nina advisory with strengthening towards moderate conditions through the coming months. We know things are more or less behaving as expected atmospherically speaking. The research mentioned above suggests that tornado potential in the SE is increased when typically La Nina conditions are present.
1638382833701.png

SSTs The GOM is warm, with average anomalies on the order of +0.5C. This again hints at some potential in the coming months that the 2022 season could start faster.


1638383065173.png
1638383096278.png

Something else to point out with the SSTs is the cool bubble in the ePAC. This is again expected with La Nina and shows up quite nicely on the current anomalies. It is also a contributor to more significant tornado activity across the CONUS.
1638383209121.png
1638383457835.png


So what does this mean? Well not a whole lot if 2021 shakes out again. However, if La Nina conditions persist and behave accordingly, the warm Gulf and expected weather pattern resulting from La Nina may contribute to a faster start to the 2022 tornado season and a mnore active witner across the SE US. There may also be some implications for spring 2022. La Nina patterns favor the southern Plains and Midwest.

1638383759430.png
While fast starts are not always indicators of future performance, it never hurts to get on the board early as opposed to playing catchup with the averages. Most of the years that do start fast hang around average longer or finish above average as opposed to slow seasons.
1638383918745.png
What do you think 2022 holds given the current patterns? Ill be verifying my 2021 forecast and issuing a full 2022 season forecast here in the coming weeks as we wrap up the year.
 
Despite how overall poor 2021 was for plains chasing, Midwest chasers actually had probably one of the better years in the last several with several regional outbreaks occurring from Iowa to Ohio from June to August. And if we are being completely honest, the overall state of the season in the plains wasn't exactly 2020 terrible. There were plenty of good storms to get on that produced nice tornadoes, it was just a matter of being there (4/23 Lockett TX, 3/13 Happy TX, 4/27 Benjamin TX, 3/27 Carthage TX, 5/3 in Northern TX, May 16-18 in Southwest/West Texas, May 22-26 stretch in CO/KS/TX/SD, and then memorial day weekend as well). Lot of marginal days that ended up panning out better than high end forecast days.

July and August being particularly active with 2 regional tornado outbreaks occurring in the Corn Belt and several other significant events for summer time (NWS La Crosse had their strongest tornado in 18 years on Aug 7th). Hopefully we can see an earlier start in 2022 versus seeing snow up until the last few days of April like we did in 2021 (Heavy wet snow in the last week of April and power outages lol). I can't remember the last time I saw a decent tornado in April lol.
 
La Nina can be such a mixed bag. It almost always means something dangerous in Dixie early but as far as the Plains/Midwest in A/M/J go, we've had both significantly above and significantly below-average tornado activity in La Nina years. I honestly don't know what to think anymore.
I think you put it best that theres more to the story for spring. Seasonal forecasting is hard! Im hoping that as more and more folks get into these long range forecasts we start to pull some more info out. Dynamical downscalling of climate and synoptic models has shown some potential but its still a crap shoot at this point.
 
Despite how overall poor 2021 was for plains chasing, Midwest chasers actually had probably one of the better years in the last several with several regional outbreaks occurring from Iowa to Ohio from June to August. And if we are being completely honest, the overall state of the season in the plains wasn't exactly 2020 terrible. There were plenty of good storms to get on that produced nice tornadoes, it was just a matter of being there (4/23 Lockett TX, 3/13 Happy TX, 4/27 Benjamin TX, 3/27 Carthage TX, 5/3 in Northern TX, May 16-18 in Southwest/West Texas, May 22-26 stretch in CO/KS/TX/SD, and then memorial day weekend as well). Lot of marginal days that ended up panning out better than high end forecast days.

July and August being particularly active with 2 regional tornado outbreaks occurring in the Corn Belt and several other significant events for summer time (NWS La Crosse had their strongest tornado in 18 years on Aug 7th). Hopefully we can see an earlier start in 2022 versus seeing snow up until the last few days of April like we did in 2021 (Heavy wet snow in the last week of April and power outages lol). I can't remember the last time I saw a decent tornado in April lol.

I certainly agree that this year was no where near as bad as 2020 "chasability" wise thankfully. Like you said, there were days smattered about that did pan out quite well. The late summer events really have been showing up with more frequency in recent years. August and July over performing in the Midwest. Not sure i can remember my last april tornado either ha!
 
Despite how overall poor 2021 was for plains chasing, Midwest chasers actually had probably one of the better years in the last several with several regional outbreaks occurring from Iowa to Ohio from June to August. And if we are being completely honest, the overall state of the season in the plains wasn't exactly 2020 terrible. There were plenty of good storms to get on that produced nice tornadoes, it was just a matter of being there (4/23 Lockett TX, 3/13 Happy TX, 4/27 Benjamin TX, 3/27 Carthage TX, 5/3 in Northern TX, May 16-18 in Southwest/West Texas, May 22-26 stretch in CO/KS/TX/SD, and then memorial day weekend as well). Lot of marginal days that ended up panning out better than high end forecast days.

July and August being particularly active with 2 regional tornado outbreaks occurring in the Corn Belt and several other significant events for summer time (NWS La Crosse had their strongest tornado in 18 years on Aug 7th). Hopefully we can see an earlier start in 2022 versus seeing snow up until the last few days of April like we did in 2021 (Heavy wet snow in the last week of April and power outages lol). I can't remember the last time I saw a decent tornado in April lol.

Possibly Rochelle? At least, that's the last time this region had a significant one.

In 2021 I managed to see tornadoes on both July 14th in Iowa and August 9th in Illinois, although of course I screwed things up both days and didn't see nearly as much as I could have. I also probably could have seen the Mineral Point, WI tornado of August 11th had I just been a little more aggressive about getting on the storm quicker.
 
Possibly Rochelle? At least, that's the last time this region had a significant one.

In 2021 I managed to see tornadoes on both July 14th in Iowa and August 9th in Illinois, although of course I screwed things up both days and didn't see nearly as much as I could have. I also probably could have seen the Mineral Point, WI tornado of August 11th had I just been a little more aggressive about getting on the storm quicker.

Off topic from this thread, but I missed that one. I was more referring to my personal chase stats. Drove 2100 miles in 2 days and did not see a single storm lol.... hell I don't even think we saw any lightning despite chasing both days (KS/OK on 4/8 and IL on 4/9). An embarrassing stretch for me to talk about. Last April tornado I saw was April 30, 2019 personally now that I've looked.

2021 was probably my better chase season since 2015. I managed to see tornadoes on March 13th (Happy, TX); May 3rd (Central IL); May 17th Sterling City TX and May 18th we saw a few brief spin ups in SW TX along with great structure; May 26th in NW KS/SW NE (briefly from a distance one of the SW NE tornadoes); June 18th about 2 miles from my house I observed a strong tornado around midnight that sideswiped our home with 100 mph RFD winds; June 20th in Pella, IA; June 25th in Central IL and then again on June 26th in Central IL; July 10-11 in Central IL on several MCVs; Aug 7th in SW WI (barely got a visual before losing my back windows in extreme RFD near that EF3); Aug 9th IL tornadofest; Aug 11th Central/Southern WI tornadoes; August 25th I observed a landspout a few miles from my house; October 11th in IL and then a brief spin up on October 24th.
 
Always great to see papers coming out of this kind of severe-focused S2S work. This type of research is definitely accelerating and receiving increased funding in recent years. Hopefully, within the next decade, the science will be able to say something qualitatively useful about the subsequent spring season from this lead time.

For now, although we have an inkling of how the scales are tipped for upcoming DJF activity over the southeast U.S., we are IMO still largely in voodoo territory when it comes to calling AMJ on the Plains in December 😆. A couple voodoo comments on that subject:

First, the ENSO plumes Andrew posted unfortunately follow the classic "weakly cool winter, warming to neutral through the spring" progression from many of our worst seasons in recent decades. Here is a plot of January-June Nino 3.4 SSTAs, color coded by my objective chase season scores (basically, MAMJ tornado activity over the S and C Plains) where blue lines are active chase seasons and red are inactive:

weekly_all.png

There hasn't been a single highly active spring on the Plains with this ENSO progression since at least 1990, and most of the real clunkers had it. In fairness: if you go back pre-1990, where I can only find more temporally smoothed data, this trend is considerably less pronounced. And strangely enough, 1965 (likely the overall most active Plains season since mid-last century) did have a weakly-cool-to-neutral progression. But, again: this is a lot of voodoo. There's so much of relevance going on outside of the mean SSTs within one ENSO sub-region, not to mention the possibility that AGW and longer-timescale phenomena are changing how ENSO affects what we care about.

Second, this year we've seen a sustained, strongly negative PDO (cool SSTs immediately offshore in the Pac NW and Gulf of Alaska) for the first time since the 2008-2012 period. This is associated with at least a weakly favorable signal for Plains activity, assuming it persists into next spring.

pdo.png

Just for fun: the combination of weakly cool ENSO with strongly negative PDO was also seen going into spring 2012. Notable SSTA differences include a warmer Gulf of Mexico and midlatitude eastern Pacific this year (the latter not surprising, as we were coming out of a stronger and longer-lasting Nina event in 2011).

ssta.daily.current.png
crw_oper50km_monthlymean_sstanom_201112_browse.gif
 
Always great to see papers coming out of this kind of severe-focused S2S work. This type of research is definitely accelerating and receiving increased funding in recent years. Hopefully, within the next decade, the science will be able to say something qualitatively useful about the subsequent spring season from this lead time.

For now, although we have an inkling of how the scales are tipped for upcoming DJF activity over the southeast U.S., we are IMO still largely in voodoo territory when it comes to calling AMJ on the Plains in December 😆. A couple voodoo comments on that subject:

First, the ENSO plumes Andrew posted unfortunately follow the classic "weakly cool winter, warming to neutral through the spring" progression from many of our worst seasons in recent decades. Here is a plot of January-June Nino 3.4 SSTAs, color coded by my objective chase season scores (basically, MAMJ tornado activity over the S and C Plains) where blue lines are active chase seasons and red are inactive:

View attachment 22367

There hasn't been a single highly active spring on the Plains with this ENSO progression since at least 1990, and most of the real clunkers had it. In fairness: if you go back pre-1990, where I can only find more temporally smoothed data, this trend is considerably less pronounced. And strangely enough, 1965 (likely the overall most active Plains season since mid-last century) did have a weakly-cool-to-neutral progression. But, again: this is a lot of voodoo. There's so much of relevance going on outside of the mean SSTs within one ENSO sub-region, not to mention the possibility that AGW and longer-timescale phenomena are changing how ENSO affects what we care about.

Second, this year we've seen a sustained, strongly negative PDO (cool SSTs immediately offshore in the Pac NW and Gulf of Alaska) for the first time since the 2008-2012 period. This is associated with at least a weakly favorable signal for Plains activity, assuming it persists into next spring.

View attachment 22368

Just for fun: the combination of weakly cool ENSO with strongly negative PDO was also seen going into spring 2012. Notable SSTA differences include a warmer Gulf of Mexico and midlatitude eastern Pacific this year (the latter not surprising, as we were coming out of a stronger and longer-lasting Nina event in 2011).

View attachment 22369
View attachment 22370
Brett this is AWESOME man thanks for posting all of this! I love your plots and chase season ranking metrics. I've tried to use some of that in my personal assessments and experiments as well. I've noted the same issues you have described with the early decaying La Nina and poor Plains years. Seems like the only time la nina really contributes is when it is strong and persistant. What is really interesting is the clustering around neutral conditions. Stagnant neutral conditions aren't all that common but seem to really favor active years.

Im curious about the 2012 analog, I see a lot of similarities. The strong - PDO is somewhat encouraging as you said. If I recall the problem with 2012 ended up being the lack of better troughs and periods of northwest flow through May.
 
Second, this year we've seen a sustained, strongly negative PDO (cool SSTs immediately offshore in the Pac NW and Gulf of Alaska) for the first time since the 2008-2012 period. This is associated with at least a weakly favorable signal for Plains activity, assuming it persists into next spring.

View attachment 22368

Just for fun: the combination of weakly cool ENSO with strongly negative PDO was also seen going into spring 2012. Notable SSTA differences include a warmer Gulf of Mexico and midlatitude eastern Pacific this year (the latter not surprising, as we were coming out of a stronger and longer-lasting Nina event in 2011).

View attachment 22369
View attachment 22370

Brett, does the science field know what the physical mechanism is by which PDO impacts planetary-scale/synoptic-scale patterns? Does a -PDO promote more troughiness over the Gulf of Alaska/NE Pacific and so on and so forth through Rossby wave teleconnections? Or is it something else?

Also, don't forget about ERTAF. It wasn't conducted in 2021, but it might come back for 2022. I don't totally understand the long-range prediction of tornado statistics, but I suspect that, for the near-future, it will remain difficult to predict seasonal tornado counts in any domain smaller than the CONUS beyond a few weeks. (Predictable...in a sense) tornadoes are a very small-scale phenomenon which are controlled by a cascade of phenomena on increasingly larger scales. Without using a proxy (i.e., assuming that correctly-forecast synoptic scale features will necessarily berth the correct mesoscale environment which will automatically berth storms that will definitely produce tornadoes), it's just very difficult to correctly predict a tornado at any length, but it gets worse the further out in time you go.
 
Back
Top