• 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.

2019-04-30 EVENT: OK/KS/AR/MO/IL/IN

Marc Gebhard

Is it too early to start talking about next Tuesday (April 30)? SPC already has a severe risk over most of OK and northern TX. Here's the 0 Z GFS Supercell Composite from the 12 Z (Thurs, April 25) run. It is worth keeping an eye on as we approach next week.


pivotalweather - Models: GFS
Last edited:
As a staff member who watches TA threads carefully, your post is borderline regarding whether it contains acceptable material for a TA thread. Can you elaborate on the nature of the threat in addition to just posting an image that anyone can easily access?
Hi Jeff. Point taken. I won't claim to be an expert at all. I was basically just hoping to get the thread off the ground. Looking a bit deeper, it does appear that SBCAPE (over 3000 in parts of the risk area) and 500 mb shear (35 near DFW and up in the 50 range up near OKC) look to be in place...at least according to the latest GFS, pointing to a significant event...but I realize we are several days out and we will have to wait to see where things go. I'm looking forward to hearing what more experienced forecasters and chasers have to add to the conversation.
Waning 850s that are out of the west are the biggest thing keeping my excitement level down for Tuesday. Of course they try to improve a little bit as the LLJ ramps up, but looks like way too little way too late. GFS and EURO are also widely different with dryline orientation and if the EURO wins out, itll be a sloppy SW-NE orientation which never works out. A severe weather day for sure, but right now I am not too excited for good supercell or tornado potential.
GFS at the dryline bulge: Veered 850mb flow, barely-southerly surface winds, only moderately-favorable lapse rates, the main upper jet streak departing the area early in the day and widespread convection breaking out in the morning from said early energy. Euro: huge moisture-scouring thunderstorm complex roaring through the area in the morning, shunting the moisture southeast to Dallas with little time for afternoon recovery. Maybe a needle-in-the haystack play if a ribbon of recovery moisture can sneak in behind the early-day stuff, but not likely.

Not very optimistic for a trip for this one as it's shown now. Especially since a warm front play might exist in the Midwest.
The threat remains, but I'd agree with a generic SLGT delineation right now. My feeling is that we won't have much clarity until Monday evening, at the earliest, as convective overturning on Tuesday could be a major issue. Still, just looking at the bigger picture, there is not a classic supercell look in the model depictions, given deep shear vectors nearly parallel to a surface boundary draped from southeastern Kansas into central Oklahoma.

Assuming that a massive MCS does not evolve, low-level hodographs may be quite favorable for tornadoes. This is assuming that surface winds remain backed and there is not too much veering of the low-level jet.

Bottom line, I'm not overly impressed at this point, but at least some severe thunderstorms will be probable from the southern Plains into portions of Kansas/Missouri. Will there be discrete, tornadic storms in the warm sector or farther southwest along the front, or more of a large MCS scenario? Stay tuned.
Tuesday looks better than the debacle progged last week; but, I agree with Quincy it's not great. Local chasers and/or people already out there have a shot in two possible areas IMO. Pls reference Quincy's surface chart fcst above...

Missouri has a slim chance of living up to its Show-Me State motto on the warm front. LLJ is forecast to make it up there. However upper winds will not be as robust as farther southwest. Huge MCS rolling down I-70 is other other risk.

I would probably target farther south. Look for (hope for) an outflow boundary south of the synoptic cold front. If not some progs have the air immediately behind the CF not too cold. Perhaps look for its intersection with the DL. That is the triple point in Qincy's chart.

Always good to have a conceptual model going before looking at numerical models. Some VBV is forecast way up high. Should not be a deal breaker. Surface features are key Tuesday IMO.
12z data is rolling in. 3km NAM has been trending more toward linear MCS development over much of the threat area. 12z HRRR is just into range now, so won't focus too heavily on specifics at 30-36 hours, but I will say that it is more aggressive with destabilization across most of Oklahoma and even into the eastern Texas panhandle.

It times like this, it's helpful to look at SREF/HREF and a blend of models. They highlight a fairly broad area under the threat, similar to the overnight Day 2 convective outlook. As @Jeff House mentioned, look for any boundaries left over by convection tomorrow morning. Assuming the front remains largely SW-NE oriented, any boundaries will help offset some of the wind profile concerns.

The triple point is always a possibility. If it's too far south, as some data suggests, weak forcing and relatively modest low-level wind fields may be tough challenges to overcome. Recent data also highlights the potential for modest to moderate destablization behind the cold front, from the eastern Texas panhandle into northwestern Oklahoma.. As we get later into the season, especially in the southern Plains, air-mass recovery behind a cold front becomes easier to achieve.

It's a sit and wait deal. Not a setup I'd advise making a long trip for, especially with convective evolution still being a major wildcard.
The GFS is much more gung-ho with the western area destabilizing by 00Z, but at this time range, I will hang my hat more on the higher-res models. This early in the season it is hard to overcome the AM storms that are sure to occur, plus there is no really good upper or low-level forcing to get redevelopment of storms back to the west. The eastern area is way too messy convective mode wise, plus the chase territory is not the best. So I am sitting this one out.
I went ahead and added IL and IN to the titular portion of the thread, particularly after looking over the 12z/29 runs of the NAM and NAM 3k, along with earlier runs of the GFS and similar trends noted on the 00z/29 ECMWF with what severe weather parameters I was able to view.

Climatologically speaking, with a sharp warm front vacillating somewhere east of the surface low, given the forecast instability, low-level vorticity and ambient mid and low-level lapse rates, there is a conditional threat for supercell tornadoes in the warm sector with any storm that interacts with the frontal boundary and takes advantage of streamwise vorticity in place. This would extend into W IN too, before the main MCS, with an attendant threat for damaging winds, crosses the Mississippi River late evening. This threat appears concentrated in a narrow corridor between I-72 and I-70/I-64, or wherever the front sets up and points south, coincident with the best deep layer shear parameters and whatever instability actually materializes.

CAMs aren't necessarily indicating a plethora of robust convection across IL into IN, but they are indicating some convection will fire INVO a rather notable mid-level perturbation overspreading the region where the warm front will lie draped as the atmosphere destabilizes throughout the morning tomorrow. There's also the threat of lingering convection from TSRA tonight across the region, and as is par for the course, how much and to what extent the warm sector south of the main WF can destabilize tomorrow, especially given the overrunning stream of mid-level perturbations that could continue to initiate broader scale forcing for ascent and keep convective overturning an ongoing caveat.
Analog data highlights two areas with a relative maximum of tornado potential, with one being over eastern Oklahoma and another near the warm front from Missouri into Illinois. Exact placement of the warm front will certainly vary based on model progs and analog graphics such as these give a rough idea of where past, similar setups have performed.

Multiple analogs from the top 15 averaged below and others show a stationary front draped across Oklahoma with a warm front lifting north through MO/IL, similar to what is expected tomorrow. Not all of these events produced, but several of them did.
Currently eyeing the corridor between St. Louis and Springfield for warm front tornado potential. Given that the warm front will be gradually moving north throughout the afternoon and evening and storm motion doesn't look too fast, there should be enough time spent along the front for a potential storm to produce before it eventually crosses into the stable air ahead of the front. This of course assumes anything goes up out in the warm sector ahead of the MCS in Missouri being progged by CAMs. Obs late tomorrow morning will determine if I bite or not.
Growing more and more concerned about a few supercells/tornadoes in the vicinity of the warm front across parts of eastern Missouri into west/central Illinois on Tuesday afternoon and evening. Potential pitfalls do exist.

There are a lot of convective implications between tonight and tomorrow morning - with one or more MCS potentially moving through/near the target area.

That being said there has been a consistent signal going back to late last week that we will see broad unstable warm sector on Tuesday afternoon featuring 0-3 km CAPE > 200 j/kg thanks in part to surface dew points in the upper 60s to near 70F underneath a broad 50 knot LLJ.
Give me those things alone along/south of an Illinois warm front in late-April and I am there every time. Additionally, high-resolution ensembles have continued to initiate surface based storms ahead of a subtle wave near St. Louis by mid-afternoon Tuesday.

Assuming there is no drastic shift in the forecast parameters, these storms should waste no time taking on supercellular characteristics and could produce a few tornadoes. Given deep moisture, ample low-level CAPE and low LCLs, a sizeable tornado would not surprise me.

What could go wrong? A lot of the potential ruiners don't seem to be present. Too much cold air north of the front? Too much WAA convection suppressing warm front movement? For now (and back to late last week) this does not seem to be an issue tomorrow.

Instead, the warm front is forecast to slowly lift through the day with the subtle surface wave, allowing storms the ideal longevity within that highly sheared environment, rather than being undercut or crossing the WF too quickly.

There is still plenty of time to go on this one with again, many convective implications in play before we get to Tuesday afternoon. The consistency and tendency for conditions to actually *improve* with each run, coupled with high-res ensemble support have me concerned.

Last edited:
I am eying Eastern MO/Western IL tomorrow as well. I like that the low level jet becomes very focused in this region by late afternoon near the warm frontal boundary with it lifting northward. CAPE/low level shear are very robust along this boundary with significant tornado parameters pegging quite high. The long range 12z HRRR hints at a very isolated convective cells firing ahead of the main MCS and a fairly robust shortwave pushing through the area should be enough to initiate convection (as mentioned before). My concerns for tomorrow is that it could end up similar to 4/29/17 in that a severe MCS pushes across the target area with damaging winds and embedded tornadoes (IE messy storm modes). Otherwise, I'm becoming concerned that a corridor from Eastern MO into West Central IL could see some tornadic activity tomorrow afternoon/evening despite storm mode concerns. Jerseyville, IL would be my starting target I think, maybe cross into Missouri if need be....but I'd rather stay on the IL side of the river where the terrain is better if I can.
I would focus on the are over southeast KS-southwest MO and eastern OK. This region is closer to the ejecting mid level impulse and 250mb speed max where there is decent CAPE. Initial storm development should include some supercells before becoming more linear before sunset. The region over eastern MO into IL might occur after dark?
Recent trends suggest less convective overturning/debris over Oklahoma this morning, meaning much of the state is probably going to see substantial destabilization by early afternoon. High-resolution model guidance also favors more discrete/semi-discrete storm modes than earlier runs, particularly during the beginning of the event. Moisture will not be a limiting factor, as 06z observations show mid to upper 60s dew-points already up to the Red River along the OK/TX border.

It's still a bit difficult to pinpoint the greatest tornado threat, as early day showers/convective activity may leave one or more outflow boundaries in place. My guess is that the eastern third of Oklahoma is at greatest risk, but if storms initiate ahead of and avoid interaction with the main frontal boundary through late afternoon/early evening, you could see a bonafide tornado farther west across Oklahoma as well.

The southern half of Illinois is still very interesting, as previously mentioned by myself and others. Near-term guidance is rather aggressive with destabilization and storms could get going as early as lunchtime. Storms that interact with the warm front will likely have the greatest potential at producing tornadoes.

I'm not quite sure what to make of the West Texas/Low Rolling Plains threat just yet, in terms of tornadoes. A few storms will probably try to initiate off the dryline, but at least through 21-22z, low to mid-level wind fields look marginal at best and this area will be displaced away from more favorable forcing over Oklahoma and points northeast. This means that any tornado threat would be highly conditional and would largely hinge off of the ability of storms to persist either far enough east and/or late enough (after 23-00z) to take advantage of a gradually intensifying low-level jet. Several convection allowing model solutions show a few isolated storms developing, but not persisting very long.
I'm mildly interested in that secondary target down along the Red River and into the Low Rolling Plains/Cross Timbers in Texas. This morning's 12Z FWD sounding has a more sizable 700 mb inversion than surrounding sites that may help temper initiation some before 18Z along with very deep moisture. What concerns me most going forward is that N/NW TX is going to be seeing some subsidence between 700 and 850 mb which will likely cause some less than ideal shear profiles for anything that goes up in the region.

That said, I'm expecting at least a few supercells with the >2500 j/kg MLCAPE and deep moisture/favorable deep-layer shear being progged in the warm sector. Some convection allowing models have been hinting at what appears to be high-based storms further west along the dryline with more persistent, likely surface-based storms initiating to the east and persisting from 21-22Z until an hour or two after sunset. Regardless, a very interesting and difficult forecast outside the main threat area of E OK/SW MO.
In reference to the northeastern target, definitely a mesoscale forecasting/NOWcasting type of situation. I don't know how far north the warm front will get, but convective debris may limit destabilization north of I-70. The area south of I-70 seems to be enjoying some sunshine now, as latest mesoanalysis showing burgeoning CAPE south of I-64.

CAMs differ in the extent, location and mode of convection. The NAM 3k seems to be locked on more of a QLCS mode following by a squall line, whereas the HRRR has shown consistent signals of some earlier, discrete convection between I-55 and I-57, as well as a few stronger, albeit less classically discrete cells to the immediate NNE of STL by 21z.

There's definitely plenty of shear INVO the warm front, though better instability dynamics may overlap ambient shear values further south in the open warm sector ahead of the main surface forcing, leaving the I-72 corridor, closer to the front, with more of a linear mode. This may be owed to shear vector and mean wind orientation to the storm initiation focus, particularly above the immediate lowest 1 km, the forcing for ascent along the initiating boundary, as well as a less favorable initial storm spacing and storm interactive environment that tends to lead to the domination of cell in cluster modes.
Another target that seems to be turning more interesting is the general area between Amarillo and Lubbock, TX this afternoon. Although low-level winds appear to be relatively weak, I could see an LP supercell or two form in this area. Given favorable deep layer shear and sufficient instability, a supercell storm mode is strongly favored. While the tornado threat may be limited due to less-than-impressive 0-1km winds, if any storm can persist far enough east, better boundary layer moisture is progged roughly east of TX-207. It's a case of a tornado not being the most likely here, but it is not completely out of the realm of possibility.
The consistent signal for a few discrete storms crossing the front along the I-70 corridor is encouraging, the limiting factor being the near-orthagonal to the boundary storm motions. Attemps at low-level organization may be short-lived.
Taking a look at the 16z special sounding for OUN. 0-1 km SRH looks good. Deep layer shear (0-6) isn't up there yet, but i expect it to increase. There are some breaks in the clouds now so Temps should bump up a few more degrees. Mesonet shows mid 70s to low 80s in the warm sector.

Farther east in Illinois storms might have a better chance to turn right on a boundary, considering upper level winds (all constant press level forecasts).

Meanwhile the parameter space looks good from SE Kansas into Oklahoma. Despite terrain, I think I'd favor it over farther west. If one strongly favors structure, west will be best. If one is seeking tornadoes, while noting terrain risk, east Oklahoma. SE Kansas might go too but I'm not as into the WF. Appears outflow made it down to I-40 in Oklahoma and is lifting north.

Norman sounding posted while I'm typing is wow. Hodo is almost perfect and that kink at 400 mb is too high to matter. Actually kind of wish I'm out there today!
Now that this event is over, I decided to go back and see which CAM did the best job and also, why there weren't potentially more tornadic storms across south central Oklahoma.

First the CAMs (12z and 18z runs)...

The HRRR favored more discrete/semi-discrete cells yesterday and an eventual line around 00-01z. Overall, it showed the areas where the storms would be. The 18z run was a bit more bullish on south central Okla. and missed out on the tornadic storms in SW Missouri. Whereas the 12z run did a pretty good job with SW Missouri and NW Arkansas. Overall, i would say the HRRR did a pretty good job. We can't exactly expect a model to literally tell us where a tornadic storm will be to the nearest km, so yep... HRRR was ok.

The 12z 4km NAM picked up on the early stuff near Ardmore, so a point there. But it quickly went linear with the afternoon storms and seemed to miss SW Missouri / NW Arkansas. the 18z run showed more semi-discrete storms and then an eventual line in Okla. However, it did miss out on MO and AR too. So, i think the HRRR did a better job than the 4km NAM.

The 12z NSSL-WRF picked up the early convection NE of OKC, and then the Ardmore Storm. It then continued to show storms into NE Okla. It show a hot mess across south central Okla during the afternoon that seems to verify quite well. Also, it kept up the discrete/semi-discrete cells in NW AR and SW MO.

So, for the CAMs, I would say the 12z NSSL-WRF did the best job yesterday.

Why weren't there more tornadic storms across central and south central Oklahoma?

We all saw the 16z and 20z OUN soundings yesterday that made our eyes pop and prompted the Moderate Risk upgrade. For the 16z see above. Here is the 20z


This thing looked great and had me licking my chops. That is a sweet looking hodograph. 0-1 km shear of 34 kt is most excellent. It all seemed there. But then the storms just didn't get cranking like we thought. Yes there was the one discrete cell NE of Wichita Falls, but overall, they never really took off. I fully expected the Lindsay-Norman-Prague storm to at least ramp up to a tornado warning, but no. So what happened???

Well looking at the 00z sounding for OUN you will see that winds at the 300-200mb level backed significantly from westerly to due south. At first, I ignored this as it looks like they literally launched the balloon into a storm. But then I decided to checked other soundings. AMA had the same thing, but not as bad. DDC, same thing, backed SW to S. TOP, you guessed it. On the 00z SGF sounding there was a bit of backing but not as pronounced. Overall the profile above 300 mb was WSW. LZK it was all due west above 300 mb.


So my theory is that as the trough rotated through the southern plains the "good" winds were pushed off toward the east, and the awesome 20z OUN sounding moved into eastern Oklahoma, NW AR, and SW MO. Central and S. central Oklahoma was left with a sounding that showed backing winds above 400 mb. This likely led anvil seeding and was possibly the reason why we had such messy storm modes in central and s. central OK.

I would love to hear anyone else's thoughts on this.

Last edited:
I'm not convinced the pristine storm environment was ever sampled anywhere yesterday, as storms were everywhere pretty much all day. All storms within 100-150 miles of OKC looked messy all day. I suspect the wind profiles were just a little too screwy in even the low-mid levels. I don't necessarily think anvil seeding had much to do with it, as even the more isolated storms still had wonky structure.