2019-05-17 EVENT: TX/OK/KS/CO/WY/NE/IA

Apr 5, 2015
85
70
11
Norman, OK
While we sit here at six days out, it appears that a significant severe weather event will be possible across portions of the central and southern Plains on Friday. Latest trends in ECMWF deterministic and ensemble guidance is for a deep wave to take hold in the west. As heights fall across the Rockies, favorable mass response will induce strong SSE trajectories across most of the S Plains. The result will be a likely corridor of low 70s F dewpoints across W Texas with 67-69 F dewpoints across much of the panhandle and western Kansas area. Lift accompanying the wave will be aided by a sharpening dry line beneath southwesterly flow aloft.

EC_Spread.png

Shear profiles will easily favor discrete supercells with a large-very large hail threat. Tornadoes will be possible as well although details are going to vary being six days out. Pattern recognition and trends in ECMWF/EC ensemble lead me to believe that Friday has a pretty high ceiling with a significant severe weather episode possible with favorable timing. GFS is not as deep with the wave nor does it have the same timing, but this is just about the Euro's sweet spot and I expect the GFS to come in line with the ECMWF by mid-week. This is a classic May dry line tornado setup. Should be a really good time out (minus the herds) if trends continue.

There will likely be a secondary northern target along a synoptic warm front in the Dakotas and adjacent area. While shear profiles will probably be enticing here, very strong warm air advection and synoptic lift will probably result in rapid upscale growth, and models show a clear signal for an MCS to develop and track along the baroclinic zone.

Targets to be refined throughout the week.
 
Sep 25, 2006
471
34
11
29
Minneapolis, MN
This will be an interesting forecast for me. The target everyone has been watching is down along the dryline in KS and OK. The GFS really wants to ramp up the northern target too along the warm front in SD and possibly MN. At this point I am not sure if there is really a wrong choice to make between the two. Forecast soundings in both areas show some weakness if H5-H3 winds. For now I would be leaning towards the northern target for two main reasons with it being closer to home for me and also it looks like LCL's might be better along the warm front. A long ways out so it'll be interesting to see how the details evolve. Unless the set up goes to complete crap, I will likely be out somewhere on Friday.
 
Jul 5, 2009
736
351
21
Newtown, Pennsylvania
That is quite a difference between the Euro and GFS as highlighted by Brandon. The operational Euro is fairly similar to the ensemble mean Brandon posted. Whereas the Euro has the 565 DAM line digging down to the Mexico border, the GFS has it down only to the latitude of southern NV / northern AZ/NM. The GFS also may be faster, as it lifts the mid-level low to WY by OZ Sat while the Euro has it at the intersection of the NV, UT and ID borders at that time. If the Euro is indeed to be favored, I would have to discount the surface features I am seeing on the GFS, which shows the synoptic front in SD as more of a cold front. Still, the important features are there, notably a sharp dryline from southwestern KS into the eastern TX Panhandle. Exact longitudinal placement will of course change between now and then anyway, and a deeper trough should only enhance the backing of surface winds east of the dryline. What’s interesting is that despite the 500mb differences, surface low placement is similar on both models and in fact the GFS surface low is actually deeper - maybe because it is closer to the mid level low on the GFS? I don’t know, I can make the observation but will leave the interpretations/explanations to the smarter people on here, that’s beyond my pay grade. CAPE is extreme on the GFS. I don’t have access to any of the pay model sites to compare any other surface features or severe parameters on the Euro with any degree of detailed resolution.
 

Jeff Duda

Resident meteorological expert
Staff member
Oct 7, 2008
2,902
1,300
21
Westminster, CO
www.meteor.iastate.edu
While we sit here at six days out, it appears that a significant severe weather event will be possible across portions of the central and southern Plains on Friday. Latest trends in ECMWF deterministic and ensemble guidance is for a deep wave to take hold in the west. As heights fall across the Rockies, favorable mass response will induce strong SSE trajectories across most of the S Plains. The result will be a likely corridor of low 70s F dewpoints across W Texas with 67-69 F dewpoints across much of the panhandle and western Kansas area. Lift accompanying the wave will be aided by a sharpening dry line beneath southwesterly flow aloft.
With all due respect, where are you seeing 70 F dews anywhere relevant to this setup? I don't see the 70-F isodrosotherm west of I-35 in either the GFS or ECMWF control member:
GFSSGP_sfc_dewp_105.png
Weather us (1).png
Weather us.png

This definitely looks like one of the first solid Plains dryline setups this year. What concerns me is how shear is impacted by the timing of the wave. As of the 12Z GFS this morning, the wave is a little further west than I would like to see it for optimal shear over the Plains. The GFS also has a tendency to slow things down at this range, so the placement could get worse as time goes on. Also, due to a pretty gnarly LLJ there isn't a ton of speed shear above the PBL. Lastly, 700 mb temps are a little on the warm side, so CI is not of high certainty at this point given the forecast Tds I'm seeing of the low-mid 60s immediately ahead of the dryline. However, if something goes, there is a high probability of it being supercellular, although bases could end up being a bit high.

I know everyone is excited after a really crappy 2018 and a slow start to 2019, and those of you who know me on this forum know of my conservative bias, but I would caution about speaking about this setup being epic and paying everyone off. It certainly does not appear to be a slam dunk at this point. The good news is we have several days to watch this come in and things could definitely change for the better.
 
Apr 5, 2015
85
70
11
Norman, OK
Jeff, you quoted a statement I made looking at 5/12 00z ECMWF/GFS data, whereas you're looking at today's 12z GFS. Nonetheless, I still anticipate mid-upper 60s F dewpoints across W KS by 00z Saturday.
 
  • Like
Reactions: Trey Thee
Aug 9, 2012
320
484
21
Galesburg, IL
www.facebook.com
I'm noticing a lack of precipitation on some of the forecast models for Friday (NAM/GFS). FV3 develops a cluster of convection along the warm front near the NE/SD border, however it seems as though there may be some capping issues? I'm not sure what to take away from this lack of convective activity when we have 45 knot shortwave pushing through around 21-00z. Some of the best soundings I could find on the 12z NAM was across Northeast NE/SD, however CIN was in excess of -100 which isn't favorable at all. Not sure what to make of the lack of convection on the models.
 
  • Like
Reactions: Trey Thee

Trey Thee

EF2
Mar 29, 2010
191
28
11
Tulsa metro
I'm noticing a lack of precipitation on some of the forecast models for Friday (NAM/GFS). FV3 develops a cluster of convection along the warm front near the NE/SD border, however it seems as though there may be some capping issues? I'm not sure what to take away from this lack of convective activity when we have 45 knot shortwave pushing through around 21-00z. Some of the best soundings I could find on the 12z NAM was across Northeast NE/SD, however CIN was in excess of -100 which isn't favorable at all. Not sure what to make of the lack of convection on the models.
Ethan, that's a noteworthy concern in most situation. I would urge you to keep in mind this is the edge of NAM capability and it's not particularly good at the end of it's ability. If the same thing is showing up at 48 hours or 36hours out, other models showing similar capping that could be a problem. At this point, I think you hit on an important piece which is the possible breaking of the cap between 21Z and 00Z due to other factors then just the cap. Shortwave timing will matter but right now the broader environment looks solid. If the cap is still ultra strong and maybe we get less or no energy ejecting, then that's a problem, at this point I'm not worried about it (could change my mind though).
 
Jun 1, 2008
419
273
11
Chattanooga, TN
www.linkedin.com
I fully expect a secondary target farther south. FV3/GFS op do convect toward 01-02Z which is still daylight. However we don't need it to tell us. Trey is also right regarding the NAM past 36 hours.

PVA is coming in; warm 700 mb is cooling with time; LLJ cranks up; surface winds are backed. ECMWF convects south. There's a model I actually care about. At the surface a slight DL bulge is forecast out of the Panhandles. ECWMF hints too with isobars. While I'd prefer a true outflow boundary, the southern stream jetlet should promote a slight DL bulge.

None of this guarantees anything. NAM hints at stubborn cloud cover east of the DL. However pattern recognition supports two targets Friday. Euro erodes those mid-level clouds which is believable in May. Even then it could be mainly hailers. If dews top 65 it'd get interesting.
 
  • Like
Reactions: Trey Thee
Sep 25, 2006
471
34
11
29
Minneapolis, MN
The aspect that has made me the most nervous is the LCLs in the forecast soundings from the GFS. Even today's 12z GFS have really high LCLs (generally abover 1400-1500m) along the DL in Kansas and this has been pretty consistent on the GFS. Once you get a little east of the dryline, the LCLs start to drop but so does the shear. Seems the overlap isn't there right now. Today's 12z NAM has better LCLs, but it's hard to tell how seriously to take it since the setup just came into range for the NAM.

If the GFS verifies with those LCLs, we are probably looking at high based supercells with little chance of a tornado, especially since shear isn't over the top.
 

Jeff Duda

Resident meteorological expert
Staff member
Oct 7, 2008
2,902
1,300
21
Westminster, CO
www.meteor.iastate.edu
PVA is coming in; warm 700 mb is cooling with time; LLJ cranks up; surface winds are backed.
There is not any appreciable cooling at 700 mb over KS during the day; I see MAYBE 1 C, but most areas see 0 C of 12-hour temp change.
PVA does not imply UVV (DVPA does).
Surface winds not forecast to back until 00Z.

Let's illustrate:
700 mb temperature
  • 1200 UTC
h7_t_f72.png
  • 0000 UTC
NAMCGP_700_temp_084.png

Yes, some of the colors in W KS change, but that's not due to synoptic scale forces. It is due to the dryline circulation, which the large-scale models struggle to resolve, but mesoscale models (like the NAM) can at least partially resolve:
NAMCGP_700_vvel_084.png

Note the corridor of moderate-strong upward velocities at 700 mb coincident with the area of apparent cooling. That band of velocity represents parcels being lifted adiabatically (which causes cooling by the time they get up to 700 mb since the environmental lapse rates are less than dry adiabatic), but does not represent differential temperature advection associated with the larger-scale trough. Note also that band of upward motion is much weaker at 21Z and is non-existent before that. Cooling does finally set in after 00Z, however, but it would be nicer if the cooling started earlier in the day.

Speaking of the trough...
12Z-20190514_GFSUS_500_spd-72-90-25-100.gif

It looks to me like the mid-level jet core weakens throughout the day, only starting to "resurge" around 06Z. That suggests to me the dynamical support fades during the day, and it appears the wave itself fills just a tad (watch the 5580 and 5640 m contours across the W US).

This to me perfectly explains why the models are hesitant to spit out precip across the dryline. The shear and instability will be there, but the triggering ingredient may end up not being there sufficiently due to lack of help from larger-scale forces. Since it remains rather warm aloft, capping will prevent CI for quite some time. Before 00Z it looks like there is very little forcing for convection outside of the dryline circulation itself along the dryline. Continued southerly flow should continue to bring in moisture throughout the day, however, which may end up being the saving grace here, as it will limit stabilization as heating goes away towards sunset.

I do think storms will form, but I suspect they will be quite isolated (which should also make them more picturesque should they go supercellular). But certainty is lacking at this point.
 
Apr 5, 2015
85
70
11
Norman, OK
NAM hardly ever resolves QPF on a dry line thanks to its convective parameterization scheme. Limited forcing will keep coverage low, but there will be storms. In the end everyone is gonna HRRR chase this one.
 
  • Haha
Reactions: Jeff House

Jeff Duda

Resident meteorological expert
Staff member
Oct 7, 2008
2,902
1,300
21
Westminster, CO
www.meteor.iastate.edu
NAM hardly ever resolves QPF on a dry line thanks to its convective parameterization scheme. Limited forcing will keep coverage low, but there will be storms. In the end everyone is gonna HRRR chase this one.
Is that not true with every chase anymore? It's the most rapidly updating operational model that exists and it contains individual thunderstorms. I'd argue that statement is a tautology.
 
Jun 16, 2015
344
719
21
32
Oklahoma City, OK
quincyvagell.com
Not to derail this topic, but a lot of people still seem to be using the NAM leading up to chase time. I can see using the 3km NAM, but when it diverges significantly from the HRRR, it's usually wrong.. Even the day before an event, for some reason, I see a lot of chasers even citing the 12km NAM...

I don't find much utility in using the 12km for convection at all. The higher resolution 3km NAM is helpful 1-2 days out, but for the general picture. It has issues with dew-points, to just name one bias. Either way, I may still weigh at least a small amount of confidence to what the 3km NAM/NSSL WRF/HREF show the morning of a chase. For me, with convection allowing models during the day of a chase, I use a much heavier emphasis on the HRRR, let's say, maybe 80/20 between it and other models. Of course, observations with your own two eyes, as well as mesoanalysis, surface observations and radar/satellite data are of the most help once you're approaching convective initiation.
 
Apr 5, 2015
85
70
11
Norman, OK
Not to derail this topic, but a lot of people still seem to be using the NAM leading up to chase time. I can see using the 3km NAM, but when it diverges significantly from the HRRR, it's usually wrong.. Even the day before an event, for some reason, I see a lot of chasers even citing the 12km NAM...

I don't find much utility in using the 12km for convection at all. The higher resolution 3km NAM is helpful 1-2 days out, but for the general picture. It has issues with dew-points, to just name one bias. Either way, I may still weigh at least a small amount of confidence to what the 3km NAM/NSSL WRF/HREF show the morning of a chase. For me, with convection allowing models during the day of a chase, I use a much heavier emphasis on the HRRR, let's say, maybe 80/20 between it and other models. Of course, observations with your own two eyes, as well as mesoanalysis, surface observations and radar/satellite data are of the most help once you're approaching convective initiation.
Depends on what you're going for. Tough to use the HRRR for assessing near-term environmental trends.. too much noise in mass fields thanks to resolving convective processes, hence why SPC prefers the coarser RAP for mesoanalysis. If we're talking convective trends and placement, then yeah using NAM is grounds for being burned.. The parameterization scheme, unless recently changed, is a tropical one that doesn't work well at all in low-forcing EML environments. Mass fields can still get you where you need to be, though.
 
  • Like
Reactions: Jeff House
Jun 16, 2015
344
719
21
32
Oklahoma City, OK
quincyvagell.com
Depends on what you're going for. Tough to use the HRRR for assessing near-term environmental trends.. too much noise in mass fields thanks to resolving convective processes, hence why SPC prefers the coarser RAP for mesoanalysis. If we're talking convective trends and placement, then yeah using NAM is grounds for being burned.. The parameterization scheme, unless recently changed, is a tropical one that doesn't work well at all in low-forcing EML environments. Mass fields can still get you where you need to be, though.
I rarely use the HRRR once the morning is over during a chase. It has issues resolving ongoing or recently initiated convection, although they've tinkered to make the short-term simulated radar fields more realistic. Still, we've had plenty of cases this season already where the 2-hour HRRR forecast was way off from reality.

Models are useful tools leading up to an event, but once it's close to game time, not so much.
 
Mar 27, 2014
145
120
11
Colorado Springs, CO
I use the GFS and European about 4-5 days out to plan when I am going to chase. This gives me time to start preparing by requesting vacation time, schmoozing the wife, and getting the car ready.

I use the NAM the day or 2 before I leave to pick a target area (in conjunction with air maps and soundings). The morning of, I look at the HRRR and radar to help me hone my target. I usually have to drive another few hours to get a better spot the morning of a chase.

That said, the HRRR has been dead accurate and equally dead wrong. I remember using the HRRR one day last year for a Colorado chase and got to an intersection in the middle of nowhere (Punkin Center) where the HRRR had forecast a supercell. Tons of chasers parked along the street just waiting. The skies were blue and not a cloud in the sky. Meanwhile, radar showed a small thunderstorm just starting to form two hours away and tracking to the southeast of where everybody was staged. I went after it, and watched it grow into one of only two tornado warned storms that day. NWS actually tornado warned it based upon my report (I was the only one chasing the storm). Meanwhile, I was watching spotternetwork dots 80 miles away moving from Punkin Center towards the storm. The storm went all the way into Kansas before they caught up to it.

Lesson: Use the HRRR to help you pick a target area the morning of, but then switch to radar and eye observations during the chase. There's a reason they call the HRRR a forecast and not a prophecy.
 
Jun 16, 2015
344
719
21
32
Oklahoma City, OK
quincyvagell.com
Back on track with Friday...

There are some concerns about convective initiation down the dryline. Looking at the GFS trends, the warm sector has been slowly shrinking, in addition to a narrowing of the favorable instability axis ahead of the dryline.
It's still early to be throwing out specific targets, but I think the highest confidence right now at convective initiation is in Nebraska. The issue is that it could get messy here. CI is more uncertain to the south, but that would allow for a more discrete storm mode. Another target could be southwest Texas... That familiar high risk, high reward with a lone supercell going to town around I-10.

I played with the latest 21z run of the SREF to look at 00z Saturday/Friday evening. The SREF severe probabilities are similar in an elongated corridor without a clear favored "target." The image below would suggest the best lift is up around the central Plains, along with the hint at more lift down near I-10. Here are the wind crossovers and mean lifted index:
sref_overlay.gif
 
  • Like
Reactions: Jeff House

Jeff Duda

Resident meteorological expert
Staff member
Oct 7, 2008
2,902
1,300
21
Westminster, CO
www.meteor.iastate.edu
Not to derail this topic, but a lot of people still seem to be using the NAM leading up to chase time. I can see using the 3km NAM, but when it diverges significantly from the HRRR, it's usually wrong.. Even the day before an event, for some reason, I see a lot of chasers even citing the 12km NAM...

I don't find much utility in using the 12km for convection at all. The higher resolution 3km NAM is helpful 1-2 days out, but for the general picture. It has issues with dew-points, to just name one bias. Either way, I may still weigh at least a small amount of confidence to what the 3km NAM/NSSL WRF/HREF show the morning of a chase. For me, with convection allowing models during the day of a chase, I use a much heavier emphasis on the HRRR, let's say, maybe 80/20 between it and other models. Of course, observations with your own two eyes, as well as mesoanalysis, surface observations and radar/satellite data are of the most help once you're approaching convective initiation.
The 12-km "mesoscale" NAM has pretty much run its course at this point. The GFS has caught up to it in terms of grid spacing, and the two run just as frequently. So unless you have a need for faster forecast latency or you prefer the NAM's physics, there just isn't much point to it anymore.
 

Trey Thee

EF2
Mar 29, 2010
191
28
11
Tulsa metro
The 12-km "mesoscale" NAM has pretty much run its course at this point. The GFS has caught up to it in terms of grid spacing, and the two run just as frequently. So unless you have a need for faster forecast latency or you prefer the NAM's physics, there just isn't much point to it anymore.
I pretty much ignore the 12km NAM now. It sucks at hours 60-84 anyway, so why bother? Resolution on the GFS is just as good as you said. I just wait for the hires NAM to be in range, which IMO is 48hrs or less.

Regarding what to use to chase? Models are helpful yes, but day of, I start off looking at mesonet here in OK and the SPC Mesoanalysis page. I then typically compare that to hi-res NAM, HRRR, RAP forecasts and look for differences that may be problematic or result in model outputs not being accurate.

HRRR is pretty impressive in its ability to resolve thunderstorms you absolutely can't chase by it a few hours out or you run a high risk of getting burned. I tend to find an area where I believe CI will occur and make sure that I have good directional roads N/S/E/W so that when convection does occur I can move towards it if I like what I'm seeing on Mesonet and Mesoanalysis from SPC.

As for Friday specifically, my biggest concern for tors are the possibility of high LCLs, capping and less than optimal shear environment. Plenty of time for these to be resolved, at least partially. I do think there will be a few supercells, and probably a couple tors south, they'll also probably be very photogenic. They'll also be close to dusk/dark which will create some problems.

The northern target area looks a bit better at this point for tor density IMO. If I could go either direction Friday and I absolutely wanted to try to get a tor, vs risking dusk structure shots, I'd head north.

You can start taking the NAM more seriously with the 18Z run today, see how that compares with GFS, GFS FV3 and Euro.

My 2 cents.
 
Last edited:
Mar 27, 2014
145
120
11
Colorado Springs, CO
Back on track with Friday...

It's still early to be throwing out specific targets, but I think the highest confidence right now at convective initiation is in Nebraska. The issue is that it could get messy here. CI is more uncertain to the south, but that would allow for a more discrete storm mode. Another target could be southwest Texas... That familiar high risk, high reward with a lone supercell going to town around I-10.
It's not too early to be throwing out targets. Many people are driving starting tomorrow, and we need to figure out where to go. Right now, I see three possible targets: Southwest NE to the middle of NE, Northeast TX Panhandle, Southwest TX.

I'm not a meteorologist, so I can't provide expert analysis, but here's what I'm thinking:

Southwest NE to the middle of NE:
This seems to have almost everything going for it right now including decent instability (~3,000 J/kg) and strong shear (65 kts). My only concern with this location is that the moisture seems to be a little too far to the north. RH is around 50% Friday, although as I'm sitting here typing, new runs are posting and that is increasing. Flooded roads are a concern, but my analysis is that the storms will fire more to the southwest and middle of the state, and won't reach northeast NE until later in the evening or overnight. Nighttime chasing will be very dangerous in this area when you combine the already inherit dangers of nighttime chasing with the flooded and closed road network.

Northeast TX Panhandle:
This area also has decent instability (~3,000 J/kg) and shear (50 kts), although not as great as NE. Moisture looks better though, as the dryline appears stronger here, which may help. Timing appears to be more favorable later in the evening though, which may result in some nighttime chasing. Road network and visibility is better than NE though, especially given the recent flooding.

Southwest TX (and into Mexico actually):
The strong dryline extends into southwest TX. Instability is strong with CAPE up to 4,500 J/kg. Shear is also good at around 55 kts. I think this setup is better than the northeast TX Panhandle play. This might be one of the best plays as it's being overlooked, and most chasers will choose NE or the TX Panhandle.

Again, I'm no expert so feel free to correct and add to anything I've mentioned. You won't hurt my feelings.
 

Jeff Duda

Resident meteorological expert
Staff member
Oct 7, 2008
2,902
1,300
21
Westminster, CO
www.meteor.iastate.edu
Staff note
Given we are looking at the arrival of a potentally extended period of chasing featuring higher-end setups, it makes sense that the TA has suddenly become very busy. There has been some slightly off-topic discussion in this thread (that I myself participated in), but those who went off on that tangent have done a great job of relating it to the event at hand. For that, the staff is very thankful and we welcome continued posting of a similar nature.

Thanks, everyone, for participating so far!