Bennington 2.0 What happened and what can we learn?

[ericjkelly]

Friends,

So I couldn't get out to chase today so I played a little virtual chasing on my laptop. My biggest bust ever was Bennington (I got too far ahead of the storm anticipating faster speeds and got caught on the north side). As I was watching today, I am almost certain I would have busted again. Can I get some help so I dont do this again?

The storms along I-70 looked like a complete mess. Shortly after, isolated storms started to fire south of 70 in clean perfect (assuming) environment. Within 60 mins all of those isolated southern storms either went completely LP or died. Not long after, the storms along 70 went nuts producing Bennington 2.0.

Attaching HRRR loops from earlier in the day when everything seemed to be shaping up as normal. STP's were off the charts, shear, moisture etc.

Radar Loop------http://weather.unisys.com/radar/hires_rad.php?image=rad&inv=0&t=l3&region=cp

Dewpoint---http://climate.cod.edu/data/forecast/animations/21Z-20180501_HRRRFLT_sfc_dewp-0-18-10-100.gif

STP---http://climate.cod.edu/data/forecast/animations/21Z-20180501_HRRRFLT_con_stp-0-18-10-100.gif

Thoughts???

 

[ericjkelly]

Adding a radar shot at the time I mentioned.


Sent from my iPhone using Stormtrack

 

[Quincy Vagell]

Regarding southern cells:
This was fairly well modeled. The farther south one went in this event, the more the location was displaced from upper level forcing. With neutral to slightly rising heights, it often takes more than just convergence along a dryline to form and sustain robust convection. One thing that I overlooked a bit was how richer boundary layer moisture (dew-points in the low to mid-60s) was displaced just far enough east of the dryline to imply that any discrete storms forming on the dryline might not even make it far enough east to tap into the moisture. This was well-modeled, as the HRRR showed generally short-lived convection dying out by 00z on the dryline. The only possible caveat is that several runs showed a long-lived UH track in the general vicinity of Great Bend. One could argue that the storm that briefly looked tornadic north of Greensburg was this one. Either way, even if it wasn't 100%, the HRRR, overall, did an outstanding job with this event. Also, reference the 00z LMN sounding to see how a stout cap was in place downstream of the southern cells around 00z.

Northern cells:
Once again, this was nailed by the HRRR. Notice how the HRRR showed messy looking storms near I-70 for most of the afternoon, only to go bonkers around the 00-02z time-frame? The number one reason is probably the low level jet. As winds increased, low-level shear improved dramatically, causing the cells to remain semi-discrete/discrete and rotate, some vigorously. Also consider that the initiating boundary in the vicinity of the triple point was basically SW to NE (maybe closer to SSW to NNE) and shear vectors, while not parallel, were not exactly perpendicular to the boundary either. This coupled with, initially modest low-level shear, resulted in a grungefest of storms. There also seemed to be some moisture pooling in the general vicinity of Salina, as more cells up north seemed to get into low/mid-60s dew-points than storms to the south, that more or less stayed in areas with dew-points in the ballbark of upper 50s.

That's my take on it. I haven't had a ton of time to analyze the exact evolution of the tornadic supercells, but I think the comments above can help partially explain what happened.

 

[JamesCaruso]

Eric thanks for asking this question and Quincy thanks for your detailed response. I have often wished there were more post-mortem discussion, instead it’s usually a lot of forecasting discussion in the Events thread, then a Reports thread on what people saw or didn’t see, maybe a small mention of an ingredient that may have been missing but very little in-depth post-mortem analysis. How cool would it be if there were a product like a Convective Outlook or an Area Forecast Discussion written about what *actually* happened instead of what was forecasted to happen...

 

[Paul Knightley]

Some might call it Bennington 3 (as the Chapman tornado wasn't *far* from Bennington).

Sitting 4500 miles away afforded me the chance of a virtual chase, any my early analysis suggested the I-135 corridor, initially because of the best moisture in that area.

The southern storm threw off a left split which ran into the more northern storm - I've seen left splits screw up a storm (but also make them briefly produce - e.g. May 29th 2012, Piedmont, OK) - but cell mergers can also help storms.

The northern storm became quite a mess for a time, as noted above, but it had a pretty good clear inflow of air from the south - and it moved into 1) better moisture, and 2) increasing LLJ.

I think if you are on a persistent area of storm development and you know better moisture is ahead and the LLJ is expected to kick in, stick with it!

Looking back to Turkey, TX, in May 2016, there was a cluster of storms which didn't look great for a time, then the LLJ started to kick in and they went nuts!