thunderbird12

TDS in Monroe Co.,WI as of 1:46 PM CDT. TOR warning out 2 minutes later.

Seattle broke their all time high (previously set in 2009) by getting to 104 on Sunday, so technically Monday's 108 broke the all-time record (set the previous day) by 4 degrees. Pretty remarkable to see many sites (including Seattle/Portland) end up breaking previously established all time highs by 5-10 degrees, and in many cases eclipsing the previous all-time highs for multiple days in a row.

First tornado watch for the OKC metro since January 10, 2020, a span of over 16 months.

Tornado watch issued for much of OK until 7 PM CDT. Tor probs for the watch are relatively low (40/20), but high probs for wind/hail (70/60 for both).

The HRRR has done a poor job with the storms now approaching central KS, but it seems to have had the right idea in limiting convection thus far across western KS. 22Z HRRR did have a very nasty UH track across northwest KS this evening, though. It's late May, so you have to get pretty deep into the evening before you can rule out surface-based initiation.

PDS TOR watch issued for western KS, southwest NE, and northeast CO.

Most of the moderate risk area is probably not going to be affected by the ongoing storms. Areas immediately to the north and east of the ongoing complex might have the threat reduced due to disrupted storm inflow, though.

It may have started north of the boundary, but based on surface obs, satellite, and its southeastward motion, it is very near the boundary now or is about to be. The boundary is pretty diffuse anyway, so it isn't easy to pinpoint a location and call that the boundary. I doubt the cell would be propagating southeastward like it is if it wasn't surface based.

Hard to imagine there won't be at least widely scattered development down the dryline in western KS by early evening, given how weak the cap is. The storm currently west of Hays looks like it is near the effective warm front and probably becoming surface based. Low-level flow is still weak, but it might have a chance to do something given the large CAPE and its location near the boundary.

Looks like Selden, KS got hit by a potentially significant tornado. Tornado appears to be occluding now to the northeast of Selden, as of 630 PM CDT.

Saturday would look a lot more interesting with upper 60s to ~70F dewpoints across the warm sector in OK/southern KS, instead of the low-mid 60s currently progged by most guidance. Cap strength is going to be a problem for storms moving off of the dryline, unless the current guidance is underestimating low-level moisture return. Even if the current progs end up verifying, there could still be an interesting storm or two near the triple point and/or warm front, though.

TDS approaching Fort Smith, AR.

The area where the apparent tornado crossed I-35E south of Waxahachie is about to get hit by more hail and wind.

Lots of golfball-sized hail reports in and around San Antonio, with a few larger (up to 3" diameter) reports mixed in. While some of the damage may be redundant after last week's storm, still probably a very costly evening down there. Also recently had baseball hail reported back west in Fisher County TX, with one of the post-frontal long-trackers that the HRRR and some of the other guidance were advertising. It'll be interesting to see how far east that cell can make it before it croaks.

Low-level wind profiles will probably improve somewhat near and after 00Z. The question is whether storms can stay discrete long enough to take advantage of that. Looks like initiation may be underway in Parker County east of Mineral Wells as of 2120Z.

Looks like roughly the northern half of Norman took some pretty serious damage from wind-driven hail last night. Lots of west-facing windows (for both houses and cars) blown out. Some people I know on the northwest side are probably going to have to replace their roof for the fourth time in ~10 years. I was about 1 mile south of the south edge of the hail core and I could hear it as it passed by to the north, even though there was no hail at my location.

GFS is now much faster compared to the rest of the global models. Today's (4/22) 12Z ECMWF continues to trend slower with the trough, though it still shows higher-end potential for Tuesday. UKMET and Canadian have slowed down so much that the main event (if there is one) would be on Wednesday, with hardly any threat (in their solutions) on Tuesday.

Today's (4/20) 12Z ECMWF definitely has the look of a classic Plains outbreak for next Tuesday (4/27). GFS/GGEM also show a powerful system, but their solutions are more amplified, which would likely lead to a messier event if it happened to verify that way. Aside from the trough evolution, a big key with next week's system will be the quality of moisture return. That will depend in part on how far south this weekend's system can push a cold front through the Gulf. The ECMWF (and other guidance) has varied from run to run regarding moisture return for early next week, so probably premature to say that it is "locking onto" anything yet. Previous runs of the ECMWF have struggled to get low 60s dewpoints north of OK next Tuesday, whereas today's 12Z run has low 60s dewpoints into IA/MN and mid/upper 60s into KS/OK. If the juicier solution becomes a trend, then next week's potential event becomes a lot more interesting.

The ECMWF only has one modeling system, which does not include anything operating at CAM resolution. The fact that they only develop and maintain one modeling system probably gives them an advantage over the plethora of U.S. models that have to be maintained and run. At some point in the future, global models will be run routinely at CAM resolution, and it wouldn't be surprising if the ECMWF ends up leading the way there. While the ECMWF is the best model overall, it was one of the models that was overdone with the deepening of the surface low during the day yesterday (as referenced above). In reality, there was only gradual deepening, and winds that quickly veered across much of central/northern MS during the day never really backed around like some guidance showed. That basically shut off the potential in most of the MS portion of the High Risk, and reduced the threat in northern AL and TN since storms had little time to mature on the effective boundary before moving into those regions. Yesterday was a fascinating event to watch, though unfortunately one with tragic consequences. Most storms didn't do much of anything, and the ones that did produce seemed to take forever to mature. The mature supercells that did finally develop were pretty nasty and long-lived, though. Also had some long-lived left-movers further north. I think one left-mover developed in AL and ended up in OH before dissipating, producing some hail and wind damage around Nashville along the way.

I'd be surprised to see a 45% tor at 1630Z, mainly because the high risk is a relatively confined area. Trying to squeeze another contour in there would be tough, unless the High risk is expanded quite a bit.

Seems to be fairly common for the ARW-based CAMs (HRRR, HRW-ARW, HRW-NSSL) to be more impressive-looking with discrete supercells/UH in these types of scenarios compared with the NMM-B based CAMs (3km NAM, NMM-B), with the HRRR and NAM usually on opposite ends of that spectrum. The HRRR is probably the most skillful of all the CAMs, and it tends to generate attention because it creates the most realistic looking storms, but obviously it is also prone to generating supercells that don't actually happen. If an event does turn out to be a supercell-driven tornado outbreak, though, the HRRR is probably the most likely to get it right. The problem is that there really hasn't been a clean supercell-driven tornado outbreak recently, so we don't know what the current generation of CAMs should look like with that type of event. April 12 of last year was a pretty big outbreak, but the mode was very messy, with only a couple of truly discrete warm sector supercells (which produced big-time tornadoes). Most of the CAMs (even the HRRR) generally forecast that the 04-12-20 event would have a messy mode, which probably prevented it from ever being upgraded to High Risk. For tomorrow, the northern extent of the high-end potential remains quite uncertain due to the potential effects of early/ongoing convection. However, the southern extent of the threat will also be a big deal, in terms of potential impact. It could be a high-impact event regardless, but if the Jackson, MS to Birmingham corridor gets involved, the potential outcomes become a lot worse than if it stays mostly north of there.

That will reduce the threat for awhile this afternoon across northern AL, but unless storms keep regenerating to the south (which is possible), the warm front will probably make it into northern AL by late afternoon or evening, with a corresponding increase in the threat.

I assume this explains the rolling power outages in TX? Over 2 million customers without power there as of around an hour ago.

In addition to the water plant issues, there have been sporadic power outages across Oklahoma since this morning. I think we peaked at ~10,000 outages across the state, and currently have ~9,000 outages. The infrastructure around here is not designed for this type of extended and bitter cold snap.

Around 3.5" on the ground in south Norman as of 330p. Probably going to bust low unless snow hangs on longer than expected this evening. Don't often see moderate/heavy snow with temps hovering around 5F around here. It's definitely an interesting and memorable event, but the relatively short duration of the moderate/heavy rates has limited our potential for a really big snowfall. Tough to get widespread 6-10 inches out of a relatively disorganized storm like this, no matter how high the SLRs get. I'd love to overperform tonight and get proven wrong, though.