High-Risk Tornado Outbreak to Rage across Southern Plains Monday

A rare combination of a summerlike air mass and a winterlike upper-level low will lead to an unusually intense, widespread, and prolonged severe weather outbreak from Monday into early Tuesday over the southern Great Plains. In its 1 am EDT Monday outlook, the NOAA/NWS Storm Prediction Center placed parts of the eastern Texas Panhandle and western Oklahoma under a high risk, the most serious of SPC’s five risk categories. The broader outbreak will extend from northwest Texas and far southeast Colorado through Oklahoma and much of Kansas into western Missouri and Arkansas. Update: At 8 am EDT, the high-risk area was extended eastward to include the Oklahoma City area (see Figure 1 below).

Conditions will be "ideal for a tornado outbreak," said SPC, "with strong tornadoes upstream and to the west of the low-level jet."  The greatest threat for long-track tornadoes will extend from the southern Texas Panhandle northeastward to parts of central and north-central Oklahoma, said SPC. Conditions will favor strong and possibly violent tornadoes (EF4 or EF5), the center said. 

"This event should result in a significant threat to life and property," they added.

Figure 1. Severe weather outlook for Monday, May 20, 2019, issued by the Storm Prediction Center at 7:55 am CDT. The threat areas may shift as the day evolves. Image credit: NOAA/NWS/SPC.

How things could unfold on Monday

The ingredients for an outbreak of severe weather that we outlined in our Saturday post are still fully on the table: a broad, powerful jet stream overlapping very warm, humid, unstable air. Sunday evening was mild and calm over Oklahoma, but a tropical air mass was moving north through Texas, as extraordinarily strong upper-level winds for late May begin to move in from the west.

00z soundings give insight to why tomorrow is a big deal.

80F dew point at Corpus Christi, 4285J/kg MLCAPE at Brownsville (h/t @JakeMulholland1), 72kt 500mb wind at Albuquerque.

These values are among records in their own right, and will be converging tomorrow in N TX and OK. pic.twitter.com/VPuGUZ8fOB

— Sam Lillo (@splillo) May 20, 2019

Computer models agree that the tropical air mass will move into at least the southern half of Oklahoma and the southeast Texas Panhandle by midday Tuesday, while upper-level winds increase. Severe storms could pop as early as Monday morning along and north of the warm front, sweeping from the Panhandle across northwest Oklahoma. In time, these could pose an increasing tornado threat. Rain-cooled air from these storms may force the warm front to stall, allowing it to serve as a boundary on which intense supercells could fire by early Tuesday afternoon.

One possibility suggested in some mesoscale models on Sunday night, including the HRRR model and Texas Tech University’s version of the WRF model (TTU WRF), is that a broken line of potentially tornadic supercells might develop by mid- to late afternoon from the eastern Texas Panhandle into western and central Oklahoma, perhaps within 50-75 miles of the I-40 corridor, and then sweep northeastward. Some other variants of the WRF model are pushing the rain-cooled air and the peak tornado threat a bit further south.

From late afternoon into Monday night, both the HRRR and TTU WRF indicate a broken line of supercells (again some potentially tornadic) could stretch from the southern Texas Panhandle across Oklahoma from southwest to northeast, possibly affecting the Oklahoma City and Tulsa metro areas. These cells could track repeatedly over the same areas, leading to a potentially serious flash flood threat extending after dark.

Finally, an intense squall line will sweep across Oklahoma and southern Kansas from late Monday night through Tuesday morning, packing widespread straight-line winds that could exceed 80 mph.

Figure 2. Forecast of storm reflectivity and updraft helicity issued by Texas Tech’s version of the WRF model on Sunday afternoon for Monday evening, May 20, 2019. The oranges and yellows show where precipitation is heaviest, while the deep red stripes show the tracks of storm rotation over the prior hour. These images are from the 18Z Sunday run, valid at 6 pm CDT (left) and 10 pm CDT (right) on Monday. Model guidance should not be taken as an exact depiction of where storms will form, but rather as one scenario of the general areas where particular types of storms could develop. Results will vary among different runs of different models. Image credit: Texas Tech University.

Mesoscale models like the ones above cannot tell us with pinpoint precision where each storm will develop. However, because these models are fine-grained enough to “allow” showers and thunderstorms (convection) to develop organically, they can be quite skillful at depicting the overall flavor of a severe weather outbreak—i.e., the kinds of storms that may develop and their general locations.

There’s certainly enough in the mesoscale model output from Sunday night to warrant concern that multiple tornado-producing storms could emerge on Monday afternoon and evening, perhaps several at the same time. The ingredients in play are sufficently strong that we cannot rule out one or more tornadoes in the EF4-EF5 range. In some ways, this event may resemble an early-spring outbreak in the Southeast’s Dixie Alley more than a typical round of severe weather in the Plains’ Tornado Alley.

Severe storms could reintensify along or near the Monday-night squall line by midday Tuesday as it heads toward the mid-Mississippi Valley. Tuesday's severe weather is not expected to be as widespread or intense as Monday’s, but damaging winds and a few tornadoes are possible. SPC has issued a Day 2 enhanced risk for much of Missouri and northern Arkansas.

See the weather.com article for frequent updates through the day and into the night as this event unfolds.

Unusual measures for an unusual event

Storm-savvy Oklahomans are taking Monday’s threat seriously. Many of the largest school systems in central Oklahoma (as well as the University of Oklahoma campus) will be closed all day Monday, which appears to be the first time such a mass closure has occurred in central Oklahoma on the night before severe weather. This step will certainly lower parents’ anxiety levels and will keep many vehicles off the roads.

It was six years ago Monday (on May 20, 2013, which was also a Monday) that an EF5 tornado ripped across Moore, just south of Oklahoma City, killing 24, including 7 children at Plaza Towers Elementary School. This was the most recent EF5 tornado confirmed in the United States.

Figure 3. The rebuilt Plaza Towers Elementary School includes a set of inlaid artwork on benches and outdoor walls that serve as tributes to the children who lost their lives in the tornado on May 20, 2013. Image credit: Bob Henson.

Less than two weeks after the Moore tornado, on May 31, 2013, the largest tornado on record (2.6 miles wide) churned near El Reno. At least four storm chasers—including the eminent independent researcher Tim Samaras, his son Paul, and colleague Carl Young—were among eight motorists killed by the tornado. Many other storm chasers had close calls from this huge, fast-evolving twister. As the tornado bore down on Oklahoma City, a weathercaster’s advice to “go south” may have played a role in sending thousands of people onto packed highways. Fortunately, the tornado dissipated before reaching the city, but flash flooding took 14 lives in Oklahoma that night, making the floods even deadlier than the tornado.

Patrick Marsh, warning coordination meteorologist at SPC—which itself is near the heart of Monday’s threat area—posted a thoughtful and heartfelt series of tweets on Sunday night on the need to be prudent without panicking. It’s a tweetstorm well worth digesting as this potentially historic storm system takes shape.

As I fly back from Atlanta to Oklahoma City I've had a few minutes to reflect upon some of what I have seen on social media the past couple of days. My thoughts (and mine only): (Thread) (1/17)

— Patrick Marsh (@pmarshwx) May 19, 2019

A flood threat that mustn’t be overshadowed

While the tornado threat on Monday must be taken seriously—not to mention the risk of giant hailstones, easily 2" in diameter and in some cases baseball-sized or larger—flash flooding will be a life-threatening concern on its own, especially on Monday night. On average, more Americans die in flash floods (often on the road) than in tornadoes. Update: On Monday morning, the NOAA/NWS Weather Prediction Center placed parts of central Oklahoma under a high risk for excessive rainfall leading to flash flooding (see Figure 4 below).

Figure 4. Excessive rainfall outlook issued early Monday, May 19, 2019, valid for the period from 7 am Monday to 7 am Tuesday. Image credit: NOAA/NWS/WPC.

Models have consistently shown that Monday’s storms will end up “training” in one or more cycles across a belt roughly from northern Oklahoma into eastern Kansas. Rainfall has not been especially heavy in recent days across Oklahoma, but localized flash flooding is certainly possible, and it could be a particular concern in the urban landscapes of Oklahoma City and Tulsa. As of early Monday, several mesoscale models were indicating that a bullseye of 10-15” of rain could emerge somewhere close to the I-44 corridor, perhaps including Oklahoma City and/or Tulsa, as the training echoes develop Monday night. For perspective, the heaviest 2-day rainfall total in more than a century of recordkeeping at Oklahoma City is 10.43" in October 1983, associated with the remnants of East Pacific Hurricane Tico.

Especially if storms end up focusing a bit further north, a major flash flood threat could emerge in southeast Kansas, where up to 400% of average rainfall has occurred in the last two weeks.

A short-lived tropical or subtropical storm may develop east of the Bahamas

In a special tropical outlook issued on Sunday night, the National Hurricane Center tagged a disturbance dubbed Invest 90L—located several hundred miles east of the Bahamas and south of Bermuda—with a 50-50 chance of developing into at least a tropical depression by midweek. Sea surface temperatures in this area are about 1°C above average for this time of year, putting them close to the benchmark minimum for tropical development of 26°C (79°F). The disturbance remained disorganized on Sunday night, with only a scattering of convection. An Air Force hurricane hunter aircraft is tentatively scheduled to investigate 90L on Monday afternoon.

Figure 5. Infrared image of Invest 90L east of the Bahamas at 0531Z (1:31 am EDT) Monday, May 20, 2019. Image credit: NASA/MSFC Earth Science Branch.

Any development of 90L would most likely be subtropical, as the system lies just east of an upper-level trough that is inducing strong wind shear of 25-30 knots. Whatever becomes of 90L will be swept northeastward ahead of the upper low, away from U.S. shores, though possibly affecting Bermuda.

In the East Pacific, NHC is continuing to watch a disturbance still simmering well south of Mexico and west of Central America. Convection around this system is limited, but in its Sunday night tropical outlook, NHC gave this area 20% odds of development between Tuesday and Friday.