(Tornado south of Parkersburg, Iowa around 12:30 pm CT yesterday. Phtoto courtesy of Jennifer Jackson/Iowa Storm Chasing Network)
Challenge for you, dear reader: Think about a semi-recent (last few years or so) severe weather forecast that completely busted. Was the first event that popped into your head a moderate/high risk SPC outlook that ultimately produced very little to satisfy the swarm of red dots on RadarScope? Yesterday’s SPC forecast was a bust of a different variety however.
Tornadoes still happened. They just happened 500 miles away from where they were supposed to. The tornadoes that didn’t occur within the 5% region hardly makes this forecast a bust, as after all, it’s only 5%. In addition, a plethora of wind reports help verify the merit of the enhanced risk.
Instead we will focus on Iowa, which is very very far away from the ArkLaTex/Mississippi Delta region. Iowa weather is known to get weird, so let’s dive into how the state’s atmosphere got drunk this time.
— UW-Madison CIMSS (@UWCIMSS) November 28, 2016
From this visible satellite image alone, we can see a very powerful low pressure circulation driving a dry region eastward into a moist air mass with thunderstorms rapidly firing along the boundary. The cloudy region ahead of the boundary shows rapid southerly winds, which from the geography, we can assume means warm and moist air advection. These observations align with yesterday’s 12z NAM solution.
The SPC’s thunderstorm forecast boundary follows the western edge of this 18z dewpoint gradient meaning their forecast didn’t go haywire in relation to the synoptic setup. To determine why no tornadic thunderstorms were predicted in this area, we must examine additional severe weather ingredients.
That’s some weak-ass CAPE bro. It’s not exactly what you look for when predicting deep convection leading to classic supercells.
The 12z NAM’s 700 mb (left) and 500 mb (right) vertical velocities agree with its CAPE solution that any convection initiating along the boundary would remain shallow. For this already unfavorable outlook to produce tornadoes, the final ingredient is shear.
Here’s a HRRR forecast sounding for 22Z ENE of Des Moines. I mean…Okay. pic.twitter.com/gBpGtPs69W
— Shawn Milrad (@shawnmilrad) November 28, 2016
This is, um, well, difficult to make sense of. It’s certainly not your classic clockwise-turning hodograph. To make the situation even more confusing, this HRRR sounding is only 2 hrs ahead of its initiation time in which tornadoes were already occurring. This model sounding is not a near-term forecast bust. This model sounding is an accurate representation of an environment that produced tornadoes despite the lack of classic ingredients.
Digging deeper into this sounding, we notice a SFC-1 km Storm Relative Helicity (SRH) of 62 m2 s-2. While that’s nothing to raise an eyebrow at, let’s examine this measure of low-level wind shear just to the east where most of the tornadoes occurred.
(We’re reverting back to the 12z NAM here since Pivotal Weather doesn’t archive more than 4 hours worth of past HRRR simulations.)
That, my friend, will get your rising air spinning. The upper level wind shear wasn’t impressive enough to produce the classic clockwise hodograph, but with only shallow convection occuring, it didn’t need to be.
— James Aydelott☂✈ (@jamesaydelott) November 28, 2016
Had the convection been deeper, the lack of directional wind shear at the upper levels would have likely disrupted the line of these nearly-discrete super cells. Instead, the shear and convection worked together in their shallow realm, and along with ample moisture and a strong frontal boundary, rattled off 9 tornadoes.
— NWS SPC (@NWSSPC) November 28, 2016
The SPC finally issued a mesoscale discussion around the same time the system produced its final reported tornado. The reluctant tone of this forecast is essentially the SPC saying, “Okay Iowa, we notice you. Now stop dancing on the table and get in the cab home we called for you.”