Chase Logs 2009
Stratiform Surprise: The Minneapolis Tornado Outbreak
August 19, 2009
1:00pm-7:00pm
Arden Hills, MNHonestly, I'm still somewhat confused by the events that transpired early this afternoon across the Twin Cities. More or less, HOW could this have happened? It seemed like everyone, even the NWS, blew this event. I guess I don't feel so bad then.
The day started off way too early for me as a passing thunderstorm disturbed my sleep shortly after 3:00am, The storms did little other than further cool the nighttime air, but I was up for an hour watching lightning--mostly in-cloud garbage. I moseyed into work at 6:00am with the remnants of the morning's storm fading into the glowing eastern skyline. The skies once again clouded up during the mid-morning as another weak system associated with a low out in eastern South Dakota took aim at the Cities. A light nimbostratus rain ensued at 10:30am and continued on throughout the lunch hour, capping temperatures below 70 degrees. The radar showed nothing impressive whatsoever, with reflectivity values barely topping 50 dBZ. I did notice that a low pressure system seemed to be forming in the shower activity in the late morning (based on a dry-slot feature forming on radar), but I thought little of it and was too busy at work to look too closely at what appeared to be a fall-like rain event.
Regional Radar Reflectivity Data for the Upper Midwest from 0000 UTC to 1800 UTC on August 19, 2009. The low begins to develop over SW Minnesota on the last few images of the set.
Shortly after 2:00pm, I wandered outside to check on the rain and hear the sirens in the distance. I initially thought that it was just a standard siren test, but within a few seconds I realized that it was not 1:00pm on the first Wednesday of the month. I then thought that it was a siren malfunction, since a similar problem had occurred within the past week in St. Paul. But as I walked back to my office, people began asking me about the tornado warning for Minneapolis. In disbelief, I quickly got back to my computer and saw that, sure enough, there was a tornado warning for Minneapolis and southern Anoka and western Ramsey Counties. Another tornado warning covered northern Dakota County.
543
WFUS53 KMPX 191911
TORMPX
MNC003-053-123-191945-
/O.NEW.KMPX.TO.W.0037.090819T1911Z-090819T1945Z/
BULLETIN - EAS ACTIVATION REQUESTED
TORNADO WARNING
NATIONAL WEATHER SERVICE TWIN CITIES/CHANHASSEN MN
211 PM CDT WED AUG 19 2009
THE NATIONAL WEATHER SERVICE IN THE TWIN CITIES HAS ISSUED A
* TORNADO WARNING FOR...
CENTRAL ANOKA COUNTY IN EAST CENTRAL MINNESOTA...
NORTHEASTERN HENNEPIN COUNTY IN EAST CENTRAL MINNESOTA...
WESTERN RAMSEY COUNTY IN EAST CENTRAL MINNESOTA...
* UNTIL 245 PM CDT
* AT 211 PM CDT...THE PUBLIC REPORTED A TORNADO JUST NORTH OF
DOWNTOWN MINNEAPOLIS. RADAR SHOWED THE TORNADIC STORM MOVING NORTH
AT 20 MPH.
* LOCATIONS IN THE WARNING INCLUDE...
ROBBINSDALE...
NEW HOPE...
CRYSTAL...
COLUMBIA HEIGHTS...
HILLTOP...
BROOKLYN CENTER...
NEW BRIGHTON...
ARDEN HILLS...
FRIDLEY...
SHOREVIEW...
BROOKLYN PARK...
MOUNDS VIEW...
PRECAUTIONARY/PREPAREDNESS ACTIONS...
IF YOU ARE CAUGHT OUTSIDE...SEEK SHELTER IN A NEARBY REINFORCED
BUILDING. AS A LAST RESORT...SEEK SHELTER IN A CULVERT...DITCH OR LOW
SPOT AND COVER YOUR HEAD WITH YOUR HANDS.Reflectivity-wise, the radar was very unimpressive. A few 40 dBZ splotches dotted the otherwise bland 25-30 dBZ rain mass that encompassed most of the radar scope. Certainly nothing even convective, much less severe. Velocity imagery likewise looked unimpressive initially. The only semi-interesting feature was a line of converging winds in the middle of the Dakota County warning. But as I zoomed in on Minneapolis, I did indeed note a few pixels of gate-to-gate velocities, which developed near the Minneapolis-St. Paul International Airport at 1:30pm and cruised up Interstate 35W and over downtown Minneapolis before dissipating near Columbia Heights at 2:16pm. Based on the location of the purported tornado reports and the radar imagery, it appeared like the tornado touched down about ten minutes before 2:00pm, a full 20 minutes before the tornado warning was issued by the NWS! Although rotation was evident on radar, it is highly unlikely that any radar algorithms caught this minute circulation center, and even so the weather conditions did not warrant a tornado warning. If I had spotted that rotation signature in time (which I don't even know if the NWS did), I would have likely held off on any warning. People began asking if we should head to the basement in a nearby building, but with the worst of the storm apparently past us, I decided not against it.
Base Velocity Radar Imagery from the KMPX Radar of the rotation signature (yellow circle) associated with the Minneapolis Tornado from 1830 UTC to 1916 UTC on August 19, 2009.
As the warning neared its expiration time, photos of damage began to appear on various local news outlet websites, and even a few pictures of what appeared to be a ragged tornado. But my attention quickly turned to the Dakota County storm, which had shifted northward into southern Washington and Ramsey Counties and prompted new warnings for my area. Again, due to the breath of the warning and since it looked like the area of strongest convergence was near the St. Croix River, I voiced that we at work were safe and that going to the basement was unnecessary. Plus, with the storm moving so fast, I was worried that that we would be caught outside in the heavy rain when the storm arrived. It was safer to stay inside than hike over to the basement. The convergence line passed over us with little more than some heavy rain and gusts of wind about 15 minutes after the warning was issued. The skies rapidly cleared as the storm departed, and I headed home at 3:15pm. If you haven't noticed yet, I didn't bring my cameras into work today and thus never took any photos of the event. Although all I saw was a blank, gray sky, so any photos would have been boring at best.
As the low tracked north and east, the NWS unleashed a slew of tornado warnings for just about every county from the Twin Cities to Hinckley and eastward towards Hayward and Eau Claire. A total of ten tornado warnings were issued for this event (each covering two to four counties). The storms weakened after 7:00pm CDT and the warnings subsided (see radar images below). However, a high wind warning was issued for the Duluth area as the center of the low passed directly overhead at sunset, bringing tropical storm force winds to the city.
NWS damage survey teams confirmed that four tornadoes developed in the Twin Cities from this complex. The Minneapolis tornado was ranked an EF0, while other tornadoes touched down in Cottage Grove (EF1), Hudson (EF0), and North Branch (EF0). The full report from the NWS can be found here.
Regional Radar Reflectivity Data for the Upper Midwest from 1900 UTC on August 19 to 0400 UTC on August 20, 2009. During this time, the low deepens and moves from SW Minnesota to Duluth.Why did this event occur? I don't know the answer for sure (and I don't think anyone does), but my guess is that strong, localized low-level wind shear, combined with a possible vorticity maximum associated with the rapidly deepening low pressure center almost directly overhead, created a unique environment that allowed for even weak updrafts embedded in the stratiform rain deck to tap into this instability tilt this strong wind shear vertically and drop brief, weak tornadoes. This would explain why it seemed like the tornadoes developed in the areas of heavier rainfall. I witnessed this firsthand when I went outside to observe the skies shortly after the Minneapolis tornado warning was issued. When looking upwards, I noticed a band of scud clouds moving northward, while barely above those clouds, another batch of scud was moving westward. There was little to no daytime heating due to the daylong rain event, so convectively induced severe weather such as hail and lightning were nowhere to be found. Below are surface observations from the Upper Midwest during the day on August 19 from 1400 UTC on August 19 to 0400 UTC on August 20. Note how the sea-level pressure in Duluth drops during this time from 1009.6mb to 997mb as the low deepens and moves overhead. A moisture tongue is also visible ahead of the trailing cold front behind the low, which may have helped aid in the development of the tornadoes.
I was unable to find any good upper-air soundings that showed the strong low-level wind shear, probably because the low spun up so quickly and was not near any RAOB sites during regular balloon launch times. Below are area soundings from that day. which do indicate somewhat of a veering wind pattern in the lower 300mb of the atmosphere in the general area of the low that day.
Upper Air Soundings from Chanhassen and Aberdeen at 1200 UTC on August 19 and Green Bay and International Falls at 0000 UTC on August 20.
In fact, the low never even showed up on the HPC surface analysis maps until 0300 UTC (below).
HPC Surface Analyses for the Upper Midwest from 1500 UTC on August 19 to 0300 UTC on August 20.
Overall, this was a relatively rare event where local conditions were prime for tornadoes, even without convective updrafts present. The day featured tornadoes that formed in the heart of the Twin Cities on a cold, rainy day, dispelling the myths that tornadoes cannot form in cites and when the temperature is below 70 degrees. I will never look at a stratiform sky the same way ever again.
