This is a report on the storm chase of Stephen Jascourt and Todd Sprinkmann, including scanned images of some of Stephen's photographs. Todd recorded video, but I have not seen it yet. Our roadmap will take us first to a review/post-case analysis of the meteorology, then some brief pre-chase logistical notes of interest, and finally we'll "cut to the chase", starting with the storms in Missouri, then following along into Illinois. The photos are embedded in the "chase" portions, with large images viewed by clicking on the thumbnail images shown on this page. The photos were scanned in from print but I could not get the colors or sharpness to quite match the prints.
If you get lost and come back to the top or insist on skipping ahead, here are links to the three
major sections:
Data analysis - what happened
The chase in Missouri
The chase in Illinois
First, a word about links in this website. All links to inside pages (data, pictures, maps) will appear in this window and have a link to return to this front page which contains the entire story (you'll even be sent directly to the part of the story you jumped out of). All links to other sources, such as NWS storm reports, will each open in a new window.
| Click on icons to view data. Beware of large loops if you have a slow network connection. Discussion below with detailed notes on some data pages. |
 Raob skew-T |
 Surface plots (loop, 160k) |
 Low-level profiler winds (loop, 175k) |
Upper-level profiler winds (loop, 175k) |
 Satellite images (loop, 900k) |
 Regional radar composite (loop, 500k) |
The ILX raob, located several counties southeast of the tornadic storms at the time of balloon ascent, shows strong straight-line shear at all levels with a shallow region of good curvature and in excess of 2000 J/kg of 100-mb mixed layer CAPE thanks to a deep elevated mixed layer above 750 hPa. Three other features to note are the dewpoint dropoff above the surface, the presence of a modest cap, and the high LFC. The storms were occurring in a region of strong deep convergence and ascent, where the moisture was probably better mixed above the surface and the cap was weaker or eliminated, with both factors contributing to a lower LFC and more CAPE within the lower troposphere (e.g. parcel buoyancy realized at low levels - which is nil in this sounding). Also, the low-level winds were slightly more backed in the tornado locations and those locations were closer to the mid- and upper level jet cores.
At the surface, the outflow boundary from the overnight/morning convection apparently helped focus and strengthen the warm front, as temperatures in central MO quickly rose into the 80s but northeast MO and IL never saw 80 ºF. Note also that the winds remained backed over western IL and eastern IA but not over eastern or central IL. The RUC and Eta both had predicted that the low-level winds would back ahead of the upper jet streak even as they veered elsewhere - so this appears to be a dynamic response to the upstream jet core rather than just due to the outflow boundary.
Above the surface, the lower tropospheric winds strengthened significantly and backed somewhat through the afternoon and evening at the Winchester, IL (WNC) profiler location, which is a few counties south and east of where the tornadoes occurred. The outflow dome was evidently shallow, as winds at 1 km above ground were southwesterly over the top of the outflow dome even early in the day. Note that the winds were 45 knots just 1 km above ground level at 0100 UTC and the winds at 2 km above ground strengthened by 10 knots per hour for three consecutive hours from 2300 UTC to 0200 UTC.
In the upper troposphere, the strong jet core associated with the potent upper wave was evident upstream across KS/OK, with some diffluence over MO and with slight directional shear extending into the mid troposphere at WNC as parcels peel off anticyclonically in the jet exit region. Note that the backed surface winds are confined to the region in advance of the strengthening deep-layer winds as the jet system moves northeastward.
The visible satellite imagery clearly shows the morning outflow boundary, the stronger cap under the hotter elevated mixed layer over southern MO, convection first firing as individual cells along the prefrontal trough and later further west in short lines on a rope cloud along the cold front, the anvil mergers of the discrete supercells, and the locations of overshooting tops. Note that as late as 2032, the surface-based instability did not extend into northeast MO yet, as the cumulus field is still upstream and the area where the tornadoes later occurred has wave clouds indicating presence of a stable layer.
The radar reflectivity shows the tornadic cells reached their peak "intensity" in the more unstable air prior to producing the tornadoes, while the tornadoes were produced as the peak reflectivities decreased after entering less unstable air. Remarkably, the tornadic cells in western IL during the evening can be traced all the way back to extreme western MO or possibly even southeast KS in the early afternoon. However, tornado production occurred primarily after the storms crossed the outflow boundary into areas with southeasterly surface winds, despite tracking for hours in an unstable airmass with strong deep layer shear across central MO.
Pre-chase notes of interest
Special airfare deal: I was able to fly from Maryland at the last minute on a "weekend special" on United for just $155. I purchased the ticket on Friday, departed Saturday morning (less than 24 hours later), stormchased all day Saturday, returned Monday morning. This turned out to be a fantastic, affordable way for an east-coast resident to make a last-second decision to chase in the midwest, and the timing was lucky.
Chase partner: I called Chris Gullikson, one of my chase partners from my Wisconsin days, and he connected me with Todd Sprinkmann. Chris wasn't sure if he was going to be able to chase that day - turns out he did and he caught a tornado from the same storm we saw but on the IL side of the big river. Todd worked out great as a chase partner for me and he was able to pick me up at the airport to start the chase and drop me off at a friend's house near Chicago on his way home. Thank you, Chris and Todd!
Airport internet connection: While leaving O'Hare airport, I discovered a little internet shop, with private rooms with PCs with fast internet connections. I forget the exact rates, but it was on the order of a dollar a minute for use of a room. It's located in the extension of the corridor between the United terminals where it connects to the baggage claim area, so if you are walking to the baggage claim area through the underground corridor, you pass right by it. This enabled me to reconfirm Quincy as the target area immediately before setting out.
The target area as of the previous night's analysis and still the same through the airport data check was Quincy, IL where the lower- and mid-level jets were going to cross by evening. We headed there by way of I-55 and decided to stop at the Lincoln (ILX) NWS forecast office, which is only a few minutes off the interstate and easily identified by the radar dome out in open fields (which was how we found it without knowing an address or directions). The folks at ILX were relaxing between severe weather waves, recovering from the overnight/morning batch and preparing for the evening batch - it was good to enter at a lull. In addition to looking at data and having a little discussion, we also were given maps outlining the County Warning Areas (CWAs) for the various offices in the area we were heading to, as we would be near the intersection of the areas served by the St. Louis, Davenport, Kansas City, and Lincoln forecast offices, so that we could call in reports to the right office whereever we happened to end up, and we were given phone numbers for calling reports in to these offices. I should point out that ground truth is always helpful for the warning meteorologist, but it is indispensible at the far reaches of the CWA, where the radar beam is too far above ground to determine whether circulations extend to low levels.
Proceeding on I-72 to Hannibal, I noticed an abrupt change from sharp visibility to hazy conditions, perhaps indicating trajectories from the Yucutan that had supplied Mexican forest fire smoke to Oklahoma the previous day, impeding chaser views of large tornadoes there. It never cleared up after that point, ruling out some local source. Shortly thereafter, we started seeing anvils, soon blocking out the sun.
Todd had on previous chases connected up with Melanie Metz and thought she was going to be available for live radar updates, so we called her and sure enough, she was there and able to help [including several more times soon thereafter for updates]. She told us there were three distinct supercells aligned northeast-southwest, all moving toward us from about 240 degrees (which was to the right of the mean wind and well to the right of the surface wind!) at about 40 knots and all sufficiently distant we could continue west and intercept any one of them. So we had to decide which cell to go after. At first we decided to proceed west on 36 out of Hannibal, hedging our bet on either the southern or middle cells, as we could proceed west or we could turn north. The middle cell was at that time the strongest with the best-developed rotation, but I then decided on the southern cell based on best inflow and an environment favoring explosive development. When chasing, you always want to project evolution up to and beyond the point of your intercept rather than assuming everything will simply translate without evolving. It turns out that the middle cell tracked further north than we would have anticipated and produced the menacing-looking Canton-Lima tornado caught by Scott Weberpal, but over the course of the evening, our cell became the dominant cell with multiple long-track tornadoes.
|
|
We proceeded on the Route 36 bypass around Monroe City and exited at the location of the camera symbol on the map above. To the north was county trunk Z, to the south the road went into Monroe City, and we had fast east and west options on Route 36. We parked in an open area immediately north of the Z overpass over 36, with a broad sweeping view to the west, and noticed a fire truck parked on 36 under the overpass. There we waited for the storm to come to us, expecting the storm core to pass to our west and north based on NOAA Weather Radio, various reports, visual cues, and calls for radar updates, the last of which was to Tim Vazquez (Todd subscribes to his service) who noted that the middle cell was about to get seeded by our cell and that the mesocyclone for our cell should cross 36 just a few miles to our west and pass slightly north of our location. We were positioned perfectly! (ah, more perfectly than we could have realized...)
We saw cumulus (not tall) inflow bands with healthy-looking updraft bases aligned southeast-northwest feeding into the storm. Surface winds at our location were out of about 150º . We watched the storm core pass to our west and north, with updraft bases aligned northeast-southwest poised to track just to our west.
Then, as the storm core had passed to the north as expected, we saw a bright area to the west that looked like a rear-flank downdraft (RFD), as we also expected, and a pronounced lowering just to the north of it, as shown in this picture (click image to enlarge). Unfortunately, the tree blocks most of the lowering, but you can still see this storm structure. |
 |
The apparent RFD was advancing upon the lowering, with relative motion indicating stormscale rotation around it, but we did not see any rotation within the lowering. We assumed that the lowering was the mesocyclone tracking as expected (it probably was, as in past tense), so we drove north on road Z to stay just to the south of it as it tracked northeast. But by the time we reached road CC, I noticed scud drifting from the northwest under this feature and actually throughout, under the entire updraft base. Could the storm be gusting out under the meso? Something wasn't adding up. Clunk! A single hailstone, large enough to make a loud noise, hits the car, but it's barely raining. We're now at the liberty bell symbol (interpreted as alarm! bell) on the map. Consider that light northwest flow relative to rapid storm motion from the west-southwest is strong storm-relative flow from the northeast (do the vector subtraction). That means we're in a bad spot, plus there's probably large ambient vertical vorticity if this is old RFD air.
We turn around and notice an updraft base crossing Z just to our south and arcing back to the southwest, and further southwest by several miles, is a rapidly surging rain sheet - a new rain shaft with obviously strong westerly winds. Apparently, this is what tracked across Monroe City. I assumed this was a new RFD surge upon a new mesocyclone which the scud was drifting toward, e.g. just to our southeast now. Instinctively, we blasted toward the south, hence the airplane symbol on the map, back to where we were originally sitting, to get to the other side of this new RFD surge before it wrapped up into a tornado, which I knew was imminent. As we headed south on Z and neared the bridge over 36, the surge passed directly in front of us and we directly into it. The rain sheet was blowing from the west (or southwest) at all times. There were no winds from the east, except that trees were moving in all directions, probably because of the strength of the wind - fortunately, we had selected an area where there were not a lot of trees beside the road. We may have driven through winds of 80 MPH as we passed by our original position, but it appeared to be straight-line RFD winds, and the rotation center it was moving toward had to be to our east where the scud were moving into and where the easternmost bulge in the updraft was that the RFD surge was catching up to. Unfortunately we don't have video or photos of this portion, because we were in decide/react/alert mode.
As we turned onto 36 eastbound and the RFD surge had moved ahead past us, a needle funnel formed to our northeast and then broadened. I tried to call the St. Louis NWS office but their number was busy for the next 15 minutes. We then proceeded along 36 to catch up as it had moved a good bit ahead of us and to the north of our route (we did not cross its path, so we did not encounter damage in this area) and was gaining distance as it moved northeast while the road goes east. We stopped at some points to take video and pictures, and I was finally able to reach the NWS office at 6:50 PM CDT, when we were at 36 and 24, due south of Palmyra (roughly 6 miles east of Ely) when the tornado was still visible to our north. I don't remember exactly where most of the pictures were taken, but I think it was in the vicinity of 36 and U, close to Ely. The picture on the upper left (with no tornado) was taken earlier, around the time the tornado formed, the picture on the lower left was taken next, and the three on the right were taken sequentially in time at one location (along with more similar pictures not posted). |
|
|
We then headed east on 36 into Illinois and north on 172 to try for another intercept, as the storm was crossing the Mississippi River between bridges, so we had to play catch-up.
Another chaser, Mike Hollingshead, provides a different viewing angle on exactly what happened with tornado genesis over Monroe City. He was further south, in Monroe City, and captured on video airborne debris and a roof piece getting ripped off and lofted. Despite his emphasis that rising debris indicates a tornado rather than straight-line winds, strong straight-line winds will loft sheet metal and other items with large exposed surface area. However, his comments in an email to me about falling debris and other details suggest there was a tornado in Monroe City shortly before he arrived on the scene, so perhaps its axis tilted to horizontal (as he describes) inside Monroe City and dissipated there while a new one formed further to the northeast, all embedded in this surge - we need the help of other chasers with different vantage points to sort it all out. I have taken the liberty of copying one of the pictures he posted that I think shows a needle funnel under a wall cloud and I enhanced the colors by adjusting the brightness, contrast, and gamma settings in the image, cropped the picture, and saved it as a bitmap (.bmp) instead of jpg in order to exactly preserve these changes (jpg compression loses some detail). It appears to me that there is a narrow funnel immediately to the left of the red arrow I inserted. This means that the tornado had already formed to the distant northeast of where the strong winds were ripping the roof off the store. The shape of this barely-perceptible funnel and the cloud it was attached to are similar to other pictures of the early stages of the tornado on his web page. This further confirms that the strong winds we drove through were straight-line winds from the RFD surge during tornadogenesis which occurred just to the east of highway Z. There may have been additional eddies and vortices within the RFD surge which may have complicated the picture. |
|
We proceeded on 172 north toward Quincy with the tornadic cell far enough to our east or northeast that we could see the backside towers (click image to enlarge). It appeared that cells were moving northeast but that the history of this storm was for new mesocyclones to form at the south end, and the backside development and region of dark skies to the east suggested this pattern continuing, shifting the track of tornadic activity more in the direction of east-northeast rather than due northeast. So we got off at Rt 104 to go due east instead of taking 24 northeast. |
 |
Heading east on 104, we made up ground on the storm and saw a suspcious lowering to the distant east-southeast. We continued due east on the same road instead of diverting to the southeast where 104 turns toward Liberty. The lowering moved northeast, crossing 104 too far ahead for us to see it, but we caught up some and it was to our northeast at the time of these pictures (click to enlarge). We were too far away to see what was underneath that low-hanging base, but based on location, appearing to be connected to the updraft of our original storm, and where we later crossed the damage path, I'm fairly certain it was the Liberty-Mount Sterling tornado which Chris Gullikson and Don Lloyd caught forming right in front of them. |
 |
 |
|
|
We continued due east, passing through a nature preserve in a valley, and turned north at Kellerville, as indicated by the car icons on the above map. At the liberty bell (again, think "alarm bell"), we encountered a vigorous-looking linear southwestward extension of rain-free updraft base with a core area to the distant northeast and another rain area immediately northwest of our position, with brighter sky behind. This was obviously flanking development trailing the original cell, but it appeared to be maturing. Thinking it could form a mesocyclone and perhaps wrap up a tornado at this location, we stayed put safely just to its south and monitored it for a while. A curious farmer driving to his nearby home inquired what we were watching and also waited a little while at our urging. Eventually the potential danger showed no signs of further developing anything other than heavy rain and possible straight-line winds and had moved to the east of this road, so we continued northward to get to 24. But we quickly were stopped by downed wires across the road. Mature trees had been cleanly snapped off but not uprooted and fences were heavily damaged. This coincides with the track of the Liberty-Mount Sterling tornado, with the track of the lowering shown in the above two photos, and also with the track of the feature we waited to pass which may have had strong winds.
After these delays, we were able to proceed north to 24, northeast along 24, and later headed east through a heavy rainshaft of another trailing development but never caught up with the tornadic storms as they moved northeast. We saw a cell intensify just to the south of the storms were were chasing and appear to merge in with them, but we didn't know it had tracked all the way from central MO as a weak cell. Looking at the radar loop (see data links above), this was the same cell we saw to the distant south of the Monroe City storm when we were still in MO. Survival of this modest cell, its rapid movement (much faster than the other cells it caught up to), its sudden strengthening prior to merger, and that this merger happened just before the F3 tornado in South Pekin is all quite curious.
Looking at the tornado tracks (click on map icon beside the above map), it appears that the Rushville tornado may have been a continuation of the Liberty-Mount Sterling tornado. Also, it appears that the Monroe City cell may have spawned most of the tornadoes across central Illinois.
