I had been watching this as a potential chase day for several days, as a warm front would move across Missouri and southern/central Illinois ahead of a strong low pressure system moving out into the plains from the Rocky Mountains. There were questions about the amount of instability and the possible effects of earlier convection, but there looked to be at least decent directional shear, and warm fronts in Illinois and eastern Missouri are often tornado producers. By around mid-afternoon, thunderstorms in the Missouri Ozarks were beginning to intensify and move northeastward. A check of the SPC Mesoanalysis pages showed that the storms were developing a little south of the warm front (which extended from a little northeast of Columbia, MO to near or just south of St. Louis). These storms were developing in an area where wind shear and instability were both maximized, with the mesoanalysis showing CAPE around 1500 and EHI (a measure of combined instability and shear) of 3 - definitely enough for tornadoes. I figured that if the storms could move northeast and become rooted in the warm front a little north of I-70, tornadoes were a definite possibility. So, around 3:50 p.m. I headed out, targeting Warrenton, MO or the area just west of there.
When I hooked up my computer and looked at radar after a pit stop in Forestell around 5 p.m., I could see a cluster of strong but poorly organized thunderstorms southwest of Warrenton. I moved west to the Flying J truck stop a few miles west of Warrenton, and sat and watched the storm and radar. At this time the storm did not look too imprssive visually (although my view of parts of it were obscured by rain), and not much better on radar. I did notice a new cell forming just south of the main cluster, or perhaps actually merged with it, but with a distinct southward extension, and this probably being the strongest radar return. In a little while the NWS issued a special weather statement for the possibility of gusty wind and hail for the area around Big Spring, High Hill, and Montgomery City, a short distance to my west. So I decided to head west on I-70 to see what I could see. By the time I was approaching the next exit, Jonesburg, I could definite updraft base to my southwest, so I decided to exit at Jonesburg and see what I could see. It was hard to find a place with a clear view, but it was evident that a large lowered area was forming under the updraft, and CG lightning was zapping down in the rain to the north or northwest of this feature. Finally I found a good viewing spot along the I-70 frontage road a mile or two east of Jonesburg. While I was looking for a good viewing spot I noticed a large inflow tail/beavertail on the north side of the lowered area. By the time I found the good viewing spot this was a little smaller but still there, and the wall cloud was very large and low, though partially obscured by rain. A picture of this feature can be seen here. Another special weather statement came out, mentioning the possibilty of nickel size hail and 40 mph wind. Everything said to me that this storm was taking on supercell characteristics.
After I watched the storm for a while, the meso began to wrap up in rain, and soon my view of it was hidden. At this point I figured the storm was going HP, and would be very diffucult to observe. But I moved back east on the frontage road to Warrenton to try to stay ahead of it and hope to somehow get a decent view. To my surprise, the rain either wrapped around to the north or dissipated, and by the time I was in the Warrenton area I could again see the wall cloud to my west. So I moved north 2 or 3 miles, until I found a parking lot with a somewhat unobstructed view. (This area, like much of Missouri, is largely wooded and somewhat hilly. There are few places with completely unobstructed views, but there are a few spots that are rather high and don't have too many trees, so if you can find one of these, you can get a fairly good, though never totally unobstructed view.) As I watched the storm from this spot, the wall cloud was huge - several miles across, and very rounded and deep on the front side. Closer to the back side, I noticed a very low area, and after a while, a rear flank downdraft (RFD) slot cut in around this area. There was no doubt I had a supercell now, but since I had the weather radio off to videotape and had my eyes on the storm, not my computer, I did not realize yet that a tornado warning had been issued by now due to radar indication of rotation.
Lowered area. A zoomed, enhanced view may be seen here
Slightly later, as the RFD became more evident
A tornado was reported 4 miles northwest of Warrenton. The local storm report gives the time as 6:40, but on video I have a SWS giving the time as 6:56. The pictures above (HD video captures) were taken around 6:48 looking west-southwest from two or three miles north of Warrenton. This is likely the mesocyclone that produced the tornado, but if either of the tornado report times is correct, it was at a different time from when these pictures were captured. OTOH, if you split the difference between the times, then it would be about the time of this video. So I don't know. Regardless, due to the distance and trees, I can't determine whether or not I have a tornado on video at this time.
Update, 8/19/10 - According to Storm Data, this tornado indeed occurred at 6:40, about eight minutes before the start of my video, and lasted for only a minute or so. So the tornado was over before the beginning of my video.
After a while I figured it was time to move north and keep up with the storm, so I did. After a brief stop or two in clearings in the woods where I had decent views, I noticed a west road option and could see a rain-wrapped meso to my west. I followed the road which zigzagged west and north for a few miles until I came to a north option. By this time the storm appeared to be cycling, with the meso to my west occluding and a new one forming much closer, to my north-northwest. I cautiously edged my way closer, until I was within a mile or so of the new meso. An RFD slot cut in from the south almost directly above me, and the wind began to blow strongly from the south. However, there was no rain or hail in the RFD, so I stayed in this spot and watched in amazement as a wall cloud formed and began to rotate rapidly. Soon multiple funnels formed under the wall cloud and rotated around one another:
Video (View this full-screen if you can):
As you can see in the video above, the rotation was intense and the funnels at times reached at least half way to the ground. This may well have been a tornado, but due to the trees and hills I could not tell what was going on at ground level. There was a tornado reported by law enforcement at almost this exact location, but the time does not match. I checked the time as I taped this rotation; it was 7:11 p.m. The law enforcement report was for 7:25, although I suppose that could be when they called it in. I doubt there was a tornado here at 7:25, because by then this meso was occluding and yet another handoff was occurring to a new meso a few miles east and a little north of this one.
Update, 8/19/10 - As I suspected, the time of the law enforcement report was initially reported incorrectly. According to Storm Data, the correct time was 7:11 - exactly the time of and in the same place where I saw the funnels rotating around one another. So as I suspected might be the case, this was a tornado, and I can now say that this is the first of two tornadoes that I captured on video on this day.
As the new wall cloud formed rapidly around 7:20-7:25, I broke off from the occluding meso and headed east and north to catch up to the new one. The wall cloud tightened and developed violent vertical motion. Around or shortly after 7:25, the tone alert on my weather radio sounded and a new TOR warning was issued based on rotation near Hawk Point. Within a minute of when I heard that, a tornado formed under the wall cloud. This was a brief tornado, or at least the condensation was brief. The condensation extended all the way to the ground, but only for about one minute - then it disappeared. The tornado was reported as a little east of Hawk Point, but I think the likely location was more likely near or just north of Hawk Point. I pushed the wrong button on my video camera so do not have video of this tornado, but another chaser, a student at St. Louis University, did get video. That video was taken from along route 47 looking north toward Hawk Point, and the tornado is nearly straight up the road, which supports my impression that this tornado was closer to Hawk Point than the estimated location in the LSR. I got my camera working again, and the wall cloud was still very tight and low. It is possible the first tornado was still on the ground at this time, but I don't think so. However, as I edged north and east to get closer, a second tornado formed under the wall cloud:
Video of second tornado near Hawk Point:
This second tornado appeared to be a little north and west of where the first one occurred, and moved toward the north or north-northwest. It lasted a little longer than the first one, between one and two minutes.
Update, 8/19/10 - According to Storm Data, a tornado was confirmed from 7:29-7:30 from one mile west of Hawk Point drifting north to a mile WNW of Hawk Point. This corresponds very well to the time and place of the tornado I captured on video, and is much more accurate location-wise than the initial report of a tornado 3 or 4 miles east of Hawk Point. Although the NWS only confirms this one tornado near Hawk Point in Storm Data, I saw (as I describe above) and another chaser videotaped (as linked above) a very brief tornado a few minutes before the one in the video above. This earlier brief tornado was associated with the same circulation as the one that produced the tornado I got on video, so I suppose you could argue they were both the same tornado, but there was a 2 or 3 minute period when there was no visual indication of a tornado. So I think there were indeed two separate tornadoes. The second one - that I caught on video - lasted longer, for a minute or two, and was slightly west and/or north of the first one.
Shortly after this tornado ended, the wall cloud became more diffuse and eventually disappeared, perhaps in rain. However, the storm remained tornadic and produced additional tornadoes after dark in the Silex area. I suspect it cycled again and that one or more new mesos were responsible for the subsequent tornadoes. This was a classic cyclic supercell - mesos would form, tighten up, and produce brief tornadoes or funnel clouds that seemed to move due north or even north-northwest as the overall storm moved northeast. Movement was slow, 15-20 miles an hour, so were it not for the hills and woods of Missouri, this would have been an easy chase. I suspect that the northward movement and frequent occlusion of the mesos as the storm moved northeast is why this storm produced a series of brief tornadoes rather than one or two longer-track ones. This is probably a good thing, given that the storm passed over or near several towns.
Total chase distance: about 200 miles.
Tornado reports for the day, from SPC
NCDC/Storm Data summary of New Truxton tornado
NCDC/Storm Data summary of Hawk Point tornado
Radar image from around the time of the first tornado (rotating wall cloud with funnels)
Radar image from around the time of the tornadoes near Hawk Point
Return to 2010 Severe Weather Observation page