April 30, 2003 Northwestern Missouri Supercell

The first tour chase day of the year was treated to a fairly intense, almost tornadic supercell. In fact, this storm may possibly have produced a tornado just before we caught sight of the low-level mesocyclone to our north near Maitland, MO. Fast storm motion and crazy roads prevented a more timely and proximate intercept.

Note: In this day's synopsis and the ones to follow, all terminology numbered (1), (2), etc... is given further elaboration below the images at the corresponding footnote number. Refer back to these explanations in synopses given for future events.

Meteorology:

The meteorological set-up for this event was a warm-up to what was to become one of the most prolific tornado-producing weather patterns in history during the latter half of the expedition. As always, we used the model forecasts the day before (and the specific observation that the low-level moisture was returning northward in the wake of a springtime cold front) to get a general idea of where to be the next day (April 30). This info suggested that the region from southern Kansas to northern Missouri would be one to scrutinize. On April 30, the observations provided more clues about where conditions favorable for supercell storms might materialize. In the wake of a cold front, low-level moisture was surging northward into the central plains ahead of the residual polar frontal boundary - hereafter referred to as the pre-dryline boundary (see Fig. 1 below).  Using the morning's soundings (1)  launched at Dodge City (Fig 2., red solid line=temperature profile, red dashed line=dewpoint profile) and Topeka, KS (Fig. 2, blue lines) and forecasting a surface-based lifted parcel possessing a surface temperature and dewpoint of 78/63 respectively, I created a composite sounding that I hoped would represent atmospheric conditions over the target area later that day. The modified "guesscast" sounding reveals CAPE eyeballed at about 2800 J/KG and CINH of 150 J/KG. Of course by mid-afternoon surface temperatures in the warm sector ranged from the mid 70's to the lower 80's (refer back to Fig. 1), suggesting that I had been a bit conservative in my surface temperature forecast used to generate a lifted parcel. Nonetheless, the composite sounding analysis (one of my favorite forecasting devices for gathering all of the relevant information onto one diagram to reveal the likely state of the atmosphere at peak heating) suggested that there would be plenty of instability throughout the warm sector wherever the capping inversion present over the area (i.e., the remaining CINH eyeballed from the forecast sounding) could be removed locally by lifting. Thunderstorm development looked to be possible along the pre-dryline boundary trailing southwestard across central/southern Kansas (black dashed line in Fig. 1) and perhaps along and just north of the warm front (the leading edge of the warm, moist air) extending eastward into northern Missouri. The final task was to forecast where vertical shear would be strongest (2) and most well-matched (4) to the CAPE because this factor is what determines convective mode (supercell or otherwise). Aloft, a current of moderately strong west-southwesterly flow was superposed over the moistening low-level airmass (Fig. 3), with stronger 500 mb winds noted to be edging into KS from the southwest. A glance at the 850 mb analysis (Fig. 4; valid 19Z) generated by the SPC mesoanalysis page (3) showed a fairly uniform 25-30 kt SSW flow at this pressure level over the plains. Hence, the one place where the vertical wind profile would be sheared more than any other (in the absence of significant low-level pressure falls farther south along pre-dryline boundary) was along and north of the warm front and downstream of the triple point (warm-front/predryline boundary intersection) where surface winds would maintain the most pronounced easterly component. Moreover, it appeared that the most convergent low-level flow (and thus the region where lifting was most likely to be sustained) was going to be located near the triple point. Thus, we targeted the area from northeastern Kansas to extreme southeastern Nebraska/northwestern Missouri. Radar downloads (Fig. 5) and visual observations revealed the show beginning east of Falls City, NE and we moved north and east to intercept.

Chase Summary:

Leaving OKC at 7:30, we made our way to Topeka, KS. My optimism regarding the potential for visible tornadoes had evaporated the evening before due to the upper-tropospheric wave responsible for surface cyclogenesis being progged to fill/weaken rapidly due to stretching deformation (a tendency apparently underestimated rather badly by most models 2-3 days prior to the event). Nonetheless, we were optimistic about the likelihood of HP supercell thunderstorms initiating along the frontal boundary from NE KS to S IA/N MO. The focus of our afternoon data analysis was on finding the likely initiation point and best environment for daytime supercells; NE KS to S IA is an unacceptably large target area when storm motion vectors are going to be 35-40 kt. We became convinced that initiation would be near Falls City, NE with rapid intensification and maturation of supercells across extreme NW MO. Our mistake was that we didn't have enough confidence in this to set up shop downstream of the expected initiation point - there have been many times when I overshoot the area of initiation and have to backtrack westward. We moved north from TOP on Hwy 75 at about 4pm CDT, observing congested (albeit mushy) CU and TCU to our NW through E. At the Hwy 36 junction we observed glaciated convective towers to our not too distant N and NNE; additionally, the developing congestus N through NW was being visibly undercut by scud apparently associated with the leading edge of the cold front, so we continued on eventually crossing into MO east of Falls City. When we turned north on Hwy 59 in east-centralHolt county we could make out much cloud base structure associated with our target supercell. We observed a well-defined dryslot with developing RFD gust front; to the NE of that was a very ominous block-shaped lowering/wall cloud that definitely had a tornadic appearance. We were too far to discern any detail. We continued stair-stepping N and E and when we were just east of Maitland on CR A we observed a fully occluded wall cloud with truncated cone shaped funnel cloud about 4-5 miles NNW of our location; the wall cloud and accessory scud tags were wildly rotating and I estimate its location to have been close to the town of Skidmore; I don't remember the exact time (too busy navigating to keep a log) but it was probably around 6-6:15pm and we were too distant to confirm/deny that the funnel cloud was actually a tornado. I would not be surprised if a large tornado occurred somewhere N or W of Skidmore. Unfortunately, we had been effectively outflanked by the storm so could not achieve the desired position due E or ENE of the meso. New mesocyclogenesis was rapidly underway to our immediate ENE and the RFD gust front was loaded with vorticity and wild, fluid vertical motion. Soon thereafter, the storm tended more to HP characteristics and developed a menacing rear-flank core with encircling shelf cloud. However, it refused to give up on trying to develop a forward flank updraft/meso as inflow stratus continued to fly into the storm from the E and SE and scud rose and attached to cloud base at the E edge of the core. East of Conception Junction in eastern Nodaway County we were briefly overtaken by the southern edge of a tightly circular precip curtain, the apparent center of low-level circulation passing just to our north. Our winds turned from SSE to W, gusting to near 50 kt. We finally got to the junction of US Hwy 169 and CR O at Gentry. The storm was now decidedly HP in character with a substantial surging RFD core/gust front bearing down on us from the SW and a less menacing FF core to the NW; we were sitting in the notch at this crucial intersection with no more east option. To our immediate SW and almost overhead a patch of cloud base was swirling wildly as the RFD core pivoted around from the W and SW to overtake us. The rotating cloud base moved quickly to our east across open pastures near Gentry, MO and we were forced north into the SW edge of the FF core,letting the more dangerous RFD core pass to our SE. After the maelstrom passed we once again attempted to outflank the storm but decided to blow it off at I-35 in favor of targeting a more isolated storm SE of STJ. Downloading radar as we drove (my goodness gracious, I never failed to get a wonderful cell phone connection yesterday), we headed directly to I-35 exit 40 at Lathrop. Too late (we heard the warming for the spotter-indicated tornado near Edgerton); it was an interesting looking sodacan LP storm (we got 1" diameter in the vault and cars were briefly blocking the interstate at an underpass) with an amusing array of sheriff deputies and other gawkers staring up at the sky. On the way south, we were mightily distracted by a ridiculously explosive convective bomb forming over STJ. As we drove south on the interstate we watched as the fat convective tower expanded, grew, overturned and went crazy. We turned around at Lathrop and stopped west of Cameron on Hwy 36 for a very entertaining lightning show (great anvil crawlers in the vault/overhead anvil). Stayed around until the storm lost supercellular characteristics and lined out.

Figures and Pictures:

Fig. 1: A surface map showing conditions at the earth's surface over the central plains 12pm CDT April 30. The leading edge of the modified tropical airmass is denoted by the warm front (conventional red line with semicircles); an analyzed boundary separating slowly modifying post-cold frontal air from the moist airmass to the east (black dashed) intersects the warm front over central KS and this "triple point" is a sensible target for storm initiation later this day. The true dryline is analyzed far to the southwest (brown stipled line).	Fig. 2: Composite sounding.	Fig. 3: 500 mb analysis valid 19z	Fig. 4: 850 mb analysis valid 19z
Fig. 5: Radar image showing first convective echoes of the day.

First view of the storm's action area north of Maitland, MO (looking north). It has strongly evident tornadic supercell characteristics, including an RFD dryslot occluding/exposing the mesocyclone; wall cloud and funnel. There might well be debris beneath the funnel (making this a tornado) but we are too far from the feature to confirm.	Another shot with wall cloud/funnel circled.	Same.
Farther east at Conception, MO looking north-northwest. Note the beaver tail streaming in from the east behind the tree in the right foreground and ragged/scuddy albeit fairly circular/compact base. Moderate cloud base rotation was noted and shortly afterwards rain/hail curtains started blasting us as they rotated around this developing low-level mesocyclone.	One last view of the mesocyclone to the north at Gentry, MO. We ran out of east option here and got munched shortly thereafter. Note the stubby tail cloud feeding the circulation.

FOOTNOTES:

(1) Soundings are balloon-borne measurements of atmospheric temperature, dew point temperature, pressure and wind; this provides a graph of the vertical structure of the atmosphere at a given point at a specific time and represents arguably the single most important dataset collected daily to probe the atmosphere and initialize the numerical weather prediction models. An analysis of individual soundings is a very important part of each day's severe storm forecast; unfortunately the paucity of this dataset makes it a challenge to forecast how the vertical atmospheric structure will change. Determining how warm and moist the input lifted parcel will become, how much CAPE/CIN will result and how much lifting is necessary to remove remaining CIN can nonetheless be a very enlightening process that forces the analyst/forecaster to focus on the data and make a best guess of how the atmosphere will evolve over the region of interest. Then, as conditions change during the day (particularly at the surface where hourly observations are collected over a much finer network than aloft) one has at least some clue as to how unstable and capped (or uncapped) a specific area is likely becoming just due to changes in surface thermodynamic properties alone.

(2) It is well established from theoretical, modeling and observational results that the interaction of nascent thunderstorm updrafts with a vertically sheared wind (i.e., a wind that changes speed and/or direction with respect to altitude) can result in a multitude of outcomes, with thunderstorm updraft rotation about a vertical axis being increasingly likely with increasing shear magnitude and the total shredding of storm updrafts occurring if the shear is too great for a given amount of CAPE and/or forcing. The way in which the shear within each layer varies with height and the strength of the mean horizontal flow in the storm-bearing atmospheric layer are the primary determinants of whether right-moving supercells are favored.

(3) The products generated on this page are analyses that blend 1-hour forecasts generated from the NCEP RUC model with the most recently available observations. The chief advantage of using model analyses rather than true observations is that the native model grid possesses much finer temporal and spatial resolution than do the observations.

(4) Some of the same studies revealing the importance shear flow-updraft interaction for the rotation and maintenance of the latter also reveal the necessity of having "well-matched" CAPE and shear; too much shear relative to the realized CAPE will destroy a storm before it gets going and too little shear will result in disorganized storms that fail to attain a rotating updraft. As a complicating factor, a storm interacting with an extant boundary (e.g., an outflow boundary left by overnight storms) oriented to the right of the flow in an environment possessing lots of low-level moisture and large CAPE may become a significant tornadic supercell in the absence of significant environmental shear.