Southern Kansas supercell and tornadoes on 6/12/04
Photos and detailed case environment study by Jon Davies

On 6/12/04, for the third Saturday in a row, a supercell produced severe weather in south-central Kansas (KS).  This storm developed in eastern Barber and western Harper county and moved eastward, eventually producing a damaging (F3) tornado near Mulvane in northern Sumner county just south of Wichita.  I was able to follow the storm through most of its life cycle and, along with hundreds of other locals and storm chasers, photograph several of the tornadoes:

      The first photo is the 1st tornado seen from the KS turnpike looking northeast from southwest of Mulvane just after 7 pm CDT.  The 2nd photo is looking due north at the Mulvane tornado in progress.  The 3rd photo shows the intense rope-out stage of the Mulvane tornado looking northeast as it destroyed a house (F3 damage) near the end of its track.  The last photo is the 3rd tornado near Rock KS as the sun set around 8:30 pm.  See the track map below and NWS Wichita's page for tornado locations:
                                                                                                 <tracks and F-scales

The following discussion and graphics are a brief case study of the setting and environment associated with this storm. 

The forecast 500 mb map for late afternoon and evening on 6/12/04 showed a strong upper wave moving through the central plains to help trigger thunderstorms in an unstable air mass.  By mid afternoon, thunderstorms were already developing in central and south central KS as seen on satellite photographs:

<Eta 12 hr fcst 500 mb winds           <Eta 12 hr fcst CAPE
<20 UTC visible satellite photo         < 2130 UTC visible satellite photos

The storm that eventually became tornadic is indicated by an arrow on the 2130 UTC satellite photo above.  This developing supercell was originally very high-based, but after moving into higher dewpoint air, the base came down significantly prior to the time of the tornadoes after 00 UTC.

A suggestion of the potential for this evolution can be seen from the surface map at early to mid afternoon.  A number of boundaries were present on the surface map, including a double dryline configuration over central KS and northern OK.  The easternmost dryline separated dewpoints in the low to mid 70s from dewpoints in the 60s further west, and winds were more backed east of this "2nd" dryline.  From the satellite photos above, it can be seen that the storm that eventually became the tornadic supercell formed between the drylines in air with larger temperature/dewpoint spreads, but was poised to move east into higher dewpoint air (the "2nd" dryline) with smaller T/Td spreads:
<20 UTC sfc map

Estimated SPC mesoanalysis parameters at early afternoon also suggested a favorable environment to the east of the developing supercell in south central KS.  The storm was tied to a moisture convergence center that had been persistent near the KS/OK border, and would have unrestricted access to air with the largest CAPE (>3500 J/kg mixed-layer) located to the south and east.  It would also move toward CAPE/shear combinations more favorable for tornadoes as indicated by the 0-1 km energy-helicity index (EHI) and significant tornado parameter (STP).  The LFC height field also suggested a higher humidity environment east of the the developing storm with more CAPE close to the ground, as well as lower cloud bases (LCL/cloud base cannot be above the LFC, by definition).  This, and the deep layer shear (0-6 km), appeared favorable for tornadoes as the developing storm moved further east:
<19 UTC sfc moisture convergence  <19 UTC MLCAPE & location of future supercell<19 UTC 0-1 km MLEHI                 <19 UTC estimated MLLFC 
<19 UTC 0-6 km shear                     <19 UTC significant tornado parameter

On satellite, it could be seen that the south-central KS storm became the dominant storm over KS and northern OK as the afternoon and evening progressed.  Other storms in central KS were less favorably positioned relative to the most unstable air over OK, and may have been hurt by the large anvil and position of the southern KS storm that increasingly dominated the environment with no real "competition" from other storms to its south.  Notice that the central KS storms died as the southern KS storm increased in intensity.  The southern KS supercell  was so dominant that it pulled another developing cell (seen just to its south) into its own circulation.

    < visible satellite at 2230, 2330, and 0015 UTC

The double dryline structure on the earlier surface map could be seen clearly on radar, with the "2nd" dryline (west edge of deeper moisture) located just west of Wichita.  As the developing southern KS supercell moved east of this feature, it encountered lower LCL and LFC heights generating lower cloud bases, and an increasingly favorable shear environment with backed southeasterly winds:

The difference in environments west and east of this "2nd" dryline feature could be seen from RUC analysis profiles estimating the local environment near Harper KS at mid afternoon (near the developing supercell west of the "2nd" dryline feature), and southeast of Wichita just before the first tornado at 00 UTC (McConnell AFB just north of Mulvane KS, east of the "2nd" dryline feature).  In the 00 UTC environment near Wichita, note the dramatic decrease in both MLLCL (cloud base estimate) and MLLFC (location of low-level buoyancy above ground), and the strong increase in shear as indicated by a clockwise looping hodograph:

<21 UTC RUC analysis near Harper KS  <00 UTC RUC analysis near Wichita-Mulvane KS

From this information, it is not a surprise that the southern KS storm became a long-lived supercell that produced multiple tornadoes.  The favorable CAPE/shear environment for this storm could be seen from the estimated 0-1 km MLEHI field at 00 UTC (tornado tracks 00-02 UTC also shown):

<00 UTC 0-1 km MLEHI

RUC forecast data 12 hours prior to the event (from Earl Barker's site) provided clues that the environment would be favorable for tornadoes over southern or central KS if storms formed and persisted in that environment.  Forecasts of 0-1 km MLEHI, MLLFC, and 0-6 km shear (below) suggested a favorable environment for tornadoes ahead of the dryline bulge in southern KS where larger EHI (3.0-4.0), lower LFC heights (< 2000-2400 m), and larger 0-6 km shear (30-40 kts) were collocated (see arrow):

    < 12 hr RUC fcst EHI, LFC, deep shear valid 00 UTC

These forecasts were confirmed with the evolution of observed surface features and estimates of environment as the afternoon progressed.  Forecasters with an awareness of these factors who monitored their evolution using observed data would have been well prepared for the tornado potential that was eventually realized this day.

- Jon Davies 6/15/04

back to Jon Davies page