Albert Pietrycha


On the late afternoon of 9 June 2003 several cyclic tornadic supercells developed over the Central Plains. At the time of this writing there are three known storms that spawned tornadoes; two storms located in north central Nebraska and one in northwest Kansas. The storm this web page primarily focuses on (storm A) initiated near a synoptic triple point and propagated eastward along a warm front. The other two storms (storms B and C) initiated along a dryline and moved generally to the east. Based on reliable chaser accounts, storm A produced four tornadoes, storm B 3 tornadoes, and storm C the Kansas storm, one tornado. There is seemingly nothing extraordinary about the day's events in terms of the synoptic and/or mesoscale setup. None-the-less, the purpose of this web page is intended to merely document the event and to provide limited data for those interested in reviewing the case. The North Platte (LBF) NWSFO damage survey was utilized for the damage ratings.

A warm thank you goes to the following individuals for involvement with this event; Paul Sirvatka (Meteorology Department head, College of Dupage, Chicago) for his outstanding and detailed nowcast support, Jared Guyer (NWSFO, Hastings), for providing the satellite, radar and plan-view profiler data contained within, and Jon Davies, Mike Umscheid (NWSFO, Dodge City), and Roger Edwards (SPC) for permission in using a few of their photos. All photos are copyrighted by either Jon Davis, Mike Umscheid, Roger and Elke Edwards, Todd Holsten (NWSFO Hastings, my chase partner this day) or myself (e.g., don't lift the photos without first seeking permission).



Beginning with the 'punch line' of the event, as previously stated there were three storms of interest. Since I chased storm A, emphasis is placed on that storm. The storm initiated immediately down stream of the triple point along the warm front near the South Dakota/Nebraska border, northwest of Springview, NE. As a young supercell it went through a series of splits with the dominant updraft (right mover) remaining anchored on/near the warm front. Once the right mover became established it propagated slowly east-southeast along the warm front. Due to the lack of in situ data under the updraft, it is not known how far within the cool side of the warm front the updraft was located. While the storm was still west-northwest of Springview it produced a brief, non-condensation tornado (F0).

Over the next hour or so, the storm continued southeast towards Stuart. During that time period, a series of occlusions were observed with rapid cloud base rotation. The base of the storm continually lowered with each occlusion as the storm modified its environment. Once the updraft was ~8 northwest of Stuart the storm produced a second tornado.

Tornado genesis appeared relatively slow as the condensation funnel and subsequent debris cloud were slow to materialize based on my vantage point, initially 3 mile southeast of the yet to develop vortex. Upon the funnel's development, its principle axis extended nearly on the horizontal and took a several minutes before becoming upright. During that time, the funnel's shaped visually went through rapid changes from thick to thin and back to thick. There was a brief debris cloud during that time period. Finally the condensation funnel become vertical and descended to the ground. The tornado crossed NE HWY 137 ~6 north of Stuart and moved southeast at 15 mph. The vortex remained mainly over wet corn fields. I was located 1.5 miles to its south and traveled nearly parallel with it. The tornado persisted for ~9 min before dissipation (F0). Concurrently, a relatively small, non-tornadic storm moving east-northeast, merged into storm A's forward flank (see radar loop below).

During and/or shortly after the storm merger, a new wall cloud rapidly developed ~10-12 miles southeast of the now dissipated Stuart tornado. Unlike the Stuart tornado, tornado genesis was very rapid as a large, multiple-vortex tornado soon developed northeast of Atkinson. The tornado moved southeast, and remained north of O'Neill. Based on the LBF damage survey, the tornado dissipated near the Holt-Antelope county line east of Page (F3). The tornado was very dusty and at times difficult to view. The storm produced a fourth tornado immediately west of Orchard. The non-condensation tornado persisted for ~1.5 min (F0).

Storm B initiated along a surging dryline, becoming tornadic in Custer Cnty, NE. Based on personal communication with Roger and Elke Edwards they witnessed two slow moving tornadoes. The first occurred ~5 miles west of Sargent at 2359 UTC and persisted for 11 min (F0). It was described as an elephant trunk. The second tornado, described as a fat cone wrapped in rain, developed at 0052 UTC 2 miles northwest of Comstock. The tornado was viewable for ~5 min (F1). There were numerous chaser reports of a third tornado 4 miles southwest of Sargent at ~0020 UTC (F0). This would have been the second of the three tornadoes.

Storm C as with B initiated along the same dryline in northwest Kansas. The storm became tornadic at ~0044 UTC northwest of Hill City. The tornado moved east and persisted for 12 min. The tornado was rain wrapped but viewable, and described as at not a major tornado but at times becoming the shape of an elephant trunk (Chuck Doswell, personal communication). Damage rating of the tornado is unkown. Click here for pictures of the tornado on Chuck Doswell's 2003 chase summary web page.

The storm A pictures below are as follows. First row: Tornado #1, 1, 2, 2. Second row: All Tornado #2. Third row: Tornado #3, 3, 3, 4.

Roger and Elke Edwards witnessed two of the Custer County tornadoes (storm B). Roger has provided an excellent narrative to accompany his pictures (click the image to the right). The page is well worth a visit for anyone learning storm, and/or tornado structure and evolution.




At 12Z conditions appeared rather quiescent. A weak lee low was located over eastern Wyoming with modest warm temperature and moisture advection over the eastern High Plains. As the day progressed, the surface low deepened in response to an approaching mid-level, short-wave trough with warm frontogenesis occurring across Nebraska. The warm front lifted north and became established along the Nebraska/South Dakota border. By 21Z the surface low was located in south central South Dakota along the warm front with a dryline extending soutward down though western Kansas and into northwest Oklahoma (one could argue the feature was not a dryline but rather a lee trough south of Interstate 70 in Kansas; an essay for another time). East of the dryline, low-level moisture surged northward within the warm sector up into north central Nebraska. The moisture advection is one of the more intriguing issues of the day. Also of interest, was the rapid dryline surge across central Nebraska prior to initiation. Through the remainder of the afternoon the surface low continued to deepen as it propagated along the warm front with the dryline continuing to rapidly mix toward northeast Nebraska.

The following links below are to unanalyzed surface data less the 21 and 00 UTC maps linked to the right. The surface obs use a standard station model with temperature and dewpoint (°F). Winds in knots with one pennant, one full barb, and one half barb equal to 50, 10, and 5 knots, respectively. Mean sea level pressure (mb), 3-hr pressure trend at the synoptic hours, cloud cover, and flight conditions; MVFR (blue), VFR (red).

12 UTC
15 UTC
18 UTC
22 UTC
23 UTC
01 UTC
02 UTC
03 UTC


I've subjectively analyzed the 21 and 00 UTC maps ( I mouse plethed the maps using Adobe Illustartor after first doing hand analysis). Standard frontal color scheme used; warm front red, cold front blue, dryline brown (dotted), troughs black (dashed). MSLP contoured every 2 mb, isodrostherms contoured every 5° greater than or equal to 60°F, and shaded greater than or equal to 65°F. MSLP pressure fall centers shaded for areas >= 3 mb 3 hr-1 (21 UTC); 5 mb 3 hr -1 (00 UTC). The purple asterisks represent the locations of the tornadoes.


The upper air data is straight forward on this day. Several pertained features to note. A mid-level trough located over the Central High Plains moved slowly east during the day. Strong low- and mid-level warm air advection persisted across the Central Plains with a pronounced elevated mixed layer over southwest Kansas. Under the elevated mixed layer a relativity narrow channel of low-level moisture advected northward out of western Oklahoma and up through central Kansas into central Nebraska (see the soundings below).

Hyper links to the 12 and 00 UTC upper air data are below. The 12 UTC data are unanalyzed charts from the College of Dupage. In addition to the RAOB data the charts also contained profiler data (blue wind barbs). The 00 UTC data are linked from the Storm Prediction Center and are objectively analyzed.

12 UTC

Click here for the 00 UTC data.


Upper air charts use a standard station model in abbreviated format; temperature and dewpoint in Centigrade. Winds as in the surface maps.

The regional upper air soundings are rather revealing in terms of the low-level moisture advection during the day across the Central Plains. Note that at 00 UTC the AMA and LBF soundings were west of the dryline.

12 UTC
18 UTC
00 UTC



CAPE/CIN calculations were derived from lifting an undiluted surface parcel and using a virtual temperature correction



Limited 19 and 00 UTC profiler data can be obtained by clicking on the image to the right. The progression of the short-wave trough and resultant increased flow are clearly identifiable in the data.



Winds as in the surface maps.



A GOES-12 visible satellite animation is to the right. Rapid scan imagery was utilized when available but there are gaps in the data. The animation is fluid if using Netscape.


Broadband will be your friend to view the animation.


The following links are to NIDS WSR-88D data. When viewing the radar reflectivity animation note the cell merger of a non-tornadic storm into the forward flank of Storm A. Numerous occlusions are also identifiable in the data. Furthermore, observe the rapid change in reflectivity character with storm A shortly after the storm produces the third tornado. The supercell quickly transitions from a classic toward an HP. Storm B's evolution can also be viewed in the animation.

In the storm-relative velocity animation, the mesocyclones are easily identifiable in both storms A and B. I hope to soon replace the NIDS imagery with 8-bit data via AWIPS.

Reflectivity animation (1.6 meg)
Reflectivity image at 0016 UTC of storm A
Reflectivity image at 0107 UTC of storms B and C
SRM Animation (1.4 meg)
SRM at 0042 UTC of storm A

All WSR-88 imgery is at the lowest radar tilt; 0.5°.


The End