|
|
|
|
| 1. Event overview
On the afternoon of 9 June 2005 several supercell thunderstorms produced tornadoes across portions of western and north central Kansas. Within the Goodland (GLD), Kansas National Weather Service's (NWS) area of warning and forecasting responsibility (CWA), seven tornadoes occurred over a three-hour period; 1 F2, 1 F1, 5 F0 on the Fujita scale. The table below, ranked by time, contains information on the tornadoes within the GLD CWA.
The information on this web page will focus primarily on the F2 tornado that occurred near the town of Hill City in Graham County. In addition to the F2 near Hill City, the storm later produce four more tornadoes, the last of which hit the town of Stockton, Kansas. Visit the Hastings, Nebraska NWS web page for more information regarding the Stockton, Kansas tornado.
|
| 2. Damage survey
The following map contains information obtained from the damage survey conducted by Mr. David Floyd, Warning Coordination Meteorologist for the Goodland NWS. The map depicts the paths of the Hill City tornado and the locations of two brief tornadoes that occurred east of the Hill City tornado. The path length of the Hill City tornado was 9.5 miles with a max width of 0.25 miles wide. |
|
Fig. 1. Tornado paths (bold blue line), damage ratings, and time of occurrance depicted. |
The following images were taken during the damage survey.
 |
 |
 |
 |
|
3. Meteorological data a. Upper air On the morning of 9 June 2005 a deep layer, long-wave trough, located over the western U.S., slowly translated east toward the Great Plains. Rotating along the periphery of the trough existed a series of mid-level, short wave troughs. As the day wore on, one of the short-wave troughs propagated toward western Kansas, increasing the mid/upper-level winds, advected colder temperatures aloft, and served as a mechanism for fostering large-scale lift in the atmosphere. As a result, strong low-level warm air and moisture advection occurred over the Central Plains. The juxtaposition of strong daytime heating with the large amounts of low-level moisture underneath the cooling temperatures aloft produced an unstable (potentially buoyant) atmosphere by the middle of the afternoon.
|
|
Fig. 2. Upper air charts use a standard station model in abbreviated format; temperature and dewpoint in Centigrade. Winds in knots with one full barb and one half barb equal to 10 and 5 knots, respectively. |
|
b. Surface data In response to the mid/upper level features described above, on the morning of 9 June 2005 a relatively broad area of surface low pressure developed over eastern Colorado. By early afternoon the area of low pressure consolidated in diameter and became located over northwest Kansas. In fact, the low became established at the intersection of several outflow boundaries produced from earlier thunderstorms. The outflow boundaries extended east from the low with the southern most boundary mixing north late in the afternoon. Additionally, a dryline extended south of the low into the Texas Panhandle. The Hill City storm initiated just south of the surface low along the dryline. Note, there existed a plethora of mesoscale surface features relevant to the event, but related discussion is beyond the scope of this web page. |
|
Fig. 3. Surface map for 5 pm CDT, 9 June 2005. Standard station model used; temperature and dewpoint (°F), pressure (mb), and sky conditions reported. Winds as in Fig. 2. Surface low (L), warm front (solid red), surface troughs (black dash), dryline (brown dash) and outflow boundaries (blue and red dash) depicted. Note, the blue dashed outflow boundaries represent cold pools that advanced away from the parent convection. The red dashed outflow boundary represents an old outflow boundary that mixed slowly north in the afternoon hours. Green shading represents surface dewpoints > 59 °F. The green arrow depicts the axis of deepest low-level moisture and the direction of largest moisture advection. The purple asterisks represent the tornado locations associated with the Hill City storm. |
|
c. Radar imagery An radar animation of reflectivity of the Hill City storm is available here. The animation spans from just after the time the storm initiated (~2:20 pm CDT) and ends at the time when the last tornado in Graham county dissipated (4:47 pm CDT). Note how the storm initial moved to the northeast, then "turned" to the east. The deviant motion is not uncommon with supercell thunderstorms. |
|
The following GLD NWS radar imagery is near the time the Hill City tornado began to develop. Both the level-II 0.5 degree reflectivity and velocity are shown to the right. The main updraft of the storm was ~80 nm away from the GLD radar with a center beam height of ~8600 ft above radar level. Note the hook shape in the reflectivity data. This configuration is often associated with tornadoes. |
|
Fig. 4. Radar reflectivity near Hill City at 4:14 pm CDT (2114 UTC), 9 June 2005. The warmer colors represent regions of greatest rainfall and hail within the storm. The T represents the approximate location of the developing tornado. |
|
Fig. 5. As in figure 4 but depicting storm-relative velocity data. The cool colors represent motion toward the radar. The warm colors represent motion away from the radar. |
| Having trouble seeing the hook? Try the following image that has been smoothed. |
|
Fig. 6. As in figure 4 but the reflectivity data have been smoothed. |
|
d. Satellite imagery Visible satellite imagery depicts the Hill City storm ~5 minutes before the F2 tornado (Fig. 7). Note the line of developing thunderstorms south of the Hill City supercell. Several of the developing storms produced tornadoes late in the afternoon.
|
|
Fig. 7. Visible satellite imagery at 4:15 pm CDT (2115 UTC), 9 June 2005. Surface observations as in Fig. 3. but for 2100 UTC. |
ACKNOWLEDGEMENTS
A sincere thank you to the law enforcement, emergency management, storm chaser and spotter communities for
