*NWSO Paducah, KY **NWSO Nashville, TN 8250 U.S. Highway 60 500 Weather Station Road West Paducah, KY 42086Old Hickory, TN 37138 (502) 744-6436 x766 (615) 754-8502 (502) 744-3828 (FAX) pat.spoden@noaa.gov
During the early evening hours on June 17, 1997, six weak tornadoes touched
down unexpectedly in the Lower Ohio and Tennessee River Valleys and produced an
estimated one million dollars in damage. Nearby soundings indicated that these
tornadoes formed in a weakly unstable (CAPE near 800 J kg-1) and low shear ( 0-3 km
storm-relative helicity near 46 m2 s-2) environment near an upper level circulation.
Magnitudes of 12 hour forecast storm-relative helicity values from the numerical
models were similar when compared with nearby soundings. However, there were
some differences when mean relative-humidity fields from numerical models were
compared with satellite imagery. The Eta and NGM output were relatively close to the
actual depiction of the location of the synoptic scale dry slot, while the RUC model was
off by several hundred kilometers.
Three storms were analyzed in this study. Radar and spotter analysis suggests
that the tornadoes developed out of mini (or low-topped) supercells. One storm was
located about 150 km away from a WSR-88D. Reflectivity data revealed a merger
between two cells followed by a storm split with a right and left mover. The tornado
associated with this storm occurred at the time of the storm split, likely associated with
the right mover. Although the reflectivity data could not verify the existence of
supercell characteristics, photographs of the tornado clearly show a wall cloud and a
dry slot wrapping around the tornado. Due to a very small vortex diameter and
distance between the storm and radar site, maximum rotational velocities with this
storm were near 10 m s-1. Another storm, located generally within 100 km of a WSR
88D, displayed nearly classic supercell characteristics (i.e., BWER, hook, etc.) in the
reflectivity data and a mesocyclone with maximum rotational velocities on the order of
19 m s-1 in the velocity data. This storm produced two tornadoes, 90 minutes apart
from each other. The evolution of the third tornadic storm significantly differed from the
other two supercells. Radar data initially revealed the development of a circulation
within an isolated cell in northern Tennessee. This cell broke up into a cluster of
smaller cells. A new cell developed northeast (downwind) of the first cell's initial
location from within the cluster. The circulation observed within the first cell continued
to maintain its identity during its dissipation stage then became part of the second cell.
Further analysis of the reflectivity data revealed a BWER and pendant with the second
cell. Storm-relative velocity data near the second cell indicated a circulation which
satisfied minimal mesocyclone criteria with maximum rotational velocities of about 13
m s-1. This circulation was responsible for spawning two F1 tornadoes within 15
minutes of each other. Discussion on the evolution each of these mini-supercells will
be presented. This paper will also suggest ways to help improve forecaster
anticipation for this type of event.