Severe thunderstorms moved across lower Michigan during the afternoon of 2 July 1997 producing significant tornadoes and widespread wind damage resulting in seven fatalities and more than 100 injuries. The storms exhibited both supercell and bow echo characteristics making the forecast and warning process particularly complex. The synoptic environment was characterized by an unusually strong upper level trough moving eastward across the Great Lakes region, which contributed to substantial changes in the vertical wind structure and thermodynamic conditions ahead of the advancing surface cold front. Eta model grid point data and forecast soundings were used operationally to provide an initial assessment of convective potential. The model predicted substantial destabilization into southern Michigan with CAPE values of 3000-5000 J/kg expected by early afternoon. In addition, model forecast wind profiles, hodographs and derived kinematic parameters (e.g., storm relative helicity, 0-6 km shear) suggested that damaging wind gusts associated with bow echoes were the most likely mode of convection, especially in view of the extreme instability forecast. This assessment was supported by the predicted synoptic pattern and hodograph structure, which were similar to those associated with warm season derechos (Johns 1993 and Johns and Hart 1993). However, for cells that might move to the right of the mean wind, model-predicted storm-relative environmental helicity (SREH) of 150-200 m2/s2 also indicated that forecasters would need to be alert for the possibility of supercells.
The model forecasts were updated during the day using real-time WSR 88D VAD wind profiles from Grand Rapids, Detroit, and Gaylord MI, and a special 1800 UTC sounding from Gaylord. SREH computed from these sources suggested that the potential for supercell development was substantially greater than earlier predicted by the eta model, with values reaching near 400 m2/s2 during the afternoon. Use of hourly VAD wind profiles also showed rapid changes in the wind structure over lower Michigan as the low-level flow strengthened and backed during the morning, resulting in SREH increasing from less than 100 m2/s2 at 1200 UTC, to 300-400 m2/s2 within several hours time. Application of storm relative flow concepts (Thompson 1998) to the observed wind profiles indicated that tornadic supercells were also possible.
This case illustrates some of the difficulties in determining the dominant mode of convection that may occur during severe weather outbreaks. Prior anticipation of the likely mode(s) of convection can improve the warning-decision process, since forecasters are more likely to correctly recognize and assess the severe potential of storms when they understand the mesoscale environment and its influence on storm type and evolution. Pattern recognition and model forecast data are often useful in aiding the initial determination. However, operational models cannot be expected to routinely resolve mesoscale features that impact instability and vertical wind structure. When possible, it is important to update model forecasts with real-time observational data, especially during rapidly changing situations when the latest soundings are not representative of conditions that develop just a few hours later.
Corresponding Author:
Steven J. Weiss
Storm Prediction Center
1313 Halley Circle
Norman, OK 73069
e-mail: Steven.Weiss@noaa.gov
telephone: 405-579-0707
fax: 405-579-0700