P5.3 NON-CONVECTIVE WINDSTORMS IN THE MIDWEST UNITED STATES: SURFACE AND SATELLITE CLIMATOLOGIES John A. Knox * University of Georgia, Athens, Georgia 1. INTRODUCTION Windstorms cause hundreds of deaths and billions of dollars in damage in the United States annually. The media and the public normally associate windstorms with tornadoes or hurricanes. But non-convective high wind events not associated with thunderstorms, tornadoes or tropical cyclones can cause a sizable amount of fatalities, injuries and property/crop damage. Event in 2000-02 Deaths Injuries Damage (M) Tornado 136 2593 $1870.1 Tropical Cyclone 77 354 $6581.1 Thunderstorm Wind 59 924 $1027.3 High Wind 68 389 $ 232.6 ALL Weather 1482 8603 $26,464.3 Table 1. Cumulative 2000-2002 weather hazard statistics for the United States, from the annual Summary of Natural Hazard Statistics, National Weather Service (2002 data is preliminary) High winds not associated with tornadoes, hurricanes or thunderstorms have caused about 5% of all weather- related deaths, 5% of all weather-related injuries, and 1% of all weather-related property/crop damage that have occurred in the United States since January 1, 2000. From 1996 through 2002, the exact percentages are: deaths: 4.6%; injuries: 2.8%; and property/crop damage, 0.9%. These statistics reveal that the human costs due to non-convective winds can rival or exceed those due to hurricanes and thunderstorm winds, even though the property damage from non-convective winds is far less extensive. This raises the possibility that non- convective windstorms may be less well forecast than other wind events. Also, official advisories for non- convective winds may not be heeded as strictly by the public as, for example, hurricane warnings because of a lack of awareness of the danger from these windstorms. A type of non-convective windstorm with relevance to the Midwest United States is a high wind event associated with an extratropical cyclone. For example, on November 10, 1998, non-convective high winds associated with a record-breaking 963-mb cyclone caused ten deaths and over $40 million in damage in the states of Illinois, Iowa, Kentucky, Michigan, Minnesota, and Wisconsin (Iacopelli and Knox 2001). * Corresponding author address: John A. Knox, Univ. of Georgia, Faculty of Engineering, Driftmier Engineering Center, Athens, GA 30602; e-mail: jknox@engr.uga.edu On this day in 1998, wind gusts up to 42 ms -1 (93 mph) blew cars and trucks off the road and into pedestrians, snapped flagpoles, and tore down water towers and interstate signs. This storm occurred on the 23 rd anniversary of a similar Great Lakes cyclone that contributed to the deadly sinking of the ore freighter Edmund Fitzgerald. The Fitzgerald sank shortly after wind gusts on Lake Superior that were estimated by a nearby boat captain to have exceeded 45 ms -1 (100 mph) (Ackerman and Knox 2003, p. 292). These “Witch of November” storms are both extreme and extremely hazardous, deadly even when forecasters anticipate their development more than a day in advance—as was true in 1998, and even to a surprising extent in 1975. It should be noted that cyclone windstorms are not limited to the Midwest. Similar events occur in the Northeast and West, and less frequently in the Southeast. The Midwest was chosen as a focus for study because of the prominence of cyclone wind events in the Midwest, and the wealth of data coverage in the region. 2. LITERATURE REVIEW Despite the impact of non-convective wind events on the Midwest, only a few climatological studies have examined Great Lakes cyclones. Even fewer studies have focused on these cyclones’ potential for producing extreme and hazardous winds at the surface. The only comprehensive study using strong winds as a primary criterion appears to have been Lewis (1987). Lewis examined 100 storms with sustained (one-minute mean) winds greater than 25 ms -1 (55 mph) that traversed the Great Lakes between 1957 and 1985. Angel (1996) reanalyzed Lewis’s results and found that 92% of the 100 storms were cyclones; 83% of the cyclones occurred in November through March. Lewis (p. 3-3), using ship observations of surface wind, determined that 0.08% of all observations were greater than 25 ms -1 (55 mph). However, these observations were limited during the late winter months due to the freezing of the Great Lakes in many years. About 0.25% of observations made on the Lakes during November, December and January exceeded 25 ms -1 (55 mph), when cyclones track near the Great Lakes. Angel and Isard (1998) compiled a 90-year climatology of Great Lakes cyclones. Among their many intriguing conclusions, the authors found that eleven strong (992 mb or less) cyclones occur in the Great