By: Matt Mosteiko and Dan Thompson
Meteorologists, National Weather Service Detroit/Pontiac
On May 8, 1964, an F4 tornado struck the township of Chesterfield in Macomb County, just north of Detroit. This tornado killed 13, making it the third-deadliest tornado on record in southeastern Michigan, and injured over 400 along its 3.3 mile-long, ½ mile-wide path. It was one of 15 violent (F4 or greater) tornadoes on record in southeastern Michigan since 1950.
CONDITIONS LEADING UP TO THE EVENT
Unseasonably warm conditions prevailed during the four days leading up to the Chesterfield tornado. The height pattern at upper levels of the atmosphere featured a prevalent, deep trough over the western U.S., with a downstream ridge anchored over much of the eastern U.S., including the Great Lakes region (Fig. 1). At the surface, a low pressure system developed in the lee of the central Rockies on the 5th and moved northeastward across the northern Plains and into far northwestern Ontario by the 8th. This system reached its peak strength on the 6th before occluding and weakening by the 8th, and its associated cold front never made it into Lower Michigan. Meanwhile, a ridge of high pressure extended along the eastern seaboard from New England into the southeastern U.S. The result of the positioning of these two surface pressure features was a continual feed of unseasonably warm air into the Great Lakes region (Fig. 2). In fact, high temperatures from the 5th through the 8th at Detroit reached the low to mid-80s, with overnight lows only in the upper 50s to upper 60s, leading to daily average temperatures of 13–20°F above normal. Despite the unseasonable warmth, low-level moisture was slow to stream into the Great Lakes region, so the atmosphere was stable enough to prevent precipitation over southeastern Michigan from developing during the days prior to the tornado.
On the 7th, a second surface low pressure developed in the lee of the central Rockies, this one a bit further south than the first, along the remnant frontal boundary from the first surface low. This surface low moved northeastward as a portion of the large-scale western U.S. upper trough ejected eastward over the northern Plains. It became clear that this pattern was conducive for severe weather and especially tornadoes, as 9 tornadoes were reported across Wisconsin and Iowa on the 7th, resulting in 43 injuries.
Click images to enlarge
MAY 8, 1964
The upper-level trough continued to move eastward across the northern Plains on May 8. An area of strong upper-level winds developed between the trough and the downstream ridge, with a core of 90+ knots at 500 mb (roughly 18,000 feet above the surface) over Kansas at 8 am that morning (Fig. 3). A large area of 50+ knots of wind at 500 mb encompassed a large part of the country, including the Great Lakes. This value is typically sufficient to generate enough vertical wind shear to support supercell thunderstorms, which can produce tornadoes. Lower in the atmosphere at 850 mb (about 5,000 feet above the surface), winds were still quite strong, with 35 knots of westerly wind recorded on the 7 am radiosonde observation at Flint (unfortunately, the morning radiosonde observation was cut short at around 650 mb, possibly due to morning thunderstorm activity; Fig. 4). There was also a plume of warm, moist air extending into the Great Lakes from Texas at this level (Fig. 5). The plume was very narrow over Texas and became broader and less concentrated over the Great Lakes. Still, the increase of low-level moisture into Lower Michigan was perhaps the biggest difference between May 8 and the days that preceded it.
Meanwhile, the surface low continued to strengthen as it lifted northeastward. The surface low was analyzed at less than 988 mb at 1 pm on May 8, centered over northeastern Nebraska (Fig. 6). It moved to the Upper Peninsula of Michigan by 1 am May 9, deepening to around 984 mb. This put Southeast Michigan in the warm sector throughout the day, as the cold front did not pass through until the early morning hours of May 9. Surface winds observed at Detroit remained from the south or southwest throughout the day at 10–20 mph. These winds ushered in moister air throughout the day, as dewpoints increased from 63°F at 2:00 am to 68°F at 4:00 pm, an hour before the tornado touched down.
While the wind profile was conducive for severe weather over southeast Michigan, it still probably was not clear to forecasters whether severe weather would develop over southeast Michigan on May 8. This was due to a few factors. First, the main forcing for large-scale vertical motion and thunderstorm development remained well to the west closer to the advancing upper level trough, surface low, and frontal boundaries. In fact, 500-mb heights actually rose over Southeast Michigan throughout the day, which usually helps to suppress thunderstorm development. Additionally, southeast Michigan saw periods of rain and nonsevere thunderstorms between 3 am and 9 am the morning of the 8th. This may have lowered confidence that the area would have enough heating to destabilize the atmosphere and allow thunderstorm development. Still, the continued feed of low-level warm air and moisture, along with some possible clearing after the morning rain moved off, likely destabilized the atmosphere just enough to allow thunderstorms to develop.
An interesting aspect of this storm was the possible influence of Lake Saint Clair. Given that it was May and the waters were likely still fairly cool, a lake breeze contribution would have resulted in the winds backing slightly. This would have enhanced the thunderstorm's ability to develop a rotating updraft and eventually a tornado. Indeed, the winds observed at Selfridge Air National Guard Base (KMTC, located just south of the tornado's path) just before the tornado were due southerly. This theory is further supported by the location of other tornadoes over Michigan during the day. One was located in Monroe County along the Lake Erie shoreline, while two others occurred in northern Lower Michigan near the shorelines of Lakes Michigan and Huron.
Click images to enlarge
THE TORNADO PATH
Several eye-witness accounts from 1964 recall that the tornado started near the I-94 and 21 Mile Road intersection at 4:59 pm. The tornado then tracked east along Jefferson Road, and skirted along Harbor Drive before exiting land into Anchor Bay in Lake St. Clair. Although there was a tornado watch in place well before the event, many people mentioned that there was little lead time on the warning, and that as soon as the tornado warning came on the television, the tornado was already audible nearby. The tornado's shape looked like a rolling dark cloud that created massive damage as it moved by. It was too large to be seen as a funnel from close-up as the damage path was1/2 of a mile wide.
Click image to enlarge
Unfortunately, 13 people died in this F4 tornado. Over 400 were injured (127 hospitalized), and 1,000 people became homeless after the storm. A survey of the damage showed that 108 houses were completely destroyed, with 46 homes damaged so badly that they would have to be tore down and rebuilt. The financial loss was estimated at $7-8 milion. $6 million of that for homes and property, $1 million on automobiles, and $500,000 on boats. From the pictures below, it is quite evident how devastating this tornado was. The public response, however, was overwhelming. Hundreds of residents, doctors & nurses, and an estimated 3,000 volunteers came forward to provide relief to the thousands of victims.
Click images to enlarge
NWS Storm Data
Plymouth State Weather http://vortex.plymouth.edu/ua-r.html
NOAA/ESRL PSD Map Room http://www.esrl.noaa.gov/psd/map/
Walter P. Reuther Library http://www.reuther.wayne.edu/node/4350
Chesterfield Township http://picasaweb.google.com/cheslib/TornadoDamageFrom581964?feat=directlink#5480783855739415730