The extreme east-central Missouri flash flood of 6–7 May 2000

During the nighttime hours of 6 May into the early morning of 7 May 2000, strong thunderstorms pounded portions of east-central Missouri with torrential rainfall. A large portion of the region (6 counties) received over 5 inches of rain, while peak rainfall amounts exceeding 13 inches were recorded in Franklin County. The greatest rainfall recorded at an official gage was 13.53 inches at Union (UNNM7). There were however unofficial rainfall amounts as high as 14-16 inches in the Union vicinity. Most of the rain fell during the period from 0500 to 1200 UTC 7 May, and included rainfall rates which at times exceeded 3.00 in/hr. Preliminary damage reports included 2 fatalities, nearly 100 homes destroyed with hundreds more damaged, and extensive damage to bridges and roads; initial damage estimates were $116 million.

From a meteorological standpoint, the precursors to the event were somewhat benign. An old upper level low was weakening as it lifted east northeast from eastern Oklahoma across southern Missouri. This low in concert with a southerly low level jet (LLJ) in excess of 40 knots was providing weak large scale ascent for precipitation. While the atmosphere was very moist with precipitable water values approaching 1.50 inches, a defined surface boundary which is characteristic of most organized heavy rainfall/flash flood events was noticeably absent.

The event unfolded late in the evening of 6 May when an area of rain with a few embedded thunderstorms moved from southwest Missouri into east-central Missouri in advance of the weakening upper level low pressure system. Between 0400-0500 UTC 7 May, the rain and thunderstorms created a mesoscale thunderstorm outflow boundary across east-central Missouri. This boundary became oriented nearly parallel to the mean wind flow aloft, and perpendicular to the LLJ. Shortly before 0500 UTC, intense thunderstorms began developing over eastern Gasconade County along the western periphery of the thunderstorm outflow boundary. These thunderstorms then moved slowly east at 10-15 mph into Franklin County. Between 0500-1200 UTC, this pattern remain nearly unaltered, with thunderstorms continuously regenerating near the intersection of the LLJ and the surface thunderstorm outflow boundary, then slowly moving to the east and training over the same areas. Around 1200 UTC the pattern broke down as the LLJ veered to a more westerly direction, diminishing the convergence and lift along the outflow boundary, and allowed the convective system to weaken and move eastward.

This paper will give an overview of the synoptic and mesoscale features attending this long-lived, extraordinary convective rainfall event. The event resembled the mesohigh flash flood type described by Maddox et al. (1979) as the thunderstorm outflow boundary served as a key focusing mechanism. Surface data, WSR-88D data, and satellite imagery will be shown to document the development of this boundary and its evolution during the event. WSR-88D data will also be used to examine the nature of the training convection, its frequency, intensity and direction to better understand how and why convection continued to regenerate over the same area for many hours.