TORNADO DAMAGE SURVEY OF NORTHERN ALABAMA AND GEORGIA: MARCH 24, 1994 by Tim Marshall, P.E.(April 1994)
On March 24, 1994, a tornado outbreak occurred in Alabama, Georgia, and South Carolina. There were many fatalities, most occurring at the Goshen United Methodist Church in Piedmont, Alabama. I was contacted by NOAA Headquarters in Washington D.C. to join the NOAA DISASTER SURVEY TEAM.
The team spent four days surveying the damage in northern Alabama and Georgia. Both aerial and ground surveys were conducted and numerous persons were interviewed. The following findings were made:
1. FINDING REGARDING TORNADO TRACKS
There were a number of tornadoes in Alabama and Georgia. All the tornado damage tracks were consistently at a 60-degree heading from north and tracked southwest to northeast. The tornado damage paths ranged from 1/8 of a mile wide to over one mile wide. Path lengths were as long as 60 miles. Due to the high translation of these tornadoes, some with forward speeds of 60 mph, much of the tornado damage was skewed to the right side of the tornado centerline.
2. FINDING REGARDING HOUSING PERFORMANCE
Permanent homes and manufactured homes performed poorly due to the lack of floor/foundation anchorage. Most homes (new and old) were supported on hollow-block concrete masonry units and the wooden floor systems remained unattached to the foundations. As a result, these homes tended to shift off their foundations and disintegrate. Although such damage was consistent with an F-5 rating on the Fujita scale, the wind velocities were no more than F-3. Most people we spoke with sought shelter in the crawl spaces and survived without serious injury.
In a few instances where anchor bolts were used to secure the floor plates to the concrete masonry foundation, no nuts were provided to secure the bottom plate to the foundation. In other instances where the floor plates were bolted to grouted cells in the concrete masonry, the CMU block was either broken or pulled loose from the foundation wall -and remained attached to the bottom plate/floor.
Wood framed walls rotated about the floor pulling out the straight-nailed connections in the bottom plate or pulling out the straight-nailed bottom plate connections out of the floor.
Interior bathrooms provided formidable occupant protection especially when the floors remained intact.
3. FINDING REGARDING MOBILE HOME PERFORMANCE
Single and double-wide mobile homes performed poorly. Many homes were not anchored or had minimal frame-ties. None of the observed mobile homes had ties over the home. Mobile homes tended to roll from their supports and disintegrate in F-1 or F-2 winds. A higher number of fatalities occurred in mobile homes than in permanent homes.
Six people died in a double-wide mobile home on Henderson Mountain near the Jerusalem, Georgia community. The home was not anchored and was struck broadside by the tornado winds estimated to be around F-2 intensity. The home rolled to the northwest and disintegrated. Occupants were tossed from the home and reportedly were crushed. Five of the occupants died instantly. One 15 year old bled to death after being gored in the neck by a tree branch. Ambulances could not reach the scene with all the downed trees across the road. Reportedly, one other occupant survived but was seriously injured.
4. FINDING REGARDING AGRICULTURAL LOSSES
Numerous chicken houses were destroyed resulting in large losses of poultry. Hundreds of acres of forests were flattened. Several hundred trees were uprooted or snapped each second as the fast translating tornadoes moved through the forests.
5. FINDING REGARDING PUBLIC WARNING AND AWARENESS
About one-half of the people we spoke with received warning of the tornado threat via radio or television. Only a few people had weather radios. The remainder of the people we spoke with did not receive any advanced warning but most heard the sound of the approaching tornado. Most people described the sound as multiple freight trains.
Several boaters on Neely Henry Lake headed for the boat ramp when it began to rain and lightning was observed off to the west. Some people remained in their vehicles to wait out the storm. Other people were caught outside their vehicles when the tornado struck. One person was crushed when a boat flipped over and other person was seriously injured.
NOAA should continue to conduct both ground and aerial surveys in the wake of tornado outbreaks. Such surveys are vital to establish the tornado tracks and F-scale damage/intensity. Aerial surveys alone would have led to higher F-scale ratings as the degree of construction (i.e. presence or lack of anchors) could not be accurately determined. Ground surveys better defined the F-scale ratings as well as established times of tornado occurrence. It is important that such surveys be conducted as soon after the event as possible before significant clean-up can be done. WCMS should recognize that unanchored homes demolished by a tornado may fit the F-5 rating on the Fujita scale but can actually be caused by winds of no more than F-3 intensity.
NOAA should continue to emphasize that people should avoid staying in or seeking shelter in mobile homes during threatening weather.