SOMETHING encouraging happened to tornado fatality statistics this year on the way to the data bank. The 30-year running average suddenly dropped from 104 to 91 fatalities a year. That dramatic drop is just one indication of what appears to be a long-term trend in reducing tornado deaths as severe weather forecasting and warning systems improve, and public awareness of what to do when danger threatens increases.
Edward W. Ferguson, deputy director of the National Severe Storms Forecast Center (NSSFC) in Kansas City, Mo., explains that data for 1953 were dropped from the 30-year average when 1984 was added in. The 1953 figure of 515 tornado deaths nationwide is by far the largest in the past 40 years, and big enough to bias the 30-year average for three decades -- an average that has, in fact, been coming down.
For tornado forecasting, 1953 marks a watershed. That was the year when the famous Worcester tornado family roared through Massachusetts without warning. It was the product of a fast-moving storm system that had sent a devastating tornado into Flint, Mich., the day before. Such a surprise would be unlikely today.
As explained in the accompanying story, the parent storm system is constantly tracked by watchful radars and weather satellites. More important, the NSSFC, established in 1954, would have placed Massachusetts under a tornado watch.
As soon as a funnel was sighted, a warning would have been flashed by the local weather service office to police, fire, and civil defense agencies, as well as to news media. The scream of National Oceanic and Atmospheric Administration (NOAA) weather radios would also have alerted the increasing number of individuals who have these inexpensive receivers.
NSSFC director Frederick P. Ostby gives much of the credit for the ensuing decrease in tornado deaths to the buildup and improvement of this watch-and-warning system, beginning with the establishment of his own center the year after the Worcester tornado. He illustrates this by citing two complementary sets of statistics -- the number of tornadoes recorded per decade and the number of tornado deaths per decade.
In the 1930s, observers recorded 1,685 tornadoes across the United States. In the 1970s, they recorded 8,575 twisters.
``It doesn't mean that there are five times as many tornadoes occurring now as then -- it just means there are more being recorded,'' Dr. Ostby explains.
With radar, improved communications, and more people watching, fewer tornadoes now sneak in undetected. This, Ostby says, translates into greater public alertness and fewer deaths.
Dr. Ostby notes that the tornado death rate dropped from nearly 2,000 per decade in the '30s to 987 in the '70s. ``And at the rate the 1980s are going so far, it looks like it will drop even further,'' Ostby told a press briefing last fall.
Nevertheless, he warns that this trend is no reason for complacency. In fact, it is a strong argument for even greater public awareness of the need to be prepared in advance to take defensive action should a tornado threat develop.
And as illustrated by the recent Venice, Fla., tornado, even sophisticated forecasting techniques don't always anticipate tornado conditions. No watch or warning was given March 17.
Tornadoes have touched down throughout North America in every month of the year. But their incidence increases significantly from late winter through early summer in the contiguous United States, especially east of the Rockies.
Nearly 75 percent of tornado-related deaths in the US occur during this period. Thus, in early March, the US National Weather Service (NWS) issued statements urging public alertness.
Two degrees of alertness are required, depending on whether an area is under a tornado watch or has been given a definite warning.
Watches are issued by the NSSFC for all parts of the continental US. They usually cover a box-shaped area 100 miles wide by 250 miles long and warn of the possibility of tornado development during, say, the next six hours.
This is a time to keep in touch with the weather situation and be prepared to seek shelter, the NWS says. But until a warning is issued, you should not interrupt your normal routine.
Tornado warnings, on the other hand, mean that danger is imminent in your local area. Radar or a spotter probably has actually seen a funnel. Local forecast centers issue these action-oriented warnings and expect them to be taken very seriously. The action to take, according to the NWS, is simple but extremely important -- seek shelter.
This means find a secure spot indoors. If you are in a car, truck, or other vehicle, leave it. You are unlikely to outrun a tornado. If you are in a mobile home, leave it. Even when tied down, mobile homes are not safe shelter.
If you are in a building, including your home, stay inside and go to the safest place within it. The NWS recommends identifying such a relatively safe spot ahead of time.
Generally, this will be an inner room, preferably in the basement, away from possible flying glass. Crouching under a sturdy piece of furniture, if possible, offers protection from falling material. An inner bathroom, where pipes in the wall reinforce the structure, also can be a sanctuary.
Joseph Minor of Texas Tech University's Institute for Disaster Research says that a small inner area -- perhaps a closet or part of a hallway or basement corner -- can be reinforced to provide rather secure shelter. Built with concrete blocks, steel reinforced doors, and other common building materials, such a shelter should add only a few thousand dollars to the cost of a house.
Dr. Minor and his colleagues have studied thousands of tornado-struck buildings. He notes that most twisters in the US have maximum winds less than 150 miles per hour.
Minor says that if a building is designed to hold on to its roof in winds of that strength, it is likely to survive even a direct tornado hit with minimum damage. But if the roof goes, everything goes, he says.
Thus he recommends securing the roof to the building frame and anchoring the frame strongly to the foundation.
Homeowners interested in Dr. Minor's tips can write to him at Texas Tech University, Institute for Disaster Research, P.O. Box 4089, Lubbock, Texas 79409.