After the deafening roar of a thunderstorm, an eerie silence descends. Then the blackened sky over Joplin, Mo., releases the tentacles of an enormous, screaming multiple-vortex tornado. Winds exceeding 200 miles per hour tear a devastating path three quarters of a mile wide for six miles through the town, destroying schools, a hospital, businesses and homes and claiming roughly 160 lives.

Nearly 20 minutes before the twister struck on the Sunday evening of May 22, 2011, government forecasters had issued a warning. A tornado watch had been in effect for hours and a severe weather outlook for days. The warnings had come sooner than they typically do, but apparently not soon enough. Although emergency officials were on high alert, many local residents were not.

The Joplin tornado was only one of many twister tragedies in the spring of 2011. A month earlier a record-breaking swarm of tornadoes devastated parts of the South, killing more than 300 people. April was the busiest month ever recorded, with about 750 tornadoes.

At 550 fatalities, 2011 was the fourth-deadliest tornado year in U.S. history. The stormy year was also costly. Fourteen extreme weather and climate events in 2011—from the Joplin tornado to hurricane flooding and blizzards—each caused more than $1 billion in damages. The intensity continued early in 2012; on March 2, twisters killed more than 40 people across 11 Midwestern and Southern states.

Tools for forecasting extreme weather have advanced in recent decades, but researchers and engineers at the National Oceanic and Atmospheric Administration are working to enhance radars, satellites and supercomputers to further lengthen warning times for tornadoes and thunderstorms and to better determine hurricane intensity and forecast floods. If the efforts succeed, a decade from now residents will get an hour’s warning about a severe tornado, for example, giving them plenty of time to absorb the news, gather family and take shelter.

The Power of Radar
Meteorologist doug forsyth is heading up efforts to improve radar, which plays a role in forecasting most weather. Forsyth, who is chief of the Radar Research and Development division at NOAA’s National Severe Storms Laboratory in Norman, Okla., is most concerned about improving warning times for tornadoes because deadly twisters form quickly and radar is the forecaster’s primary tool for sensing a nascent tornado.

Radar works by sending out radio waves that reflect off particles in the atmosphere, such as raindrops or ice or even insects and dust. By measuring the strength of the waves that return to the radar and how long the round-trip takes, forecasters can see the location and intensity of precipitation. The Doppler radar currently used by the National Weather Service also measures the frequency change in returning waves, which provides the direction and speed at which the precipitation is moving. This key information allows forecasters to see rotation occurring inside thunderstorms before tornadoes form.

In 1973 NOAA meteorologists Rodger Brown, Les Lemon and Don Burgess discovered this information’s predictive power as they analyzed data from a tornado that struck Union City, Okla. They noted very strong outbound velocities right next to very strong inbound velocities in the radar data. The visual appearance of those data was so extraordinary that the researchers initially did not know what it meant. After matching the data to the location of the tornado, however, they named the data “Tornadic Vortex Signature.” The TVS is now the most important and widely recognized metric indicating a high probability of either an ongoing tornado or the potential for one in the very near future. These data enabled longer lead times for tornado warnings, increasing from a national average of 3.5 minutes in 1987 to 14 minutes today.

4 Comments

1. 1. msteviet 04:51 PM 4/19/12

   
   As an expert in the field, I can say unequivocally that the claims of the authors on the value of satellite data, especially from the polar orbiters, for warnings of severe weather 5-7 days in advance are excessively hyped and not sufficiently substantiated.
   
   Moreover, the authors would have you believe that forecasts of this sort require just plugging in more and more satellite data into computer weather prediction models and, wa-la, early warnings of extreme weather. The fact is that for a fraction of the cost of a single satellite, increased investments focused on accelerating improvements in weather models and systems would almost certainly yield greater gains in forecast accuracy and reliability (more bang for the buck).
   
   
   

   Reply | Report Abuse | Link to this

2. 2. Bruce Voigt 01:28 AM 4/22/12

   New to me are Weather Radar Circles.
   I can see were these could be used in deciphering the path and strength of Tornado.
   Guessing, but I think the understanding of these would give a 24 to 48 hour Tornado warning.
   http://www.physforum.com/index.php?showtopic=11462&st=0

   Reply | Report Abuse | Link to this

3. 3. Kevin Martin 06:09 PM 4/23/12

   If grid spacing is so important for tornado prediction, why not use dynamic grid spacing? If most of the atmosphere is relatively calm, and can be modeled with a large grid spacing, then volumes containing potential tornadoes could have their grid spacing dramatically reduced for a very small computational cost.

   Reply | Report Abuse | Link to this

4. 4. dcary3133 04:52 PM 5/21/13

   It looks to me like better decision making on the part of people would help too. When a resident says on the news that they all took shelter in the bathroom and we see pictures of a former house that is less than a foot high, does it not seem appropriate that the old-fashioned storm cellar would be a better idea? The tornado didn't dig holes; it only took away what was sticking up from the earth (like homes, cars, people). Am I missing something?

   Reply | Report Abuse | Link to this