• Eli is a contributing correspondent for Science magazine.

Experts Sifts 'Tens of Thousands of Photos' in Post-Catastrophe Tornado Analysis

10 May 2011 11:05 am


Doppler radar helped quantify the initial assessment of the impact of the historic outbreak of 305 tornadoes in the last week of April from Texas to New York that killed more than 325 people in the U.S. Southeast. But as crews continue to dig out from the horrific event, teams of hurricane scientists and structural engineers are conducting post-storm surveys that offer crucial clues about tornadoes that radar misses.

Doppler radar arrays offer local snapshots of precipitation and even moving debris. But they are aimed a half-degree angle above the surface of the ground, so they miss details near the surface where tornadoes are most deadly. Radar even misses many tornadoes completely, says Weather Channel severe storm meteorologist Greg Forbes, adding that "quite a few" false alarms are also raised when upper atmosphere circulation gives a false clue.

Tornado warnings based on current weather data are very effective—National Weather Service (NWS) warnings gave residents as much as half an hour's notice before the twisters. But understanding more about how tornadoes form and how radar data can be better interpreted requires careful forensic detective work after the storms wreak their havoc.

Dozens of experts on survey teams have already assessed 201 tornado tracks, and teams are still completing other assessments.

Clues from the damage, says structural engineer and meteorologist Timothy Marshall of Haag Engineering in Dallas, Texas, have helped researchers to determine the width of the tracks, the speed of the winds, and the precise path of the twisters. The roofs, walls, and foundations of buildings also provide data: Joints that fail provide a lower bound of wind force, whereas structures that survive offer an upper bound in any one spot. Marshall has collected more than 1000 pictures from an assessment team stint that surveyed tracks of three tornadoes in Birmingham, Alabama.

After NWS completes the initial survey of the event, researchers will try to piece together a more detailed step-by-step picture of the outbreak. "We'll go house by house," says Marshall, calculating inflow, the speed of winds moving toward the tornado, and upflow, the force of winds moving from the ground into the sky (tossed cars are a useful clue). Teams need to move quickly in areas where reconstruction is already beginning. "I'm sure there are tens of thousands of photos that were taken" of the damage, says Marshall of the effort.

These data can yield fundamental lessons about the storms, says tornado meteorologist Roger Wakimoto, who didn't participate in surveys this time around. An analysis by tornado science pioneer Tetsuya Fujita of a deadly 1974 outbreak in Alabama, Illinois, and Ohio showed that one storm can produce a line of tornadoes and that tornadoes can take right and left turns before they dissipate, says Wakimoto, director of the National Center for Atmospheric Research in Boulder, Colorado.

But doing the data-gathering is also an emotional experience, he says. "You can see people who have lost their homes, maybe their loved ones. That is really horrifying," says Wakimoto: "What keeps me going, and many of my colleagues, is the knowledge that what we're doing can lead to better [tornado] warnings and hopefully save lives."

Posted In: