Using a car-top video camera, USC researcher Andrew Curtis tracks disaster recovery over time.
When a series of tornadoes rumbled through the Dallas-Fort Worth area earlier this month, Andrew Curtis booked a flight. He's kind of like a stormchaser, except with the good sense to wait until the storm is already gone before checking things out. It's not the storm he's interested in, anyway. He wants to see the destruction.
So, with his backpack-sized equipment, Curtis set off for Dallas to see just how much destruction the multiple tornadoes had caused. What he saw was not as dramatic as how the media portrayed it.
"They were pretty mild," says Curtis. "They were nothing compared to Joplin and Tuscaloosa."
Joplin, Missouri, on October 6, 2010 (left) and on June 7, 2011 (right) after an EF5 tornado swept through town.
Those two cities were of course ravaged by tornadoes last spring, and Curtis went shortly after each disaster to see the damage. And then he went again. And again.
A professor in the Department of American Studies and Ethnicity at the University of Southern California, Curtis has been using a car-mounted video camera to record, street-by-street, the devastation wrought by these natural disasters, returning periodically to re-record and track recovery. He's trying to track what he calls the longitudinal change in these places to better understand how recovery takes place and what may be holding it back.
Like a nerdier and less menacing Google Street View, his car-top video technique uses GPS to link all the imagery he captures to locations, the data of which can be used with geographic information systems to create detailed maps of the destruction. He says he's collected video of about 90 percent of the damaged area in Joplin, Missouri, and has video footage from three different trips – June, September, and December 2011. He's recorded about 60 percent of the damaged area in Tuscaloosa, Alabama, on three separate occasions and plans to go back later this month.
The recovery process of one house is tracked over time in these stills from the videos captured by Andrew Curtis and his team. In the upper left, a house in Joplin in June 2011, then again during reconstruction in September (upper right), and nearly complete in December (lower right).
Post-disaster situations like those faced in Joplin and Tuscaloosa can be hectic, especially in the early days. Collecting data isn't always a priority for locals, which means that what is collected can be spotty. "A lot of the data that is collected is quite often lost," says Curtis.
He says his video technique offers an easy way to gather a lot of information with relatively little effort. It's basically a Sunday drive, only through a completely devastated neighborhood. Scenic to say the least.
The video below is one of the data collection efforts conducted in Joplin shortly after a category EF5 tornado hit the town in May 2011. (Muting is recommended.)
With each video, Curtis and his team of researchers at USC apply codes to each property to help track changes over time – whether it's blighted, re-occupied, under reconstruction, or demolished.
It's an approach Curtis former colleague William Fagan developed while teaching at Louisiana State University. An early application: Hurricane Katrina. Over the years, Curtis has returned regularly to the city to collect more video. He says he's got six years worth of detailed video data for about six different neighborhoods.
"And we're going to keep on doing it," says Curtis. "There are not many places in the U.S. – whether it’s a post-disaster environment or not – where you've got that level of visual data that you can see how places have changed."
(Check out this feature from The New York Times that simultaneously plays four years of video from the same street in the Lower Ninth Ward. That's Curtis's work.)
And though this technique has obvious applications for tracking physical recovery, Curtis says it can be used for a variety of research projects not necessarily related to natural disasters. For example, his grad students have used the car-top camera technique to explore gang graffiti and pedestrian conditions in neighborhoods with high rates of diabetes. Recording commentary from locals riding in the car during filming can add another layer of information.
But neighborhoods don't need a tornado ripping through town or even a little graffiti to make this approach useful. Curtis says video recordings like these can be almost like time capsules, recording a place in time to create a robust history that might be used in any number of ways in the future.
''Literally just put it on a shelf so that in years to come, we can go back to it or potentially others can go back to see how changes have occurred in the neighborhood," says Curtis. "The limiting factor is only the imagination of use."
Images courtesy Andrew Curtis, Google Earth