Debris from the Columbia shuttle catapulted through the skies over at least four southwestern states on Feb. 1, creating a puzzle spread over thousands of square miles. GIS and GPS made a whole picture from this extensive landscape and slowly created maps showing the retrogressive pattern of debris.
The initial concern was recovering the remains of the seven astronauts who lost their lives in the event. At the same time, an intense effort was launched to find the cause of the shuttle's midair disintegration. Collection of debris was a key element in this effort.
The need for accurate mapping and GIS technology was immediately apparent -- calls for GIS support were issued on day one. The state was ready to respond to that need, according to Drew Decker, director of Texas Natural Resources Information System (TNRIS).
"What we have been doing for the last four to five years is collecting a lot of base map information," he said. "Therefore, we already had the resources."
As an agency within the Texas Water Board, TNRIS had already gathered statewide data. The agency also collaborated with Stephen F. Austin University in Nacogdoches, Texas, which has an extensive database of maps, and surrounding counties to form a data pool -- the foundation for reconstruction of the shuttle's final minutes.
This repository of information provided teams throughout eastern Texas with data about roads, terrain, infrastructure and topography -- all plotted with coordinates that allowed field personnel to document what debris was found and where. Decker said some of the search teams were equipped with GPS receivers, allowing them to immediately record the data.
Some of the debris was logged using a 24-satellite array, which provided accuracy within 100 feet or more. Software from Trimble Navigation Ltd. fine-tuned the GPS data to within 3 feet. This information was then transmitted to ArcInfo GIS software from ESRI where it was plotted on TNRIS' data-rich maps.
Decisions made early in the recovery effort also helped the extensive process of mapping debris. Officials in Nacogdoches developed a "data dictionary" to create standards before any logging of debris was done, according to Mick Garrett, Trimble's North American sales manager, mapping/GIS products. The initial effort, however, was decidedly low-tech.
"A grid was laid out, about a kilometer square ... people would walk the grid and look for pieces," Garrett said. "When a piece was found it was flagged. If it was a significant piece, we would take a GPS location, photograph it and flag it so someone would come and recover it. There were so many pieces of tile out there -- some the size of a silver dollar."
Reports of debris came from as far away as California.
Narrowing the Search
"Teams were based out of regional efforts by county," Decker said, and search teams were sending information to Lufkin, Texas. He said layers of data on GIS maps indicated where debris was most likely located and what natural impediments -- such as water or rugged terrain -- might hamper the search.
"There are some places where you might expect to find things and don't," he said. "That might mean you have to go out and look again."
One week after the disaster, the search refocused in eastern Texas, around Lufkin and Nacogdoches, where teams from the university's GIS program and its Forest Resources Institute found thousands of parts and pieces. The search area eventually narrowed to a 10-mile by 240-mile corridor along the shuttle's path. The landing gear was found in this area in mid-February
The terrain in eastern Texas ranges from rugged -- thick woods and impenetrable underbrush -- to open pine forests, marshy streambeds and 238 bodies of water. The search was often hampered by weather, which turned from warm and sunny to rainy and near-freezing.
Information transmitted from the ground to a satellite receiver is easily affected by the environmental factors, such as weather and terrain. High-quality electronics can compensate for those variables, according to Trimble's Garrett.
"Certain errors occur when GPS signals travel through the ionosphere and troposphere that can affect the accuracy of the position," he said. "What we can do is use a method called differential correction that removes that error. Instead of being 30 or 40 feet out of position, we are within 3 or 4 feet."
Accuracy is critical in a disaster that covers multiple counties, said Joseph B. Bowles, senior technical marketer for ESRI. "If you are off just a half a degree, your line goes off by miles," he said.
Garrett said this kind of accuracy depends, in part, on good firmware -- generally defined as software embedded in hardware. The tool used in Nacogdoches and other sites in eastern Texas was the company's GPS Pathfinder Pro XR receiver that runs TerraSync software.
A Collaborative Effort
More than 40 counties in Texas reported finding shuttle material. Information was drawn from multiple sources across jurisdictional lines ? from all three levels of government.
The effort included 17 state agencies, 17 federal agencies and numerous local governments. Texas has a history of cross-boundary collaboration, said Sheila Sullivan, ESRI's San Antonio regional manager.
"It's a real community here in Texas, and people see the need to share data," she said. "They recognize we have to share data to achieve things." Sullivan also said there was very little turf struggle over who would take the lead in the days following the disaster.
Adequate funding from the Texas Legislature created a platform for cross-jurisdictional sharing long before a disaster demanded it, Decker said.
Communities throughout the state had a head start on collaboration because of TNRIS' five-year statewide mapping project. The Legislature also recognized the ongoing need to update maps and allocated subsequent funding so TNRIS could update information.
"We happen to have a lot of data here because the Legislature provided funding for four years," Decker explained. "We could invest in partnerships with local government."