| Floodplain Management Program Spawns Massive Data Collection Network
By J.D. Wilson Nearly two years after the waters subsided, the clean-up is still in progress throughout the upper Mississippi River Basin. The nation's eyes have turned to other events and in the heartland life is slowly - but never completely - getting back to normal.
The Great Midwest Flood of 1993 entered the record books as the largest and most destructive hydrometeorological event in modern times. It set records for amounts of precipitation, upland runoffs, river levels, flood duration, area of flooding and economic loss.
To make sense of this mammoth disaster, the multi-department, cross-discipline Interagency Floodplain Management Task Force (IFMTF), began collecting information from as many sources as possible. They gathered meteorological, biological, environmental, geological, critical infrastructure, social and economic information that will serve as the baseline data for future floodplain management plans and disaster responses. Herculean Task
Measuring and understanding the multiplicity of variables that conspired to create a disaster of the magnitude of a 100-year flood is something akin to trying to count the hairs on your head. After all, the upper Mississippi River Basin covers nearly 714,000 square miles in 13 states - and nearly 12 million acres were under water at the flood's peak.
The IFMTF's first step was to gather the volumes of data on the event and compile them into a useable form - everything from floodplain boundaries and wetlands, to soil types and flow rates, to climate patterns, precipitation and temperatures, to buildings and businesses and farms, even location and nature of toxic wastes.
The team gathered data from approximately 20 federal agencies, 13 state governments and hundreds of regional, county, and local government entities, as well as insurance companies, banks and other private organizations (see page 33; Creating a Super-GIS).
Earth Satellite Corp. (EarthSat) was among the private GeoTechnology companies enlisted to help compile the data. "This was a major effort for our organization," said Roger Mitchell, vice president of business development for EarthSat. "We ran three shifts, nights and weekends, to meet critical deadlines."
Mitchell said EarthSat produced nearly 130 gigabytes of raster and vector data in just 55 days, including a 200-square-mile digital mosaic of the basin using LandSat images from before, during and after the flood. The firm also digitized information on floodplain and wetland boundaries, flooding extent and high water marks.
Tight deadline pressures compounded common data collection problems that occur on any large data collection and synthesis projects. "One of our biggest challenges was locating and gaining access to special data sets," according to John A. Kelmelis, Ph.D., chief of the USGS Science and Applications Branch and head of IFMTF's Scientific Assessment and Strategy Team (SAST). "We didn't want to go out and create new data where useable information already existed."
SAST found that some data vitally important to making informed decisions on the floodplain were not available or were not in a useable form or level of quality. When data were available, the data sets were often incompatible and, in the absence of widely-accepted standard transfer formats, migration was complex and tedious.
Kelmelis said that organizations were generally cooperative and many went far beyond what was asked of them, because they understood the future value of the database. Accurate Data, Sound Analysis
As data became available, scientists began conducting their analyses. The task force performed more than 100 different studies and analyses, Kelmelis said.
Some of the studies were relatively simple, like analyzing flood insurance activity and responsiveness patterns, according to Kelmelis. Other studies were far more complex, like developing a model to determine the pre-flood land use and land cover in the flooded areas and identifying sedimentation and scour patterns. Other notable studies included:
• Hydraulic models of the uplands and floodplains.
• Literature reviews for relevant information on hydrology, wetlands, geology, ecology, economics and other topics.
• Field studies on energy flows in selected sites on the floodplain.
• Analysis of where and why levees broke.
• Locations and impacts on toxic sites that were inundated by the flood.
As a result of the SAST project, scientists and government decision-makers know more about the '93 midwest flood than any other natural disaster in history. The combined data and analyses reveal its causes, its energy and destructive force, its impact on human lives and commerce.
Moreover, the studies make major strides in identifying how human actions within the water system and our attempts to prevent flood damage affect the nature and results of a flood, both positively and negatively.
"The most important thing we learned is that we must treat the whole river as a system, instead of a series of independent projects and activities," Kelmelis explained. "We must recognize that every action we make changes the nature of the risk," Kelmelis admonished. "It's not enough to know that a particular land treatment works for a certain area. We must ask if this is the best treatment for this area and how it will affect other areas." A New Approach
Consequently, the task force is attempting to change an approach that has persisted for as long as people have lived in the river basin.
It's report, presented to the White House this spring, suggests a far-reaching approach to floodplain management which "seeks to balance competing land uses in a way that maximizes the net benefits to society."
The report recommends redefining federal-state-local relationships and responsibilities and identifying ways to improve coordination and efficiency of planning and management activities at all levels. At a time when federal purse strings are tightening and budget cuts are a virtual certainty, Kelmelis emphasizes that multi-level government collaboration is more important than ever.
"The resources we have to work with are limited and shrinking," he cautioned, "but the problems we face will only get worse, if we don't act decisively."
The report proposes a combination of strategic and operational goals:
• Reduce the vulnerability of the nation to the dangers and damages that result from floods.
• Preserve and enhance the natural resources and functions of floodplains.
• Streamline the floodplain management process.
• Capitalize on technology to provide information required to manage the floodplain.
This fourth goal, to make full use of information technology, may prove to have the most far-reaching impact of all of SAST's actions. Information as a Hedge Against Uncertainty
The SAST's first full year of existence was devoted primarily to gathering and processing the available information upon which their analyses and subsequent conclusions and recommendations were based. These data now form the basis for a computerized regional geographic information system (GIS) of the entire upper Mississippi River Basin.
This 240-gigabyte "super-GIS" contains both spatial and aspatial (textual, or descriptive) information, organized in more than 100 separate data sets all integrated through an ARC/Info server which can be accessed on the Internet. The GIS enables scientist from remote locations throughout the U.S. to access, download and use data for their own research.
Historically, data of this kind has been collected for a single purpose, used once and then discarded. "Most agencies develop data for a specific need and then put it on shelf until someone remembers it and knows where to find it," said Charles Trautwein, project manager for the SAST database at the EROS Data Center in Sioux Falls, S.D. "There is a real need for integrating and making available data that exists for anyone who might need it."
But if compiling the data and creating the original database was a Herculean task, ongoing management must seem a job like for the gods. Housed at the Eros Data Center, Sioux City, S.D., the SAST database is being established as a clearinghouse on the Internet. It will serve as a prototype for the Federal Geographic Data Committee (FGDC) and will help promote the National Spatial Data Infrastructure (NSDI).
The distributed clearinghouse model returns responsibility for updates and maintenance of data to the various departments that originally collected or most commonly use a particular data set. Consequently federal, state, local and private groups all manage various parts of the overall database.
Despite its apparent complexity and logistical obstacles, the clearinghouse approach actually provides the best potential for long-term success. It does not require any new departments, agencies or levels of government and, once the initial development is complete, it will require virtually nothing in the way of additional expenditures to maintain.
"Each agency needs to retain responsibility for its own data," Trautwein said, noting that if the data were kept centralized, it would not be updated regularly and would lose its value over time. If agencies retain ownership of their data, and continue to maintain it based on their use, the database will remain current, dynamic and much more useful.
To that end, Trautwein and a team of 20 computer professionals at Eros are working to develop standards, protocols, query and retrieve tools, and documentation for the distributed database approach.
He explained the clearinghouse model relies on two key elements. The first part is the metadata - a database of directories, query tools, descriptions and instructions which enables users to find out what data sets are available, what they contain and where they can be found. The second element incorporates dynamic links that provide direct access to the individual databases.
Using the capabilities of the Internet, Trautwein explained, providing access is relatively simple. "It's just a matter of adapting existing tools - putting them together in the right way. The most important thing now is the documentation," he said, explaining that each agency is preparing on-line documentation for its data sets. A Database for All Reasons
By the end of the year, Trautwein expects the distributed clearinghouse model to be operational. It will be a seamless database which users can access and extract data based by quadrangles, county boundaries, watersheds or whatever definition their research dictates.
Already users are lining up to access the data, even though only a sampling of the data is on-line, pending documentation. Furthermore, potential uses go beyond floodplain management, although scientists involved in the flood analysis will continue to be the major users.
EarthSat's Mitchell, for example, said his firm is already using SAST data for other applications, in addition to its floodplain work. "We're using some of the data to analyze the pesticides, farmland run-off and the effect on endangered species," he said.
"Producing a database of this magnitude and quality has never been done before. The fact that it's there means people will use it," he said.
Between July and December of last year - the first six months the database was on the Internet, and while only a limited portion of the data was accessible - some 2,500 individuals in 266 organizations accessed the SAST database.
"And the database has not yet been formally announced or advertised," Trautwein said.
Of course, with full implementation just months away, there is still much work to be done, and some obstacles will persist for years. Different groups which share data have different hardware and software systems, different requirements for data quality and detail, different applications for their own analyses and use of the information.
Many of the problems that arose in compiling the first database will recur with each successive update. Implementation of final data standards is years away. Spatial incompatibilities will continue to thwart efforts for consistent analyses. Different computer systems still cannot share data as easily and seamlessly as users expect.
Nevertheless, the SAST database is pioneering new territory in distributed spatial data that will ultimately overcome one of the biggest problems facing the GeoTechnologies industry - access to consistent, detailed geotechnical information.
EarthSat's Mitchell attributes a dynamic, almost living character to the SAST experiment. "Its own progressive evolution and growth will increase its utility over time," he said. "The fact that it's available and easy to use will keep it alive."
For SAST team leader, Kelmelis, such statements from outside parties is affirmation of the team's concept and recommendation. The very dynamic nature of the data will help drive the kind of communication, cooperation and sound decision-making that will help solve the floodplain risks the upper Mississippi River Basin still faces. As for the rest of the applications, they are a bonus. About the author:
J.D Wilson is a freelance writer in Denver, Colo., specializing in the GeoTechnologies. He may be reached at 303-751-7636. Back |
