GIS Training for Disaster Relief
Continued...
After determining which products are available, data mining begins. The key is to determine what decision makers will need so the correct data can be gathered, and extraneous and irrelevant data is not introduced into the geospatial analysis so that the maximum time and resources are allocated and devoted to supporting the disaster relief operation. Queries should be structured based on these needs while keeping in mind that derived products will eventually be distributed to a number of end users.
Effective Data Searches
One strategy is to narrow the initial searches to trusted Tier 1 databases such as proprietary and government databases (e.g., municipal, state, National Geospatial-Intelligence Agency (NGA) [formerly the National Imagery Mapping Agency or NIMA], United States Geological Survey (USGS), and Federal Emergency Management Agency (FEMA)). The effectiveness of data mining can be simulated through geospatial data fusion and the use of the Raster Calculator in ArcMap and the ArcScene application in the ArcGIS 3D Analyst extension.
Geospatial analysts must be able to find the data that best supports a specific disaster relief operation. The data must be in a structure and format accessible to many users. ArcGIS can import and export data in raster, vector, digital elevation model (DEM), and text formats and access data stored in geodatabases, shapefiles, coverages, and CAD files.
Both FEMA and United States Army Corps of Engineers' CADD/GIS Technology Center are in the process of developing a disaster relief product that has two Unified Modeling Language (UML) logical data models for FEMA readiness and response. The submodels will be Spatial Data Standard for Facilities, Infrastructure, and Environment (SDSFIE)-compliant and will reside within a relational database. The goal is to provide a GIS model that eliminates duplication, improves data quality, and saves time in disaster response situations in which critical public assistance data needs to be shared. ArcGIS supports UML for the development of a geodatabase with full topological relationships, integrity rules, and behavior as well as raster, surface, and locational representations.
However, it is unlikely that such a robust data model will be available over a disaster-stricken area in the near future, but the principles required to generate this data are applicable when selecting suitable data for disaster relief operations. For example, if a DEM of the disaster area is required, NIMA Digital Terrain Elevation Data (DTED) Level One, RADARSAT imagery, and LIDAR data could be used as well as contour layers derived from triangulated irregular networks (TINs) generated from known spot heights. A geospatial analyst must be able to determine which product will satisfy the initial requirements and any future requirements.
Quality and Appropriateness
There are many measurements of accuracy for geospatial data, particularly raster data. Digital maps usually have an associated accuracy statement included in the product or in its metadata. Although imagery typically has an accuracy statement, it is always a good practice to compare imagery against a map source with the same datum, projection, and coordinate system.
With this information, the geospatial analyst can determine if the data is accurate enough for disaster relief related GIS analysis. If the information accompanying the data is the only reference source, then a qualitative comparison of the imagery and the existing information should be performed and the differences between the two sources should be analyzed. A truly rigorous quantitative assessment cannot be achieved with the dataset and is probably not attainable given the time constraints of a disaster relief operation.
Another consideration is scale and how it describes geographic data. Geospatial products are created at specific scales and can be displayed at various scales in software viewers. Geospatial analysts should understand how products are developed for an intended scale and potential use at other scales. In most cases, a product derived at a 1:100,000 scale will still be useful with products developed at a 1:50,000 scale, if displayed at a scale of 1:50,000.
Generating the Right Products
Geospatial analysts must understand the different products required for different types of disaster relief operations. For example, in the event of a flood, emergency planners will want to know where the dry areas are located so command/control facilities can be established. Potential products may be maps merged with elevation data and layered with transportation networks and hydrology data. If emergency supplies are being transported to an area, decision makers will want geospatial products that show viable routes into the area; potential heliport locations; and the locations of ports, beaches, and bridges. In the latter stages, planners may want to merge imagery with digital elevation models and vector networks to determine the extent of flooding and postflood modeling. For example, a quick assessment can be done using the elevation data, the transportation network, and a 1:50,000-scale map.
Disaster relief operations provide a range of challenges for geospatial analysts, from getting data in a format that is usable to creating quantitative analysis for the distribution of relief funds. A trained geospatial analyst equipped with ArcGIS and a quality dataset can provide decision makers with the information needed to restore order to a stricken area. The methods and techniques described here will provide a geospatial analyst assigned to a disaster relief operation with some basic tools for developing a training plan that addresses the wide range of operations, from floods to fires, and while taking into account the time-constrained nature of disaster response and relief.
Acknowledgments
 |
| A trained geospatial analyst equipped with ArcGIS and a quality dataset can provide decision makers with the information needed to restore order to a stricken area. |
The author would like to thank Steve Sarigianis of the Joint Precision Strike Demonstration Project Office's Rapid Terrain Visualization Program for the LIDAR data. The author would also like to thank Nancy Towne of CADD/GIS Technology Center for providing information about SDSFIE development for hazard and disaster entities.
About the Author
Jared Ware is a United States Army engineer officer who teaches geospatial intelligence fusion at the National Geospatial Intelligence School at Fort Belvoir, Virginia. His military assignments have included combat engineering, systems engineering, electrical engineering, and geospatial engineering. He holds a bachelor's degree in geography from the United States Military Academy, West Point, New York; a master's degree in engineering management from the University of Missouri, Rolla, Missouri; and a master's degree in defense geographic information from Cranfield University, Shrivenham, England. He is also a 2001 graduate of the Royal School of Military Survey's Army Survey Course in Hermitage, England. He worked with the United States Army Corps of Engineers and the Federal Emergency Management Agency in disaster relief training and operations for emergency power management from December 1998 to July 2000, when he was the commanding officer of A Company, 249th Engineer Battalion (Prime Power) at Fort Lewis, Washington.
|
