Immediately after Hurricane Katrina struck the Gulf Coast, first responders were working on ways to supply aid to residents in the hardest-hit areas. Helicopters had the ability to rapidly arrive in isolated locations, but with the terrain altered from flooding, pilots needed updated information on where they could land. In Mississippi, GIS technology allowed these responders to quickly identify potential helicopter landing zones (HLZs) and expedite relief efforts.

U.S. Army Space and Missile Defense Command/U.S. Army Forces Strategic Command (SMDC/ARSTRAT) Measurement and Signature Intelligence/Advanced Geospatial Intelligence (MASINT/AGI) Node, which is located at Peterson Air Force Base, Colorado, was tasked by United States Northern Command (NORTHCOM) to provide a potential HLZ map that reflected the posthurricane conditions along the coast of Mississippi. To complete this task, the MASINT/AGI Node integrated advanced spectral analysis of commercial satellite imagery with GIS analysis of supplemental datasets, such as elevation and land cover, and produced a map delineating potential HLZs that met specific criteria:

• Certain land-cover classes, such as urban/built-up areas, agricultural areas, forested or heavily vegetated areas, wetlands, and water, must be excluded. Those land-cover classes with the greatest HLZ potential, such as grassland and barren/sparsely vegetated areas, need to be identified.
• New data reflecting current flooded and saturated areas must be incorporated into the analysis.
• Slope must be less than 15 percent.
• Vertical obstructions, such as towers and power lines, must be mapped as avoidance areas.

This helicopter landing zone map of coastal Mississippi shows open grassy areas in green, wetland or saturated areas in light blue, and standing or open water in royal blue.

To render an analysis that met these criteria, two software packages were utilized: ArcView was used for geodatabase development, analysis of raster and vector datasets, and final map production, and ITT Visual Information Solutions ENVI 4.1 was used for spectral analysis of newly collected satellite imagery. NORTHCOM then disseminated the up-to-date maps to various recovery teams, including the 82nd Airborne Division and the Mississippi National Guard.

Meeting the first two criteria proved to be of the utmost importance, since recovery efforts required a rapid but accurate situational analysis of the posthurricane conditions along the Mississippi coastline. IKONOS four-meter resolution multispectral imagery (MSI) of the area was collected by Space Imaging, Inc., and delivered to the MASINT/AGI Node for analysis. Using ENVI 4.1, the imagery was isolated into its individual blue, green, red, and near-infrared (IR) bands. By ratioing the blue band with the near-IR band, areas along the coast that had been flooded and/or saturated were easily identifiable. Using ArcView, this ratio image was then thresholded and displayed so that only the flooded and saturated areas were delineated as a single thematic layer. To further refine this analysis, a supervised classification of the new imagery was generated in ENVI.

Land-cover classes reflecting up-to-date ground conditions were then incorporated as a layer in GIS for use in the HLZ analysis. Supplementary land classification information was provided by MDA Federal's EarthSat GeoCover data, both to check the validity of the new classification and to fill in areas farther inland not covered by IKONOS. Finally, the land-cover classes were separated into "exclusion" and "potential HLZ" categories based on the land-cover types specified in the criteria.

The next stage in the analysis required the ArcGIS Spatial Analyst extension. First, a Digital Terrain Elevation Model (DTEM) was brought into ArcView and rendered into a slope file. Next, using Raster Calculator, only those potential HLZ land-cover classes that fell within areas where slope was less than 15 percent were isolated as a layer. To further aid in terrain visualization, a hillshade was generated from DTEM for use as a background in the final map.

The final product that was delivered to NORTHCOM was fully created in ArcView. The end goal was to provide the user with a highly readable, current map to aid in locating HLZs during the recovery effort; to accomplish this goal, the data layers were stacked in the following manner: the hillshade was placed as a backdrop for topographic rendering, followed by the potential HLZ land classes that met the slope criteria; overlaid on this was street, city, and vertical obstruction data.

Because of the methodology of integrating spectral and GIS analysis, the MASINT/AGI Node was able to perform the analysis and produce a highly informative map reflecting posthurricane conditions within a matter of hours. This rapid turnaround time enabled recovery teams to receive, disseminate, and utilize the map during the most crucial initial days of the recovery effort.

More Information

For more information, contact Tara Burkey, SMDC/ARSTRAT MASINT/AGI Node (e-mail: Tara.Burkey@smdc-cs.army.mil).