Feature A three-dimensional view of the flooded area showing stream centerlines and elevation In September 2009, torrential rains in the northern portion of Georgia caused excessive runoff, triggering massive flooding in several areas. Flooding principally affected Cobb, Fulton, and Douglas counties and caused nine deaths and approximately $250 million in property damage. Cobb County, in northern Georgia, with a population of approximately 700,000, experienced heavy flooding along the banks of Chattahoochee River and various major creeks. Hundreds of houses and families in southern and southeastern Cobb County were affected. County and federal governments and relief organizations played an important role in rehabilitating affected areas by providing emergency response, shelter, and postdisaster support. Part of that support came in the form of geospatial modeling within an enterprise GIS environment. Established in 1999, Cobb County’s enterprise GIS was an integral part www.esri.com of the county’s Emergency Operations Center (EOC). At the EOC, GIS staff collected data from damage assessment teams and staff members supporting the center, created data layers, performed analyses, and produced maps in response to several requests for information. Away from the EOC, GIS staff in different departments coordinated with the EOC to create data layers and produce maps, such as the road closure maps generated by Cobb County Department of Transportation. To focus immediate assistance where needed most, identification of the flood-affected properties was important. Flood analysis carried out by the GIS group and the tax assessor’s office benefited different departments that were working to provide a fast and effective response and mitigation after the event. The flood inundation analysis was carried out with stream gauge measurements from the United States Geological Survey (USGS) and spatial data contained in the county’s enterprise GIS. The main objectives of the analysis were to identify the approximate extent of floodwater coverage; identify and locate the affected parcels; determine the minimum, maximum, and mean water heights in the parcels; and identify and document completely and partially flooded properties. The analysis used three data sources: maximum stream gauge readings at select locations, a high-resolution elevation surface derived from the county’s digital terrain model (DTM), and the parcel database. All vector and raster data analyses were carried out in ArcGIS Desktop because it has the vector, raster, and three-dimensional capabilities needed to complete the analysis. Apart from the routine analysis, the hydrologic analysis tools helped in the faster delineation of watersheds, whereas the 3D capabilities helped in visualizing the output. Continued on page 28 ArcUser Spring 2010 27