Delineating the wildland urban interface using publicly available geospatial data

The Wildland Urban Interface (WUI), a human-designated area between undeveloped wildlands and urban areas, has been identified using many different kinds of data. The most common data used have been census housing densities to determine urban areas and a vegetation layer from the National Land Cover Database (NLCD) to identify wildlands (Theobald and Romme 2007, Radeloff et al. 2005, Stewart et al. 2003, and Haight et al. 2004). Knowing the location and area of a WUI is important for federal land agencies because federal legislation (Federal Register Notice 2001, 66-3) has provided parameters to identify WUIs and has directed agencies to mitigate the possibility of a catastrophic wildland fire that may reach urban areas. Many studies have looked at WUIs on large scales (e.g. the entire U.S., or part of a state) but have only used one datum input to determine urban areas and one to delineate wildland areas. The objective of this study was to (1) look at whether publicly available geospatial data could be used to determine WUIs for small tracts of land and (2) compare WUI areas resulting from the combination of different urban and vegetation datasets. Four national parks were studied: Badlands National Park, Wind Cave National Park, Pea Ridge National Military Park, and Wilson’s Creek National Battlefield. Urban areas were identified using 2010 U.S. Census Block housing densities or from points identifying individual structures accessed from State web sites. The vegetation layers used were the NLCD, LANDFIRE, and a USGS Vegetation Characterization. In addition, a protocol, “Procedures for Delineating the Wildland Urban Interface at Your Site,” was developed using ArcMap 10. Results showed that either census housing densities or GPS points identifying structures, along with any vegetation classification can be used to determine WUIs for small tracts of land. WUIs varied in size depending on the combination of datasets used but the only factor that appeared to result in larger WUIs was using a detailed vegetation dataset