Toward the rapid characterization of the built environment within the wildland-urban interface: A soft classification strategy

Abstract

Rapidly increasing human populations and spreading cities and suburbs are a common phenomenon throughout the United States. Urban spread, and the desire to leave the subdivision for a more natural setting in the country, create both opportunities and challenges for natural resource managers in the path of urban expansion. Perhaps no challenge is as great as those related to wildfire risk within the lands describing the urban-wildland interface. The need to gain a better understanding of the wildland-urban interface (WUI) is critical to policymakers charged with risk-reduction responsibilities. In this paper a soft classification approach based on Dempster-Shafer Theory of Evidence was employed with data derived from the Landsat TM system to improve the problem of WUI characterization. A 6,000 km2 study area encompassing seven counties in southeastern Ohio was selected to evaluate the applicability of the Dempster-Shafer classification for WUI characterization. Results of this study demonstrate that the Dempster-Shafer model is a useful land cover classification procedure capable of delineating complex land covers and delivering customized data products detailing the form and extent of the WUI in a timely fashion. Copyright © 2006 by V.H. Winston & Son, Inc. All rights reserved.