Spatio-temporal trend analysis of long-term development patterns (1900–2030) in a Southern Appalachian County

• Published in: Landscape and urban planning Vol. 104; no. 1; pp. 47 - 58
• Main Authors: Kirk, Ryan W., Bolstad, Paul V., Manson, Steven M.
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier B.V 01.01.2012; Elsevier
• Subjects: Biological and medical sciences; Exurban development; Forest development; Forestry; Fundamental and applied biological sciences. Psychology; General forest ecology; Generalities. Production, biomass. Quality of wood and forest products. General forest ecology; GIS; Housing patterns; Land use change; Quantile regression; USA, North Carolina; GIS; Exurban development; Quantile regression; Housing patterns; Land use change; Forest development; Forests; Regression; Housing; Long term; Land use; Geographic information system; Trend analysis; Development; Quantile; Southern Appalachians
• ISSN: 0169-2046; 1872-6062
• DOI: 10.1016/j.landurbplan.2011.09.008

► We reconstruct long-term building development patterns in a scenic mountain county. ► We evaluate development changes in relation to forest cover and terrain variables. ► Development in forested areas has increased steadily over the past 45 years. ► Most forecast future growth is at suburban densities rather than exurban. Many mountainous areas in the U.S. have experienced significant changes in residential development patterns over the past century, in large part due to changing economies and in-migration from outsiders seeking the scenic resources of these areas. To evaluate changes in the Southern Appalachian mountain region, we reconstructed building and road locations in Macon County, North Carolina, since 1906 from historic maps, aerial photographs, county parcel records and emergency response program data. We used these data to analyze growth trends using spatio-temporal classification and quantile regression analysis. We also forecast future development patterns through 2030 using a multi-step model incorporating population growth projections, extrapolations of recent development density trends, and spatial logistic regression models. We characterize new development in the county as primarily rural from 1906 to 1960, exurban from 1960 to 1975 and increasingly suburban since 1975. While the rate of population growth and new building construction peaked in the 1980s, the total road length and rate of development in forested areas continued to increase from 1990 to 2009. We forecast that through 2030 approximately 75% of new buildings will be constructed at urban and suburban densities and that 67% of all new buildings will be constructed in forested areas. In addition to establishing the utility of using spatial analysis to draw together disparate data sources to understand urbanization over long time scales, this analysis identifies trends that have important implications for land-use planning, hydrology, forest management, and wildlife conservation.