Urban land teleconnections in the United States: A graphical network approach: Fid-Bau Portal

Urban land teleconnections in the United States: A graphical network approach

McManamay, Ryan A. / Brinkley, Catherine / Vernon, Chris R. / Raj, Subhashni / Rice, Jennie S.

Abstract Cities influence land use change on neighboring and distal areas through sociopolitical or infrastructural connections between urban and non-urban regions, termed teleconnections. While teleconnections are generally recognized as important to land cover dynamics, many land use and land cover change (LULCC) modeling efforts do not explicitly account for non-contiguous spatial urban-land relationships. Here, we quantify and map urban land teleconnections in the US using information theory in concert with graphical networks. Evidence of teleconnections relied on long-term (1950–2016) changes in urban land cover and urban population intensity in urban areas (census defined “urban area”) relative to land dynamics in other cities and rural lands (rural portions of counties), located both proximate and distal to each city. While there are numerous definitions of rural and urban (see for review), we rely on national census definitions in this study for replicability. We find that optimal urban-rural land network complexity and network size ranged dramatically and depended on the information theory measure and variables under consideration. City-city networks were very large, suggesting the complex and simultaneous effects of globalization on cities. However, we also find that proximate urban-rural networks are small and influenced directly by individual cities, though each city may, in turn, be influenced by complex teleconnections with other cities. Distal rural areas providing agricultural commodities were characterized by decreasing agriculture land use yet increasing agricultural production intensity. Hence, distal rural commodity areas may support multiple cities, incurring increased land stress in those areas. Our results suggest that changes in population intensity, rather than urban land cover alone, induce more numerous effects on proximate and distal rural lands, as well as other cities. Predictive models of teleconnection strength explained 4.5% to 98% of variation in teleconnection measures and suggested that rural land characteristics were equally, if not more, important than variables characterizing city land dynamics. This indicates that local dynamics in rural areas are potentially more important drivers of land change relative to teleconnections with cities. Ultimately, the methodological approach presented in this paper holds promise in incorporating teleconnections in LULCC modeling efforts, providing important considerations for later studies on emissions modeling, economic supply chains and future land-use planning.

Highlights • Cities influence land dynamics in other urban and rural areas via teleconnections. • Urban land teleconnections mapped using information theory and graph networks. • Teleconnections modeled as changes in land, population, and agriculture production. • City-city networks were larger and indicative of simultaneous globalization effects. • Urban-rural networks were less complex with rural areas influenced by few cities. • Predictive models explained 4.5% to 98% of variation in teleconnection measures.