Unveiling Urban Complexity: Integrating OpenStreetMap to enhance representation of fine-scale landscape heterogeneity

Abstract

Landscape heterogeneity has an impact on wildlife behavior, their interactions, and their persistence. Urban landscapes are among the worlds most heterogeneous landscapes, yet current global landcover maps classify developed land in a single landcover type. This limits the spatial scale at which urban ecologists can approach research questions. OpenStreetMap (OSM), an open-source mapping platform, can be leveraged to enhance the representation of landscape heterogeneity in developed areas. For this, we extracted OSM features with attributes representing infrastructure, land use and green cover, integrating these into a continental landcover map through a globally applicable computational framework. We validated our OSM-enhanced landcover layer against existing remote sensing, aerial photography, and local governmental maps for 33 cities in North America. Our frameworks output provides an 89% accurate representation of landscape heterogeneity. We discuss caveats, potential improvements, and ecological applications. Our OSM-based landcover enhancement framework will facilitate the use of open-source landscape information for improved ecological modeling and urban planning.