ABSTRACT: Urban landscapes are heterogeneous matrices of residential, recreational, and commercial development, as well as managed and un-managed natural and agricultural lands. While heterogeneity at sub-city scales has been well studied, the heterogeneity among cities remains poorly characterized across diverse physiographic regions and population sizes. This knowledge gap inhibits our ability to identify the scope of inference for case-studies and design multi-city studies. In other heterogeneous systems, such as biology or soil science, taxonomic frameworks facilitate research design and interpretation. No such classification system is available for urban ecology. To work towards this larger goal, this study: 1) quantified the land cover composition using the 2011 NLCD, population density, and biophysical characteristics of urban areas in the conterminous United States from 2500 to 18 million people, 2) separated cities into similar types using a grouping analysis and identified which attributes differentiated cities, and 3) determined if the city types had different amounts of natural capitol (biomass) and environmental impacts (CO2 emissions, impervious surface, and impaired stream length). Four types of cities emerged that were differentiated by medium intensity development and undeveloped land, followed by low intensity development and population density. The city types were geographically concentrated in, but not limited to, different regions of the US. Notably, natural capitol and environmental impacts differed significantly among the city types. These results provide the basis of a classification system for cities that enhances our ability to design research that considers and captures differences among cities and their effects on the local, regional, and global environment.
