The association between urban greenspace and health is well known, but less is known about how the spatial arrangement of greenspace affects population health. The relationships between urban greenspace distribution, mortality, and morbidity risk were investigated via cross-sectional studies of major cities in the US, based on accepted landscape metrics (i.e., greenness, fragmentation, connectedness, aggregation, and shape), using geographical spatial pattern analysis programs. Study 1, utilized negative binomial regression, focused specifically on all-cause and cause-specific mortality (related to heart disease, chronic lower respiratory diseases, and neoplasms) recorded in the city of Philadelphia. Study 2 adopted spatial regression models, focusing specifically on morbidity risk related to poor mental health, coronary heart disease, stroke, diabetes, chronic obstructive pulmonary disease, and physical inactivity in five major cities in the US (i.e., Seattle, Los Angeles, San Antonio, Miami, and New York). Overall, census tracts with more connected, aggregated, coherent, and complex shape greenspaces had a lower mortality and morbidity risk, although the magnitude of these effects varied across health outcomes and cities. The results support the proposition that environment-based health planning should consider the shape, form, and function of greenspace. Study 3, based on the results of the prior two studies, explored a novel health evaluation tool using machine learning and spatial gaussian process models. The tool automates the extraction of greenspace morphology from landscape and city planning master plans in the service of predicting health indicators. This tool is designed to be used routinely by landscape and city designers as well as policymakers to help estimate the likely health consequences of a design plan prior to implementation.
