Urban forest and per capita income in the mega-city of Sao Paulo, Brazil: A spatial pattern analysis

The integration of natural and social sciences in multidisciplinary research may be the key to improving public policies for urban greening. Analyses on the distribution of urban forests in relation to income are scarce for cities in developing countries, which is a crucial problem in view of the problematic and unplanned urbanization process common in these cities. Our study aims to fill this gap, focusing on the mega-city of Sao Paulo. We analyzed the spatial correlation between tree cover and per capita income and the interaction of these variables with other land covers (lawn, buildings and shadows). The city's land cover was estimated from a digital infrared orthoimage and Exploratory Spatial Data Analysis techniques. We found that Sao Paulo presents two patterns for the spatial correlation between tree cover and income, which may represent other mega-cities with similar characteristics. When considering the entire area of the city, the spatial correlation is negative, result contrary to most of the literature and strongly influenced by the presence of reminiscent forests in the peripheral regions, where lower-income population lives. When considering only more urbanized areas, the spatial correlation is positive, indicating an uneven distribution of urban forest that favors the wealthier neighborhoods.