Potential drivers of urban green space availability in Latin American cities

Urban green spaces (UGS) are crucial for regulating air temperature, reducing pollution, and improving well-being. However, research on UGS in Latin America is limited. This study addresses this gap by examining 371 large cities across 11 Latin American countries, representing a diverse range of climates and socioeconomic conditions. The researchers aimed to describe UGS quantity and spatial distribution, analyze variability across different city environments (natural, built, and socioeconomic), and assess the relative importance of climate versus built and socioeconomic factors in explaining UGS availability. The study's significance lies in its contribution to understanding the determinants of UGS availability and distribution in a region characterized by rapid urbanization and high biodiversity. This information is critical for informing sustainable urban planning and achieving the United Nations Sustainable Development Goal 11.7, which aims to provide universal access to safe, inclusive, and accessible green and public spaces. The introduction also highlights the existing knowledge gap regarding UGS research in Latin America compared to the Global North and China, emphasizing the unique challenges and opportunities presented by the region's rapid urbanization, diverse climates, and high levels of biodiversity. It further points out the inequitable distribution of UGS, often correlating with socioeconomic disparities, emphasizing the need for a comprehensive study to understand these patterns and inform policy interventions to address environmental justice.

Existing research from the Global North and China extensively explores urban green spaces, but studies focusing on Latin America are comparatively scarce, although this is changing. Previous work has largely concentrated on a few large cities in Brazil, Argentina, and Chile, leaving a significant knowledge gap for other countries in the region. The literature review highlights the rapid urbanization experienced in Latin America since the 1960s, often without adequate policies to manage urban growth, leading to the loss of green spaces. Some cities have implemented successful greening strategies, but UGS provision generally receives lower priority compared to basic infrastructure projects. The review also notes the significant presence of informal settlements in Latin America, characterized by poverty, inadequate housing, and limited access to essential services, which often lack access to quality green spaces. The study mentions previous research that suggests a socioeconomic gradient in UGS, with wealthier areas generally having more abundant and better-maintained green spaces, although some studies reported contrasting findings. Overall, the review underscores the need for a large-scale, representative study of UGS in Latin America to fill existing research gaps and provide a more comprehensive understanding of the factors influencing its availability and distribution.

The study employed a multi-method approach using data from diverse sources to analyze urban green space (UGS) in 371 large cities across 11 Latin American countries. The methodology involved three primary data sources: (1) Satellite imagery (MODIS and Sentinel-2) for assessing greenness (NDVI), green space area, and spatial configuration; (2) Web-mining techniques (Google Cloud) to identify and map urban parks; and (3) Census data and other datasets for city-level characteristics, including climate zone, population, population density, street intersection density, GDP per capita, educational attainment, and unemployment. The study sample comprised cities with at least 100,000 residents, covering a broad range of geographical locations and socioeconomic contexts within Latin America. The definition and measurement of urban green space involved multiple metrics to capture quantity, quality, and spatial configuration: greenness (NDVI), percent green space area, green space area per capita, patch density, isolation, aggregation, number of parks per capita, and park area per capita. Statistical analysis included ANOVA tests to compare UGS characteristics across different city environments (natural, built, socioeconomic), and random forest regressions to assess the relative importance of various city-level variables in explaining UGS availability. The study also conducted sensitivity checks using linear regressions, considering both continuous and categorical variables. Time periods varied for different datasets (UGS data were from 2017, and urban parks data were from 2020, while socioeconomic data varied by country, with 2005 being the earliest year). The methodological section also contains detailed information about the selection and processing of satellite imagery, the mapping of green spaces, the identification and validation of urban parks, and the methods for handling temporal and spatial misalignments in data.

The study revealed significant heterogeneity in UGS characteristics across the 371 Latin American cities. Greenness (NDVI) varied from 0.10 to 0.80, green space coverage ranged from 4% to 72%, and green space area per capita varied considerably. Analysis of UGS variability across city environments showed that climate zone had the largest influence on quantity-related metrics (NDVI, percent green space area, green space area per capita), with arid cities exhibiting lower amounts of green space and more fragmented patterns. In contrast, the built environment (population density, street intersection density) and socioeconomic factors (education, GDP per capita, unemployment) played a larger role in explaining urban park availability. Specifically, higher population and street intersection densities were associated with less green space overall but potentially greater park density. A noteworthy finding was the nonlinear relationship between some socioeconomic indicators and UGS, suggesting that both high and low levels of GDP per capita and unemployment were associated with higher NDVI compared to cities with moderate levels. Random forest regressions revealed that climate zone accounted for a substantial portion of variance in explaining green space quantity, while education, population density, and street intersection density were the most important drivers of urban park availability. Partial dependence plots and linear regression analyses supplemented these findings, providing further insight into the complex relationships between city characteristics and UGS.

The findings highlight the complex interplay of environmental, built, and socioeconomic factors shaping UGS availability in Latin American cities. Climate significantly constrains the quantity of green space, particularly in arid regions, implying that cities in different climatic zones may have different capacities for nature-based solutions. The relatively greater importance of socioeconomic factors for urban park availability suggests that urban planning policies and investments can play a larger role in influencing the distribution and accessibility of these green spaces. The study's findings challenge the common assumption that socioeconomically disadvantaged cities always have less green space. The observed nonlinear relationships between socioeconomic indicators and UGS necessitate a more nuanced and context-specific approach to greening strategies. The unequal distribution of green space within cities, coupled with the prevalence of informal settlements, emphasizes the importance of considering environmental justice in urban planning. Further research is required to understand intra-city variability and disparities in UGS access.

This study provides the first large-scale assessment of UGS availability in Latin American cities, revealing significant heterogeneity and highlighting the complex interplay of climatic, built, and socioeconomic factors. Climate plays a dominant role in determining overall green space quantity, while urban planning and socioeconomic factors are more influential in shaping urban park distribution. The study emphasizes the need for context-specific greening strategies that account for diverse climate conditions and address environmental justice concerns. Future research should investigate intra-city variations in UGS, the temporal dynamics of green space change, and the multifaceted social and ecological impacts of UGS across different socioeconomic contexts.

The study acknowledges several limitations, including the use of broad vegetation metrics that did not capture specific vegetation types, temporal misalignment between green space and socioeconomic data, and slight spatial mismatch between data units. The reliance on Google data for urban parks might introduce geographical bias. The study's focus on city-level analysis prevents easy generalizations to smaller, within-city scales. These limitations should be considered when interpreting the findings, which primarily provide a regional overview of UGS availability across Latin American cities rather than a detailed intra-city analysis of access and equity.