Urban areas, despite covering only ~3% of Earth's land surface, house over half the global population. Rapid urbanization presents various environmental challenges, including heat stress, habitat loss, air pollution, and water scarcity. Achieving sustainable development goals (SDGs) necessitates a focus on sustainable urban planning and management. Vegetation plays a crucial role in mitigating these challenges, providing cooling effects, improving air and water quality, supporting biodiversity, and enhancing human well-being. While studies show an upward trend in urban vegetation growth (urban greening), significant disparities exist in greenspace exposure between the Global North and South. Previous research has highlighted inequality in overall greenspace exposure, but less is known about the specific differences in exposure to tree and nontree vegetation and how this changes over time with ongoing urban greening. This study aims to address these gaps by examining the variability and drivers of urban tree and nontree vegetation, and their associated changes in human exposure globally over the past two decades.

Existing literature documents widespread urban greening, largely relying on satellite-derived greenness indicators. However, these lack the resolution to differentiate between tree and nontree vegetation, which have unique ecological characteristics. Trees offer significant cooling and air quality benefits, while nontree vegetation provides habitat and other social benefits. A recent study revealed substantial disparities in greenspace exposure between the Global North and South, with the Global South exhibiting significantly higher inequality. The extent to which widespread urban greening mitigates this inequality remains an open question. This paper seeks to expand upon previous research by specifically analyzing the separate trends of tree and nontree vegetation and their implications for human exposure inequality.

This global analysis employed multiple satellite-derived datasets on vegetation, climate, and socioeconomic factors to examine the spatiotemporal variability of urban tree and nontree vegetation cover from 2000 to 2020. Data sources included MODIS Vegetation Continuous Fields (VCF) for tree cover (TC) and nontree vegetation (NTV), Landsat data for tree canopy cover, and MODIS Enhanced Vegetation Index (EVI) for greenness. Urban extent data were obtained from combining the global urban land database with the LandScan population database. Climate data were sourced from TerraClimate and ERA5 datasets, encompassing parameters such as precipitation, temperature, and solar radiation. Socioeconomic data included the Human Development Index (HDI), population density, and an Urban Development Index (UDI) calculated from impervious surface area data. The Mann-Kendall method assessed vegetation cover trends. Generalized additive models (GAMs) were used to analyze the relationships between urban tree cover and climatic and socioeconomic drivers. A population-weighted exposure model, similar to that of Chen et al. (2022), was used to quantify human exposure to urban trees (HET) and nontree vegetation. The Gini coefficient measured inequality in greenspace exposure.

The study revealed that approximately 90% (1320 out of 1464) of cities globally showed an upward trend in tree cover, while 48.8% (467 out of 957) exhibited an increase in nontree vegetation (p<0.05). Urban tree cover exhibited a stronger association with urban greenness (EVI) than nontree vegetation. The increase in tree cover was most pronounced in high-latitude areas (e.g., Eastern Russia, Northern Europe), while cities near the equator showed a declining trend. GAM analysis indicated that climate factors, particularly climate water deficit, vapor pressure deficit, and downward surface shortwave radiation, significantly influenced urban tree cover. Socioeconomic factors, such as HDI (positive correlation above HDI=0.7), population density, and UDI (negative correlations), also played a role. Mean human exposure to urban trees (HET) significantly declined in the Global South (5.22% in 2000 to 4.13% in 2020), while remaining relatively stable in the Global North (15.3% to 15.4%). Inequality in HET, as measured by the Gini coefficient, increased in both the Global North and South. Analysis of specific countries (China, India, USA, Russia) revealed a general decline in HET and an increase in inequality, largely due to population growth and uneven development. Exceptions were noted where depopulation (e.g., Eastern Russia) coincided with increased tree cover, leading to increased inequality despite higher HET.

The findings confirm widespread urban greening driven primarily by increased urban tree cover. This greening is influenced by both climatic and socioeconomic factors. High-latitude regions benefited from reduced population density and increased temperatures, facilitating tree growth. Conversely, warmer regions faced challenges due to increased water deficits. Economic development, particularly above an HDI of 0.7, appears to positively correlate with urban tree cover. While overall urban greening has increased, the disparity in human exposure to urban trees between the Global North and South has widened. This highlights the inequitable distribution of greenspace benefits, despite the overall increase in greenspace. This study offers a more nuanced understanding than previous work by separating tree and nontree vegetation, revealing the unequal distribution of the benefits associated with each type of vegetation.

This study demonstrates a global increase in urban tree and to a lesser extent nontree vegetation cover, primarily driven by factors such as climate change and socio-economic development. However, this increase in green cover does not translate to increased equitable access to greenspaces for all populations, especially in the Global South. This emphasizes the urgent need for policies and strategies that prioritize equitable access to urban greenspaces, addressing both the ecological benefits and social equity challenges. Future research should focus on incorporating more detailed socioeconomic data and analyzing the effectiveness of specific interventions aimed at creating more equitable distribution and access to urban greenspace.

The study acknowledges several limitations. The use of constant urban boundaries may introduce biases, particularly given rapid urbanization in many regions. The accuracy of vegetation cover estimation from remote sensing can be lower in urban areas due to heterogeneity. The GAMs, while robust, only explain a portion of the variability in urban tree and nontree cover. The socioeconomic factors used are surrogates for more direct measures of urban green infrastructure investments. Further improvements could be made through integrating higher resolution data, dynamic urban boundaries, and more detailed socioeconomic variables.