Â鶹ÒùÔº

There is a concerning decline in street-level greenery in cities, raising important questions about urban resilience, climate adaptation, and environmental equity. Monitoring exactly how fast trees and other greenery are being lost in the world's rapidly changing and expanding urban areas is, however, a challenge.

Led by IIASA researcher Giacomo Falchetta and in Environmental Research: Infrastructure and Sustainability, the study introduces a novel, open-source method that enables continuous and large-scale monitoring of vegetation along urban streets using satellite imagery and machine learning.

"Our aim was to develop a method to monitor how much green infrastructure exists along streets and how it changes over time," explains Falchetta. "Street greenery has well-documented benefits—it reduces urban heat, improves mental health, and can also absorb carbon. But until now, it's been difficult to track it globally in a consistent and timely way."

The authors developed a machine learning model to estimate the Green View Index (GVI)—a canopy coverage indicator derived from labeled street-level photographs. The model was trained and validated using data from multiple world cities and applied to 190 large urban areas across 20 global regions.

The results show that on average, during the eight-year period investigated, street greenery has declined globally by 0.3 to 0.5% per year. Some regions—particularly urban areas in Asia and Oceania—saw even sharper declines, with median GVI drops of 1.7% and 2.6% per year, respectively. Meanwhile, cities in Europe and North America showed moderate increases of about 1% annually. In African and Latin American cities, changes were smaller and more inconsistent.

"Not only is greenery decreasing in many places," notes co-author Ahmed Hammad, "but it's also distributed unequally. In many cities, green space is sparse where people live. This disparity raises serious questions of fairness and access, especially as heat waves and other climate stressors intensify."

The model allows real-time updates using freely available Sentinel-2 and climate data, offering a scalable tool for city planners and policymakers and supporting the ambitions of UN Sustainable Development Goal 11 (Sustainable Cities and Communities).

The model and data are publicly available and the output data can be combined with other key indicators of urban resource use and impacts including local temperature records from weather stations, energy demand for thermal regulation in buildings, records of the impact of extreme weather events on people and urban infrastructure, climate-related adverse impacts on human health such as morbidity and mortality, as well as subjective well-being indicators.

As climate extremes intensify, the authors emphasize that protecting and expanding urban greenery must become a core component of sustainable city planning.

"Our results can support targeted greening policies. With climate extremes on the rise, ensuring equitable access to green space is more urgent than ever. Our work can be a valuable resource for cities to make more informed, just, and targeted greening decisions," Falchetta concludes.