Urban Greenery Under Threat: New AI-Powered Satellite Tool Reveals Stark Global Declines in Street-Level Vegetation
Across the world’s expanding metropolises, the trees, shrubs, and plants flanking streets and sidewalks are more than mere decoration—they are vital components of urban ecosystems, delivering measurable benefits in climate resilience, public health, and social equity. However, quantifying these green assets and tracking their trajectories over time in fast-changing cities has proven a major scientific and logistical challenge. A groundbreaking new study led by the International Institute for Applied Systems Analysis (IIASA) now fills this critical gap by harnessing satellite imagery and cutting-edge machine learning to map and monitor urban street greenery at an unprecedented scale and resolution.
The research, published in Environmental Research: Infrastructure and Sustainability, introduces a novel, open-source methodology that leverages advanced algorithms trained on street-level canopy coverage data to generate global, continuous metrics of green infrastructure along urban streets. This innovation is crucial given the increasingly urgent need for accurate, timely, and comparable information on urban vegetation, which plays a central role in countering the intensifying effects of climate change in cities worldwide.
Urban greenery mitigates the urban heat island effect by providing shade and evapotranspiration, reducing the need for energy-intensive cooling. It also contributes to improved mental health through aesthetic and recreational benefits, while trees and shrubs serve as natural carbon sinks, absorbing atmospheric CO2. Nonetheless, the spatial and temporal monitoring of these green assets has historically suffered from data scarcity and inconsistency, particularly across diverse urban contexts and rapidly evolving landscapes.
The IIASA team, led by researcher Giacomo Falchetta, developed a machine learning model to estimate the Green View Index (GVI), a metric that quantifies the fraction of the visual field occupied by greenery as captured in street-level photos. Utilizing an extensive training dataset drawn from multiple global cities, the model was rigorously validated to ensure robustness and transferability across different urban morphologies and climatic zones. The researchers then applied this approach to 190 major urban areas spanning 20 world regions, yielding a comprehensive picture of how street vegetation volumes have shifted over the past eight years.
Results indicate a troubling global decline in street-level greenery, with average annual decreases in GVI ranging from 0.3% to 0.5%. Particularly stark are the declines detected in rapidly developing cities across Asia and Oceania, where median yearly drops reached 1.7% and 2.6%, respectively. In contrast, many urban centers in Europe and North America showed modest annual increases in green coverage of approximately 1%, likely reflecting sustained greening initiatives and stricter urban planning regulations. Cities in Africa and Latin America exhibited smaller, more heterogeneous trends, suggestive of complex socio-economic and environmental dynamics.
Coauthor Ahmed Hammad highlights a critical equity dimension: “The distribution of urban green space is often unequal, with greenery sparsely available in densely populated, lower-income neighborhoods. This disparity exacerbates vulnerability to climate extremes such as heatwaves, disproportionately impacting marginalized communities.” The study underscores the need for equitable urban greening policies that prioritize accessibility and inclusion to avoid deepening environmental injustices.
Integral to the study’s potential impact is the model’s ability to incorporate real-time data streams from freely accessible Sentinel-2 satellites combined with localized climate variables, enabling up-to-date monitoring at city and street levels. This capacity offers a scalable, cost-effective tool for urban planners, policymakers, and researchers to track green infrastructure and integrate it with other critical urban indicators, ranging from temperature records and energy consumption to health outcomes and well-being metrics.
The research supports the implementation and monitoring of Sustainable Development Goal 11, which aims to make cities inclusive, safe, resilient, and sustainable. It provides actionable intelligence that can inform targeted urban greening strategies to maximize environmental and social benefits. For example, combining the GVI data with energy use statistics could help optimize tree planting efforts to reduce cooling loads in buildings, while integrating health data can reveal connections between access to street greenery and morbidity or mortality patterns.
As climate change fuels more frequent and severe heatwaves and extreme weather events, safeguarding and expanding urban green spaces must be a cornerstone of sustainable city design. The study’s authors call for urgent action, emphasizing that maintaining and enhancing street-level vegetation will be indispensable in protecting urban populations, particularly the most vulnerable, from escalating climate hazards.
Falchetta concludes: “Our findings can drive more informed, just, and effective urban greening policies. With climate extremes intensifying globally, ensuring equitable access to street greenery is not only an environmental imperative but a social one. Our open-access model and data empower cities to make smarter, evidence-based decisions that enhance resilience and quality of life for all residents.”
The innovative methodology and openly available datasets present a transformative leap forward in urban environmental monitoring. They exemplify how integrating cutting-edge remote sensing technology with artificial intelligence can illuminate critical aspects of the rapidly changing urban fabric, enabling responsive and equitable governance in the face of 21st-century challenges.
For city officials and urban ecologists alike, this research offers a powerful resource to understand and combat the stealthy erosion of vital green infrastructure. By revealing the nuanced spatial and temporal patterns of tree and shrub loss—alongside pockets of greening—stakeholders gain a clearer path to sustainable urban futures, where green streetscapes become resilient sanctuaries that nurture people and planet alike.
Subject of Research:
Article Title: Tracking green space along streets of world cities
News Publication Date: 27-May-2025
Web References: https://iopscience.iop.org/article/10.1088/2634-4505/add9c4
References: Falchetta, G. and Hammad, A.T. (2025). Tracking green space along streets of world cities. Environmental Research: Infrastructure and Sustainability. DOI: 10.1088/2634-4505/add9c4
Keywords: urban greenery, street-level vegetation, climate resilience, machine learning, satellite imagery, Green View Index, urban heat island, environmental justice, sustainable cities, remote sensing, AI, urban planning
Tags: AI-powered satellite toolsclimate resilience in citiesglobal city vegetation mappingmachine learning in environmental researchopen-source urban vegetation methodologypublic health and green infrastructuresocial equity and urban green areasstreet-level vegetation monitoringtracking urban greenery changesurban ecosystem benefitsurban green spacesurban heat island effect mitigation
