Researchers have introduced a groundbreaking framework called Urban Pulse, published in the Proceedings of the National Academy of Sciences on June 8, 2026, which utilizes satellite imagery to monitor the physical development of cities in near real-time. By analyzing decades of data from NASA's Harmonized Landsat and Sentinel-2 (HLS) datasets, the team can identify construction, demolition, and infrastructure changes as they occur, rather than relying on retrospective census or zoning data. The study, led by Zhe Zhu of the University of Connecticut and Karen C. Seto of Yale University, posits that cities function like living organisms with measurable vital signs. The authors argue that understanding these dynamic rhythms is essential for effective urban planning, as it allows for the detection of growth patterns that are otherwise invisible to human observers until they are already established. The core of the methodology involves the CAPES deep learning and time series analysis technique, which tracks urban transformations across diverse global cities including Seattle, Shenzhen, Lagos, Mumbai, Dubai, and Mexico City. This approach reveals that urbanization is characterized by spiky, cyclical, and asynchronous growth patterns, providing a diagnostic tool for urban health. This framework offers significant practical value as an early warning system for governments and a decision-making aid for individuals and businesses. By identifying shifts in urban activity, stakeholders can better anticipate economic vitality, manage infrastructure loads, and mitigate the environmental impacts of rapid, unplanned urban expansion.
