Using multi-source big data to identify “double-aging” neighborhoods for urban retrofitting: A case study of Beijing

Building stock aging has long been a major concern in urban retrofitting programs. In particular, as the person-environment (P-E) mismatch becomes increasingly severe for older adults, identifying “double-aging” neighborhoods has become the main challenge for age-friendly city development. Taking Beijing, China as a case study, we used Baidu Map location-based service data and Anjuke imagery data to depict the spatial patterns of the aging population and building stock and applied the competence-press model to measure the degree of double aging in neighborhoods. The results showed that the distribution of the aging population follows a “high-low-high” gradient radiating outward from the urban core. The old and dilapidated building stocks are mainly concentrated within the 2nd Ring Road. Neighborhoods experiencing severe double aging form distinct clusters within the 4th Ring Road. These spatial patterns are shaped by the interplay of historical, economic, and governance-driven factors. Notably, approximately 30 % of neighborhoods severely affected by double aging are currently excluded from municipal retrofitting plans. By using multi-source big data and deep-learning methods, this study provides a timely and innovative approach to identifying double-aging neighborhoods in urgent need of urban retrofitting in currently aging cities.