Constructing visually appealing public spaces has become an important issue in contemporary urban renewal and design. Existing studies mostly focus on single dimensions (e.g., vegetation ratio), lacking a large-scale integrated analysis of urban color and visual elements. To address this gap, this study employs semantic segmentation and color computation on a massive street-view image dataset encompassing 56 cities worldwide, comparing eight machine learning models in predicting Visual Aesthetic Perception Scores (VAPSs). The results indicate that LightGBM achieves the best overall performance. To unpack this “black-box” prediction, we adopt an interpretable ensemble approach by combining LightGBM with Shapley Additive Explanations (SHAPs). SHAP assigns each feature a quantitative contribution to the model’s output, enabling transparent, post hoc explanations of how individual color metrics and visual elements drive VAPS. Our findings suggest that the vegetation ratio contributes the most to VAPS, but once greening surpasses a certain threshold, a “saturation effect” emerges and can no longer continuously enhance visual appeal. Excessive Sky Visibility Ratio can reduce VAPS. Moderate road visibility may increase spatial layering and vibrancy, whereas overly dense building significantly degrades overall aesthetic quality. While keeping the dominant color focused, moderate color saturation and complexity can increase the attractiveness of street views more effectively than overly uniform color schemes. Our research not only offers a comprehensve quantitative basis for urban visual aesthetics, but also underscores the importance of balancing color composition and visual elements, offering practical recommendations for public space planning, design, and color configuration.
29 August 2025 in Frontiers of Urban and Rural Planning
In rapidly urbanizing cities, historical neighborhoods often experience drastic spatial transformation, leading to the erosion of urban form, memory, and identity. This study examines the morphological transformation of the Siwenli Lilong neighborhood in central Shanghai, tracing its evolution from 1948 to 2021. Drawing on a 70-year fine-scale GIS dataset at the lane-block level which is a rare longitudinal resolution in related urban research, the study integrates historical cartography, urban morphology, and heritage interpretation to identify three key phases: wartime densification, socialist consolidation, and market-driven redevelopment. Each phase reflects distinct governance rationales, cumulatively producing a shift from spatial continuity to fragmentation. The research introduces the concept of “interface rupture” to capture the disjunction between old and new typologies, particularly in façade logic and public–private transitions. Rather than treating transformation as incidental, it proposes a conceptual model linking governance regimes, development logics, and spatial consequences. While symbolic heritage elements are selectively retained, most morphological memory is weakened or erased. By integrating urban morphology with the Historic Urban Landscape (HUL) framework, the study contributes to heritage-led urbanism by moving beyond site-specific diagnosis toward transferable explanatory mechanisms. It calls for adaptive conservation frameworks that recognize spatial memory as a planning asset, promoting continuity during inevitable change. The Siwenli case thus serves as both empirical evidence and a theoretical lens for understanding structural dynamics behind morphological rupture in East Asian cities.
Extreme temperature variability (ETV) is a key dimension of climate risk for billions of residents. However, how urbanization shapes global ETV divergence remains unclear. Here, we assemble a 1950–2020 panel of 10,522 cities and demonstrate that ETV trajectories diverge by development status as measured by the human development index (HDI). ETV intensifies in high-development cities, whereas it slowly weakens in low-development cities, resulting in a current gap of roughly 1.66 °C. Decomposing ETV into event frequency and intensity reveals that cumulative ETV is driven mainly by intensity. Using an interpretable machine-learning framework, we find that aerosols are the urban factor most strongly associated with ETV after controlling for climate and geography. Blue-green space is consistently associated with lower ETV, whereas urban morphology has a smaller and context-dependent effect. These findings link global ETV inequality to urban governance and support targeted management that focusses on limiting volatility under climate risk.