Extreme heat increasingly threatens public health and urban sustainability, creating a need for fine-scale assessments that connect spatial risk patterns with climate-resilient planning. This study assessed heat risk across 1639 blocks in Tianjin’s six central districts, China. A block-scale heat-risk index (HRI) was constructed within the Hazard–Exposure–Vulnerability–Adaptability (HEVA) framework by integrating remote-sensing, population, built-environment, socioeconomic, and public-service data using a modified CRITIC weighting method. Differences in HRI among Local Climate Zones (LCZs) were examined, and three machine-learning models were compared. Among the three evaluated models, Random Forest showed the best overall test-set performance, and supplementary five-fold spatial block validation retained the same model ranking and major feature-importance ordering despite lower absolute predictive performance. Feature importance, partial dependence plots, and SHapley Additive exPlanations were subsequently used to interpret feature contributions and nonlinear associations. HRI was lower in the resource-rich urban core and higher in peripheral transitional zones, accompanied by clear spatial differences in adaptive-resource provision. The urban core generally exhibited higher adaptive-resource provision, whereas peripheral areas showed more limited provision. Under the adopted HRI formulation, higher values of this component contributed mathematically to lower composite HRI, rather than demonstrating a realised risk-reduction effect. Built LCZs generally exhibited higher HRI than land-cover LCZs; LCZ 5 (open mid-rise) had the highest median HRI, whereas LCZ G (water) had the lowest. Higher sky view factor and vegetation cover were generally associated with lower HRI, while larger blocks, taller buildings, and greater distances from water bodies were associated with higher HRI. These findings support differentiated block-scale interventions, heat-response resource allocation, and climate-resilient urban regeneration.

Rapid urbanization has intensified the demand for street designs that reconcile ecological quality with positive human experiences, particularly in high-density cities such as Tianjin, China. Streets function as key interfaces where ecological processes, social activities and human perception intersect. However, existing research tends to emphasize the amount of greenery while overlooking its structural characteristics, to treat perception as a psychological response decoupled from spatial context, and to make limited use of fine-grained functional data to examine how ecology and perception interact. This study develops an integrated analytical framework that combines the DeepLabV3+ model to extract the Urban Street Greenery Generalized Structure (USGGS) from Baidu Street View imagery with a vision transformer model trained on the Place Pulse 2.0 dataset to derive multidimensional perceptual metrics. Functional diversity is represented using point-of-interest (POI) data, and an enhanced Light Gradient Boosting Machine (LightGBM) model is employed to explore associations among greenery structure, perceived qualities and functional characteristics. Analyses of six urban districts in Tianjin indicate that ecological and perceived street qualities are closely related to the degree of coupling between vegetation structure and functional diversity. Streets characterized by multi-layered greenery and diverse, active functions tend to exhibit higher perceived aesthetics, safety and vitality, whereas streets with single-layer vegetation or functionally monotonous environments generally do not perform as well. Functional patterns appear to mediate relationships between greening and perception by shaping how ecological form is experienced through everyday social activities. Overall, the results suggest that closer coordination between ecological design and functional organization is important for fostering urban streets that combine environmental resilience with strong perceived appeal.