The impact of urban landscape patterns on land surface temperature at the street block level: Evidence from 38 big Chinese cities

Existing literature has made substantial efforts to examine the relationships between land surface temperature (LST) and urban landscape patterns (ULPs). However, the inconsistent findings from studies on LST conducted in different cities lead to concerns about the significance and importance of ULPs. Moreover, insufficient attention has been paid to vertical ULPs and variations in their thermal effects over space and time. This study conducts a comparative analysis in 38 Chinese megacities across different seasons at the street block level to identify regularities and differences in ULP-LST linkages using geographical open data. The study quantifies ULPs with an amount of widely used and new two- and three-dimensional spatial metrics from three aspects—city plan patterns (CPPs), building patterns (BPs), and land use patterns (LUPs)—based on Conzen's townscape analysis framework. Results reveal that the consideration of overall or specific aspects of ULPs can enhance the explanation of spatial variations in LST, particularly during summer and spring. The improvements are highest for LUPs, followed by CPPs and BPs. Regardless of seasons and cities, building arrangement, built-up areas, greens, water bodies, elevation, slope, and road density are the most influential ULP indicators, whereas block size, sky view factor, building density, and building height present limited or unintended effects. Furthermore, our results indicate the time- and place-varying relationships between ULPs and LST, and some ULP indicators demonstrate two-sided effects on LST across different seasons or cities. We suggest that optimizing building layout and land use composition to increase green-blue spaces and urban shading zones may be more effective for alleviating the urban heat island effect than changing urban density.