Impact of the built environment on local urban heat island intensity in cold regions of China: A comparative study of old and new urban blocks in Xi’an

Although urban-form optimization is an important strategy for mitigating the urban heat island effect, cross-seasonal field comparisons of the relative and combined effects of two- and three-dimensional urban morphology on pedestrian-level urban heat island intensity (UHII) across different block types remain scarce. We focused on representative old and new blocks in Xi’an, China. Pedestrian-level UHII measured during winter and summer mobile transects was paired point by point with urban morphological indicators extracted from buffers around each sampling location. Kriging interpolation, random forest modeling, and three-dimensional response-surface fitting were used to characterize spatiotemporal UHII patterns, assess the relative contributions of morphological factors, and identify key combinations of two- and three-dimensional urban form. UHII was generally higher and more spatially variable in the old block than in the new block. The relative influence of morphological factors varied by season: three-dimensional urban form had a greater influence in winter, whereas two-dimensional planimetric layout was more influential in summer. In the old block, response-surface analysis associated higher UHII with either of two combinations: road density above 31 % and mean building height below 12 m, or road density below 18 % and mean building height above 21 m. In the new block, UHII tended to decrease when building density exceeded approximately 17 % and mean building height was below approximately 27 m. These findings provide field-based evidence that morphology–UHII relationships vary by season and block type. The identified morphological combinations associated with UHII variation can inform the design of block-scale thermal environments.