Effects of building morphology on land surface temperature in Xi'an City – insights based on different settlement types and seasons

This study analyzes the seasonal impact of 2D and 3D urban morphology indicators on LST across 2,300 residential areas in Xi’an, China, using Boosted Regression Tree (BRT) modelling. Results reveal that LST follows a stepwise decrease from the city centre to the periphery. Among different layouts, the traditional layout (T-L) has the highest LST, while point (P-L) and point + row-by-row residential (PR-L) exhibit better thermal comfort conditions. Building coverage ratio (BCR) and mean building height (MBH) are the dominant factors influencing LST across all seasons. In summer, MBH has a cooling effect across most layouts, except for P-L. A 5% increase in BCR raises temperatures by 0.1 °C–0.3 °C in summer, while in autumn, MBH becomes the most significant cooling factor, with a 3-meter increase reducing temperatures by 0.01 °C–0.06 °C. Findings reveal the thermal characteristics of different residential types and offer insights for optimizing urban thermal environments through seasonal adaptive planning.