This study demonstrates the applicability of the PALM(Parallelized Large Eddy Simulation Model) model in simulating urban thermal comfort in Suzhou, subtropical monsoon climate city, during hot periods. Using an observation-corrected downscaling method for meteorological forcing, validated simulations yield reliable 2 m air temperature with RMSE <1.6 °C in hot weather and daytime RMSE ≤1.8 °C in extreme heat. Spatial patterns show vegetation coverage offer the best comfort, bare pavements the worst, with water surfaces in between. Under hot weather, Universal Thermal Climate Index (UTCI) ranges from 28 °C to 44 °C, corresponding to ‘moderately hot uncomfortable’ to ‘very hot uncomfortable’ levels in the hottest hour. In extremely hot conditions, UTCI exceeds 35 °C citywide, spanning ‘hot uncomfortable’ to ‘extremely hot uncomfortable’ categories. Notably, building shading significantly reduces UTCI by an average 3.4 °C in hot weather and 7.4 °C in extreme heat. Net radiation dominates UTCI variation Conversely, while wind plays a secondary cooling role. These findings highlight priority cooling strategies: increasing shaded vegetation cover is most effective, whereas unshaded grass and static water offer limited mitigation. Results support targeted urban planning and heat mitigation for Suzhou and similar high-humidity subtropical cities. • The validated PALM model is used to simulate thermal comfort under hot/extreme hot weather conditions in subtropical Suzhou with 4 m resolution. • Tree shade offers the best thermal comfort, bare pavement the worst, and water bodies an intermediate effect. • Shading reduces UTCI by up to 3.4–7.4 °C under different heat conditions, while water bodies only mitigate nearby thermal comfort locally.
Building similarity graph...
Analyzing shared references across papers
Loading...
Bufan Cao
Ziqiang Huo
Yang Li
Urban Climate
Nanjing University of Information Science and Technology
China Meteorological Administration
Chinese Academy of Meteorological Sciences
Building similarity graph...
Analyzing shared references across papers
Loading...
Cao et al. (Sat,) studied this question.
www.synapsesocial.com/papers/69dc88303afacbeac03ea179 — DOI: https://doi.org/10.1016/j.uclim.2026.102892