Quantifying Key Design Factors for Thermal Comfort in Underground Space Through Global Sensitivity Analysis and Machine Learning
Shisheng Chen, Nyuk Hien Wong, Chao Cen, Ruohan Xu, Lei Xu, Zhenjiang Shen, Zhigang Wu, Jiayan Fu, Zhongqi Yu

TL;DR
This study uses global sensitivity analysis and machine learning to identify key factors affecting thermal comfort in underground spaces during high temperatures, providing insights for sustainable design in hot climates.
Contribution
It introduces a novel combination of sensitivity analysis and machine learning to quantify the impact of various factors on thermal comfort in underground environments.
Findings
MRT is the most influential factor for aboveground spaces.
Underground PET is mainly affected by metabolic rate, humidity, and wind speed.
Underground spaces offer higher thermal tolerance and passive cooling during heat waves.
Abstract
This study identified the key design factors related to thermal comfort in naturally ventilated underground spaces under high temperature conditions (outdoor Tmax = 42.9 C) in Fuzhou, China. Fuzhou has a humid subtropical climate and is one of the three hottest cities in China in 2024. Daytime measurements indicated reduced air temperature (AT), mean radiant temperature (MRT), and wind speed (V), together with elevated relative humidity (RH) in the underground space. Physiological Equivalent Temperature (PET) in the underground was consistently lower during peak hours (08:00-16:00), with the maximum difference in PET between pedestrian and underground levels being 11-11.9 C. Higher pedestrian-level PET at L1 was attributed to reduced greenery and shading, and decrement factors indicated greater thermal dampening at L2 (0.197) than at L1 (0.308). Sensitivity analysis showed that MRT was…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
