Jaue2025-038 Winter thermal comfort evaluation of urban blue-green infrastructure space based on explainable machine learning

Abstract

Greenery and water features are components of urban blue-green infrastructure (UBGI). However, human thermal comfort measurements in the winter have received less attention than the summer thermal standards of UBGI. A questionnaire survey on the thermal comfort perceptions of users of the four UBGI venues was used in this study to determine the thermal comfort benchmarks of users throughout the winter. The findings demonstrated that when inhabitants were thermally neutral throughout the winter, they felt at ease in the UBGI environment. On the slightly hot side of TSV, nevertheless, was the most pleasant setting. This study established an explainable machine learning model with high reliability. Meanwhile, the significance of PET was greater than UTCI on TSV based on this model. PET was more suitable than UTCI to evaluate thermal comfort in the UBGI space in winter of Tianjin. The thermal neutral PET of people using the UBGI space of Tianjin was 11.7°C. The thermal neutral range of PET for the UBGI users was 4.2–19.2°C. This investigation furnishes a theoretical foundation and empirical evidence for advancements in thermal environment enhancement and assessment of UBGI outdoor communal areas, so as to realize the breakthrough and innovation of outdoor thermal environment enhancement of UBGI space.