Spatiotemporal variations in indoor thermal environments and dynamic comfort of passengers in airport terminals during winter

Clarifying passengers’ dynamic thermal comfort needs is essential for sustainable civil aviation and for progressing toward zero-carbon airport terminals. This study adopts a passenger “whole-journey” perspective. It examines spatiotemporal variations in thermal conditions and subjective responses in a major hub terminal during winter in a mild (Cwb) climate. Field measurements in key functional zones revealed clear thermal non-uniformity. Large temperature gradients occurred between boundary transition zones and interior areas. These step changes often exceeded commonly used comfort limits for transitional exposure. We also observed strong zonal differences in thermal neutral temperature (TNT), ranging from 17.6 °C to 20.2 °C. In most zones, the measured operative temperatures were higher than the neutral requirement. This indicates an “overheating margin” and suggests energy-saving potential. Based on zone-level setpoint alignment with TNT, heating loads could theoretically be reduced by 7 %–18.5 % in some zones. The weighted average saving across primary passenger areas is 7.7 %. Waiting time further shaped comfort requirements. It served as a practical proxy for metabolic-rate decay. TNT increased from 18.3 °C to 20.3 °C when dwell time exceeded 60 min. Finally, we explored an LLM-based predictor as a complementary tool for thermal sensation inference, achieving 72.8 % accuracy (within ±1 TSV). This work proposes a passenger-centered framework for zone-specific control, demonstrating the exploratory potential of AI-assisted prediction in low-carbon operations.