A Comparative Study of PET and mPET models in Outdoor Walking Scenarios

Outdoor walking plays a critical role in daily commuting and public health, yet thermal comfort assessment remains challenging due to inconsistent indices. Previous research has not comprehensively evaluated Physiologically Equivalent Temperature (PET) and modified PET (mPET) for physiological responses in hot and humid regions. This study conducts a systematic assessment of PET and mPET performance by examining physiological parameters (i.e., skin temperature and core temperature) across two age groups in three distinct outdoor environments.

A comprehensive sensitivity analysis investigates the impacts of individual inputs, including age, sex, weight, height, and clothing index, on thermal comfort model predictions. By applying multiple evaluation indices, the research detects nuanced variations in physiological responses and model sensitivities. The systematic approach allows for a detailed comparative analysis of thermal comfort models under varying environmental and personal conditions.

Preliminary findings reveal the mPET model's superior performance across different walking environments and age groups. Statistical analysis showed significant improvements in thermal parameter estimations: For skin temperature (Tskin), the root mean square errors (RMSE) of PET and mPET models were 3.66 °C and 1.04 °C respectively, while for core body temperature (Tcore), PET and mPET models presented RMSE of 2.02 °C and 0.24 °C, separately. The PET model consistently overestimated both Tskin and Tcore across all outdoor walking scenarios, demonstrating its inherent limitations. In contrast, the mPET model provided more accurate and precise temperature estimations, highlighting its potential for improved thermal comfort assessment.

This research provides an effective comparative analysis of thermal comfort models, offering feasible guidance for selecting appropriate indices in walking experiments and informing the future development of thermal-physiological modeling approaches. It highlights the potential for refined physiological modeling in outdoor walking research, and indicates the potential for refined physiological modeling in outdoor walking research.