An experimental investigation of the dynamic relationship between weighted mean skin temperature and metabolic rate in outdoor pedestrian activities

Human thermal comfort and heat balance are critical factors in environmental control and health research. Traditional models, such as Fanger's Predicted Mean Vote (PMV) and Gagge's Standard Effective Temperature (SET*), are built on steady-state assumptions, while computationally intensive models like Fiala require core temperature inputs, limiting their use in dynamic field applications. To address this gap, this study proposes a dynamic human heat storage model that incorporates three innovations: (1) real-time infrared thermography-based measurement of spatially representative weighted mean skin temperature (WMST), (2) estimation of time-resolved metabolic rate using heart rate surrogates, and (3) introduction of a time-lag function to capture the physiological delay between metabolic rate (M) and skin temperature response. The results demonstrate that WMST is a robust indicator of thermal comfort, showing strong correlations with subjective mean thermal sensation votes (MTSV) and thermal pleasure levels (Tp). Additionally, a significant time-lag effect between WMST and metabolic rate was observed, highlighting the need for dynamic modeling. The improved model, which incorporates transient variations in WMST and M, shows reductions in discrepancy of up to 88.2% compared to the original steady-state formulation. This study provides a theoretical foundation for adaptive thermal comfort models and real-time environmental control strategies, with potential applications in wearable technologies and urban thermal safety monitoring. © 2026 Elsevier Ltd.