Effect of local skin blood flow during light and medium activities on local skin temperature predictions

Human thermophysiological models: Quantification of uncertainty in the output quantities of the passive system due to uncertainties in the control equations of the active system via the Monte Carlo method

Uncertainty propagation analysis in the Fiala thermophysiological model is performed by the Monte Carlo Method. The uncertainties of the output quantities of the passive system, due to imported uncertainties in the coefficients of the control equations of the active system, caused by the variation of the experimental data, are computed. The developed and implemented in-house code is accordingly validated. The effect of the input uncertainties, in each of the four main responses (shivering, vasodilatation, vasoconstriction, sweating) of the active system, is separately examined by simulating the human exposure from neutral conditions to cold and hot environments. It is predicted that the maximum output uncertainties of the response mechanisms may be of the same order of magnitude as the imported ones, while the corresponding maximum uncertainties in core and skin temperatures always remain less than 2%. The maximum absolute deviations of the rectal (core) temperatures from their estimated mean values may be up to 0.72 °C and 0.22 °C, due to input uncertainties in shivering and sweating respectively, while the corresponding deviations due to uncertainties in vasomotion processes are negligible. The deviations, particularly the ones due to shivering, are significant, since differences of a few tenths of a degree may have large impact in human health. The maximum absolute deviations of the skin temperatures are 0.42 °C in the hands due to uncertainties in shivering and 0.69 °C in the feet due to uncertainties in vasodilatation. These deviations are less significant than the core ones, but they may still affect human thermal sensation and comfort. The present analysis provides a better insight in the dynamic response of the model and indicates which response mechanism needs to be further investigated by more accurate estimates in order to improve model reliability. It can be also applied in other human thermophysiological models.