Comparison of mean radiant temperature from field experiment and modelling: a case study in Freiburg, Germany

Mean radiant temperature (T-mrt) based on two measurement methods and outputs from three models are compared in this study. They are the six direction radiation method, globe thermometer method, RayMan model, ENVI-met model and SOLWEIG model. The comparison shows that globe thermometer method may overestimate the T-mrt since wind velocity is a key variable in the estimation based on this method. For better estimation, T-mrt measured by the globe-thermometer method be corrected by the imported wind speed (stable, low and assuming wind speed) and validated by the six-direction radiation method. The comparison of models shows that the RayMan model's evaluation of T-mrt involving global radiation with fine time resolution was better than the corresponding evaluations under the other two models (ENVI-met and SOLWEIG) in this case. However, the RayMan model can only assess T-mrt for a one-point one-time context, whereas the other two models can evaluate two-dimensional T-mrt. For two-dimensional evaluations of T-mrt, SOLWEIG have a better prediction of T-mrt than ENVI-met, and ENVI-met can simulate several different variables, which are wind field, particle distribution, CO2 distribution and the other thermal parameters (T-a, surface temperature and radiation fluxes), that SOLWEIG cannot.