Enhancing urban microclimate simulations : validating ENVI-met's accuracy in modeling multi-directional radiative fluxes and mean radiant temperature in subtropical Hong Kong

Abstract

Urban microclimate modeling requires precise simulation of shortwave (SW) and longwave (LW) radiative fluxes to accurately determine thermal comfort via mean radiant temperature (MRT). This study evaluates the performance of ENVI-met in simulating six-directional radiative fluxes and diurnal MRT in a high-density city under its subtropical cloudy conditions, specifically at Oi Man Estate in Kowloon City, Hong Kong. Measurements were conducted on three separate days of intense heat – July 29, August 15, and August 23, 2022 – from 09:30 to 17:00. The model configuration was calibrated for different site settings, including unshaded urban areas and areas with various shading types. Through iterative refinements of empirical albedo calibration and cloud approximation significant improvements in model accuracy were achieved. For incoming SW radiation, RMSE in Open Space decreased by 78 % (from 201.82 to 44.29 Wm⁻²) between initial and optimized simulations, with R² reaching 0.984. Using site-specific inputs and optimized configurations, including calibrated surface albedo and a mid-level cloud approximation, the model’s MRT estimates improved markedly across all sites and scenarios (mean RMSE = 2.62 K; R² = 0.73). These findings offer actionable insights for refining urban climate models and highlight the critical importance of input parameter calibration in complex urban morphology. Contrary to earlier studies, this study showed no systematic MRT underestimations in shaded areas. The results demonstrate ENVI-met’s capability to resolve spatially variable radiative fluxes when critical inputs are empirically constrained, enabling reliable area-wide diurnal MRT analysis.