A simplified correlation for predicting air temperature at rooftop unit inlet height based on Green’s above-roof temperature model

Predicting the inlet air temperature of heating, ventilation, and air-conditioning (HVAC) rooftop units (RTUs) is important because rooftop microclimate effects can cause the actual RTU inlet condition to differ from ambient weather-station data. Our earlier study verified that the semi-empirical model developed by Green et al can reasonably predict the vertical air temperature profile above roofs, although its computational procedures remain relatively complex. To facilitate practical engineering applications, this study developed a simplified correlation based on Green’s model, hereafter referred to as the simplified model, for predicting air temperature within the near-roof thermal boundary layer. In this study, 0.75 m above the roof was selected as a representative RTU inlet height and used as the primary case for model development and validation. For the 0.75 m case evaluated here, the simplified model uses only roof surface temperature, ambient temperature (i.e., measured at 8 m above the roof) and ambient wind speed as inputs. The model was validated using the field measurements from two rooftop sites, respectively, located in Wuhan, China (warm and humid), and Boulder, Colorado, USA (cold and dry). Because local ambient air temperature may not be directly measurable such as for the Boulder site, an inference method was proposed and used to calculate the ambient air temperature at 8 m above the Boulder roof. The simplified model using both measured and inferred ambient temperature was proven to provide predictive accuracy comparable to Green’s model while remaining applicable across different climatic conditions.