Solar UV radiation has both adverse and beneficial effects to human health. Using models (a radiative transfer model coupled to a building shading model), together with satellite and surface measurements, we studied the un-obstructed and obstructed UV environments in a sub-tropical urban environment featured with relatively high pollution (aerosol) loadings and high-rise buildings. Seasonal patterns of the erythemal UV exposure rates were governed by solar zenith angles, seasonal variations of aerosol loadings and cloud effects. The radiative transfer modelling results agreed with measurements of erythemal UV exposure rates and spectral irradiances in UVA and UVB ranges. High-rise buildings and narrow road width (height to width, H/W, ratios up to 15) reduced the modelled total UV (UVA+UVB) radiation and leave 10% of the un-obstructed exposure rate at ground-level at noon. No more than 80% of the un-obstructed exposure rate was received in the open area surrounded by 20-storey buildings. Our modelled reduction of UVB radiation in the urban environment was consistent with similar measurements obtained for Australia. However, our results in more extreme environments (higher H/W ratios) were for the first time reported, with 18% of the un-obstructed exposure rate remained at the ground-level center of the street canyon.
Urban air pollution has aroused growing attention due to its associated adverse health effects. A model which could promptly predict urban air quality with considerable accuracy is, therefore, important and will benefit the development of smart cities. However, only a computational fluid dynamics (CFD) model could better resolve the dispersion behavior within an urban canyon layer. A machine learning (ML) model using the Artificial Neural Network (ANN) approach was formulated in the current study to investigate vehicle-derived airborne particulate (PM10) dispersion within a compact high-rise-built environment. Various measured meteorological parameters and PM10 concentrations were adopted as the model inputs to train the ANN model. A building-resolved CFD model under the same environmental settings was also set up to compare its model performance with the ANN model. Our results showed that the ANN model exhibited promising performance (r = 0.82, fractional bias = 0.002) when comparing the > 1000 h PM10 measurements. When comparing the diurnal hourly measured PM10 variations in a clear-sky day, both the ANN and CFD models performed well (r > 0.8). The good performance of the CFD model relied on the knowledge of the in situ diurnal traffic profile, the adoption of suitable mobile source emission factor(s) (e.g., from MOBILE 6 and COPERT4), and the use of urban thermal and dynamical variables to capture PM10 variations in both neutral and unstable atmospheric conditions. These requirements/constraints make it impractical for daily operation. On the contrary, the ML (ANN) model adopted here is free from these constraints and is fast (less than 0.1% computational time relative to the CFD model). These results demonstrate that the ANN model is a superior option for a smart city application.