Multidimensional evaluation of solar energy on urban buildings for driving the energy transition: Insight from Hong Kong, China

Abstract

In the global transition to low-carbon energy, the multidimensional evaluation of solar energy utilization on building surfaces is crucial for sustainable urban development. Existing studies lack an understanding of the complex interactions between multiple periods, spatial orientations, radiation types, and dynamic thresholds, which hinders a revelation of the spatial–temporal variability of solar energy utilization. This study presents a novel urban-scale multidimensional utilization indicators prediction model with a multi-source heterogeneous data fusion method, employed in Hong Kong, China, with a coefficient of determination (R²) of 0.948 and a root mean square error (RMSE) of 0.228. Results demonstrate that the total solar energy (TSE) reserves of urban building surfaces reach 170 TW·h, with over 60% deemed usable. The mean shading ratio (MSR) is 39.97%, with the roof being the lowest at 10.76%, the south facade at approximately 40%, and the other facades at around 50%. Multiple coupled regional, seasonal, and orientation variability in the mean ratio of energy (MRE) over threshold is captured by combining a reasonable baseline utilization threshold (BUT) and a dynamic threshold field (DTF). Model interpretability and parameter sensitivity analyses reveal key variables that influence MRE across various orientations. The practical utilization potential analysis further uncovered spatial and temporal heterogeneity, offering new insights into optimizing installation deployment.