Predicting daylight availability dynamically based on forecasts of a weather observatory

Abstract

By using measured data of the research class IDMP station, a subset of three sky types and their probability of occurrence has previously been established to represent sky conditions of Hong Kong. This was dubbed the Hong Kong Representative Sky (HKRS). The HKRS allows a better prediction of daylight on the vertical surfaces of buildings than the CIE Overcast Sky. The HKRS was still a static representation. This paper investigates the possibility of using the weather predictions of the observatory to establish dynamically the HKRS. A methodology is proposed, and the results indicate that the approach is feasible. It is established that the weather prediction information issued by the Hong Kong Observatory (HKO) provides a reliable means for estimating probabilities of the sky types at a particular moment. By using this information, a better prediction of vertical sky component (VSC) on building surfaces, with a reduction of 20–30% mean absolute error, can be achieved. Engineers and architects may use this more accurate information to design more dynamically.