Evaluation of Perez slope irradiance and illuminance models against measured Hong Kong data

This paper presents the validation of the Perez inclined irradiance and illuminance models for Hong Kong using three-year (1996-1998) measured data. The simplified circumsolar (α = 25°) and point source versions using the local coefficients and those given by Perez have been assessed. A statistical analysis of the Perez models has shown that predictions for solar irradiance provide more accurate results than daylight illuminance. All versions tend to have less statistical errors for shaded vertical surfaces. For both vertical solar irradiance and daylight illuminance estimations, the circumsolar ((α = 25°) version produces the largest errors when using coefficients proposed by Perez but it shows the best performance when local coefficients are fitted. It may suggest that the large number of empirical coefficients used in the Perez models make them site-specific.