An image-based analysis of solar radiation for urban settings

This paper describes a novel approach for evaluating the total annual/monthly irradiance incident on building facades in urban settings. The analysis is founded on a physically-based rendering approach and uses datavisualisation techniques to generate ‘maps’ (i.e. false-colour images) of annual/monthly irradiance. The irradiance ‘maps’ are derived from hourly time-series data for one year and take accurate account of shading by, and inter-reflection from, other buildings and surfaces. The sun and sky irradiance are evaluated separately. The sky contribution is calculated using realistic, non-isotropic models for the sky radiance distribution. The ‘maps’ can be used to confidently identify facade-locations where there is high irradiance, for example to aid the siting of photo-voltaic panels. The technique can be applied to scenes of arbitrary complexity from a single building to fully ‘worked-up’ city models. The results of the analysis have been linked to a GIS-based solar energy planning system. The system is targeted at city planners and one of its aims is to encourage the consideration of solar energy in the urban planning process.