A framework for predicting the non-visual effects of daylight - Part II: The simulation model

This paper describes a climate-based simulation framework devised to investigate the potential for the non-visual effects of daylight in buildings. It is part 2 of a study where the first paper focused on the formulation of the photobiological underpinnings of a threshold-based model configured for lighting simulation from the perspective of the human non-visual system (e.g. circadian response). This threshold-based model employs a static dose-response curve and instantaneous exposure of daylight at the eye to estimate the magnitude of the non-visual effect as a first step towards a simulation framework that would establish a link between light exposure at the eye in an architectural context and expected effects on the non-visual system. In addition to being highly sensitive to the timing and duration of light exposure, the non-visual system differs fundamentally from the visual system in its action spectrum. The photosensitivity of the retinal ganglion cells that communicate light exposure to the brain is known to be shifted to the blue with respect to the photopic sensitivity curve. Thus the spectral character of daylight also becomes a sensitive factor in the magnitude of the predicted non-visual effect. This is accounted for in the model by approximating yellow' sunlight, grey' skylight and blue' skylight to three distinct Commission Internationale de l'Eclairage (CIE) illuminant types, and then tracking their circadian-lux' weighted contributions in the summation of daylight received at the eye. A means to condense' non-visual effects into a synthesised graphical format for the year, split by periods of the day, is described in terms of how such a format could inform design decisions. The sensitivity of the simulation model's predictions to prevailing climate and building orientation is demonstrated by comparing results from eight European locations.