Daylight is dynamic and rich in the blue part of the spectrum. To date, the spectral composition of daylight is ignored in sky models used in Radiance. Spectral sky composition is particularly important when non-image-forming (NIF) light effects are concerned, since the action spectrum for these effects differs from that of the image-forming (IF) effects as it is shifted towards the blue. Recently, Stock has implemented the analytic spectral sky model proposed by Preetham and colleagues approximating daylight’s full spectrum in Radiance for a clear sky condition (Preetham, Shirley, & Smits, 1999; Stock, n.d.). However, this spectral sky has not been used in Radiance for irradiance simulations. This study is focusing on the accuracy of daylight simulations in Radiance implementing the spectral sky model. The aim is to investigate the usability of the spectral sky model in simulations for both IF and NIF calculations. Simulations are performed using the CIE clear and the spectral sky model for two locations (Eindhoven, the Netherlands and Utah, USA). Spectral data (in RGB bands) throughout working hours were estimated and compared for different sun positions (March 15th, June 15th, and December 15th). For the Eindhoven location, simulated data are compared with small-scale measurements. To assess the NIF effects of daylight using simulated data the method suggested by (Geisler-Moroder & Dür, n.d.) is implemented. The findings show that, when the two sky models are compared, results are comparable for the higher sun angles, but significantly different for low angles.
|Title of host publication||International Radiance Workshop 2016|
|Publication status||Published - 26 Aug 2016|
|Event||15th Annual International Radiance Workshop 2016, 29-31 August, Padua, Italy - Padua, Italy|
Duration: 29 Aug 2016 → 31 Aug 2016
|Conference||15th Annual International Radiance Workshop 2016, 29-31 August, Padua, Italy|
|Period||29/08/16 → 31/08/16|