PNNL

Abstract

Using a simulation based on a real, five-channel tunable LED lighting system, we show that Correlated Color Temperature (CCT) is not a reasonable predictor of the biological potency of light, whether characterized with CIE melanopic Equivalent Daylight Illuminance (mel-EDI), Equivalent Melanopic Lux (EML) (a scalar multiple of mel-EDI), or Circadian Stimulus (CS). At a photopic corneal illuminance of 300 lux and Rf = 70, spectra can vary in CS from 17% to 41% across CCTs from 2500 K to 6000 K, and up to 23% at a single CCT, due to the choice of spectrum alone. The CS range is largest, and notably discontinuous, at a CCT of 3500 K, the location of the inflection point of the CS model. At a photopic corneal illuminance of 300 lux and Rf = 70, mel-EDI can vary from 123 lux to 354 lux across CCTs from 2500 K to 6000 K and can vary by up to 123 lux at a fixed CCT (e.g., 196 lux to 319 lux at 5000 K). The range of achievable mel-EDI increases as CCT increases and, on average, decreases as color fidelity as characterized with IES TM-30 Rf increases. These data demonstrate that there is no easy mathematical conversion between CS and mel-EDI when a spectrally diverse spectra set of spectral power distributions is considered.