A pilot daylighting field study: Testing the usefulness of laboratory-derived luminance-based metrics for building design and control

This paper documents a field study that collected visual comfort feedback and High Dynamic Range (HDR) luminance maps to test the validity of laboratory-derived metrics in field settings. This study included 13 subjects for 149 daylighting evaluations. Similar to previous laboratory-based research, simple luminance-based metrics outperformed complex glare indicators in their predictive ability in this pilot field study. Also similar to lab studies, metrics derived using luminance data within the horizontal 40° band of human vision and a mask of just the window represented the majority of the top 20 squared correlation coefficients, considering all questionnaire items. While standard deviation of window luminance, a highly predictive metric derived in a laboratory setting, did not maintain similar predictive ability in these field studies, the field study supported previous laboratory findings that this metric may be useful in discerning visual discomfort in spaces that are very dim. Two new metrics, the ratio of the 98th percentile to mean window luminance and the coefficient of variation (COV) of luminance in the horizontal 40° band, yielded the strongest squared coefficients in field studies (adjr2 = 0.35 and adjr2 = 0.32 respectively). The COV 40° band was more stable across multiple positions within a scene than other luminance-based methods. These findings suggest that some metrics are better suited for use within a single room over time (for purposes of dynamic daylighting scene control), whereas other metrics are better suited for evaluation between spaces and from alternate positions within a room (for purposes of design guidance).