Andrew Crumey has written a paper that you might be interested in:
"Human Contrast Threshold and Astronomical Visibility" (accepted in
MNRAS). It presents a model which gives visibility predictions somewhat
different from the standard Hecht-Schaefer model, but in better
agreement with the laboratory data (Blackwell, Knoll et al). In the
conclusions, Andrew makes some proposals regarding sky quality
indicators for public use.
http://arxiv.org/abs/1405.4209
