As clear as the benefits of amber lighting are, why don’t others like the International Dark-Sky Association advise its use?

They do… sort of.

If you read the recommendations carefully you will see wording like this:

International Dark-Sky Association – Fixture Seal of Approval application guidelines [emphasis added]

…
For luminaires with integrated LED light sources, the light source shall have a listed correlated color temperature (CCT) configuration of 3000K or below
…

and

American Medical Association recommendations – Report of the Council on Science and Public Health Report 2-A-16, Human and Environmental Effects of Light Emitting Diode (LED) Community Lighting [emphasis added]

…
2. That our AMA encourage minimizing and controlling blue-rich environmental lighting by using the lowest emission of blue light possible to reduce glare.

3. That our AMA encourage the use of 3000K or lower lighting for outdoor installations such as roadways.
…

So if read carefully these authorities support the use of amber lighting, but they are disappointingly unclear on the benefits of doing better than the limit of 3000 K CCT white LED. Unfortunately most readers are misunderstanding these recommendations to mean 3000 K is the de facto recommendation, missing the important “… or less.” This lack of clarity means communities are choosing LEDs with sky glow impacts 100% greater than HPS, 200% greater than good PCA LED, and more than 400% worse than NBA LED; these increases are even worse when considering potential health impacts (i.e. melatonin suppression). It is a consequential misunderstanding arising from poor choice of words.

We think it is vital that guidance be expressed more clearly, and do so in our recommendations.

Visibility/Safety

Many in the lighting industry claim that white light provides better vision and by implication improved safety on roadways. (For a general if brief discussion of lighting and safety on roadways – outside of the spectral issues discussed here – we recommend the Illinois Coalition for Responsible Outdoor Lighting (ICROL) Outdoor Lighting and Safety)

These claims are based on misinterpretation of research results (e.g. those by Lewis (1998; 1999) and more recently Gibbons et al. (2016)). These researchers have placed persons in test setups (such as driving simulators or driving cars on special test roadways), and investigated visibility with different kinds and amounts of outdoor lighting. For example, test subjects in a driving simulator were presented with a picture of a person standing at the side of the road – the experimenters measured how long it took the subject to determine whether that person was facing toward or away from the roadway. Another study had test subjects drive an automobile on a test roadway (with no other vehicles), and told them to press a button when they saw a person standing at the side of the road – detecting that “pedestrian” from a greater distance is considered a better performance, naturally.

Some of these tests showed differences in performance with different kinds of lights, but we consider it important that the more recent work, based on more realistic test conditions (a real car on a test road vs. a person in a driving simulator) show small or even no effect of spectrum.

But the conclusion that white lighting is better is not supported for the following reasons:

1. All performance measures improve when lighting levels are increased, for all types of lights. Even if these performance measures were actually meaningful indicators of safety on the roadway (which see below), then equal performance can be had with all lighting types by simply adjusting lighting levels accordingly. This point was made long ago (e.g. Adrian, 1998). Recent results indicate that the adjustments necessary to achieve equivalent performance are actually not very large – 30% or less in conditions relevant to outdoor lighting design (see e.g. IESNA, 2010, Table 4.2). (The dramatically greater effects reported in the oldest research (Lewis, 1998) have not been confirmed by other researchers.)
2. The experimental conditions in these studies raise serious questions about the applicability of the results to real-world safety. Test subjects are presented with dramatically simplified driving situations and are informed a priori of what they are to be looking for (unlike a real driver responding to the unexpected). The drivers are not tired, inebriated, or distracted – critical, even dominant, influences in accidents. These shortcomings severely compromise the reliability of conclusions about real driving circumstances.
3. The largest and most often quoted effects showing benefits for broad-spectrum lighting are observed only at extremely low brightness levels that are very likely irrelevant to roadway environments.
4. Many of the claims for improved visibility with white sources revolve around the ability to perceive color, but color perception is provided in nearly every circumstance by vehicular lighting (headlights), independent of any fixed roadway lighting, and is extremely limited or even nonexistent at low illumination levels where the largest effects are found.

For general lighting for parking areas and even on roadways, given the huge sky glow and other environmental benefits of doing better than 3000 K CCT white LED (and even 2700 K CCT), the way forward is clear and our recommendation is likewise clear.

References

• Adrian, W., 1998,  “The Influence of Spectral Power Distribution for Equal Visual Performance in Roadway Lighting Levels,” Proceedings of the 4th International Lighting Research Symposium
• Illuminating Engineering Society of North America (IESNA), 2010, Lighting Handbook, 10th ed.
• Gibbons, R., et al., 2016, “Applicability of mesopic factors to the driving task,” Lighting Research & Technology 48:70
• Lewis, A, 1998, “Equating Light Sources for Visual Performance at Low Luminances,” J. Illuminating Engineering Society 27:80
• Lewis, A., 1999, “Visual Performance as a Function of Spectral Power Distribution of Light Sources at Luminances Used for General Outdoor Lighting,” J. Illuminating Engineering Society 28:37