FS fixtures are vital for reducing sky luminance

[Jan Hollan <jhollan@amper....muni.cz>]

Hello quality lighting advocates,

I just overcome a long lag in reading the dslf. David Keith's remarks on 
a bit higher uplight from a new installation of a road lighting using
flat-glass FCOs compared to the best non-FCOs (with a still very low
direct upward flux) may be true, at least if the fixtures are new and
luminous intensity maxima around 70 or even glaring 75 (``useful'' for
rising the road luminance by semispecular reflection)  degrees hold
(deprecation may concern namely those maxima, not the entire geometry
homogeneously). The differences are not very large, however, using best
available technology. 

(I wonder if it could evolve by lowering the luminaire efficiency a bit by
mirroring the direct downlight from the lamp (which cannot be properly
directed by the aluminium-surface cavity) partly upwards to the cavity,
getting a better coefficient of utilization (and lower
luminance/illuminance maxima). Or fresnel-refracting it, even when using a
flat bottom glass. Are there such road fixtures?  With the IP66 fixtures
it might be a persistent solution.)

(Another niche is using low-iron glass and antireflection coatings, is
there such an example?)

However, for the sky luminance, the uplight amount is not much
informative. As I have mentioned some time ago, I wrote a program
(originally for making human-readable tables from ies files) which
computes the sky LP increment from direct light from the luminaire.
Typically, 1 per cent of direct light going low above horizon adds some
30 per cent to the sky pollution. So FS fixtures ARE vital for reducing
man-made sky luminance.

This ratio depends only slightly on the ground albedo, assuming no snow
(my program accepts just one single value as an input presently, 0.10 is
default value, and supposes flat lambertian surface), a bit more on the
model of dispersion of light in the atmosphere (my programme uses the
model as in Pierantonio's works as a default, but three other indicatrices
can be used, as published in
 Kittler, R.: Relative scattering indicatrix: Derivation from regular
radiance/luminance sky scans. Light Res. and Technol., 25, pp. 125-127,
1993). 
 Absorption by aerosols is neglected by the program, so is absorption by
obstacles as trees etc. See the results in an old posting available at
  http://amper.ped.muni.cz/darksky/a/msg00027.html
 or directly data in the directory 
  http://amper.ped.muni.cz/light/ies2
 (the binaries for the program ies2tab are available there as well).

As FS or rather FCO (or generally, least glaring) fixtures are a must for
restoring the night sky, it would be very good, if recommendation of the
suitable models for road installations would be published. Which are the
best FCO fixtures for large pole spacing / pole height ratio, from your
experience (David?)? The answer is vital even for helping the Czech
anti-LP law to pass, not just the NY one. We have to demonstrate that FCO
is never too expensive, if properly chosen.  

My idea is to assemble the data on the best performing luminaires at one
site (as is the italian one for the images), encouraging producers to
submit their values. Why should everybody do the same long (never ending)
search? Remarks from the users could be added too (concerning performance
after 10 and more years, e.g.).

sunny skies,
Jenik Hollan
 
PS.
 ies2tab can be used for getting an overview of an illumination of a road
as well, by hundreds of different fixtures at once (an automated choice of
the best performing ones could be added too). As luminances are concerned,
the problem is I have no BRDF data for various asphalts/concretes -- do
you know where they are available?
 Another (new) feature of ies2tab is computation of the glaring flux
between 75 and 90 degrees. It can be even below 2 per cent, for the best
fixtures. Why should another be used?