[Darksky]Re: [OL-Forum] Uplight Examples Report

Jan Hollan
Tue, 27 Nov 2001 23:26:43 +0100 (CET)

Going from uplight to polluting downlight:

the example is really interesting. However, the light coming down again
from the sky is what matters. In this respect, two of the luminaires in
the report differ quite a lot: the light going above horizon amounts some
100 percent more light pollution in one case and some 30 per cent increase
in another case.

There are two reasons for it. First, the almost horizontal light travels 
a longer path in the atmosphere, so a large fraction scatters. Second, the
scatter is mostly forward-directed due to aerosols, making horizontal
directions more harmful (for steep ones, most of the dispersed light
escapes to the space). 

To be able to make a qualitative estimate, I have supplemented my
new programme for viewing and sorting ies files by a ``light pollution
increase ratio'' computation. Four alternative angular
scattering functions are available: 
 the one used in Cinzano et al, 2000, 
 as offered at 
   Linkname: Papers and preprints about light pollution
        URL: http://debora.pd.astro.it/cinzano/papers.html )
 and the three for CIE clear skies (4,5,6) (I took them from Darula et al,
private comm.). Single scatter is considered, simply in such a way, that
from the total light not escaping to the universe, a down-pointing
scattered part is computed. Transparency of the atmosphere and albedo of
the terrain (lambert scattering is considered there) can chosen.

Here is the overview output of ies2tab for the six luminaires:

#  % of Increase of Light Pollution due to light going
#   directly from the luminaire above horizon
#   concerns the following situation:
#  Albedo = 0.10
#  Zenith Extinction = 0.30 mag (i.e., direct light remaining  76 %)
#  Indicatrix type =0 (0: acc. to P.Cinzano, 4..6: CIE sky types)

#Fi Out >= 80degrees    >= 90degrees      CutOff?        filename
#  %    cd/klm % Out  cd/klm % Out % I.L.P.
  67.3   36.8   0.4    0.0   0.0    0   Full_CutOff     samp0187.ies
  76.9   90.5   6.8   21.7   3.6  103   IES_CutOff      samp0191.ies
  73.0   58.6   2.6   16.0   1.0   32   IES_CutOff      samp0195.ies
  79.7   38.4   0.7    0.0   0.0    0   Full_CutOff     samp0209.ies
  73.4  208.6   1.1    0.0   0.0    0   Fully_Shielded  samp0478.ies
  73.8  316.4   2.7    0.7   0.0    0   Non-CutOff      samp0479.ies

David's result for the Uplight Unit Density, that the first FCO and the
last ,,non-cutoff`` lights perform best, are encouraging. It's true that
the classification itself says little, esp. without the existence of a
simple fully-shielded class (that luminaire ranked third best for the
purpose). It fails just on the 80 degrees limit to be FCO, and that's true
also for the Non-Cutoff case (a closer look on the table made by
  ies2tab samp0479.ies > samp0479.tab 
 shows, that just one single value at >=90 degrees is over 0.5cd/klm). 

Even better, the first FCO one (with the least output flux) performs best
in the UUD sense and limits glare in the same time. The other two with
over 200cd/klm may give enough (uniform) illuminance and/or luminance, but
scarcely a good lighting (even if they pass a glare index criterion - it's
perhaps too loose).   

The programme and sample data are available at 
 (I should stress, that the ``Light Pollution Increment'' computed there,
e.g., for thousands of GE ies files, has no meaning in many cases -- for
luminaires not meant to illuminate roads etc. So the computed illuminances
of an arbitrary rectangles -- luminaires work mostly together (but an idea
on ``distribution type'' may be got this way and, eventually, suitable
luminaires might be batch-found).)

clear skies, 
Jenik Hollan

>    Date: Tue, 20 Nov 2001 16:36:38 -0700
>    From: "David Keith" <david.keithAmindspring...>
> Subject: Uplight Examples Report
> In response to requests, I put together a report on the uplight
> examples in my SALC 2001 presentation (available at
> http://resodance.com/mdi/SALC2001.html).  The report is on the web at
> http://resodance.com/mdi/UUDcalc.html and the associated photometry is
> available at http://resodance.com/mdi/SALC_exs.html and any comments,
> suggestions etc are welcome.
> David Keith