Re: ...skyglow (fwd)

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

> >  Today I've finished a draft image trying to explain why the light
> > going
> > almost horizontally is the most polluting one, even if just skyglow
> > is concerned.
>
> I like it very much.  Where did the numbers come from?

Thanks, Steve, I'm happy the graphics works.

As given rather shortly at the bottom of the page listing,
  http://amper.ped.muni.cz/light/drafts/graphics/
 the exact directions where the light is scattered from the original
direction are those of ``CIE sky No.5'', the non-Internet reference is in
that long draft of mine, lp_quant3 within
  http://amper.ped.muni.cz/light/drafts/
 and the formulas are within that paper.

> Is there any data at the horizontal plane of the fixture and just before?

The proportion of the light which is scattered to the ground should be not
very different from the case when it goes 5 degrees upwards, so roughly
one half again. In fact, for horizontal light, it can be over 60 per cent,
as the light which is scattered down hits the ground indeed and the light
scattered just a bit upwards is scattered once more for almost sure,
almost half of it down to the ground again. The CIE indicatrix accounts
for some multiple scattering, but rather just for directions over 10
degrees upwards. It is a simplification. The downward scattered proportion
is higher in fact even for the bottom case in the graphics, but if it is
45 % or 50 % makes no large relative difference. My model is rather
simple, the only detailed computations have been made by Pierantonio
Cinzano, but I think not published in this form.

> The NY state bill changed from full-cutoff (FCO) to fully shielded (FS)
> which displeases me greatly.  This type of diagram can be used to show
> that
> radiation just below the horizontal plane can contribute much to skyglow.

I think it holds for luminaires on top of a hill. On a flat terrain, even
when going just one degree downwards, the ray hits the ground at 57
times the height of the fixture, so in a clear air just a tiny fraction
scatters (in another words, the luminance of the luminaire, being not
dependent on its distance in a fully transparent medium, is not much
diminished by extinction). Of course, if the luminaire itself is hidden
behind an obstacle, say a balcony of a nearby house, even the downward
going rays would produce an observable bright spot on the sky around it, of
course it would be just very low in the sky.

The difference between FS and FCO on a flat terrain is very little as the
skyglow is concerned, but may be very large in another aspect, namely the
glare. In many cases this is the limiting factor why the people don't
notice even bright planets low in the sky -- they are bright enough to be
conspicuous even on a city sky, but when they are thousand times fainter
then the nearby luminaires, just those experts who seek them will see them
indeed.

Still, I'm in favour of having FS rather then FCO as a primary criterion
in any legislation. FS property can be mostly judged at a very glance on a
fixture, whereas FCO needs complicated measurements and the data given by
producers may be questionable. Some minor new producers don't have them at
all, and would have a clear disadvantage over the established ones. For
many purposes it's not needed to know the exact photometric data for a
luminaire. It is quite sufficient to see that the mirror inside has a
low-absorbing surface and to try which pattern it produces on the ground
when used at night. From that knowledge, it may be easy to judge if it
fits the given poles or not, to fulfil the illumination task efficiently.

Another reason is, that FCO has a very loose limit for the light at 80
degrees from nadir: 100 cd/klm, whereas ``CIE cut-off'' category demands
for <=30cd/klm. This is already a value which limits glare effectively,
being used e.g. on some European railways. So, if demanding more then just
FS, it should be rather ``FS and CIE CO'' (CIE CO itself is not enough, as
it allows 10cd/klm at 90 degrees and above, CIE has no category for
no-uplight luminaires...).

Such a more stringent demand to prevent skyglow would be really suitable
for luminaires, whose light can travel more than one kilometre when going
say 5 degrees downwards -- those on slopes or hilltops. But glare is more
important in these cases again then the rise of skyglow.

> -Steve

jenik

PS.
 I'm unable to follow the dslf and olf, so I avoid to write there now.
However, I put this letter to
  http://amper.ped.muni.cz/jenik/letters/public,
 so it's accessible for interested people.