Author Topic: Introducing DarkSiteFinder.com  (Read 648 times)

knigabretta

  • Active Astronomer
  • ***
  • Posts: 107
  • Activity:
    0%
  • Reputation: +0/-0
    • View Profile
Re: Introducing DarkSiteFinder.com
« Reply #30 on: February 02, 2018, 05:24:49 PM »
Quote
My first suggestion is to use the 255 grayscale map to create a new color mapping. It uses 7 colors:

clear - no light pollution, natural pristine skies (e.g. 22 magsas)
gray - 0-1 magsas brighter than natural, pristine skies (21-22)
blue - 1-2 magsas brighter (20-21)
green - 2-3 magsas brighter (19-20)
orange - 3-4 magsas brighter (18-19)
yellow -  4-5 magsas brighter (17-18)
white -  5+ magsas brighter (17+)

Aside from the practical problem of calibrating it, there's a more fundamental problem with this idea.

What's interesting is the amount of artifical skyglow, not the total skyglow. That's precisely the problem with the SQM; it measures total skyglow, and at the darkest levels (21.5 - 22.0) the variation is caused almost entirely by variations in natural rather than artificial sources.

Put another way, the difference between 18.0 and 19.0 is quite substantial, but by no means overwhelming. The difference in observing experience between 20.0 and 21.0 is much bigger, because at 21.0 quite a lot of the total light is coming from natural sources. And the difference between 21.0 and 21.5 (where less than half the glow is artificial) is even bigger than the difference between 20.0 and 21.0.

That's why the system of the original Light Pollution Atlas, which works in terms of artificial rather than total skyglow, makes more sense to me.

James Merrill

  • Active Astronomer
  • ***
  • Posts: 128
  • Activity:
    0%
  • Reputation: +0/-0
    • View Profile
Re: Introducing DarkSiteFinder.com
« Reply #31 on: February 08, 2018, 07:44:04 PM »
Perhaps. First, the website creator's intention is to help people find dark sites, not demonstrate the level of light pollution specifically, as was the intention of the World Atlas, So let's keep that goal in mind, in case it makes a difference.First, natural skyglow does vary, as we have both said above, so the map is necessarily relative to the natural sky brightness. I hesitated putting fixed numbers with the scale because of this. If we should take away the direct association with integer mpsas values, that's OK with me.The World Atlas works not in terms of artificial sky brightness alone, but in terms of a ratio between natural sky brightness and artificial sky brightness. Evidently they took this scheme exactly from Garstang's work, except for the addition of 0.1-0.10 dark gray. My understanding is that the key to deep-sky observing is the contrast between extended objects' surface brightness and the background sky brightness. I suspect that is what the current scale attempts to capture, and the main thrust of your objection. Yet it cannot hurt to review that scale to see if it remains the most effective way to indicate a significant difference between areas. I do not understand why the divisions used since the 1960's accomplishes that. The "3x" factor seems completely arbitrary to me.I'm happy to work on the basis of changes in contrast, if that is the way to go; I suspect we can refine your example values.I will PM Glenn LeDrew and ask him to join in, since he will certainly say something if he chooses to on the subject of contrast, and then maybe some of us can decipher what he might have meant. His mind works differently than mind (yeah, and better, too, I admit).  Regards,Lee

bermordliro

  • Active Astronomer
  • ***
  • Posts: 114
  • Activity:
    0%
  • Reputation: +0/-0
    • View Profile
Re: Introducing DarkSiteFinder.com
« Reply #32 on: February 09, 2018, 03:06:44 AM »
At Lee's behest... First, I should put it out there that I don't put quite so much stock in the vaunted LP maps, which is why this topic did not catch my eye. Why my aloofness regarding this map? I live near the edge of a 'red' zone in west Ottawa, yet on a really good night, from my *stupendously* light-awash apartment balcony I can just glimpse the 'Mexico' portion of the North Amarica nebula with unfiltered 10X50 binos. This object has a surface brightness of perhaps 24 MPSAS, implying that the sky SB must be no more than 3-4 magnitudes brighter, or 20-21 MPSAS. Further, my NELM on such a night is about 5.8m. In (admittedly near the edge of) a *red* zone?!Near sizeable populated centers, someone not too near the center of the light dome will often have a not insignificant gradient in sky brightness. For that reason it's hardly worthwhile to more finely quantize the brightness steps, at least for the brighter 1/2 or 2/3 of the scale (green--certainly yellow--and brighter.) But it might be worthwhile to more finely establish the divisions for the darker levels (due to the usually gentler brightness gradient), *if* the data permit this by not having significant error or noise.Now, I'm not dismissing the LP map out of hand. It is a most useful indicator of at least relative sky brightness variation. And it permits to identify populated places of even quite small size (when seeking to avoid all light sources possible.) What I 'object' to is that it's too easily taken as a precise indicator, resulting from not appreciating sky brightness gradients and the inherent variability of atmospheric aerosols/transparency.About the scaling of the map. It seems to quantize in factors of 3, where a factor of 2.5 (more precisely 2.512) would correspond to 1magnitude; more directly applicable to astronomers.Could a correlation between LP map brightness, Bortle scale, SQM and (possibly??) NELM be feasible? As long as certain assumptions are made and clearly stated, perhaps. For one thing, a standard baseline for a pristine sky is a given, which presupposes a reasonably clean and dry airmass. For another, the mean air glow brightness by latitude should be established (it does brighten toward the poles.) And for NELM, the most variable and contentious 'standard', some acceptable mean must be established.I don't use the Bortle scale, although I have a barely passing familiarity with it. But I do use my own variation of it, under dark skies, where certain milky way features are used as a gauge of sky quality. I find this more reliable than the SQM, as the latter merely measures brightness, which is not the same as clarity.This is admittedly something of a jumble of first thoughts on the matter. Perhaps I can focus more coherently on specific aspects...

reaipasjime

  • Active Astronomer
  • ***
  • Posts: 107
  • Activity:
    0%
  • Reputation: +0/-0
    • View Profile
Re: Introducing DarkSiteFinder.com
« Reply #33 on: February 09, 2018, 04:00:07 AM »
Thank you Glenn. I thought of you since I think the optimum display scale will have a known relationship with the contrast of the sky brightness v extended objects, and I think I have learned from your chart that it is not linear with magnitude.

This site has DaveL's 15-color map (minus snow), overlaid on Google Maps, just in case you wanted to double-check your exact location.

The Bortle Scale
----------------
The Bortle Scale does include NELM as one of its several attributes of each class, so that is done. Other than that, John Bortle himself has said his scale is often substantially off compared to the World Atlas (or the other way around), and his scale was not intended for use as a map input. He appeared to not want to be associated with the World Atlas (his posts are in another recent thread here, judge for yourself). He made similar remarks regarding SQM readings. So, I think we respect his preferences if we leave his scale as a standalone measure, to be used on-site each night.

I myself would like to be able to compare the LP maps with SQM readings and the iPhone app (mentioned elsewhere on this LP forum). We know they won't match, but by doing better work with more, and more recent, data, each tool can be improved. IMHO it is worth the attempt.

The statement that the existing scale quantizes in a factor of 3 is, I think, an oversimplification. I suggested just using magnitudes (and decimal fractions thereof) and we can read Tony's feedback.

The equation that is used to assign color in the World Atlas is:
(nsb = natural sky brightness, asb = artificial sky brightness)
color  range
black/clear...1.0nsb
gray...1.0 nsb to 1.10nsb (asp < 0.10 nsb)
blue...1.1 nsb to 1.33 nsb (asp < 0.33 nsb)
green..1.33nsb to 2.00 nsb (asp < 1.0 nsb)
yellow.2.0 nsb to 4.0 nsb (asp < 3.0 nsb)
orange.4.0 nsb to 10.0 nsb (asp < 9.0 nsb)
red....10.0 nsb to 28 nsb (asp < 27.0 nsp)
white..>28 nsb (asp >27 nsp)

Please correct me if I am wrong.
I don't think these come close to the same scale as magnitudes. i don't know how to convert this chart to magnitudes to find out.

The original research paper stated it used 21.6 magasa I believe as an average nsb.

BTW, there is no reason there should be just one chart, if it is helpful to have several or have one that is configurable.

Regards to all,
Lee