13 replies to this topic

[AstroVPK]

What is the darkness of the night sky between stars? I ask because Roger Clark's Visual Astronomy book suggests 24.0 mpas as being the sky brightness at zenith from an ideal location. Is this the sky brightness corrected for all stars, or is the value just a misquote?

[GlennLeDrew]

The practically universally accepted, darkest terrestrial sky SB is 22 MPSAS. This principally comprises airglow and zodiacal light (sunlight scattered/reflected by interplanetary dust), with unresolved stars adding a bit to it.

 

To obtain 24 MPSAS one would have to above Earth's atmosphere, and perhaps in the outer region of the solar system as well.

[ABQJeff]

 

 

To obtain 24 MPSAS one would have to above Earth's atmosphere, and perhaps in the outer region of the solar system as well.

I always knew there was a trick to Roger Clark’s ability to see his objects with his C8!  He had a spaceship!

[AstroVPK]

The practically universally accepted, darkest terrestrial sky SB is 22 MPSAS. This principally comprises airglow and zodiacal light (sunlight scattered/reflected by interplanetary dust), with unresolved stars adding a bit to it.

To obtain 24 MPSAS one would have to above Earth's atmosphere, and perhaps in the outer region of the solar system as well.

Yes, that's what I thought too. However, the puzzle is that Clark's book gives 24.0 mpas at zenith and it's too straightforward a mistake for a person of his caliber to make. Hence my wondering if the subtlety lies in 24 mpas being the value without stars, i. e. just airglow, etc....

[JohnTMN]

 I ask because Roger Clark's Visual Astronomy book suggests 24.0 mpas as being the sky brightness at zenith from an ideal location.

"An Ideal Location"

That's the key,, where is that exactly?

Theory, Reality,, 

You choose,

[Tony Flanders]

I think the honest answer is that it varies tremendously. Not so much by the observer's location as by time. In any case the optimal brightness between stars is certainly quite a bit darker than 22 mpsas, which is the figure usually cited when stars are included.

The two main non-stellar natural light sources in the night sky are airglow and the zodiacal light. Airglow varies hugely depending on the solar cycle, the time of night, and the latitude. It also has large-scale structure that's not well understood. In a recent thread in the Light Pollution forum, LP-measurement veteran Dan Duriscoe showed an all-sky brightness panorama taken from Mauna Kea where airglow appeared to be almost 21.0 mpsas across much of the sky. But at its darkest, airglow is likely less than 23.0 mpsas.

Zodiacal light is presumably temporally constant, but varies greatly depending on the distance from the ecliptic plane and the distance from the Sun. It famously has a second peak directly opposite the Sun called the Gegenschein.

Edited by Tony Flanders, 14 August 2024 - 05:56 AM.

[Redbetter]

What is the darkness of the night sky between stars? I ask because Roger Clark's Visual Astronomy book suggests 24.0 mpas as being the sky brightness at zenith from an ideal location. Is this the sky brightness corrected for all stars, or is the value just a misquote?

Assuming pristine sky, even nearly overhead it will depend on the portion of sky one is examining (e.g. near the galactic poles or in the Milky Way glow.)  The answer also depends on the magnitude limit one is assuming for a cutoff, which is aperture/observer dependent.  I might choose 17+ mag for the cut off for resolving stars in a 20", and 15+ in an 8".   

 

The practically universally accepted, darkest terrestrial sky SB is 22 MPSAS. This principally comprises airglow and zodiacal light (sunlight scattered/reflected by interplanetary dust), with unresolved stars adding a bit to it.

 

To obtain 24 MPSAS one would have to above Earth's atmosphere, and perhaps in the outer region of the solar system as well.

I doubt that is true.

 

I think that if you try to calculate it you will get a far different answer.  100% for certain the mean darkness between stars in pristine skies is darker than 22 mpsas V mag.  The question is how much and in which portion of sky?  In the Milky Way glow and not in a dark nebula zone, the sky between stars will be relatively bright even in large aperture.  But looking away from the galactic disk into regions with sparse star counts the answer will be far darker.  Clark could be correct for some portions of the sky.

 

I have worked through some calcs doing different star count binnings just considering the overall sky.  Unfortunately the completeness has been poor past 10th magnitude.  This is far too bright of a cut off for determining star contribution to the overall sky glow.  Gaia is supposed to be complete between 12 to 17 G magnitude, but one would have to convert all of the data to V mag.  From that we could get a better estimate of how the rate of star count incremental rise declines for each additional magnitude.   Even though higher magnitudes will have less completeness, they could be useful for contributing to estimates and making extrapolations.

 

Related to this I am hoping that a Vmag catalog will be created from the Gaia data as away to show the sky brightness contours.

 

If my calcs are not off, my estimate of the contributions of all stars to ~15+ magnitude is about 22.7 mpsas. vs. 23.7 mpsas at mag 9+.  But again that is the average across the sky.

 

In the past two years, natural air glow has clearly been brighter visually than it was in the several preceding years.  I have also found it much more variable.

 

And of course the zodiacal glow stretches across the entire sphere of the sky, even if it is brightest along the ecliptic.  It isn't until one reaches Voyager type distances that it is almost entirely absent. 

[Tony Flanders]

Assuming pristine sky, even nearly overhead it will depend on the portion of sky one is examining (e.g. near the galactic poles or in the Milky Way glow.)  The answer also depends on the magnitude limit one is assuming for a cutoff, which is aperture/observer dependent.  I might choose 17+ mag for the cut off for resolving stars in a 20", and 15+ in an 8".

I assumed that the question meant the darkness between stars in the absolute sense, i.e. how bright the sky would be if all stars in our galaxy were eliminated, including ones too faint to resolve in a particular telescope by a particular person.

In that case the position in the Milky Way becomes much less important, though not completely irrelevant due to the large nebulae that pervade our galaxy.

[Redbetter]

I assumed that the question meant the darkness between stars in the absolute sense, i.e. how bright the sky would be if all stars in our galaxy were eliminated, including ones too faint to resolve in a particular telescope by a particular person.

In that case the position in the Milky Way becomes much less important, though not completely irrelevant due to the large nebulae that pervade our galaxy.

I assumed the same, but you can't ignore the magnitude limit or what is beyond that. and therefore unresolved.  Rather than your interpretation, the Milky Way position is far more important.  Looking into less obstructed (star dense) portions of the MW, the unresolved stars are a huge factor past some magnitude limit.  This is why relatively small portions of the total light dome have such a negative impact on the surface brightness of the wider sky sampled when birther portions of the MW are overhead--without even accounting for the stars within various bins.  

 

Simply put, you have to assume a magnitude limit somewhere, or the whole exercise is meaningless/unanchored.  Everything dimmer than X magnitude in a field is assumed to be unresolved and contributing to the overall surface brightness.  This very much depends on the aperture used.  In pristine conditions we might see the difference as ~22mpsas overhead when the MW is on/near the horizon, and 21.65 mpsas or so with it directly overhead.  

 

In brighter skies these sort of considerations don't matter much because light pollution dominates the light between dim stars, but as one approaches the pristine the actual unresolved star density can have a big impact.  It is all in the math (e.g. doing the light pollution ratio with 22.7 mpas, 23 or even 24 as the pristine level between stars.)  

 

And that is also why seemingly small differences in mpsas can matter so much in skies that are near pristine.  One is not comparing the contrast to 22mpsas (an average that includes stars), but to 23 or more.  Light pollution can wash that out easily.  

 

I appreciated this more when I realized how much easier it was to pick up certain Terzan globulars and the like when they happened to be placed within a notch in a dark nebula..  This made them much easier to detect than I had anticipated, despite the wider sky in the region being much brighter overall.  It all comes down to local contrast. 

[Starman1]

Well, the 24.0 sky brightness figure could be true for certain spots in the sky, but it certainly isn't true for any multi-degree sections of sky anywhere on Earth.

I just consider this one of Clark's many errors, along with his "Optimum Detection Magnitude".

[timelapser]

Simply put, you have to assume a magnitude limit somewhere, or the whole exercise is meaningless/unanchored.  Everything dimmer than X magnitude in a field is assumed to be unresolved and contributing to the overall surface brightness.  This very much depends on the aperture used.

In general, yes.  But we can also ask how dark can it get when measuring with specific equipment, especially a de facto standard like SQM-L.

 

Slightly tangentially, outside of most of the zodiacal light New Horizons has made measurements of the glow between stars, after subtracting scattered light and unresolved stars.  They also needed to subtract diffuse Galactic light, ie Milky Way light scattered from interstellar dust.

 

They find a result consistent with that expected from galaxies.  Their numbers are given in W/m^2/sr but correspond to something like 26 V mpsas for the total glow out there.  The sky from space near Earth is very roughly 10 times brighter due to the zodiacal light, so something like 23.5 V mpsas.  So a dark sky site on the ground at 22 mpsas would be roughly 4 times brighter still, due to airglow.

[Tony Flanders]

This is obviously one piece of a much bigger question -- the total Light of the Night Sky. A famous pamphlet by that title was published in 1973 by F. E. Roach and Janet L. Gordon; much of the data originally appeared in a 1964 paper by Roach which you can read here.

 

Perhaps that deserves a different thread, however.

 

Redbetter's question about how the total stellar brightness breaks down by star magnitude is also a very interesting one. If the total brightness of all stars mag 15.5 and brighter does indeed average out to 22.7 mpsas, that suggests that stars fainter than mag 15.5 don't contribute much, since 22.7 mpsas is almost half the nominal grand total of 22.0 mpsas including airglow, the zodiacal light, and more exotic sources like distant galaxies.

 

I imagine it depends greatly what part of the sky you're examining. There aren't a huge number of ultrafaint stars in the Cygnus Star Cloud, due partly to its proximity and partly to the fact that stars in galaxies' spiral arms tend to be young. The Great Sagittarius Star Cloud, by contrast, seems full of unresolved stars even in very big telescopes.

[Redbetter]

Well, the 24.0 sky brightness figure could be true for certain spots in the sky, but it certainly isn't true for any multi-degree sections of sky anywhere on Earth.

I just consider this one of Clark's many errors, along with his "Optimum Detection Magnitude".

I suspect that it is true or nearly so, for multi-degree sections of the sky, once stars to some magnitude level are subtracted.  This is more likely near zenith well away from the ecliptic plane (minimizing the glow from solar system dust.)   And this is where natural air glow will be at its weakest as well. Keep in mind that we are talking about how dark the sky is between stars, not the net including stars.

 

I am less certain what that star magnitude limit would be to make it work.  Naked eye the change in the sky subtracting visible stars, even assuming 7th mag NELM, works out to an average flux of the stars that would be ~24.17 mpsas by my calculation/estimate.  So just subtracting the naked eye stars to that level should reduce the total flux by less than 14%, making the sky about 0.16 mpsas darker for the average of the sky  Sounds small, but there is a catch...

 

The problem here is working from averages for the sky.  Star density varies so greatly, so the unresolved star glow (far deeper than 7th mag) is missing in portions of the sky while it is intense in the Milky Way.    I took advantage of the very sparse regions doing my tests of how dark was effectively black (actually more of an indistinguishable dark gray) by using tiny exit pupils until the field stop could no longer be resolved.  I had to keep any field stars away from the field stop, but also have one somewhere near the middle of the field, just bright enough to anchor the exit pupil and hold focus. 

 

The sky I did this in was very dark (21.6+ mpsas for two different refractors), but not pristine, so I was using the SQM-L at the time for the average surface brightness that was then dimmed.  The actual surface brightness in this nearly starless stretch would have been somewhat dimmer, although actual light pollution limited how much dimmer it could be (along with natural airglow during solar minimum, and solar system dust far from the ecliptic.)   So the 28 mpsas calculated darkness of the field at the transition was a conservative estimate.  The actual field would have been somewhat dimmer.

 

 

In general, yes.  But we can also ask how dark can it get when measuring with specific equipment, especially a de facto standard like SQM-L.

 

Slightly tangentially, outside of most of the zodiacal light New Horizons has made measurements of the glow between stars, after subtracting scattered light and unresolved stars.  They also needed to subtract diffuse Galactic light, ie Milky Way light scattered from interstellar dust.

 

They find a result consistent with that expected from galaxies.  Their numbers are given in W/m^2/sr but correspond to something like 26 V mpsas for the total glow out there.  The sky from space near Earth is very roughly 10 times brighter due to the zodiacal light, so something like 23.5 V mpsas.  So a dark sky site on the ground at 22 mpsas would be roughly 4 times brighter still, due to airglow.

As illustrated in my response above, the SQM-L provides only an approximate naked eye sky brightness.  Beyond that it will only get one in the ball park if the sky is relatively dark.  Roger Clark's between stars value would be for pristine sky.  However, somewhere in the sub 21 mpsas level, light pollution is swamping everything else in terms of how bright the sky is between stars, so little else matters in that case, except of course when natural air glow is intense.

 

Where did you see the Earth based zodiacal light value in the paper?  I couldn't find it.  23.5 mpsas sounds somewhat bright for regions well away from the ecliptic and gegenschein (and Zodiiacal light), but I haven't seen any recent papers that made a careful look at these far off axis, so I don't have a firm basis.  The 1964 paper values for the gegenschein central intensity look considerably brighter than the 2013 measures.  Unfortunately the 2013 intensities are only viable to about 20 degrees off center in what is presented, the far off axis data isn't provided.

 

If the dimmest portion of the zodiacal glow is indeed only 23.5 mpsas, then it will set an absolute limit for how dark the area between stars will appear.

 

. 

[timelapser]

Where did you see the Earth based zodiacal light value in the paper?  I couldn't find it.  23.5 mpsas sounds somewhat bright for regions well away from the ecliptic and gegenschein (and Zodiiacal light), but I haven't seen any recent papers that made a careful look at these far off axis, so I don't have a firm basis.  The 1964 paper values for the gegenschein central intensity look considerably brighter than the 2013 measures.  Unfortunately the 2013 intensities are only viable to about 20 degrees off center in what is presented, the far off axis data isn't provided.

 

If the dimmest portion of the zodiacal glow is indeed only 23.5 mpsas, then it will set an absolute limit for how dark the area between stars will appear.

The numbers I quoted are from the precursor New Horizons paper, at the end of the intro to section 4 on pg 12.  Their measured value of 26 mpsas from over 40 AU is at high galactic latitudes, excludes resolved sources (stars and galaxies), but includes significant scattered starlight from outside the fields (see Fig 9 in the newer paper).  The glow from unresolved stars is subdominant (Fig 9) and the remaining main components are starlight scattered from interstellar dust and the background of unresolved galaxies (which is what they were trying to measure).

 

In the precursor paper they only say that 26 mpsas is "more than 10x fainter than the darkest sky available to the Hubble Space Telescope", but don't give any HST references.  Hence my value of 23.5 mpsas, and this presumably is for high ecliptic latitude (hence the darkest HST sky).  I haven't tried to track down any HST references.  But indeed this would represent the floor for the sky between stars from the ground, with the difference from 22 mpsas due to airglow.

 

(Technically the sky in Earth orbit would be a bit brighter than that from a dark site on the ground minus airglow, due to atmospheric extinction, but that's a fairly small effect.)