[Chaos-l] What's wrong with this picture?

[D Gary Grady]

Jeff,

You raise an interesting point. A greatly overexposed crescent moon 
might well bloom out into a blob (especially in silver halide 
photography). On the other hand, it seems unlikely to me that it would 
look like a perfectly round disk.

After giving the matter some more thought, I think I can say with a fair 
amount of confidence that what we're seeing in the picture is the Moon 
illuminated by reflected light from the Earth. In fact, I'm convinced 
that at the exposure used, the lunar disk illuminated only by earthlight 
is necessarily overexposed. Here's my reasoning:

If the Moon and the Earth had the same albedo, earthlight on the Moon 
would be about 16 times as bright as full moonlight on Earth, simply 
because the Earth's radius is about 4 times that of the Moon and hence 
the area of the apparent disk about 16 times as great. So nighttime 
photography on the Moon by full earthlight would require four stops less 
exposure than nighttime photography on the Earth by full moonlight.

Or to put it a little differently, camera settings that would produce a 
nicely exposed moonlight photo on Earth would overexpose an earthlight 
photo on the Moon by four stops (not even counting the Earth's greater 
albedo).

The clouds visible in the photograph are lit only by the light of a 
crescent moon -- which, to belabor the obvious, is a lot less than you'd 
get from a full moon. In the meantime, the nighttime lunar surface is 
lit by a nearly full Earth. The difference in lighting levels is 
probably at least six stops.

(Cloud tops are of course brighter than the surface of the Moon for a 
given lighting level, but then again the Earth has a higher albedo than 
the Moon. Together they partly cancel out.)

Even if the clouds are underexposed by, say, two or three stops at this 
exposure, the nighttime lunar surface would be overexposed by three or 
four stops, producing the bright white disk we see.

The sunlit crescent is of course even brighter, but past a point 
photography (especially with CMOS and CCD imagers) registers everything 
as undifferentiated white.


-- 
D Gary Grady
Durham NC USA
dgary at mindspring.com


Jeff Polston wrote:
> The exact date according to my software is December 4, 2005.
> 
> The moon phase was about 16%.  Now, would earthshine be bright enough to look like a full disk, if exposed long enough?  I'm thinking the bright crescent "bleeds" enough to make it look like a disk, especially given the wide angle.  In other words, I don't think earthshine had a part.  I've taken shots of the moon that were over exposed enough for me to lose the phase and have it appear more full (granted, never at this small a crescent).
> 
> Jeff
> 
> 
> -----Original Message-----
> From: chaos-l-bounces at rtpnet.org [mailto:chaos-l-bounces at rtpnet.org] On Behalf Of D Gary Grady
> Sent: Wednesday, April 23, 2008 12:53 AM
> To: Chapel Hill Astronomical Observation Society
> Subject: [Chaos-l] What's wrong with this picture?
> 
> NASA's latest Astronomy Picture of the Day is a bit surprising:
> 
> http://apod.nasa.gov/apod/ap080423.html
> 
> It appears to be a full moon setting next to Venus, which if you think
> about it isn't possible.
> 
> The caption says the photo was shot in "early December of 2005," and per
> http://aa.usno.navy.mil/data/docs/MoonPhase.php the new moon was
> December 1, 2005, so one would expect the phase to be waxing, somewhere
> in the neighborhood of the first quarter. Hence there's nothing at all
> implausible about Venus being near the moon (and if I were feeling
> better I might try figuring out the exact date using planetarium software).
> 
> So the real mystery is why the moon in the photo looks like a round
> disk. I think the explanation is that in exposing enough to pick up the
> Milky Way, we're seeing the dark side of the moon illuminated by enough
> reflected earthshine to make it appear white. The part of the moon
> illuminated by sunlight is of course far brighter, but in photography,
> past a point white is white and adding more light doesn't make the image
> any whiter.
> 
> 
> --
> D Gary Grady
> Durham NC USA
> dgary at mindspring.com