Telescope Reviews: Searching for thin crescents

Martin,

It seems extremely doubtful to me that the edge of the terminator would be located differently at 1 micron vs. in the visible (0.5 micron); and, if anything, I would think that at a trailing edge like this, the lingering surface warmth might extend it very slightly into the area that has just been plunged into darkness.

The problem, I think, is with the Guide 8 software which is simply showing the terminator in the wrong position relative to the lunar surface features. I'm not at all familiar with Guide 8, but the image from it that you show puts the terminator at 10:03 UT on Feb. 3, 2008 running through Aristarchus, in the north, and roughly tangent to the east rim of Mersenius, in the south. This is really the theoretical terminator position some 4 to 5 hours earlier. By 10:03 UT the terminator (for a spherical Moon) would be well onto the Aristarchus Plateau, in the north, and roughly down the middle of Mersenius, as well as tangent to the east rims of Billy and Schickard in the south. That looks pretty close to where you saw it in your photo.

I'm not sure why Guide 8 would make this mistake. It's not a sophisticated calculation that requires knowledge of the observer's location or the Moon's libration: it's simply a matter of drawing a circle 90° from the Sun's position in the lunar coordinate system, which happens to have been 139.47°W/0.41°S.

If you need an accurate representation of the theoretical terminator position relative to the lunar surface features. I might suggest the freeware Lunar Terminator Visualization Tool, or the LIGHT version of VMA 3.5 or later (this popular program also comes in "Basic", "Expert" and "Pro" versions but strangely, the higher versions represent the terminator by a kind of huge sawtooth graphic that moves in large steps. It is not only less clear, but also far less accurate than the simple line used in the LIGHT version. Also, all versions of VMA prior to 3.5 use a flawed ephemeris calculation, although that is probably primarily of interest if you are making predictions for the distant past or future.) There are probably many other programs that show the lunar terminator accurately. Guide 8 just doesn't seem to be one of them.

Of course, all these programs only tell you where the terminator would lie if the Moon were a perfect sphere. Until the detailed topographic information from the new crop of lunar space probes becomes available, the only way to tell what the terminator will actually look like is to look at a photo taken at a time when the theoretical terminator matches the predicted position at the time you're interested in.

--

The other matter of even more interest (at least to me) in connection with your work on thin crescents is how far the crescent extends at its tips. Since the current generation of simulation software assumes a spherical Moon, it is always going to show a perfect 180 deg crescent. The French astronomer Andrè Danjon thought the length became systematically shorter as the crescents got thinner. While this may be true to some extent, I would guess he might have based his conclusions on the inadequate reports and photos available to him in the 1930's. The sequence from Feb. 18, 2007 on your 2007 Thin Crescents page illustrates the problem nicely. When first spotted in the late afternoon sky at 15:40 local time, the crescent looks extremely short (53° or so?). When the Sun sets at 17:48, we see it is much longer than it at first appeared (at least 122°), and by 18:26 (when the sky has darkened enough to see the Earthshine), we see it is longer still (at least 162°). Obviously, the crescent is not really changing during this time, only our ability to see it.

Since extremely thin twilight crescents (observed visually) are never seen in fully dark skies, and since the thinner they are the brighter the sky they will be when they are observed, it's easy to see how Danjon could have been fooled into thinking the crescent was shortening (or at least overestimated the amount by which it shortened if it really does shorten).

The question arises: is 162° the "real" length of the 18:26 crescent, or if the photo were exposed more deeply, or if it were possible to extract greater contrast (e.g. in the infrared), would it extend still farther? Might it possibly go the full 180°??

Perhaps someone here has made observations of this type: trying to see just how far the tips of the crescent really go. I myself spent some time his morning looking through my (quite good optical quality) window at the roughly 7% illuminated Moon as it rose in the pre-dawn sky. With my 4-inch (aperture) Maksutov I found it extremely difficult to decide with certainty where the crescent ended. Especially in the north, I thought I might be seeing a vague hint of a razor thin extension of the cusp beyond what I would at first have taken to be its "end". Perhaps a camera that could look more deeply and with better contrast than my eyes (and, hopefully, with larger aperture/better resolution and better seeing) could have seen more?

-- Jim