Position of Moon on Cumulus

[Asidonense]

OK , So Steve and everyone else can ignore the suggestion.

I suspect, but I do not know, that libration and topocentric effects are a not relevant to the original post, but that north/south, left/right, and up/down are the problem.
I think that the poster was puzzled by the disparity between what he saw on the screen in Cumulus as a representation of the Moon and what he saw in the sky that night, and that he wanted to know what to do to get things right. For example, seen from New Zealand the Crater of Tycho is near to the top of the Moon while seen from UK it is near to the bottom - unless you are looking through an astronomical telescope. I think that the phases of the moon are portrayed in Cumulus as they would be seen by a hypothetical observer floating in space some thousands of kms above/ north of/ perpendicular to the orbit of the earth. Although very few of us have seen the moon from such a position we know what it would look like.
What a person considers to be the "top" of the moon will depend to some extent on what he/she knows about its orbit around the earth, and the night sky in general.
I was trying to suggest politely, to those who had ears to hear, that reversing a sequence of photographs of the moon in its different phases taken in the northern hemisphere would not automatically produce a sequence of equivalent photographs taken in the southern hemisphere. At least not to a person who knows where to look for the Sea of Tranquility, the Crater of Copernicus, or whatever.. E.g. if I see that the right side of Tranquility is illuminated tonight should I see more or less of it tomorrow?

[beteljuice]

.. but surely it's been mentioned before elsewhere, if you want the near enough true observed image use eg. http://www.fourmilab.ch/cgi-bin/Earth/a ... nTopo.evif

There are others ...

[steve]

I didn't write the component which draws the moon phase. It does just simply the reverse the position of the illuminated portion according to North or South hemisphere. The source code is available, but it's not something I want to go poking around in. Cumulus is primarily a weather program, not an astronomy program, so while I accept that the image isn't perfectly accurate, I would rather concentrate on the hundreds of weather-related enhancements which have been requested.