Partial fix to lunar rotation axis

[granthutchison]

I've just noticed that the Moon's libration in latitude is much to great - when viewed from the surface of the Earth, you can see far too much of the lunar poles at some points in the Moon's orbit.
The reason is that Celestia currently leaves the Moon's equatorial plane in the default position for satellites, aligned with that of its parent planet (in this case, of course, Earth). In fact, the Moon's equator is more closely aligned with the ecliptic - its equator is tilted just a degree and a half off the ecliptic plane, and it precesses in that plane, making one rotation every 18.6 years, keeping pace with the precession of the nodes of the Moon's orbit.
We can't model that precession until Celestia provides the option of "precession poles" (hint, hint), but it's certainly possible to tighten up the rotation axis a bit - add this to your Moon definition in solarsys.ssc:

Code: Select all

   Obliquity      23.45
   EquatorAscendingNode   0


That pushes the Moon into alignment with the ecliptic, and cuts the observed libration in latitude from ~28 degrees to a much more realistic ~5 degrees - a degree and a half less than the real value, but that small angular error is much less noticeable.
It also makes the apparent movement of the Earth in the lunar sky more realistic and less extreme. (In November last year ccm01 remarked on how much the Earth was whanging around up there - http://www.shatters.net/forum/viewtopic.php?t=1238 - but I never went and looked for myself until now .)

Grant

[chris]

I've just noticed that the Moon's libration in latitude is much to great - when viewed from the surface of the Earth, you can see far too much of the lunar poles at some points in the Moon's orbit.
The reason is that Celestia currently leaves the Moon's equatorial plane in the default position for satellites, aligned with that of its parent planet (in this case, of course, Earth). In fact, the Moon's equator is more closely aligned with the ecliptic - its equator is tilted just a degree and a half off the ecliptic plane, and it precesses in that plane, making one rotation every 18.6 years, keeping pace with the precession of the nodes of the Moon's orbit.
We can't model that precession until Celestia provides the option of "precession poles" (hint, hint),

I was afraid that this was going to be a problem . . . So, to make precession general enough to work for the moon, Celestia needs to have the equivalent of obliquity and equator ascending node for the precession axis?

--Chris

[granthutchison]

So, to make precession general enough to work for the moon, Celestia needs to have the equivalent of obliquity and equator ascending node for the precession axis?

I think so - the Moon is precessing with reference to its mean orbital plane, which is different from the plane of any of its instantaneous orbits.
The same, I think, would be useful for better modelling of satellite orbits - if Celestia could precess the nodes of an orbit with reference to some defined plane (and also independently precess the pericentre) it would allow much more realistic modelling of (say) the outer satellites of Jupiter.

Grant