ecliptic line is for J2000.0, not for current date

[gzotti]

Someone mentioned this in the forum, I only confirm it and express a wish.
The currently shown ecliptic appears to be the VSOP87 or J2000.0 ecliptic only, not the eclipticof current date.
For ancient dates, the sun thus is off the ecliptic!

It would be desirable to have not only equatorial grids "current" and "J2000.0", but the same for the ecliptic.

[gzotti (georg-zotti)]

(After discussion)

"Ecliptic plane" is always only Earth's orbital plane. Currently, Stellarium shows the projection of the VSOP xy plane, which is more or less the same as the ecliptic line for the present time. For times far from now, however, there is a slow shift, which causes most visibly the sun to stand besides the displayed ecliptic line. Therefore it would be desirable to call the currently visible line "Ecliptic 2000.0", and add another line to show the ecliptic "of current date".

Maybe whole ecliptic coordinate grids can be added, not just the ecliptic line?

For observer locations on other planets, an optional feature might also show the orbital plane of the planet where you are located, in addition to the ecliptic lines. However, extraterrestrial astronomy is few observers' concern.

[Fabien Chéreau (xalioth)]

Actually I have checked the code and it seems that the ecliptic lines does follow the earth orbital plane, so we actually display the ecliptic line "of date". I also couldn't reproduce the fact that the sun is displayed beside the ecliptic line.

What could be wrong is that VSOP plane could itself be moving in time, which is not taken into account now. Is this what you meant?

Fabien

[gzotti (georg-zotti)]

On Do, 25.03.2010, 14:51, Fabien Chéreau wrote:
> Actually I have checked the code and it seems that the ecliptic lines
> does follow the earth orbital plane, so we actually display the ecliptic
> line "of date". I also couldn't reproduce the fact that the sun is
> displayed beside the ecliptic line.

This is strange. Have you tried dates before about AD50, or (more
impressive) e.g. -2000?

> What could be wrong is that VSOP plane could itself be moving in time,
> which is not taken into account now. Is this what you meant?

I admit I have not thought through the VSOP for a few years now, and
cannot give you a reason for this strange behaviour at this moment. I
guess I should look through the interplay of accurate code and the simple
model from ssystem.ini that also causes e.g. the moon rotation problem,
given some spare time...

Yours, Georg.

--
DI Dr Georg Zotti
Archaeoastronomy / ASTROSIM
VIAS-Vienna Institute for Archaeological Science
University of Vienna
Franz-Klein-Gasse 1/V
A-1190 Wien
[phone]

[gzotti (georg-zotti)]

I just tried the following: The ecliptic obliquity was larger (eps>24deg) in prehistoric time, say around JD=0. Sun should reach a declination equal to the obliquity on summer solstice, and -eps at winter solstice. Currently, declination will stay within 23.5 deg, which is certainly not correct. I am now afraid there is a fundamental flaw in coordinate systems, which only manifests itself in times far from today, so I am aware that >99.7% of users will not notice. I hope it's just a small thing to correct, once we know where...

Display of solar ecliptical coordinates is available in the refraction branch, use ecl. longitudes 90 and 270 to find maxima (or use "RA of date" with 6 or 18 hours.) I also added computing and display of ecl.obliquity here (select Sun or a planet).
The disturbing effects are the sun displayed off the ecliptic line in times before about year0, and wrong rising/setting azimuths for every user wanting to use stellarium for long-time (say, within 10000 years) changes of earth's axis, e.g. for building alignment studies or archaeoastronomy in general.

[Amrane TOLBA (amrane-tolba)]

Dear all

I think that the Stellarium ecliptic line is NOT THE ONE for J2000 at all!!! because:

In a past (or future) year, the ecliptic line SHOWN AS THE ONE FOR J2000 should not have "remarquble" changes during THAT YEAR.
In the case of Stellarium the displacement is too large(respectively) in ONLY THREE MONTHS OF THAT YEAR!!
Example: The attached pictures show the sun positions for June and September for the year 20 AD!
If you want to see more about this strange case:
Fix (lock) the sun at screen center, zoom, set year to an old one say 123 AD and change the month from 6 to 12. You'll see the ecliptic vibrating between two large positions!! ...
For me in addition to the famous softwares: Starry Night, Red Shift,....I see clearly the accurate ecliptic lines for J2000 and for the actual date displayed together , in an old, free, small, simple, but nice and accurate software:
SkyGlobe for Windows (versin 4) from Klasm software (free software):
http://www.sidewalkastronomy.com/skyglobe.html
Thinks..

[Alexander Wolf (alexwolf)]

Amrane, are you understand that the inclination ecliptic for current dates is 23d26'15,6" and for 20 A.D. is 23d41'32,5"? If you need check ecliptic then you need check it for 2000 year because this line for J2000 epoch - not 20 year!

[Amrane TOLBA (amrane-tolba)]

Yes, Alexander I know the equation calculating that angles and used it years ago, they call that in astronomy "The inclination of the earth's axis with respect to the 'normal' of the ECLIPTIC TAKEN AS REFERENCE".
But, I think there is a misunderstanding:
Anyway, my essential question then is: for 20 AD? In STELLARIUM WHY WE SEE MANY ANGLES FOR THIS YEAR BY ONLY CHANGING THE MONTH?
Have you seen the attached pictures in my last comment in this bug and in bug #1013865?
Thanks.

[Martin Lewicki (martin-lewicki)]

Version 0.12.4 still no fix

It should be locked to the precessing equinox not stuck at J2000.

 In addition the longitudes should be in degrees not hours.

[Alexander Wolf (alexwolf)]

Georg, I had an interesting observation - current implementation of ecliptic is fun and the Sun has stepped movement around ecliptic on some date in past (around -4000) and in future (around 8000). For very big dates in past and in future the Sun has correct position on the ecliptic.

[gzotti (georg-zotti)]

Oh, yes, that's interesting!

You know VSOP is valid approx J2000+/-6000 years. Whatever happens
outside? Most likely this is one effect! Fine that it is "benign", and the
sun does not run away.

One of the things possible to extend validity of our time range is using
DE430/431. One of the many plans and ideas...

G.

[gzotti (georg-zotti)]

OK, see r7604 in branch
https://code.launchpad.net/~georg-zotti/stellarium/gz_fix-ecliptic-obliquity

I think I now have all angles applied correctly, at least for earth-based observers, but it still needs comparison with some reference. Precession angles are from Vondrak 2011 (200.000 years model!).
I can confirm Alex's observation now in pictures: Sun runs nicely on the ecliptic of date downto -4000, then migrates to ecliptic of J2000 by -4100. This must be an implementation artifact of VSOP87.

What do we learn from this? At least our implementation of VSOP87 is indeed applicable only for -4000...+8000. Beyond that, some numerical dampening forces earth's orbit back to the ecliptic plane of J2000 within a 100 years buffer. I.e., outside of -4000...+8000 planetary coordinates are guesstimates at best. On the positive side, as analytical solution at least we get "something" even millennia beyond its range of validity.

We should IMHO not push earth off the VSOP87 plane to put the Sun back onto ecl. of date. Just take note that the real sun likely runs along the displayed ecliptic of date.

For better accuracy in Meso-/Neolithic times, DE431 may hopefully be available.

[Alexander Wolf (alexwolf)]

A fix has been committed as revision 7748 of the trunk branch of Stellarium's Bazaar repository at Launchpad: http://bazaar.launchpad.net/~stellarium/stellarium/trunk/revision/7748