The march equinox is defined as the 0 decree geocentric sun longitude, but I cannot seem to see the ecact time in Stellarium. If I uncheck topocentric coordinates in configuration tools (to see the apparent geocentric sun longitude coordinates) the time in Stellarium do not match the time of the official march equinox. For the 2017 equinox Cartes du Ciel, and others, gives the date and time 2017-20-03 10h28m40s UT, for the apparent geocentric sun longitude 00 00' 00'', but for that same geocentric date and time Stellarium gives the sun longitude 00 00' 37.2'' (on date). Why is that?
You seem to have CSS turned off.
Please don't fill out this field.
Yes, I have noticed this. My previous reference solution agrees well with Stellarium's time, but we have a difference to other published values for season starts (esp. Horizons) by about 15 minutes. Who solves this conundrum will be honorably mentioned in the manual. ;-)
I have one idea...
There was a previous thread on this awhile ago
I have the idea that the ligth speed is computed well, as the solar eclipses are depicted properly and at the correct time (so for both distances Moon and Sun it looks ok). The results fo VSOP and DE431 also seem to be the same. So it must be something about knowing/having the correct definition of equinox. If we know that we can check it and compare with other sources that use the same/correct definition. Sorry, I don't kow what the correct definition is...
I wish you all a good 2017,
No, I rather think one part of the solution does indeed have to do with light time correction.
We speak about a positional error for the sun of about 1/2 arcminute. As negligible as this has been in previous versions (when computation accuracy was lower), going for higher accuracy we should go on to improve this. Vernal Equinox is defined as the apparent geocentric position of the sun passing 0 degrees ecliptical longitude, so this includes light time, nutation and other corrections.
We have a bug report about a small error with the 2017 eclipse, and indeed the sun is apparently not impressed when light time correction is enabled.
A known part of "apparent" coordinates correction is aberration which is so far not applied. Aberration shifts solar coordinates by about 20 arcseconds, so the major source of discrepancy may indeed be aberration. It should be possible to add aberration correction to the coordinate output of the selected object. (it may be hard/slow to compute to shift all displayed objects by an insignificant amount of usually less than a pixel. However, plotting a highly zoomed star group with arcsecond grid but without aberration and giving coordinates of a selected star with aberration may also look stupid.). The proper technical solution would require another vector-bending matrix operation (like refraction).
Hehe, I just read Solar aberration correction and light time correction are practically the same...
But that's bad, it will leave a few arcseconds unexplained. :-(
Given that aberration shifts everything in a region, it does not affect eclipses. But it does affect the seasons definition. So, let's see how we can model this. Maybe start with aberration for the solar coordinate output only?
BTW, according to the Explanatory Supplement, Helio-/barycentric coordinates should differ by milli-arcseconds only, so this should be truly insignificant.
I understand now it was foolish of me to think 5:46 AM UTC that the difference in apparent ecliptic longitude could have something to do with the difference between computing sun barycenter or heliocenter. Thank you for the information, and thanks to all the Stellarium team for a very, very fine program. But it would be really nice to be able to see the true ecliplic longitude, so maybe "start with aberration for the solar coordinate output only".
The problem in Stellarium seems to be the apparent geocentric ecliptic coordinates. For example the "official" apparent geocentric ecliptic coordinates for the date and time 2017-01-01 0h0m0s UTC was about 280 45' 19'', but the same time in Stellarium 280 45' 53.1'' (with DE 430). Maybe Stellarium is computing the apparent geocenric ecliptic coordinates from the sun barycenter, and not the heliocenter?
By the way this last post and the first was from Preben, Denmark.
Happy New Year!
Sign up for the SourceForge newsletter: