The earth isn't at 6h 00m 00s during the winter solstice?

[dave1235]

I’ve been trying to "observe" the winter solstice from the viewpoint of the sun but even when I use the script below to position myself perfectly something still seems to be wrong.

{
# Strictly speaking, this script assumes the Inclination of the sun has been set to zero in the following file. (File C:\Program Files (x86)\Celestia\data\nearstars.stc) However if you don't change it, the difference is small. (You get 89.53 degrees instead of 89.51 degrees.)

select { object "Sol" }
gotolonglat { time 0.1
distance 1.1
up [0 1 0]
longitude 0
latitude 90 }
wait { duration 0.1 }

# I got the below date from the below URL and it agrees with other sources.
# http://www.timeanddate.com/calendar/seasons.html
time { utc "2010-12-21T23:38:00" }
timerate { rate 0.001 }

select { object "Sol/Earth" }
center { time 0.1 }
wait { duration 0.1 }

set { name "FOV" value 1.0 }

}
(Sometimes I have have to run the script twice)

After running the script the earth is at the longitudes
89.51 degrees
and
5h 59m 20s
(Its at 89.53 degrees if you don't change the inclination of the sun's rotation.)

You have to increase the time to Dec 22 03:15 before it lines up with
90 degrees
and
6h 00m

If you look directly down on the sun with the ecliptic grid on, it seems perfectly centered. (If the inclination of the sun is set to zero)

Does anyone know what I am doing wrong?
I got similar results for the other equinoxes and the other solstice.
I’m thinking that maybe the definition of the solstice uses the position of the Solar System Barycenter and not the sun.(Celestia\data\nearstars.stc) Could this be it?

Thanks in advance.

Dave

[jogad]

Hi,

I’ve been trying to "observe" the winter solstice from the viewpoint of the sun but even when I use the script below to position myself perfectly something still seems to be wrong.
...
After running the script the earth is at the longitudes
89.51 degrees

Ok you find 89° 51' (and not 89.51 degrees) instead of 90° and the difference with the expected value is 9' (0.15°)
One may think the difference is very tiny and celestia is not precise enough.
But it is not the case and I take this opportunity to demonstrate the accuracy of Celestia.

This is not a scripting problem but a problem of coordinates.
Celestia (and not only Celestia) works with a system of coordinates based on January 1st 2000 mean equinox.
So the ecliptic longitude is shifted by 50.3" each year because of the axial precession of the Earth.
Other little variations around the mean value may also occur.

From the beginning of 2000 to the end of 2010 the elapsed time is 11 years and the shift in longitude is:
50.3 * 11 = 553.3" = 553.3 / 3600 = 0.1537 degrees
Thus the expected longitude is roughly
90 - 0.1537 = 89.8463 degrees = 89° 50' 47"

solstice.png

In Celestia I find 89° 50' 50" and this time the difference is only a 3" (0.0008°).
Actually Celestia computes the position much more precisely than my appoximative and simplified calculation.

I did not use any script and I would never have dared to change the tilt of the Sun. It's much too dangerous.
Instead, the point of view is only one meter from the center of the Sun (with the goto item of the navigation menu).

[dave1235]

OK I got it, thanks a one thousand million.

I realize that they froze the celestial grid in time for measurement reasons but there is something unholy about it not pointing in the direction of the earths spin.
I’m going to go over to the Development forum and make a pitch for a second “real time” celestial grid option so that other trusting souls like myself are not lead astray.
In hind sight, to position myself, I could have also traveled back and forth between the earth and sun a couple of times.

Thanks again Jogad.
Dave.

P.S. For the record, I never doubted Celestia.
P.S.S Except for that misplaced M90 Galaxy that was suppose to be in the Virgo cluster, which Fridger fixed.

[selden]

You can provide your own grid which is tied to the Earth's actual orientation. A spherical model that's gigantic and which uses the same SSC parameters as the Earth would be one way to do it. An ancient example is available at http://www.lepp.cornell.edu/~seb/celest ... les.html#4