Celestia Forums

I have been trying to update my model of Stonehenge recently to place it in exactly the right position and also include the surrounding countryside,and to test the lunar and solar alignments have had to reconstruct the original henge.Although little is known about its original purpose,there seems little doubt that it was intended to mark the solstices and lunarstices-these I can see in Celestia perfectly for the current epoch.


The trouble is that when I rewind time to 2000BC the alignments are off by about three hours.At first I thought it was because of precession or the VSOP limitations in Earths orbit,but now I think it is the value for Earths RotationPeriod .I think I need to change it to the tropical year rate with precession turned on,but am unsure of the values,also changing the rotation value seems also to affect the rotationOffset (or at least the orientation of the models are affected when I change it).Can anyone help me with the correct values.Jestr

Maybe it's because you've set it to the wrong DATES. remember, the earth has precession, and even a 4000 year difference (at least, i THINK so...) can go a long way, so try setting the stonehenge to be aligned with the 2000 bc solstice instead of the 2005 solstice?

I could be wrong, though.

Lack of Precession can explain the approx. 3 hour error. Taking Earth's precession period as 26,000 years, then the 4,000 years back to 2,000 B.C. would equate to 4000 / 26000 * 24 hours = 3.7 hours.

I was surprised when you said maybe you should switch precession on, because I thought that would already be in the earth SSC. I just checked minbe and it's not after all.

On rotation period, there is the matter that the Earth's rotation has been slowing down, but I think that's a red herring here. It's true that ancient Chinese eclipse observations seem to reveal that Earth local times were several hours behind from where they should have been given a constant rotation rate, but the fact that your Celestia uses such a constant rate means you shouldn't notice. No-one from Stonehenge, 2,000 B.C., is here to tell you that the solstice sunrises he saw then actually happened a few hours later than Celestia says. They didn't have atomic time to tell them TAI and ET or UTC were so different then.

Spiff.

Spaceman Spiff wrote:Lack of Precession can explain the approx. 3 hour error. Taking Earth's precession period as 26,000 years, then the 4,000 years back to 2,000 B.C. would equate to 4000 / 26000 * 24 hours = 3.7 hours.

I was surprised when you said maybe you should switch precession on, because I thought that would already be in the earth SSC. I just checked minbe and it's not after all.

I think there's a reason about orbits being referenced to the equatorial plane of the parent body and precession messing this up, which is why precession hasn't been implemented for Earth yet.

So,anyone know what the rate is in Celestia format?If this is switched on though doesn't this also imply using the tropical rotation period instead of sidereal?Are there any other orbital parameters which would be affected?Jestr

Try:



If you trust Wikipedia (http://en.wikipedia.org/wiki/Year) on the matter, then (for Epoch 2000.0):
- Tropical year = 365.242 189 67 days.
- Sidereal year = 365.256 363 051 days.

The precession period is found by:



which gives 25,770.59 years, and the number to put in Celestia is converted to degrees per day:



gives 3.824617e-005.

chaos syndrome wrote:I think there's a reason about orbits being referenced to the equatorial plane of the parent body and precession messing this up, which is why precession hasn't been implemented for Earth yet.

At first, I wondered if you are thinking of the reference of orbits and general co-ordinates in real astronomy, where the R.A. and Dec are referred off the first point of Aries, which drifts westwards along the ecliptic every 26,000 years...? This won't apply in Celestia as it's X, Y, Z reference frame is perpetually fixed to epoch 2000.0.

Now I think you mean that satellite orbits would precess with the parent body. In case people think this is a reason to defer precession, I can tell you this isn't a problem!
1. the Earth's moon's orbit is part of VSOP-87, so isn't affected by this,
2. precession of satellite orbits with the parent planet is what happens in real life anyway! Mars, Jupiter, and Saturn all precess, yet their regular satellite orbits and rings are obviously tracking the equatorial planes!

I inserted the above precession rate into my Earth SSC and went back to 21 Mar 2001 B.C (-2000 03 21), and I see no disruption to the orbits (that I can tell, because the celestial grid is aligned to the Earth at Epoch 2000.0 and won't precess, because Celestia happens to be designed to assume it's fixed to the general X, Y, Z reference frame...). Instead I can see:
1. the Earth's axis is 'twisted' w.r.t. this grid,
2. the Sun is at about 2.2 hours R.A. and +15?° Dec. (Epoch 2000.0) so the First Point of Aries does indeed seem to have drifted along the ecliptic well into Aries,
both of which are much as I'd expected.

I'd say it 'works' and is worth a try.

The reason Celestia doesn't do the Stonehenge part properly without precession is, I suspect, because the Earth's orbital period is defined as one tropical year, and that is based on precession. You can't redefine the orbital period to sidereal unless you abandon VSOP-87, and if you try and fudge the Earth's rotation period to be 1/26,000 shorter, then solar eclipses will start to happen at the wrong places on Earth at times far from 2,000 A.D. The correct way should be to set the precession rate as above.

Spiff.

Thanks for the help Spiff.So you set the PrecessionRate to -0.00003824617,and leave the RotationPeriod at 23.9344694.This works for Stonehenge,but the Heel stone now is directly aligned with the Sun as it rises at summer solstice,though I had assumed it was one of a pair of stones which framed the sunrise,in 2000BC however the Avenue (a 3km earthwork leading from Stonehenge in the direction of the sunrise)no longer aligns with the view from the centre over the Heel stone.
To check the lunar alignments I would have to add PrecessionRate (nutation?) to the moon declaration but in this thread
http://www.shatters.net/forum/viewtopic.php?t=6670&highlight=precession
Chris states there is a problem with the moons orbit with precession added.Cheers Jestr

Sorry Spiff,one more question,when calculating a PrecessionRate for the moon should it be with reference to the Earth or to the Sun-is this the Draconitic Year on the same Wiki page above?Jestr

Hullo jestr,

jestr wrote:So you set the PrecessionRate to -0.00003824617,and leave the RotationPeriod at 23.9344694.

Yes, at least I find Celestia understands exponent notation like 3. 824617e-005 in SSCs just as well as -0.00003824617 (in case people didn't realise that).

Yes, leave the RotationPeriod alone for reasons I described above (you can't model the drift of Earth's rotation period through time since Celestia allows only one (constant) value for it...).

jestr wrote:This works for Stonehenge,but the Heel stone now is directly aligned with the Sun as it rises at summer solstice,though I had assumed it was one of a pair of stones which framed the sunrise,in 2000BC ...[snip]

OK, good so far! Actually, I was a little surprised when I saw there are two Heel Stones in your model, but I thought maybe you knew more than me. Check this thouroughly, because I thought there was only ever one Heel Stone, and the sun is supposed to rise on the horizon exactly where the left side of it meets the ground.

jestr wrote:[snip]... however the Avenue (a 3km earthwork leading from Stonehenge in the direction of the sunrise)no longer aligns with the view from the centre over the Heel stone.

I don't follow your problem here, isn't that just saying your model isn't straight with itself? Also, wasn't the Avenue just the road approaching Stonehenge from the Avon? I think it's not aligned with anything particular, it curves heavily further away.

jestr wrote:To check the lunar alignments I would have to add PrecessionRate (nutation?) to the moon declaration but in this thread [url="http://www.celestiaproject.net/forum/viewtopic.php?t=6670&highlight=precession"]http://www.celestiaproject.net/forum/viewtopic.php?t=6670&highlight=precession[/url]
Chris states there is a problem with the moons orbit with precession added.Cheers Jestr

jestr wrote:[snip]...one more question,when calculating a PrecessionRate for the moon should it be with reference to the Earth or to the Sun-is this the Draconitic Year on the same Wiki page above?Jestr

Now, I feel a little uncertain here, but doing my best, my understanding is this: you don't need to precess the moon's orbit, VSOP-87 takes care of that.

Note that MKruer's post was about his artifical solar system where all planets and moons have orbits defined in the unevolving Keplerian way. That means his moons' orbit planes are left fixed, stranded, while his planets precess like he defined them to.

The fix of introducing another keyword: OrbitPlanePrecessionRate, which lets the moons' orbital planes precess, might already have been done! Using the same value for planet PrecessionRate and moon OrbitPlanePrecessionRate will do that. IIRC, Chris was trying to do this for Phobos' orbit, but abandoned that in favour of VSOP-87.

jestr wrote:Sorry Spiff,

Ah, no worries!

Spiff.

In 2.5.5 here they say that a stone hole was found ten feet to the left of the Heel Stone, so maybe there was a second stone there. The sun rose on its right edge and went to the top of the Heel Stone on summer solstines in around 3000 BCE. Or there was one stone that was moved.

http://www.megalithicsites.co.uk/Calendar5.html

I dont think anybody can be certain about how many stones were there when,in fact there seems to have been a number of different layouts of differing numbers of stones at different times.Around the time the Sarsen stone circle was erected it seems there was a kind of three pillar entrance to the site (in line with the slaughter stone-lying down now between the circle and the heel stone).If this is the layout at the time of use of the circle these would block the view of the heel stone altogether.Any how heres a picture of the midsummer sunrise,over the heel stone in -1832 BC (atmosphere off).





A little later,(atmosphere on)




Heres the view a little further back outside the circle,




And a couple of the midwinter sunset




Heres the view from the Avenue




Theres still a lot of work to do on the model,and I cant get the lunarstices to line up right in -2000 BC,because of the problem of adding precession (nutation?) to the moon,but they match OK in the present time frame.
Cheers Jestr