Celestia Forums

The standard asteroids list for Celestia only has the first four asteroids discovered (Ceres, Pallas, Juno, Vesta), plus an additional 12 of special interest.

The first four asteroids discovered are not the four largest asteroids. Ceres, Pallas and Vesta are three of the four largest asteroids, but Juno is a substantially smaller body that only ranks ninth in size (about seventh in mass) of the asteroids in the main asteroid belt.

The ten largest asteroids in size order are (source: Wikipedia):
1 Ceres
4 Vesta
2 Pallas
10 Hygiea
511 Davida
704 Interamnia
52 Europa
624 Hektor
3 Juno
87 Sylvia

Juno's ranking ninth in size would surprise many people. This is why I believe the creation of an addon that includes the other asteroids that are larger than Juno would be a valuable addon. It could even be incorporated into Celestia's main asteroid list.

I have tried creating orbits for a couple of the asteroids (I think I used the database at http://ssd.jpl.nasa.gov/sbdb.cgi) but I could not be sure I had the orbits right. When I created Ceres as a test, my version of Ceres was offset from the real Ceres by 30 degrees.

I found my error in the ssc file for Ceres. I used LongOfPericenter and it should have been ArgOfPericenter. When I made that correction the two copies of Ceres were much closer together.

I also used the database at http://www.alpheratz.net/murison/calcul ... ements.php

Here are two asteroids I have done so far. Are there any improvements I can make before including the orbital elements for the remaining bodies?

bdm,

Rather than using "fan" sites, I'd suggest that you get current ephemeris data from the IAU's Minor Planet Center's database at the Harvard-Smithsonian Center for Astrophysics. That way you reduce the possibility of transcription errors or out of date information.

See http://cfa-www.harvard.edu/iau/mpc.html
and http://cfa-www.harvard.edu/iau/MPEph/MPEph.html

The original site I used did not give Julian dates for the epoch so I'm not using it any longer. I've just used this JPL site for orbital elements, with good results: http://ssd.jpl.nasa.gov/sbdb.cgi

With the orbital elements subject to change due to planetary perturbations, long-term accuracy is difficult to achieve.

To test the accuracy, I've included the JPL and IAU orbital elements for the first four asteroids in one file. The JPL and IAU elements give orbits that are fairly close together, and the Celestia asteroids are relatively far away. Perhaps Chris needs to update the orbital elements for the main asteroids?

The orbital elements for the first four asteroids are given here. The IAU elements are given as-is, whereas the JPL elements were rounded to 8 decimal places.

This sounds like an exciting project! I've always thought that the asteroids received something of a less-than-fair shake, and will be glad to see some accurate orbital representations of these subjects!

...John...

I've got orbital elements for the 13 asteroids that have a radius larger than 125 km. A few notes:

• This list has nine asteroids, not six.
• I have not yet verified these for correctness. My method involved manual cut and paste, and it's possible I've made one or two errors.
• All asteroids in this list are given with their name only, except for Europa, which is given as "52 Europa". This is necessary because this asteroid shares a name with a well-known moon of Jupiter.
• I don't know how to assign colours to the asteroids. This is why all of them are grey. The source I have has U-B and B-V color indices but I don't know how to convert these to RGB, nor whether such conversion is possible with any degree of accuracy.
• The source for the orbital data was the "JPL Small-Body Database Browser", http://ssd.jpl.nasa.gov/sbdb.cgi
• Asteroids are given with radii, but most of the asteroids are not spherical. The actual shapes of the asteroids vary, but I lack the expertise to create suitable models.
• All orbital elements are subject to significant change over time. This will be most noticeable with 624 Hektor, which is a Trojan asteroid.

If anyone is up to the challenge of preparing asteroid models, here is a scientific paper that models the shapes of some of these asteroids. This paper also provides axial tilts for some of these asteroids.

http://www.rni.helsinki.fi/~mjk/IcarPIII.pdf

UPDATE: I have checked the orbit figures given above. 511 Davida had the figure for the semimajor axis copied to the ascending node; the other figures were correct. I have corrected the figures.

All we need are textures and colours and it's good to go. Models would be good as well but they are not essential.

Can anyone offer suggestions on what would be suitable textures to use for these asteroids? I also don't know what colour they should be other than most of them being very dark.

I have edited the list to include textures for the asteroids. Without knowing their true textures, I've chosen to give them the usual "asteroid.*" texture.

Most of these asteroids require models but I cannot make models for myself. So these will render as spheres, even though that is not really accurate. Juno is treated this way in the Celestia asteroids.ssc file so I think that should be OK.

I don't know what colour to set the asteroids, so I have left them with 50% grey. I notice that many of the asteroids in the Celestia asteroids file also skimps on colour information so I guess that's OK.

Would this be good enough for release?

Bdm, for your asteroids model you can always use the default asteroid.cms; open it with a text editor and play with the values. You can do irregular spheres or whatever shape.

Concerning the whole addon, you should ask someone well skilled on those topics for a revision of your work before releasing (Grant Hutchinson?); this just to be sure all is 100% ok...

Anyway, Tx for the effort.

In early 2002 I have extracted already 7309 asteroids from the
official published scientific resources and produced a corresponding .ssc
file.

Here is a funny image from the time, where I artificially enhanced the
albedo of all these asteroids such that all of them could be seen at the
same time!



Bye Fridger

t00fri wrote:In early 2002 I have extracted already 7309 asteroids from the
official published scientific resources and produced a corresponding .ssc
file...

I suppose this mean the asteroids Bdm has extracted are already part of those 7309 bodies?

ElChristou wrote:

t00fri wrote:In early 2002 I have extracted already 7309 asteroids from the
official published scientific resources and produced a corresponding .ssc
file...

I suppose this mean the asteroids Bdm has extracted are already part of those 7309 bodies?

Certainly. The Universe is full of Asteroids. 13 are a rather small fraction...

Seriously: mass rendering of asteroids, say 250000 (!), are high on my todo list. But this is among the hardest mass-rendering tasks. The reason is that a proper octree structure (required for superfast rendering) is difficult to realize, since the asteroids are moving (unlike galaxies, for example!) . Yet I have some ideas to explore. We had that topic before, so I'll shut up now...

http://www.celestiaproject.net/forum/viewtopic ... sc&start=0

Bye Fridger

ElChristou wrote:Bdm, for your asteroids model you can always use the default asteroid.cms; open it with a text editor and play with the values. You can do irregular spheres or whatever shape.

Concerning the whole addon, you should ask someone well skilled on those topics for a revision of your work before releasing (Grant Hutchinson?); this just to be sure all is 100% ok...

Anyway, Tx for the effort.

I didn't know that cms files were text files, thanks for the tip. Now all I need is to figure out how they work.

I have bookmarked on my other computer a scientific paper that gives approximate shapes for the majority of these asteroids based on light curve studies. Who knows, I may yet have reasonably accurate shapes for these asteroids.

Toofri's big list of asteroids would be useful for people who require many asteroids, but that's not what I need. My needs are a complete list of the largest asteroids which Celestia currently doesn't provide. In particular, I needed 10 Hygiea to be included because it's the largest asteroid that's not included in Celestia's standard list of asteroids. I chose to include other asteroids as well to show 3 Juno in its proper place as the ninth largest asteroid.

I have a preliminary version of the addon available for download and evaluation at the following location: http://www.bdm.id.au/download/large%20asteroids%201-0.zip

Size of addon: 5 kb
Includes CMS models for 10 Hygiea, 15 Eunomia, 16 Psyche, 52 Europa, 87 Sylvia. (being CMS models, these models are very approximate)

I have checked the positions of 10 Hygiea and 15 Eunomia against asteroid ephemerides at http://cfa-www.harvard.edu/iau/MPEph/MPEph.html#summary

The SSC descriptions include rotation periods but not axial tilts and nodes.

bdm wrote:Toofri's big list of asteroids would be useful for people who require many asteroids, but that's not what I need. My needs are a complete list of the largest asteroids which Celestia currently doesn't provide.

My big list of course included also ALL the largest asteroids and the PERL script can easily sort and cut that list in size or magnitude according to individual needs! The list represented a complete, official asteroid catalogue. In addition an upgrade of respective ephemerides is only a CLICK in my case! Finally, no numbers were touched manually which greatly helps to avoid typos...

In general it is always a good idea to start Celestia projects from the /general/ side and specialize later. More people can profit this way and the solution is usually more versatile & "elegant"

If I'll find some spare time, I'll get back on this interesting culling task with /many/ moving asteroids. An appropriate octree structure in this case is kind of a math & coding challenge.

Bye Fridger

Toofri, how detailed is your large asteroid addon? My addon includes rotation periods for all nine asteroids, and rough CMS models for five of them. At present I don't have Color, EquatorAscendingNode or Obliquity, nor do I have different textures.

The approach I have taken for my addon was to provide a complete inventory of the large asteroids, and provide some detail where I could find it. I wanted to have a small addon that could provide some extra asteroids that I believe should be in Celestia by default. I'm hoping that the addon could be developed to the point where it could be incorporated into the default installation of Celestia. If not, it will provide a choice of addons for those that may need it.

Oh, and Toofri, could you kindly post a download link for your addon if you haven't already?

bdm wrote:Toofri, how detailed is your large asteroid addon? My addon includes rotation periods for all nine asteroids, and rough CMS models for five of them. At present I don't have Color, EquatorAscendingNode or Obliquity, nor do I have different textures.

But why did you leave out EquatorAscendingNode and Obliquity? To reconstruct the complete orbits /precisely/ you need ALL orbital parameters and MUST transform them from the skyplane frame to the Celestia frame as usual, using quaternions. I suppose you did not ignore these required transformations?

The orbital parameters in the *.ssc files refer to the Celestia frame, while the ephemerides in the catalogs refer to the skyplane frame!

The approach I have taken for my addon was to provide a complete inventory of the large asteroids, and provide some detail where I could find it. I wanted to have a small addon that could provide some extra asteroids that I believe should be in Celestia by default. I'm hoping that the addon could be developed to the point where it could be incorporated into the default installation of Celestia. If not, it will provide a choice of addons for those that may need it.

The lack of asteroidal data in the Celestia distribution is just due to lack of (qualified) developer time. Like we have recently incorporated the /complete/ revised NGC/IC catalog of galaxies, Celestia's data base will certainly be expanded to also include entire published catalogs with planetary nebulae, galactic clusters...and asteroids.

We certainly would not deal with just a few at this advanced point of code development. As I said, if you have any clever ideas about how an octree might be constructed for asteroids that would be extremely helpful!

Oh, and Toofri, could you kindly post a download link for your addon if you haven't already?

[ note my username is t00fri NOT Toofri , but never mind]

I did that work on asteroids ~4 years ago and the actual .ssc file + PERL script must be dumped somewhere on some media. All I can readily find now was my asteroid image from that time.

Meanwhile we have changed the respective code structure completely (octree culling!) such that these lists have to be regenerated anyhow (which is very easy with PERL).

Also I am usually not involved with making add-ons. As a Celestia developer, my work will be rather integrated into the official distribution when everyone is content with the state of the art.

Bye Fridger

t00fri wrote:But why did you leave out EquatorAscendingNode and Obliquity?

I think you have confused EquatorAscendingNode and Obliquity with Ascending Node and Inclination here. EAN and Obliquity defined the rotation axis of the body, and Ascending Node and Inclination are orbital parameters.

I don't ever intend to release any addon, no matter how incomplete, without a full set of orbital parameters.

On the other hand, I have omitted the rotation parameters because I have not yet found good data for these.

t00fri wrote:[ note my username is t00fri NOT Toofri Wink , but never mind]

Noted ... if you're particular about using zeroes instead of oes then I will take care to include them.

t00fri wrote:To reconstruct the complete orbits /precisely/ you need ALL orbital parameters and MUST transform them from the skyplane frame to the Celestia frame as usual, using quaternions. I suppose you did not ignore these required transformations?

The orbital parameters in the *.ssc files refer to the Celestia frame, while the ephemerides in the catalogs refer to the skyplane frame!

What is the difference between the skyplane frame and the Celestia frame? I thought Celestia used orbital parameters in relation to the ecliptic.

I have checked the methodology by rendering Ceres, Pallas, Juno and Vesta in the same way and comparing the positions of these asteroids. The test asteroids were very close to the Celestia versions of the asteroids. I also checked two asteroids by using online ephemerides to compute the positions, and then comparing the positions in Celestia with the ephemerides. These also checked out to good accuracy. I intend to check the other seven in the next few days.