16

Where can I find the best values to use for standard gravitational parameters of solar system bodies for numerical simulation experiments?

Here are two sets I've put together. The first set is from Wikipedia, the second set is from the a JPL progress report listing the masses used for the calculation of the DE421 ephemeris.

In addition to the Sun and planets, I've included Earth's Moon, and the three largest asteroids. I am not sure how I should handle the masses of the larger moons of the large planets; I assume they are not counted here and I should deal with them separately.

Is there a significantly better "master table" than the DE430/431 values, or is that about as good as it gets? 

NOTE: units are MKS

            Wikipedia:        JPL: DE421          JPL: DE430/431
           (1) unless         (2) unless            (6) unless 
         otherwise noted     otherwise noted       otherwise noted
        -----------------  --------------------  --------------------
Sun     1.32712440018E+20  1.32712440040944E+20  1.32712440040944E+20
Mercury 2.2032E+13         2.203209E+13          2.203178E+13
Venus   3.24859E+14        3.24858592E+14        3.24858592E+14
Earth   3.986004418E+14    3.98600436233E+14     3.98600435436E+14
Moon    4.9048695E+12      4.902800076E+12       4.902800066E+12
Mars    4.282837E+13       4.2828375214E+13      4.2828375214E+13
Ceres   6.26325E+10        6.2178E+10            6.28093938E+10
Pallas  2.11E+20*G(4,5)    1.3402E+10            1.3923011E+10
Vesta   2.59076E+20*G(3,5) 1.7630E+10            1.7288009E+10
Jupiter 1.26686534E+17     1.267127648E+17       1.267127648E+17
Saturn  3.7931187E+16      3.79405852E+16        3.79405852E+16
Uranus  5.793939E+15       5.7945486E+15         5.7945486E+15
Neptune 6.836529E+15       6.836535E+15          6.83652719958E+15

G = 6.67408E-11 (5)
uhoh
  • 148,791
  • 53
  • 476
  • 1,473
  • 1
    Looking at the sources on the Wikipedia page, I see that the Ceres figure is based on Dawn data from 2015, Mars from 2011, while some of the other sources date to 2005, so the data straddles that of DE 421. – Russell Borogove Sep 08 '17 at 14:03
  • @RussellBorogove I've just found the report for DE430 and DE431 dated 2014, and I've added it to the table. I've dropped the ratios because it doesn't seem to matter here how much Wikipedia is really off. – uhoh Sep 08 '17 at 14:42
  • DE430/431 (we're actually up to DE436 now, but I don't think the values have changed) is what NASA uses, so I'd go with those. All the official US astronomical information (HORIZONS, SPICE, etc) is based on the DE ephemerides. Perhaps see also http://astronomy.stackexchange.com/questions/13488 –  Sep 08 '17 at 14:51
  • @barrycarter ya DE430/431 is the latest one for which I can find reports or documentation, and when I download planet positions from Horizons they are all calculated using DE431. – uhoh Sep 08 '17 at 15:06
  • 2
    https://naif.jpl.nasa.gov/pub/naif/JUNO/kernels/spk/de436s.bsp.lbl but, again, not really different. –  Sep 08 '17 at 15:07
  • @barrycarter Oh that's easier to copy/paste from than a PDF, thank you! :-) I'll have to track down the major moons of the outer planets separately. – uhoh Sep 08 '17 at 15:13
  • 1
    https://naif.jpl.nasa.gov/pub/naif/generic_kernels/spk/satellites/ –  Sep 08 '17 at 15:14
  • 1
    Better values for Pallas are difficult to obtain. No orbit of a moon of Pallas was observed and there was no space craft in an orbit around Pallas yet. Therefore the error of its mass is big: (2.11±0.26)×10E20 kg. I would suggest to include the error values to the mass values.
    The mass value for Vesta is much better: (2.59076±0.00001)×10E20 kg, Dawn was in orbit to Vesta for more than one year.     – Uwe Sep 10 '17 at 10:28
  • @Uwe thanks for your comments about Pallas. I did not include errors on purpose, because I do not intend or want this table to be uses as a primary source or to be quoted. I'm showing the values that I found and checking if these are as good as it gets right now (it seems that it is). I believe that the last column probably is as good as it gets right now, and would probably accept an answer that suggests so. – uhoh Sep 10 '17 at 11:28
  • I agree with @barrycarter, the latest/greatest DE's are the best information available, since they're computed using the best models and tracking data available. In addition to NASA, most commercial software packages use the DE files as the source of planetary data. – DuffBeerBaron Sep 12 '17 at 18:31
  • @DuffBeerBaron your comment is non sequitur. The question is about standard gravitational parameters, not the ephemerides. barrycarter did not say that there is a source of masses related to DE436 that has different masses, but that it probably uses the same masses as DE431. So actually you don't actually seem to be agreeing at all. If you re-read the chain of comments you'll see I answered with "DE430/431 is the latest one for which I can find reports or documentation, and when I download planet positions from Horizons they are all calculated using DE431" – uhoh Sep 13 '17 at 00:35
  • 5
    @uhoh, My apologies for not being clear in my comment, so let me offer two clarifications:
    1. Gravitational Parameters and planetary ephemerides are inherently linked, so that they can change for any iteration of DE. I'm not sure if they are estimated every time, but they can be.
    2. I didn't mean to imply that DE436 was the latest and greatest. JPL produces DE files specifically for missions at times, and the fact that the file path for the DE436 contains JUNO implies that it was produced for that mission, and not meant for general consumption. DE431 may be the official latest/greatest
     – DuffBeerBaron Sep 14 '17 at 20:39
  • 3
    I'll note that my understanding of the generation of the DE kernels is the same as that of @DuffBeerBaron. They are built for specific missions, but they encompass all of the objects known to be useful to that mission, and possibly more. However, these values can be refined over time: the more probe we send out, the more orbital determination solutions we can compute, and therefore the better the resolution of the GM values. I'll also note that DE438 has been released a few weeks ago, for JUNO as well. – ChrisR Jul 21 '18 at 17:19
  • 1
    @ChrisR There are (at least) two groups of bodies to consider in ephemeris-building. The group of sources of gravity that influence trajectories of other bodies, and the group of bodies whose positions are reported by the final, released ephemeris. The two groups might be identical but they don't have to be. I don't quite understand what "encompass all of the objects known to be useful to that mission" means though. – uhoh Jul 22 '18 at 01:20
  • 1
    @uhoh that actually makes sense, thanks! I thought until now that there was a perfect intersection between the two sets. – ChrisR Jul 23 '18 at 01:50
  • 4
    One reason for the discrepancy between the wikipedia vs DE values for Jupiter, Saturn, and Uranus is that the wikipedia values are for the planet only while the DE values are for the planetary system (the planet plus its moons). Even then, the wikipedia values appear to be off, and except for Uranus, those values are unreferenced (and the reference for Uranus is an old 1992 paper). OTOH, the wikipedia value for Neptune appears to be an unreferenced value for the Neptunian system as a whole. – David Hammen Oct 21 '18 at 11:28
  • But those are just a few, for example I have jup365.bsp and jup344.bsp. From the second I just got some masses with gm_de431. Id's for 551 to 569 also havent' name with bodc2n_c(naifid, LENNAME, sname,&bfound ). And 55501 to 55507. – Luis ALberto Dec 13 '21 at 19:34

1 Answers1

10

If you consider the ephemeris files put out by JPL as the final say in terms of simulations, then the data they put in the PKC files alongside the SPK files are the authoritative value, because that's the data used in the simulations.

These values are available at https://naif.jpl.nasa.gov/pub/naif/generic_kernels/pck/gm_de431.tpc

For completeness, JPL has a list of planetary satellite physical parameters at: https://ssd.jpl.nasa.gov/?sat_phys_par that is pretty comprehensive.

As a side note, I was looking for the GM value for Nerid, and it's listed on that page, but in the files found on the ftp site along with the SPK files, Nerid's value is set to 0. This suggests that there are discrepancies between the value used in the simulation and what you can find in this table.

uhoh
  • 148,791
  • 53
  • 476
  • 1,473
Kaushik Ghose
  • 990
  • 1
  • 8
  • 14
  • Hi @uhoh, that depends on what/who you consider an authority. Or, in your case "The" authority. If you consider the ephemeris files put out by JPL as the final say in terms of simulations, then the data they put in the pkc files are the authoritative value, because that's the data used in the simulations (https://naif.jpl.nasa.gov/pub/naif/generic_kernels/pck/gm_de431.tpc) – Kaushik Ghose Jul 22 '18 at 03:17
  • 1
    Thanks for all the helpful links! It's great when someone takes the time to revive an old question and then track down and post links to sources of helpful information. – uhoh Jul 22 '18 at 15:53
  • 1
    @uhoh happy (and surprised) I could help, being a new comer and all :) – Kaushik Ghose Jul 23 '18 at 00:03
  • 5
    For reference, the numbering system in the first file linked by @KaushikGhose is as follows: Bodies 1 to 9 are Mercury, Venus, the Earth and the Moon, the Mars system, the Jovian system, the Saturnian system, the Uranian system, the Neptunian system, and the Plutonian system. BODY10 is the Sun. The file also lists values for the planets themselves and for key moons. Moons are prefixed by the 100 times the system number plus an index for the moon (e.g., the Moon is BODY301). The planets themselves are the 100 times system number plus 99 (e.g., the Earth is BODY399, and Pluto is BODY999). – David Hammen Oct 21 '18 at 11:50