Orbiter and Earth rotation

[Ajaja]

I'm working now on tuning some celestial body rotation parameters in Orbiter, trying to make them closer to current IAU model using Python and pygmo2 optimization.
For some bodies there is some space for optimization. For example Ganymede:

Some rotation parameters are very close to optimal, e.g. for Mars:

For some bodies like Moon there is nothing to optimize al all:


But for our home planet Earth I'm getting this:



Any idea what's happening? Are default rotation parameters of Earth in Orbiter so inaccurate?

Green (optimized) line is:

SidRotPeriod = 86164.10041642643
SidRotOffset = 4.889487574817284
Obliquity = 0.409092790995108
LAN = 0
LAN_MJD = 51544.5
PrecessionPeriod = -9413040.4
PrecessionObliquity = 0
PrecessionLAN = 0

 

[n72.75]

...woah, that is a good question...to which I do not have the answer.

@indy91, would this affect RTCC calculations? Or is something else going on here that we might have run accross?

 

[indy91]

We use a modified Earth config file in NASSP. I always assumed the difference was mostly due to the Earth rotational rate changing over time, where in Orbiter it is constant. So we mainly needed to modify the SidRotPeriod and SidRotOffset parameters to get the best parameters for around 1970, as opposed to 2000 in the standard Earth config file. Here are our numbers in NASSP:

PrecessionLAN = 0
PrecessionObliquity = 0
PrecessionPeriod = -9413040.4
LAN = 0.00001553343
LAN_MJD = 51544.5
SidRotOffset = 4.894942829
SidRotPeriod = 86164.098904
Obliquity = 0.4090928023

To be honest, these numbers were empirically adjusted to better agree with hardcoded numbers in AGC software. Especially that tiny LAN value is probably unnecessary.

Most of the difference between the standard Orbiter and your optimized numbers seems to be the change to SidRotPeriod. But why that might be a bit wrong, I don't know. Our SidRotPeriod in NASSP agrees with the Earth rotational rate used in most AGC software versions at least.

 

[Ajaja]

It turned out that IAU_EARTH is not accurate too:

https://naif.jpl.nasa.gov/pub/naif/JUNO/kernels/pck/pck00010.tpc

NAIF strongly cautions against using the earth rotation model (from [1]), corresponding to the SPICE reference frame name IAU_EARTH, for work demanding high accuracy. This model has been determined by NAIF to have an error in the prime meridian location of magnitude at least 150 arcseconds, with a local minimum occurring during the year 1999. Regarding availability of better earth orientation data for use with the SPICE system:

Earth orientation data are available from NAIF in the form of binary earth PCK files. These files provide orientation data for the ITRF93 (terrestrial) reference frame relative to the ICRF.

I did as they say and used ITRF93.
There is deviation of [1,1,0] vector (north pole + prime meridian direction in Orbiter’s left-handed system) in radians from ITRF93 :


Optimized parameters are:

SidRotPeriod =  86164.09983253977
SidRotOffset =  4.890656175683156
Obliquity =  0.409101796432498
LAN =  0,
LAN_MJD =  51544.5
PrecessionPeriod =  -9413040.4
PrecessionObliquity =  0
PrecessionLAN =  0

There is current deviation of Erath rotation in Orbiter from ITRF93 separetly the north pole and prime meridian vectors:


So, the main error is from rotation.
And it's not a joke. 0.004 radians = ~55 seconds of Earth rotation.

 

[Ajaja]

Most of the difference between the standard Orbiter and your optimized numbers seems to be the change to SidRotPeriod. But why that might be a bit wrong, I don't know. Our SidRotPeriod in NASSP agrees with the Earth rotational rate used in most AGC software versions at least.

Yes, and optimization based on ITRF93 is a bit closer to SidRotPeriod in NASSP.