[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Orekit Users] Changes to FramesFactory in 6.0

Hi Matt,

Matt Edwards <matt.edwards.issinc@gmail.com> a écrit :

I am in the midst of upgrading our implementation of Orekit from v5.0.3 to
6.0. One of the methods we use is getTIRF2000(), however in 6.0 this now
has an additional parameter for the IERS conventions, either 2003 or 2010.

I am not familiar with this convention, and as such not sure which option
to choose, or what the difference would be. Ultimately, I want the same
behavior as I had with the empty constructor.

You should probably use 2010. The difference between the various options are the underlying precession/nutation models. Up to conventions 1996, IERS used Lieske 1976 precession and IAU 1980 Wahr nutation. For conventions 2003, the precession/nutation model has been changed to Mathews, Herring and Buffett model IAU 2000A. Since conventions IERS conventions 2010, the model has been changed again to Nicole Capitaine's P03 model, also known as IAU 2000/2006. Newer models are more accurate than former models, but this is leveraged as IERS publish corrections to apply to these models, in the form of Earth Orientation Parameters.

The problem is that IERS cannot publish the correction values for all their past precession/nutation models. For now, they publish only the correction for the 2010 conventions (i.e. with respect to IAU 200/2006 precession/nutation) and with respect to the legacy IAU-76/IAU-80 used in the 1996 conventions. The intermediate model used for conventions 2003 has no published corrections I am aware of.

The net result is that the best choice for now is to choose the most recent 2010 conventions.

We are currently in the process to add back older conventions (read IERS-1996), as can be seen in the repository here: <https://www.orekit.org/forge/projects/orekit/repository/show?rev=IERS>. However, this is work in progress. It is only in the Git repository, and as you may notice from the URL not in the master banch but in a dedicated IERS branch that will be merged into master only when stable enough.

Hope this helps

Matt Edwards

This message was sent using IMP, the Internet Messaging Program.