Tutorial

EarthOrientation.jl downloads, parses, and interpolates weekly-updated tables from the IERS that contain the following Earth Orientation Parameters (EOP):

  • Polar motion:
    • x-coordinate of Earth's north pole: $x_p$
    • y-coordinate of Earth's north pole: $y_p$
  • Earth rotation
    • Difference between UT1 and UTC: $\Delta UT1$
    • Excess length of day: $LOD$
  • Precession and nutation based on the 1980 IAU conventions
    • Correction to the nutation of the ecliptic: $d\psi$
    • Correction to the obliquity of the ecliptic: $d\epsilon$
  • Precession and nutation based on the 2000 IAU conventions
    • Correction to the celestial pole's x-coordinate: $dx$
    • Correction to the celestial pole's y-coordinate: $dy$

These parameters are required for precise transformations between quasi-inertial and rotating terrestrial reference frames.

Getting Earth Orientation Data

When the package is imported for the first time the required data will be automatically downloaded from the IERS servers. After that the data needs to be updated manually like shown below.

using EarthOrientation
EarthOrientation.update()

Loading Earth Orientation Data

The downloaded data is parsed into an EOParams object:

eop = EOParams()

By default the files downloaded by EarthOrientation.update() will be used. It is also possible to manually pass the required finals.all and finals2000A.all files in CSV format.

eop = EOParams("finals.csv", "finals2000A.csv")

This is useful if the data should not be managed by EarthOrientation.jl but by a different system instead.

Interpolating Earth Orientation Data

Get the current Earth orientation parameters, e.g. for polar motion:

xp, yp = polarmotion(eop, now()) # arcseconds

Or the current difference between UT1 and UTC and the associated prediction error:

ΔUT1 = getΔUT1(eop, now()) # seconds
ΔUT1_err = getΔUT1_err(eop, now()) # milliseconds