1 / 32

The IAU 2000/06 changes to celestial reference systems a Bluffer’s Guide

The IAU 2000/06 changes to celestial reference systems a Bluffer’s Guide. Patrick Wallace Rutherford Appleton Laboratory, UK P.T.Wallace@rl.ac.uk. Presentation outline. What has changed, and why? What do we have to stop saying, and what do we say instead?

harry
Télécharger la présentation

The IAU 2000/06 changes to celestial reference systems a Bluffer’s Guide

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. The IAU 2000/06 changes to celestial reference systemsa Bluffer’s Guide Patrick Wallace Rutherford Appleton Laboratory, UK P.T.Wallace@rl.ac.uk Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

  2. Presentation outline • What has changed, and why? • What do we have to stop saying, and what do we say instead? • Consequences for data analysis software? Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

  3. Why some astronomers are suspicious What has the IAU been up to? • Sidereal time seems to have gone... • ...and so have equinox and ecliptic: why? • The new zero point for right ascension is something elusive (and French) called a “non-rotating origin” that you can’t draw a diagram of. Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

  4. Things were happening even before 2000 • B1950 star positions had already died out: everyone was using something they called “FK5 J2000” or “equinox J2000” (or, if they were ignorant, “epoch 2000”). • But in fact by the end of the century we were all using the International Celestial Reference System (ICRS). • ICRS is defined by a list of coordinates for 212 extragalactic radio sources; this list is called the International Celestial Reference Frame (ICRF). • The optical realization of the ICRS is at present the Hipparcos Catalog. Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

  5. Mean J2050? • Former star catalogues contained coordinates with respect to the mean (i.e. precessing) equator and equinox of a given epoch that was not too far from the present – B1900.0, B1950, J2000… • But ICRS is permanently frozen, for convenience close (~23mas) to the J2000 dynamical system but in other respects disconnected from precession, equators, ecliptics and equinoxes. • There will never be a change to J2050, and data analysis software need not prepare for such a change. Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

  6. IAU 2000/2006 Resolutions 2000 B1.1 Maintenance and Establishment of Reference Frames and Systems 2000 B1.2 Hipparcos Celestial Reference Frame 2000 B1.3 Definition of Barycentric Celestial Reference System and Geocentric Celestial Reference System 2000 B1.4 Post-Newtonian Potential Coefficients 2000 B1.5 Extended relativistic framework for time transformations and realization of coordinate times in the solar system 2000 B1.6 IAU 2000 Precession-Nutation Model 2000 B1.7 Definition of Celestial Intermediate Pole 2000 B1.8 Definition and use of Celestial and Terrestrial Ephemeris Origins 2000 B1.9 Re-definition of Terrestrial Time TT 2006 1 Precession Theory and Definition of the Ecliptic 2006 2 Supplement to the IAU 2000 Resolutions on Reference Systems 2006 3 Re-definition of Barycentric Dynamical Time, TDB Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

  7. BCRS and GCRS • Two GR-compliant celestial reference systems were introduced: BCRS for barycentric uses and GCRS for geocentric uses. • They have (in effect) the same orientation as ICRS, but the BCRS-GCRS transformation includes light deflection and stellar aberration. • Note that although GCRS is geocentric, it does not precess. GCRS is the quasi-ICRS arena in which the precession-nutation models operate. Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

  8. Earth orientation • New precession (IAU 2006) and nutation (IAU 2000A) theories, much more accurate than before, have been adopted. • There is a new definition of the celestial pole, called the celestial intermediate pole (CIP). • Earth rotation angle (ERA) is the new sidereal time and is simply proportional to UT1. It refers to a point on the sky called the celestial intermediate origin (CIO) and a point rotating with the Earth called the terrestrial intermediate origin (TIO). • The CIO and TIO define the “intermediate” celestial and terrestrial reference systems (CIRS, TIRS). Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

  9. Why new precession-nutation models were needed BIG errors in old 1976/1980 precession-nutation model: mas ! Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

  10. The surprisingly complicated pre-2000 expression for sidereal time: GST1982 (0h) = 24110.54841 + 8640184.812866tUT + 0.093104 tUT2 - 6.2e-6 tUT3 + Δψ cos ε + correction terms (2) The even more baroque IAU 2000 expression: GST2000 (0h) = 24110.5493771 + 8639877.317376 tUT + 307.4771600 tTT + 0.0931118 tTT2 - 0.0000062 tTT3 + 0.0000013 tTT4 + Δψ cos ε + correction terms (34) The much simpler Earth rotation angle formula: ERA (0h) = 24110.54841 + 8639877.317376 tUT Why getting rid of sidereal time was a good thing Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

  11. CIO: the new zero point of right ascension • Classical: zero point defined geometrically, by intersection with ecliptic. But messy: intersection of two moving planes. • New: zero point defined kinematically. Keeps precession-nutation and Earth rotation separate and allows use of ERA. But you can’t draw a picture of it. Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

  12. The CIO is merely a point on the moving celestial equator that stays as still as it can. Obviously, it has to move north-south in the sky as the equator precesses… …but the CIO doesn’t move along the equator: from moment to moment it moves only at right-angles to the equator. In fact it does creep along a bit, but very slowly. The CIO’s present αICRS is about 00h 00m 00s.0001; by the end of the century it will have drifted only as far as αICRS = 0h 00m 00s.005. Where is the CIO? Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

  13. The ecliptic remains important in a qualitative and descriptive sense… …and is part of constructing a precession theory… …but is no longer needed to define the zero point of right ascension. The ecliptic is in any case a slippery concept: Is it defined by the Earth-Moon barycenter’s path, or the orbital angular momentum vector? n.b. Difference ~0.1”. Does the ecliptic plane go through the Sun? The solar-system barycentre? The Earth-Moon-Sun barycentre? What about very-long-period nutation terms? So what happened to the ecliptic? Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

  14. Old and new ways of calculating h,δ ICRS  etc. space motion parallax light deflection aberration old new frame bias GCRS CIP,CIO precession nutation Greenwich mean sidereal time Earth rotation angle equation of the equinoxes polar motion h, Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

  15. Predicting h,δ, old and new: example 2004 May 31, 22h UTC λ = −104.9950, φ = +39.7427 fictitious star polar motion omitted Old-style prediction using sidereal time ICRS 23 32 55.171 +52 16 38.29 Apparent RA,Dec 23 33 06.176 +52 17 43.50 Local HA,Dec + 8 05 50.276 +52 17 43.66 New-style prediction using Earth rotation angle ICRS 23 32 55.171 +52 16 38.29 Intermediate RA,Dec 23 32 53.329 +52 17 43.50 Local HA,Dec + 8 05 50.276 +52 17 43.66 Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

  16. Why none of this matters much to ordinary users • When did you last compute an apparent place? Telescopes nowadays are always pointed by a computer. • Star positions are always ICRS now. • Planet positions can be astrometric places, i.e. ICRS as well. • “Intermediate places” (the new apparent places) are really just an internal detail. Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

  17. Consequences for data analysis software • When identifying the celestial coordinate system, there isno longer any need for a numerical “equinox” parameter. Just identify the reference system with a string (currently “ICRS”). But there still needs to be an “epoch” – for proper motion. • Avoid using apparent (or intermediate) place for image WCS. Either resample to astrometric (i.e. ICRS) RA,Dec or leave as observed place (i.e. what the camera saw). • Have nothing to do with sidereal time. Record UTC date and time and leave it for the data analysis programs to calculate Earth rotation. Helpful to record the predicted UT1−UTC (and xp,yp) though. (But always handle UTC leap seconds properly.) Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

  18. Bluffer’s guide to time scales • Know what the time scales UTC, TAI, GPS, TT and TDB are used for, how much they differ by (roughly) and how to get hold of them. • At least have heard of the coordinate time scales TCG and TCB. • Know that TDB is a coordinate time scale and is not about “a clock at the barycentre”. • Know that UT1 and sidereal time aren’t really time scales but measures of Earth rotation, and that ERA is the new GST. • Know you mustn’t use UTC without explicit provision for leap seconds and that it’s better to use TAI except for “user interface” things. Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

  19. Bluffer’s guide to celestial reference systems • Know that precession just makes the Earth’s pole move round the sky, and forget about ecliptics and equinoxes. • Know that precession and nutation are merely the low-frequency and high-frequency parts of the pole’s motion. • Don’t say “FK5” or “J2000” or “mean place” any more: just say “ICRS”. • Never say “precessing to current coordinates”. It’s more complicated than that, and anyway is something the TCS will look after,… • …but do know that the new “apparent place” is called “intermediate place”… • …and that the RA zero point for intermediate place isn’t the “first point of Aries”: it’s now the “CIO”, which is “a point on the celestial equator close to αICRS= 0”. It is permissible to wave your arms about a bit when saying this. Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

  20. END Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

  21. Additional material Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

  22. Precession-nutation and Earth rotation • Astronomers have traditionally talked of the “precession of the equinoxes” and also have distinguished between luni-solar precession, planetary precession and general precession. • But to the layman precession is simply about what the Earth’s rotation axis does. • This is also the IAU 2000 picture: precession-nutation describes the motion of the pole with respect to the stars. The ecliptic has only a “behind the scenes” role. • In the IAU 2000 “new paradigm”, the clean separation between the formulations forthe pole’s motion and Earth rotation respectively makes things simpler still by eliminating crosstalk from the former to the latter, as can be seen by comparing GST and ERA. Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

  23. GST and ERA compared Apart from a ~3s per year slope (precession) and some ~1s wiggles (nutation), they are very similar... Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

  24. ...so set your sidereal clock to ERA! ERA Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

  25. The real challenges • The new ways are easier, not harder... • the celestial sphere is ICRS for ever • precession-nutation simply describes where the CIP is • the CIO (zero point of α) stays as still as possible • h−α = Earth rotation angle = a + b∙UT • ...so the challenge is to persuade people to unlearn the old ways: • mean places and the precession of the equinoxes • equinox of date, and apparent places • mean and apparent sidereal time, and the equation of the equinoxes • ...and then to educate the next generation. Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

  26. References • IAU Resolutions and implementation: • IERS Conventions (2003) • Kaplan (2005) USNO Circular 179 • NFA WG explanatory documents • Kovalevsky & Seidelmann (2004), Fundamentals of Astrometry • GCRS/BCRS: • Soffel et al. (2003) AJ, 126, 2687–2706 • IAU 2000 precession-nutation: • Herring et al. (2002) , JGR, 107, B4, Issue B4, ETG 4-1 • CIP/CIO: • Capitaine et al. (2003) A&A, 400, 1145–1154 • Capitaine & Wallace (2006) A&A, 450, 855–872 • UT1/GST: • Capitaine et al. (2000) A&A, 355, 398–405 • Capitaine et al. (2003) A&A, 406, 1135–1149 Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

  27. Four generations of precession • IAU 1976 - Lieske precession, in z,ζ,θ form. • IAU 2000 - Pole offsets and precession rate corrections absorbed into new nutation model. • SOFA/IERS 2003 - Bias extended to include Δα and kept separate from nutation. Precession uses the Lieske ε0,ψA,ωA,χA angles with rate adjustments applied to ψA, and ωA. IERS offers CIP X,Y directly through series, SOFA through the classical NPB matrix. • IAU 2006 – New precession model, consistent with dynamical theory but otherwise matching IAU 2000. Includes small adjustments to IAU 2000A nutation because of change to ε0 etc. New SOFA routines, mainly using angles (now γ,φ,ψ,εA) but also providing direct X,Y series. To come in service at the start of 2009. ‾ ‾ ‾ Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

  28. “Angles” method TT X,Y ERA UT s GCRSRTIRS GCRSRCIRS Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

  29. “Angles” method– equinox option GCRSRtrue EO GST TT X,Y ERA UT s GCRSRTIRS GCRSRCIRS Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

  30. Free core nutation: the “noise floor” Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

  31. Accuracies now available • Earth orientation known to 300 μas RMS, ~ 1 cm • Smallest terms in nutation model ~ 1 μas • Size of nutation model = 1365×2×2 coefficients • Error in former IAU 1976 precession ~ 3 mas/y • Uncertainty in IAU 2000 precession ~ 30 μas/y • Nutation-model noise floor ~ 500 μas (from FCN) n.b. • 1 mas (~Hipparcos) is the aberration you get from walking pace • 1 μas is about 30 μm at Earth’s surface Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

  32. Barycentric Dynamical Time, TDB defining linear relationship 4D spacetime transformation TAI GCRS BCRS TT TCG TCB TDB TDB keeps roughly in step with TT, on the average Astronomical Data Analysis Software & Systems XVII London, 23-26 September 2007

More Related