3. Considerations on the main issues

The main issues (see Capitaine 1999 for more details) concern the predictable high frequency motions in nutation (Bretagnon et al. 1997) and polar motion (Herring & Dong 1994) and the new strategy of observations in which the part of sub-daily observations will increase in a near future.

1) The recent models, at a microarsecond accuracy, include :

-

prograde diurnal nutations (in $\Delta\psi\sin\varepsilon_0$ and $\Delta\varepsilon$) with amplitudes of the order of 15 microarseconds,

-

prograde semi-diurnal nutations with amplitudes of the order of 15 microarseconds.

-

prograde diurnal variations in polar motion with amplitudes of the order of 150 microarseconds,

-

prograde semi-diurnal variations in polar motion with amplitudes of the order of 100 microarseconds,

-

retrograde diurnal variations in polar motion with amplitudes of the order of the order of 800 microarseconds,

-

retrograde semi-diurnal variations in polar motion with amplitudes of the order of 300 microarseconds.

The overlapping between these different motions is such that :

- the prograde diurnal nutations in space appear in the Earth (Bizouard et al. 1999) as long periodic prograde and retrograde variations in polar motion,

$\Rightarrow$ it seems therefore appropriate to extend the definition of the CEP such that this motion be rather considered as a part of the motion of the pole in the TRS (which can be given by a model),

- the semi-diurnal prograde nutations in space appear as prograde diurnal variations in polar motion and it is necessary to choose where to consider more logically this motion, (it must be noted that such prograde diurnal variations in polar motion have in fact already been computed, for a non-rigid Earth, by Chao et al (1991) as resulting from the ``polar libration" and are thus included in their model for diurnal polar motion).

$\Rightarrow$ such a motion being dependent on the Earth model, it seems more appropriate to consider it in the TRS (which can given by a model).

- the retrograde diurnal terms in the tidal polar motion variations are in fact included in the most recent models of nutation for a non-rigid Earth,

$\Rightarrow$ such terms must therefore be excluded of the model of polar motion,

- the effect of the purely diurnal oceanic tide K1 must be considered to be included in the constant part of the celestial pole offsets,

$\Rightarrow$ such a term must therefore be excluded of the model of polar motion.

2) For sub-daily observations, the two frequency intervals of polar motion and nutation are no more disjoined and the frequency domain decomposition into polar motion and nutation becomes unclear. This prevents to estimate simultaneously polar motion and nutation as it is the case in the current VLBI daily estimates.

The major part of the polar motion being not a predictable motion, which has therefore to be estimated by the observations, it seems reasonable, in this case, to specify the precession-nutation by a model and to estimate only the terrestrial part.

References

Bizouard, C., Folgeira, M., Souchay, J., 1999 a, in the Proceedings of the IAU Colloquium 178.

Bretagnon, P., Rocher, P., and Simon, J.-L., 1997, Astron. Astrophys. 319, pp 305-317.

Brzezinski,A., 1999, to be published in the Proceedings of the IAU Symposium 178.

Capitaine, N., 1999, to be published in the Proceedings of the IAU Symposium 178.

Chao, B.F., Dong, D.N., Herring, T.A., 1991, Geophys. Research Letters 18, 11, 2007-2010.

Herring, T.A. and D. Dong, 1994, J. Geophys. Res. 99, pp. 18051-18071

Mathews, P.M., 1999 a, in the Proceedings of the Journées Systèmes de Référence Spatio-Temporels 1998, N. Capitaine ed, Observatoire de Paris, pp 161-163.