Conclusions of the discussion concerning a modern definition of the CEP and proposal

A summary of the main points of the comparison between two possible options for a modern definition of the CEP and preliminary proposal is given below.

Option (1): Mathews's proposal (Journees 1998) was to define the CEP by keeping the symmetry in the frequency band between terrestrial motion and celestial motion in extending the definition outside the current frequency interval; the proposed procedure for its realization is to estimate simultaneously, in processing the observations, the current celestial pole offsets and current polar motion as well as the high frequency signal in a symmetric way in the frequency domain in the CRS and TRS. In fact, the essential content of Mathews's proposal at the Journees 1998 was the method for estimation for the higher frequency components, whether in the CRS or TRS. The suggestion for distribution of the different frequency components between the CRS and TRS has now be modified such as the conventional model would imply that the series proposed for $\delta \psi$ and $\delta \epsilon$ would be modified to include also terms with negative n : $n=-1$ for the prograde diurnal nutations as well as the long period ocean tide terms (assuming that the latter can be reasonably well modelled); $n=-2$ for the prograde semidiurnal nutations and for the prograde diurnal ocean tide terms; $n=-3$ for the prograde semidiurnal ocean tide terms; $n=1$ for the retrograde semidiurnal ocean tide terms. All these are, of course, in addition to the long period nutations comprised under $n=0$, and like the latter, will be fixed be determined from theory.

In any case, I am of the firm opinion that the diurnal and semidiurnal nutations (in space) which are indeed predictable, should be part of the conventional model. It would be illogical not to include them.

Option (2): the CEP is defined by separating the motion into the retrograde diurnal component of the motion wrt the TRS, which is considered as being long periodic motion in the CRS (periods larger than 2 days mainly due to the effect of the external torque on the Earth) and the motion wrt the TRS consisting of the whole complementary part of the motion; the proposed procedure for the realization of the CEP is to use the best model for precession nutation (periods longer than 2 days) in the estimation of the EOP (estimating the celestial pole offsets if possible) and then to extract the high frequency signal (or rather corrections to an empirical model for this signal) from the pole coordinates only.

Both choices (1)and (2) correspond to a clear concept not dependent on further improvements in the model. Both of them extend the current definition in order that the current definition is an approximation of the new one and has minimal impact on users. Both can be realizable by a model as accurately as possible.

The advantage of (1) is its clear mathematical basis.

The advantage of (2) is its clear relation with a deterministic approach.

The preliminary proposal presented in the Newsletter 3, which appears to be the "preferred" option, corresponds to the choice (2).

Concerning its realization, precession and nutation are specified by the best available model including the components with periods larger than 2 days and the high frequency signal in the TRS is specified by a model including the tidal variations in polar motion and the terms corresponding to the sub-diurnal nutations. The corrections to these models are estimated in two steps : the first step corresponds to the current procedure for estimating the EOP (including when it is possible the celestial pole offsets); the second step is the estimation of the high frequency motions from the currently estimated coordinates of the pole.

The reasons for this choice, which considers the whole high frequency motion in the TRS are that: - the prograde diurnal nutations (ampl in Deltapsi x sin eps and Delateps : 15 microarseconds) appear as long periodic polar motion which cannot be separated from the polar motion itself of much larger amplitude, - the semi-diurnal prograde nutations (ampl in Deltapsi x sin eps and Delateps : 15 microarseconds) appear as prograde diurnal variations in polar motion and cannot be separated from the much larger components (150 microarseconds) of the tidal variations of polar motion.

The proposed realization is such that it can be applied to all the techniques, whereas it is not the case of (1).