Errors due to adopting the TDT/TT framework enter at the 10^-17 level, about 7 orders of magnitude below
the current geodetic capability but only ~2 orders below current frequency standards. The TCG frame will be
required in the near future by those working on frequency standards under development.
Nonetheless, despite considerable and enduring confusion on this point, it does not appear that any space
geodetic analysis group which contributes to ITRF uses TCG, nor are there any known users who do. Thus, the
change made in ITRF94 (and the subsequent related realizations ITRF96 and ITRF97) to shift the global scale
by applying an a posteriori rescaling of [1 + (0.6969... ppb)] has resulted in uniform inconsistency with
the entire user community, including the IGS and its orbits. Since the effect is small (~4.5 mm at the
Earth\'s surface; ~18 mm at GPS satellite altitude), this inconsistency has been nearly negligible.
However, with ITRF2000 the time scale effect is significant at the level of a few formal sigmas for the
global combination and should therefore be considered carefully.
Based on this discussion, the majority view of the attendees was that ITRF2000 should revert to the conventions
of ITRF realizations before ITRF94, consistent with TDT/TT time. All coordinate results should be distributed
in this framework and clearly described as such. To satisfy any users needing an ITRF consistent with TCG time,
as recommended by the IAU/IUGG, two alternatives were proposed:
- 1. Provide the formulation needed to rescale the ITRF2000 frame by
[1 + (0.6969... ppb)]
- 2. Provide a secondary set of coordinates with the rescaling already applied but only
if these are kept in a separate location with very carefully controlled user access.
Note that a consequence of this change, together with the natural evolution of the contributed data, is that
the shift in scale from ITRF97 to ITRF2000 will be about -2 ppb at epoch 2000.
|