[IGSMAIL-7125] Re: Updates to the IERS Conventions (2010)

[Jim Ray]

Author: Jim Ray

Concerning the message below, I urge users to exercise great caution
before implementing the recommended new Mean Pole model.  All IERS
Conventions updates should contain three components: 1) a description
of the model change; 2) a justification for why the change is needed;
3) a quantification of the impact of the change on user results.  The
updated Chapter 7 text is misleading in why a new Mean Pole is needed
and no attempt at all is made to quantify its impact on user results.

We are told that the "model is intended to remove the effects of
rotational deformation at the  annual  and  Chandler  periods".  But
it is easy to see that the new model is materially no different from
the previous linear or cubic+linear models in doing that.  The real
difference among mean pole models is in their responses to inter-
annual polar motion changes and therefore inter-annual pole tide
deformations -- as explained by King and Watson (GJI, 2014, 199,
1161-1165) -- not in the annual+Chandler band.  In my view, it is
irresponsible for the IERS Conventions to shift responsibility for its
consequences onto users (especially a Class 1 model recommended a
priori for all data reductions), as the new text does: "it may be that
the model is deficient with respect to longer-term variations".

Indeed, by tracking the motion of the mean pole even more closely than
before, adoption of the new Mean Pole model implies that even less of
the longer-period pole tide signal will be removed from geodetic station
position estimates.  Following the recent paper by Wahr et al. (JGR,
2015, 120, 4597-4615) for the impact on the geopotential C21 and S21
coefficients, one expects their recommendation of a linear mean pole
trajectory due to GIA alone (computed from observations over 1900-1978)
would be more consistent with the linear ITRF definition given in
Chapter 4 of the Conventions.  That would be very similar to the IERS
Conventions 2003 Mean Pole model.

A potential difficulty with following such a linear mean pole arises
for the alignment of the Earth's geopotential model figure axis (C21,S21)
with its rotational axis, which currently uses the same Mean Pole model
(Chapter 6 of the Conventions).  It is not clear to me why the latter
could not rely instead on daily mean polar motion values.

No disrespect is intended towards the highly appreciated efforts of the
IERS Conventions Product Center nor its volunteer contributors.


- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
From: <central_bureau at iers.org>
To: <messages at iers.org>
Subject: IERS Message No. 274: Information on the updates to the IERS
 Conventions (2010)
Date: Mon, 22 Jun 2015 05:43:32 +0000

************************************************************************
IERS Message No. 274                                       June 22, 2015
************************************************************************


Information on the updates to the IERS Conventions (2010)


The IERS Conventions Center informs of the following updates:
19 June 2015
Chapter 7 "Displacements of reference points" has been updated, see
details in http://tai.bipm.org/iers/convupdt/convupdt_c7.html .

The main changes are:
- Modifications  to account for the next leap second to occur
  June 30, 2015.
- Corrections of errors.
- Introduction of the new version 2015 of the IERS Conventional Mean
  Pole, realized by a provided subroutine.

The USNO web site http://maia.usno.navy.mil/conv2010/convupdt.html
is similarly updated.


Gerard Petit and Brian Luzum
IERS Conventions Center
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.igs.org/pipermail/igsmail/attachments/20150718/6ce8da77/attachment.html>