[IGSSTATION-476]: Change to discontinuity table format

"Ferland, Rémi" RFerland at NRCan.gc.ca
Mon May 16 08:31:18 PDT 2005


******************************************************************************
IGS Station Mail          16 May 08:31:28 PDT 2005      Message Number 476
******************************************************************************

Author: R. Ferland

Dear IGS Colleagues,

	As you most likely know, files summarizing station coordinates
time series discontinuities are being used to keep track of events
affecting coordinates time series. To improve the usefulness of such
file, it was recently suggested within IGS to also include the type
of modeling as well as the axis that may be affected by the discontinuity.
The proposed changes were first suggested during an IGS meeting at
the EGU, and then circulated within the IGS Reference Frame Working
Group and to IERS.

	The proposed changes could eventually also be used to document
the models used in a combined file (assuming that SINEX is also
expanded).

	In the recently accepted discontinuity format, which is
intended to keep track of station coordinates time series
discontinuities. The discontinuity block includes for each
station (code + pt + soln) a validity interval. ex:

+SOLUTION/DISCONTINUITY
*CODE PT SOLN T _DATA_START_ __DATA_END__
_______________COMMENTS_______________
 ALBH  A    1 P 00:000:00000 94:104:00000 Antenna Change
 ALBH  A    2 P 94:104:00000 95:011:80100 Antenna & Receiver Change
 ALBH  A    3 P 95:011:80100 00:000:00000 
......................
-SOLUTION/DISCONTINUITY
 	
	For proper modeling of time series coordinates, it may be
required to have more than coordinate discontinuity. Velocity
discontinuity, as well as other models discontinuity may be needed
along one two or all 3 axis. To address those, it was proposed,
within IGS, to extend the format to include modeling refinements
as well as  1D 2D or 3D application of the model. This can be achieve
by adding one column to specify which model is used/activated/reset
and another column specifying which axis the model is being applied to.

Commonly used models for time series analysis are:
 - Position / Velocity (P , V)
 - Periodic signal, namely Annual (A) and semi-annual (S)
 - Exponential decay (E)
 etc...
M=models (P,V,E,A,S,...)
More models can potentially be added without a need to
updated/change the format again. In the current procedures,
only P&V are used.

By also adding one axis column, one could refine the use/reset
of a Model to a given axis, local and geocentric coordinates
are proposed (E,N,U) or (X,Y,Z) (similar to eccentricity).
The character "-" could mean all axis (3D). e.g.:
A=Axis  (X,Y,Z,E,N,U,-)

An example for the updated discontinuity table including models
and Axis may look like:

+SOLUTION/DISCONTINUITY
*CODE PT SOLN T _DATA_START_ __DATA_END__ M A
_______________COMMENTS___________
 ALBH  A    1 P 00:000:00000 94:104:00000 P - Antenna Change
 ALBH  A    2 P 94:104:00000 95:011:80100 P U Antenna & Receiver Change
 ALBH  A    3 P 95:011:80100 00:000:00000 P -
 ALBH  A    1 P 00:000:00000 00:000:00000 V -
 ALBH  A    1 P 00:000:00000 00:000:00000 A - 
 ALBH  A    1 P 00:000:00000 00:000:00000 S - 

...........................
-SOLUTION/DISCONTINUITY

The existing format can be easily extended by implying that if no
velocity (V) model interval is given, that a constant velocity is
assumed.

Best Regards,
Remi,



More information about the IGSSTATION mailing list