[IGSMAIL-311] Deweighted JPL Orbits in Orb. Comparisons

Geoff Geoff
Sat Aug 14 12:31:56 PDT 1993


IGS Electronic Mail       14-AUG-1993 12:31:56       Message Number 311
***********************************************************************
 

>From: Geoff Blewitt
Subj: Deweighted JPL Orbits in Orb. Comparisons
-----------------------------------------------

Product files should never be used without first reading summary files.
We remind IGS and users to take note of lines in our summary report that
begin with the word "WARNING". For example, IGS mail#417 states:

WARNING: Data from PRN07/GPS37 have been deweighted for days
178-181. (Jun 27 through Jun 30, 1993).

We do not recommend the use of products specified in WARNING messages.
Recent IGS orbit comparisons (IGS mail #418 and #431) show anomolously
large RMS values for intercomparisons with JPL orbits simply because 
deweighted satellites have been included in the RMS.
Note that in IGS mail #431, our orbit RMS has a dramatic improvement
going from Day 4 to Day 5 of Week 703. In reality, our normally weighted
orbits prior to Day 4 were probably as good as Day 5.

Other analysis centers simply remove suspect satellites, and we are 
considering doing the same.  This raises the more general question: 
should analysis centers assume that customers will NOT read summary 
files?  Part of the problem is that there is no standard way of formatting
the summary files, however, we have made it easier for users by using
the key work "WARNING" at the beginning of the line.
The disadvantage of removing all suspect orbits entirely is that it may be 
appropriate to use a deweighted satellite in certain circumstances (e.g., 
rapid static survey over 30 km, where the orbit solution is degraded, but 
is nevertheless better than the broadcast ephemerides.)

      		Sincerely,

      		Geoffrey Blewitt
      		JPL IGS Analysis Center
      		geoff at logos.jpl.nasa.gov



More information about the IGSMail mailing list