The predictions for COMPASS-MS1 and COMPASS-MS2 have been consistently poor since the start of their laser tracking operations. This is strange, for the COMPASS-M3 predictions are of the regular good quality we expect for this kind of satellite.
We are talking about time biases of seconds, plus up to hundreds of milliarcseconds cross-track offsets. This makes daytime tracking simply impossible, and requires substantial additional time and effort to obtain returns during night time operations.
COMPASS-MS1 has been recently included in the high priority list of the LARGE campaign. The intensive tracking requirements for this campaign are obviously harder to fulfill if the predictions are sub-standard. Two questions come to mind: 1) Why are predictions so poor for these objects? 2) When should we expect them to improve?
In the interim, perhaps it would be useful to include COMPASS-MS1 in the time bias prediction service run by Potsdam?
(http://sgf.rgo.ac.uk/media/compass_tb.png)