From: slr@slrb.rgo.ac.uk (SLR Herstmonceux) 
Date: 18 May 2000 20:00:19 
Subject: [IGEXMail-0763] Glonass Outages and eclipsesBACK/EDIT 

Precedence: list

******************************************************************************
IGEXMail                18-May-2000 20:00:19                   Message No 0763
******************************************************************************

Author: slr@slrb.rgo.ac.uk (SLR Herstmonceux) 
Subject: Glonass Outages and eclipsesBACK/EDIT 


We have come across a rather puzzling relationship between eclipses and outages
on GLONASS satellites and wondered if anyone in the IGEX community could help 
us understand what is happening.

As participants in the IGEX campaign we are currently running a 3S receiver
to collect GLONASS/GPS data.  During the campaign we have been using NAGUs to 
determine the status of the satellites. The information in the NAGUs is useful
in determining if gaps in our data coincide with satellite outages, and also 
for determining when satellites are withdrawn from the system.

We have noticed that NAGUs generally report satellite outages after they have 
taken place, and very few satellite outages are forecast in advance. With the 
exception of the satellite in Slot 18 most of the forecast outages are of 
relatively short duration, typically of less than an hour. Slot 18 was rather 
exceptional.  This satellite used to have outages lasting several weeks 
forecast in advance.  These outages corresponded to periods when part of the 
satellite's orbit was situated in the Earth's shadow.

We noticed that the forecast NAGUs often occurred in groups, with several 
outages occurring on the same day.  By chance we noticed that the clusters of 
satellite outages happened to occur round about the time of a new moon. On 
looking more closely we found many of the outages that are forecast in advance
coincide with the satellite being eclipsed by the moon. We have tried to find 
the reason why some satellites have an outage and others do not, but our 
analysis has not found anything conclusive.  For example one satellite had an 
outage during a six-minute eclipse whereas many satellites can undergo 
eclipses of 30 minutes or more without having an outage.  Events such as this 
rule out any dependency of the depth of the eclipse (and the resulting drop in
solar intensity) and suggest that it is unlikely to be related to problems 
with the batteries.  Does anyone have additional data or information that may 
assist us?

We noticed that in IGEXMAIL No 0086 there was a comment from NPO PM related
to a decrease in accuracy of the orientation of the solar panels when a 
satellite is passing in and out of eclipse by the Earth.  Does anyone know if 
this problem could also relate to eclipses by the moon?  Even so we do not 
understand why this would cause a satellite to be switched off.

If anyone knows of any reason why these outages occur we would be very 
interested in hearing from them.


                                                               |
    ========================================================= -O- =========
    |                                                         /|          |
    |  NERC Space Geodesy Facility                           /            |
    |  Herstmonceux Castle                                  /             |
    |  Hailsham                                     (\    / /  /)         |
    |  East Sussex                                 (  \  /_/  /  )        |
    |  UK   BN27 1RP                              (    \ |O| /    )       |
    |  Tel:  +44 (0)1323 833888                   =================       |
    |  Fax:  +44 (0)1323 833929                   ==             ==       |
    |  Email: slr@slrb.rgo.ac.uk                  =               =       |
    =======================================================================