Index Home About Blog
Date: 11 Feb 1982 18:58:40-PST
From: ihnss!karn at Berkeley
To: ucbvax!space@Berkeley
Subject: Eclipses in geostationary orbit

This is a topic that is well understood in the communications
satellite business.  Solar eclipses for spacecraft in geostationary
orbit occur only during the spring and fall seasons (equinoxes) when
the sun appears to pass behind the earth.  During the summer and
winter seasons (for the Northern hemisphere), the sun appears to
pass above the north and south poles at local midnight,
respectively, and there is no eclipse.

The longest eclipses occur at midnight on the equinoxes, where
they are 72 minutes long.  Eclipse durations on days preceeding and
following equinoxes decrease to zero in about + or - 20 days.

A common practice in communications satellite operations is to place
them to the west of the areas they serve.  This causes eclipses to
occur after midnight in the service area, when the traffic has
fallen to lower levels.

A more serious problem for communications are sun transits, in which
the sun passes behind the satellite.  The radio noise from the sun
blots out the satellite signal.  Of course, this would not be
a problem for the SPS;  the sun would even contribute (very slightly) to
the received microwave energy.

Phil Karn
Bell Labs Indian Hill

From: (Henry Spencer)
Subject: Re: space station batteries
Date: Sat, 4 Mar 2000 19:58:19 GMT

In article <>,
David Buchner <> wrote:
>> ....Government satellites .... can operate in higher inclinations (hence
>> much less stationkeeping fuel required), so they could conceivably hit
>> their design lives with a lot of fuel left.
>Okay, I'm coming back to this discussion late, but now I have another
>question. Why does an equatorial orbit require more maintenance than a
>higher-inclination one?

It's not so much a matter of a higher inclination, as of being more
tolerant of substantial inclinations.  Commercial comsats have to stay
pretty exactly where they're supposed to be, so that fixed antennas will
work.  But observation satellites, or comsats for mobile customers who
need tracking antennas anyhow, can tolerate some slop in their orbits.  In
particular, most of a comsat's stationkeeping fuel is used up keeping the
orbital inclination precisely zero, so a satellite which can drift some
needs much less fuel.

>didn't somebody say that geosynchronous satellites only experience
>eclipses during a small part of the year? I don't have a model here, and
>I'm having a tough time imagining, but is this by any chance around the
>equinoxes, because of the inclination of the equator being turned
>"sideways" to the Sun?

Correct.  Think of it from a non-rotating Earth-centered point of view.
The satellite goes around the Earth every 24 hours in a fixed plane.  The
Sun goes around the Earth every 365-odd days, in a different plane.  The
Earth's shadow is always precisely opposite the Sun, so it goes around in
a one-year cycle too, in the Sun's plane.  There are two places where the
Sun's plane crosses the equator, so twice a year, Earth's shadow slowly
crosses the plane of the satellite's orbit, and so the satellite passes
through it once a day during those "eclipse seasons".  And yes, they're
around the equinoxes.

The good part of this is that the eclipses happen only during short
seasons, and only once a day.  The bad part is that the eclipses are
rather longer than what a low-orbit satellite sees.  So a comsat's
batteries get cycled less often, but rather more deeply.

Some comsats, doing direct TV broadcast to areas slightly east of their
own position, simply accept this, and shut down most of their operations
during eclipse -- after all, it happens at maybe 0200 in their coverage
area, when demand for TV is low.  They still have small batteries to keep
their housekeeping and command systems alive.  However, most comsats have
to operate through eclipse, and that means quite a bit of battery mass.
Computer disaster in February?  Oh, you |  Henry Spencer
must mean the release of Windows 2000.  |      (aka

Index Home About Blog