Henry Spencer wrote:
In article ,
Christopher M. Jones wrote:

...However, a telescope in an orbit with such a
low lifetime would also need frequent reboosts, which would
require a lot of onboard propellant and frequent usage of some
sort of thruster. Things which are not terribly compatable
with high precision optical astronomy equipment...

No, it just means you have to choose propulsion systems carefully,
avoiding orthodox hypergolics and other systems which spew out condensible
garbage that can easily contaminate optics. Xenon Hall-effect thrusters
or an ammonia arcjet would be good choices.

True, which is why I did not *completely* rule out the
option. Of especial note is that such efforts come with
rather sizeable down sides (and price tags). Low thrust
propulsion requires nearly continual operation to maintain
a low orbit. This comes at a cost in either
sophisticated scan platforms, inconvenient pointing
constraints, degraded duty cycles or all of the above
(plus additional mass and complexity in power generation
and processing, e.g. solar arrays and PPUs et al).
Somehow, NRO manages to keep Hubble class optical
surveillance satellites in pretty low orbits for a
substantial period of time, so there might be a few
tricks to getting away with that. Though the constraints
for spysats are much different than those for space
science observatories, so the experience may not be
applicable (spysats never create long exposures like the
Hubble deep field, for example).

Of course, none of this comes for free, no matter how
well it works. If manned spaceflight were significantly
more routine and robust than it is today, it might be a
worthwhile tradeoff. However, if it were then it would
likely have the capabilities needed to service more
remote installations, so the concept has something of a
catch-22 built into it.