Re: Imaging Spacecraft [See August Sky & Telescope]

John Pike (
Fri, 28 Jun 1996 13:56:41 -0400

At 01:33 PM 6/28/96 -0400, Larry Klaes <> wrote:

>Is the following true:  That in the 1980s certain aspects of the US
>military intelligence network had ground-based telescopes that could
>count the rivets on an orbiting Soyuz?

This is *probably* not the case.

1 - As I recall, if one does the math on Hubble looking down, one is left
with the impression that such a thing would provide ~~10cm resolution at
ranges of interest.

2 - Until the advent of the Starfire facility at Phillips Lab in NM a few
years back, there was no Hubble-class telescope [known to be] used for
satellite characterization [and no particular reason to believe that there
was such a facility in the black world, though I could be wrong, I am at a
loss to imagine where it might be found].

3 - The Hubble comparison ignores the whole atmospheric compensation
problem, which is non-trivial, though addressable.

Work was done using laser guide "stars" in the late 1980s, which was able to
achieve nearly diffraction-limited seeing to overcome issue #3, but I don't
know of anything that would get you past objections #2 and #1 [there is no
particular reason to believe that CIA had a more competent spacetrack
capability than did the Air Force].

>Apparently they checked for
>NASA the number of missing tiles on Columbia's first space flight in 

Urban legends relate this episode both to using a KH-11 in space-to-space
imaging mode, as well as instruments at the Malabar Test Annex in Florida [I
have no direct knowledge in either event, and find neither account incredible].

But the issue with STS-1 was whether entire sections of tiles had ripped off
[loss of a single tile would not have been catastrophic] creating roughly
meter-size dark features that would have had very high contrast against
white tile surfaces. And in any event the *individual* STS tiles are of the
order of ~10-20 cm, and thus within Hubble equivalent capability, whereas
Soyuz bolts would seem to be an order of magnitude smaller.

Of course, the issue is much more complicated than this, given the wide
range of circumstances under which sub-resolution detection is possible [eg,
high contrast linear features such as the stripes down the middle of a road
are regularly detected at sub Ground Sample Distance scales] -- so one could
imagine that specular reflection from a regular array of bolts might be
interpreted as an array of bolts, given prior knowledge of the spacecraft

>And would it be wrong to conclude that they can do even better
>these days?

I don't *think* that they could do this well back then, but they are
certainly doing things with Starfire [and will do things with ADEOS] that
were previously only dreamed of.


John Pike
Director, Space Policy Project
Federation of American Scientists
307 Massachusetts Ave. NE
Washington, DC 20002
V 202-675-1023,   F 202-675-1024,