Space & Astronomy/Space and Astronomy (October 1993).iso/mac/TEXT/SPACEDIG/V11_5/V11_597.TXT

Internet Message Format | 1991-07-08 | 20KB

Return-path: <[e-mail]>
X-Andrew-Authenticated-as: 7997;andrew.cmu.edu;Ted Anderson
Received: from beak.andrew.cmu.edu via trymail for +dist+/afs/andrew.cmu.edu/usr11/tm2b/space/[e-mail] (->+dist+/afs/andrew.cmu.edu/usr11/tm2b/space/space.dl) (->ota+space.digests)
          ID </afs/andrew.cmu.edu/usr1/ota/Mailbox/oaXNjsi00VcJ43Fk51>;
          Sun,  1 Jul 1990 02:45:15 -0400 (EDT)
Message-ID: <[e-mail]>
Precedence: junk
Reply-To: [e-mail]
From: [e-mail]
To: [e-mail]
Date: Sun,  1 Jul 1990 02:44:36 -0400 (EDT)
Subject: SPACE Digest V11 #597

SPACE Digest                                     Volume 11 : Issue 597

Today's Topics:
        Re: The HST was risky from square one
         Re: Bringing the HST down -- even possible?
         Re: Bringing the HST down -- even possible?
    Re: Image reconstruction.  (Was Re: HST focus problem)
               Re: model rocket staging
        Re: Testing of Hubble Space Telescope
         Re: Hubble Space Telescope Update - 06/28/90
         Re: Hubble Space Telescope Update - 06/28/90
                  Palo Alto
             Re: Space telescopes
           Re: NSS protests Chinese launch pricing
        Re: Testing of Hubble Space Telescope
    Re: Image reconstruction.  (Was Re: HST focus problem)

Administrivia:

    Submissions to the SPACE Digest/sci.space should be mailed to
  [e-mail].  Other mail, esp. [un]subscription notices,
  should be sent to [e-mail], or, if urgent, to
             [e-mail]

----------------------------------------------------------------------

Date: 30 Jun 90 15:43:23 GMT
From: uwm.edu!zaphod.mps.ohio-state.edu!brutus.cs.uiuc.edu!ux1.cso.uiuc.edu!aries![e-mail]  (Doug McDonald)
Subject: Re: The HST was risky from square one

In article <[e-mail]> [e-mail] (Greg Kuperberg) writes:
>It may seem now as if the Hubble Telescope is a priceless asset to this
>nation and to science, and some bozos down at Perkin-Elmer have nearly
>ruined it.  It may look like the proper course of action is to form a
>marauding lynch mob that will terrorize NASA for the next two years.

The bozos aren't at Perkin Elmer - they are at NASA itself. It was
just plain stupid to not test the components that could not be fixed.

>
>But let's face it:  Launching satellites is a risky business.

Then why didn't they test the components? In a way that would have discovered
a very serious very  large error?

The answer does not need a fancy review board. It doesn't have to be qualified
by some sort of mealy mouth "hindsight is clearer than foresight" statement.
It is obvious and straightforward: pure stupidity and or incompetence.



Doug McDonald

------------------------------

Date: 30 Jun 90 16:55:25 GMT
From: cs.utexas.edu!titan![e-mail]  (Donald Heskett)
Subject: Re: Bringing the HST down -- even possible?

That's a very good point; coming back down, the only delta-v burden
imposed by the HST is that of slowing it enough to get it into the
atmosphere for reentry.

Much of the concern about recovering it involves other details.
Getting it back into the cargo bay and lashed down is almost certainly
going to be tougher than deploying it was.  Also, there is some
justified concern about landing with this very heavy and delicate
payload. Optical contamination by the shuttle's thrusters and other
volatile junk it outgasses is also a problem.  Apparently, long before
the current problems cropped up, the HST people had decided not to
bring it down for periodic refurbishment, as originally planned, and
to keep on-orbit servicing down to the absolute minimum.

My guess is that a 'patch-up' mission is a year to a year-and-a-half
away. In the mean time, I'd guess, emphasis will be on the non-imaging
instruments, which will get a much larger fraction of the observing
time than originally planned, to maximize the science return.  I'd
guess, from the comments of the image deconvolution experts that we'll
still get some impressive images, enough to quell the public outcry,
but at the expense of longer exposure times, much computer processing,
and 'objectivity'.  I emphasize that all of this is just 'educated' guesses.

------------------------------

Date: 1 Jul 90 00:18:37 GMT
From: usc!cs.utexas.edu!news-server.csri.toronto.edu!utgpu!utzoo![e-mail]  (Henry Spencer)
Subject: Re: Bringing the HST down -- even possible?

In article <[e-mail]> [e-mail] (Steve Nuchia) writes:
>>Two dedicated shuttle missions.  (You won't be able to piggyback another
>>payload for a maximum-altitude mission, and HST fills the *whole* payload
>>bay for all practical purposes.) ...
>
>Seems to me that the "maximum altitude" is the altitude reached carrying
>the HST up.  You have to have enough fuel to get back with the extra
>mass again, but you should be able to carry up a large fraction
>of HST's mass on the retrieval mission...

You can carry a fair bit of mass, *if* it involves essentially no
obstructions in the payload bay.  Most substantial payloads also have
substantial cradles that can't easily be moved out of the way.

On reflection, though, I can think of one big mission that doesn't need
a cradle, isn't enormously massive, would sort of like a high altitude,
and may have trouble getting a launch slot in the ordinary run of events:
deploying LDEF a second time.
--
"Either NFS must be scrapped or NFS    | Henry Spencer at U of Toronto Zoology
must be changed."  -John K. Ousterhout |  [e-mail]   utzoo!henry

------------------------------

Date: 1 Jul 90 00:26:31 GMT
From: bfmny0![e-mail]  (Tom Neff)
Subject: Re: Image reconstruction.  (Was Re: HST focus problem)

In article <[e-mail]> [e-mail] (Greg S. Hennessy) writes:
>In my opinion, an investagation should be made, but not to
>run someone out of town on a rail, but to learn from the mistake, and
>prevent it from happening again.

Oh, at these prices, I don't think we need to worry ourselves about it
happening again... :-(

--
"We walked on the moon --       ((      Tom Neff
     you be polite"              ))     [e-mail]

------------------------------

Date: 30 Jun 90 21:39:20 GMT
From: unhd![e-mail]  (Robert M. Kenney)
Subject: Re: model rocket staging

In <[e-mail]> [e-mail] writes:
>        The primary problem is going to be building a model that can withstand
>the aerodynamic pressures without being reduced to its component parts.  If you
>use metal in the structure, I believe it ceases to become a model in the eyes
     ^^^^^
>of the NAR (but I'm really just guessing, so no flames, please.)  If anyone can
>build and fly a paper and balsa rocket past the speed of sound, I'll print this
         ^^^^^^^^^^^^^^^
>message and eat it.  :-)


    How about fiberglass and/or carbon-fiber composites?

>                                -Thomas D. Blersch
>                                 [e-mail]  (I think I've got
--
-----------------------------------------------------------------------
Robert M. Kenney                uucp:    uunet!unhd!rmk
USNH Computer Services                Nearnet: [e-mail]
Kingsbury Hall, UNH                BITNET:  R_KENNEY@UNHH

------------------------------

Date: 30 Jun 90 15:39:03 GMT
From: maytag!looking![e-mail]  (Brad Templeton)
Subject: Re: Testing of Hubble Space Telescope

I can believe that a full sytem test of the HST would have been too expensive.

But I refuse to believe that a test that would have detected real BONEHEAD
errors like a 1/2 wave curvature error would have been expensive.  In fact,
any astronomer with mirror making experience could have done it for $10, given
a hallway and a chance to sit at the focus, couldn't they?
--
Brad Templeton, ClariNet Communications Corp. -- Waterloo, Ontario 519/884-7473

------------------------------

Date: 30 Jun 90 19:03:59 GMT
From: portal!cup.portal.com![e-mail]  (Paul L Schauble)
Subject: Re: Hubble Space Telescope Update - 06/28/90

>  -- the spooks seem to have space telescope testing under control.

Maybe we just haven't heard about their failures.

    ++PLS

------------------------------

Date: 30 Jun 90 06:02:05 GMT
From: vsi1!zorch![e-mail]  (Phil Gustafson)
Subject: Re: Hubble Space Telescope Update - 06/28/90

In article <[e-mail]> [e-mail] (George William Herbert) writes:
>In article <[e-mail]> [e-mail] (Robert McGwier) writes:
>>I don't believe this for a second when I KNOW of testing facilities that
>>EXISTED at the time and from yesterday's papers they were offered and
>>the SNOTS at NASA said `Not invented here.'  I haven't been this livid
>>since I watched seven people die on television.
>
>Nice idea, but it's not quite that simple.  You have intrinsic problems
>testing a flexible mirror in a gravity well when it's supposed to be
>operational in space.  I.E. it wouldn't have looked right even if we
>DID test it down here.

This morning's New York Times says that the spooks offered their
spy-telescope test facility to NASA, but the latter turned the offer
down.  Various unnamed spokesmen say that the facility could not have
verified the the HST was within spec, but could have detected the
gross aberration currently present.

I don't see much difference between testing a telescope designed to
point down and one designed to point up  -- the spooks seem to have
space telescope testing under control.  I'm just surprised that they
loosened up enough to deal with NASA.

The NASA claim that testing HST would cost "hundreds of millions of
dollars" boggles the mind.  Tens, maybe.
------------------------------------------------------------------------------
    Opinions outside attributed quotations are mine alone.
------------------------------------------------------------------------------
--
  |  [e-mail]          | Phil Gustafson
  |  (ames|pyramid|vsi1)!zorch!phil      | UNIX/Graphics Consultant
  |                     | 1550 Martin Ave., San Jose CA 95126
  |                     | 408/286-1749

------------------------------

Date: 30 Jun 90 03:49:17 GMT
From: attcan!utgpu!utzoo![e-mail]  (Henry Spencer)
Subject: Palo Alto

In article <[e-mail]> [e-mail] (john reece) writes:
>> I note that you're from somewhere called "Palo Alto".  What's the
>> American translation of that (Spanish) placename?
>
>It stands for "tall trees", which is rather strange, since the place was mostly
>grassland before it was settled.  It's about half an hour drive south of
>San Francisco...

I know where it is, actually.  My point was that its name is "Palo Alto"
in any language, even American, and OV-105's name similarly is "Endeavour"
even in countries that don't speak English. :-)
--
"Either NFS must be scrapped or NFS    | Henry Spencer at U of Toronto Zoology
must be changed."  -John K. Ousterhout |  [e-mail]   utzoo!henry

------------------------------

Date: 30 Jun 90 17:58:09 GMT
From: cs.utexas.edu!titan![e-mail]  (Donald Heskett)
Subject: Re: Space telescopes

Roughly five or six years ago a group announced a project to launch an
amateur space telescope.  It was to have been a 17", I think, and was
to use a shuttle "getaway special" launch.  The group was based at
Rutgers U., if I remember correctly, with scattered support at one or
two other universities.

The group appeared to be mostly composed of
undergrads, a few grad students, and a faculty advisor or two.  They
seemed to have problems with attrition due to graduation, and I'm
afraid the delay caused by the Challenger disaster did them in
(largely by graduating and scattering the most devoted members of the
organization, I'd guess).  In any case, I haven't heard from them or
seen anything about them in a couple of years.

The original announcement occured in "Sky and Telescope" and was
followed by one or two progress reports, if I remember correctly.

------------------------------

Date: 30 Jun 90 17:53:18 GMT
From: thorin!vangogh![e-mail]  (Jonathan Leech)
Subject: Re: NSS protests Chinese launch pricing

In article <[e-mail]> [e-mail] (Craig E. Ward) writes:
>In economic terms, what China is doing is called dumping.

    And we all know how effective anti-dumping action by the Feds is;
look at the DRAM situation over the last few years.
--
    Jon Leech ([e-mail])    __@/
    ``Scientific progress goes "Boink"?'' - Hobbes

------------------------------

Date: 30 Jun 90 21:08:52 GMT
From: uoft02.utoledo.edu![e-mail]
Subject: Re: Testing of Hubble Space Telescope

In article <[e-mail]>, [e-mail] (Brad Templeton) writes:
> I can believe that a full sytem test of the HST would have been too expensive.
>
> But I refuse to believe that a test that would have detected real BONEHEAD
> errors like a 1/2 wave curvature error would have been expensive.  In fact,
> any astronomer with mirror making experience could have done it for $10, given
> a hallway and a chance to sit at the focus, couldn't they?
> --
>

There are not that many astronomers that have made 2.4 meter mirrors.  You
are drastically oversimplifying things.  If it turns out that some
of the reports are true and there was available some very sophisiticated
militray testing equipment this still shoots a zillion holes in your statment.
If any astronomer could do it for 10$ why need such sophistcated equipment?
The mirriors were rigourously tested, but never in tandem and two things
that may look good alone don't necessarily look good in tandem.

Robert Dempsey
Ritter Observatory

------------------------------

Date: 30 Jun 90 22:11:54 GMT
From: uvaarpa!murdoch!fits![e-mail]  (Don Wells)
Subject: Re: Image reconstruction.  (Was Re: HST focus problem)

In article <[e-mail]> [e-mail] writes:
>.... 1. There will be less photons per pixel so that even perfect image
>reconstruction will only work with brighter objects.

Right conclusion but wrong reasoning.

The problem is not the number of photons per pixel, but the noise per
pixel. HST sees a light_of_the_night_sky that is only about 3x lower
than on ground in the visual range (I think sky may be darker in UV).
The limitation of all faint object imaging in optical astronomy is the
*noise* of the night sky background, more than shortage of photons.
The purpose of a big telescope is as much to get more night sky
photons so that you reduce relative noise by square root of N (Poisson
statistics) as it is to get more photons from the object. The really
big win for an HST with proper optics is that the photons of point
sources are *concentrated* in a smaller area and therefore compete
with the noise from fewer night sky photons.  HST is only slightly
better (due to 3x darker sky) than a 2.4m groundbased telescope for
extended objects with sizes greater than the typical seeing size of
groundbased telescopes.  So, with a PSF comparable to groundbased
telescopes HST has lost most of its advantage over groundbased imaging
in the visible. (Note, however, that imagig in UV, and maybe nearIR,
still has a special advantage even with resolution of 1arcsec.)

Your conclusion that deconvolution will help (it won't be "perfect",
of course) only for brighter objects is right (and they should be
spatially confined also); that is what Stockman told Senator Gore
yesterday on behalf of the STScI. But the reason is that the total
amount of sky noise competing with the signal in the deconvolution is
immensely greater because the object photons have been spread over
such a large area. You need lots of object photons (i.e., bright
object) to make headway against this noise.

>2. deconvolution inherently adds noise compared to data that don't need
>deconvolution.

Again, right conclusion (sort of), but wrong reasoning for modern
*nonlinear* deconvolution techniques.

It is, of course, true that deconvolution will inevitably produce
poorer results than would an instrument with proper PSF. So, I agree
with your conclusion (sort of). (It is, however, also generally true
that deconvolution can improve the results derived from the proper
PSF!)

The greatest challenge for any deconvolution algorithm is that it must
somehow separate the fragments of object signal from the noise, so
that it can (mostly) restore the object rather than the noise.  Good
algorithms do this. The really elegant ones (e.g., MEM) have proper
models of the noise and produce estimates which can be characterized
as "most likely" (in some statistical sense) to have produced the
data. One approach is to select the "maximally smooth" answer from the
infinity of possible answers to any deconvolution problem. It is
improbable that you would characterize an MEM estimate for an optical
image as having "added noise"; likewise, CLEANed radio synthesis
images are generally smoother than the raw "dirty map" images.

The classical linear techniques amplify high spatial frequency noise
at the same time that they amplify attenuated high frequency object
signals. I.e., they add noise, as you say. But nonlinear techniques
invoke physical constraints, usually positivity, to rule out a very
large fraction of the infinity of possible solutions. I.e., THEY
REDUCE NOISE RATHER THAN ADDING IT. For this reason only nonlinear
algorithms are used to any extent in astronomy.

>[e-mail] >James Jay Klavetter >Astronomy Program >U of MD
>College Park, MD 20742

The problem with HST's optics is *so* frustrating because it means
that we will have to wait for several more years for the science to be
produced from the high resolution imagery. But for deconvolution
specialists in optical astronomy I expect that the next 12-24 months
are going to be the glory years! I hope that the experience will finally
establish the credibility of deconvolution among optical astronomers.
If so, then in the later years of the HST mission we will see even
better imaging results than we would have if this stupid mistake had
never occurred.

My bet is that STScI will order its first vector/concurrent computer
before Christmas... [*]

Donald C. Wells, Associate Scientist | NSFnet: [e-mail] [192.33.115.2]
National Radio Astronomy Observatory | SPAN:   NRAO::DWELLS    [6654::]
Edgemont Road                        | BITnet: DWELLS@NRAO
Charlottesville, VA 22903-2475 USA   | UUCP:   ...!uunet!nrao.edu!dwells
Lat: 38:02.2N  Long: 78:31.1W        | Tel:[phone] Fax:[phone]

[*] Radio astronomers, whose science is critically dependent on
nonlinear deconvolution and self-calibration, own about 25 vector
computers worldwide.  The highest performance CPU is the Fujitsu VP-50
at Nobeyama Observatory.  My own institution, NRAO, has three Convex
C-1s now and anticipates replacing them with a 10-20x greater power
during the next two years with the post-processing budget of the Very
Long Baseline Array [VLBA] which is now nearing completion.  As far as
I know, no UV/optical/X-ray/IR astronomy organization has yet procured
a vector/concurrent CPU specifically for image processing (a few of
them have APs, but I have seen essentially no science results from
them).

------------------------------

End of SPACE Digest V11 #597
*******************