Why Hubble was cancelled, and what to do now

Since Ken Arromdee fairly asked what really killed Hubble, I found an
honest-looking answer in the notes from a meeting between Sean O'Keefe
and Hubble's managers at the Space Telescope Science Institute:

http://www.stsci.edu/institute/sm4meeting.html

My summary: O'Keefe cancelled the Hubble service mission to satisfy two
masters, President Bush and the CAIB, neither of which were anti-Hubble.
Bush set a limit to the shuttle program, while CAIB demanded new safety
requirements. If you fit ISS service flights between Bush and CAIB,
there is no room left for the Hubble flight. Hubble would have required
special safety rituals and NASA can no longer afford them.

Even though everyone realizes that it sucks, the Hubble service flight
was sacrificed to politics. The space station is a political caricature
of science, the shuttle EVA repair rules are a political caricature of
safety, and the Bush plan is a political caricature of a space agency
plan. Together they killed Hubble.

So what to do now? If you like space telescopes, do not cry for
Hubble. Instead of asking Washington for a reprieve, you should ask
for guaranteed funding for the James Webb Space Telescope. After all,
it used to be called the Next Generation Space Telescope, meaning the
next generation after Hubble. Each of the 18 main mirrors of JWST
will be about the same size as the entire main mirror of Hubble.
Here is a rendering:

http://www.stsci.edu/jwst/overview/i...lls_SK_sm1.jpg

This is the reflection telescope that astronomers really want.

Granted, the "save the Hubble" camp argues that JWST is an infrared
telescope, not optical. But why should it be? The optical band is
one of the few narrow bands for which the atmosphere is transparent.
(It's opaque to near infrared, incandescent in far infrared, and badly
translucent beyond near ultraviolet.) Astronomers are planning a
ground-based mirror, the Thirty-Meter Telescope, with 25 times the area
of JWST. It should be better than Hubble in the optical range.
--
/\ Greg Kuperberg (UC Davis)
/ \
\ / Visit the Math ArXiv Front at http://front.math.ucdavis.edu/
\/ * All the math that's fit to e-print *

In article ,
Greg Kuperberg wrote:
Granted, the "save the Hubble" camp argues that JWST is an infrared
telescope, not optical. But why should it be? The optical band is
one of the few narrow bands for which the atmosphere is transparent.

Kind of, sort of, loosely speaking. There are still major advantages to
being above it.

And I notice you don't mention ultraviolet.

Astronomers are planning a
ground-based mirror, the Thirty-Meter Telescope, with 25 times the area
of JWST. It should be better than Hubble in the optical range.

Better for some things, not for everything.

And it won't do UV.
--
MOST launched 30 June; science observations running | Henry Spencer
since Oct; first surprises seen; papers pending. |

In article ,
Henry Spencer wrote:
In article ,
Greg Kuperberg wrote:
Granted, the "save the Hubble" camp argues that JWST is an infrared
telescope, not optical. But why should it be? The optical band is
one of the few narrow bands for which the atmosphere is transparent.
Kind of, sort of, loosely speaking. There are still major advantages to
being above it.

Sure, loosely speaking. I guess I was speaking around the real
point: The narrow optical band is very relevant for human physiology,
and therefore for popular astronomy. It isn't very special for
research-level astronomy. Moreover, it's one of the more favorable
bands for ground-based telescopes, because people and telescopes are
adapted to the same atmosphere. In this sense, the popular and practical
attributes of a space telescope are contradictory.

And [TMT on the ground] won't do UV.

Yes, Hubble is a great UV telescope. I suppose that that capability
might not be replaced for a long time. Still, the infrared JWST is
what the STScI people chose as the "next generation". I'm sure that
they had good research reasons for that.

(Actually "infrablue" might be a better term for JWST. Supposedly its
detectors will extend through green light.)
--
/\ Greg Kuperberg (UC Davis)
/ \
\ / Visit the Math ArXiv Front at http://front.math.ucdavis.edu/
\/ * All the math that's fit to e-print *

(Greg Kuperberg) wrote


(Actually "infrablue" might be a better term for JWST. Supposedly its
detectors will extend through green light.)

600 nm as of last word.

In article ,
Allen Thomson wrote:
(Greg Kuperberg) wrote
(Actually "infrablue" might be a better term for JWST. Supposedly its
detectors will extend through green light.)
600 nm as of last word.

600 nm is labelled as yellow he

http://www.gamonline.com/catalog/col...y/spectrum.gif

According to this spec page, the different JWST cameras will be sensitive
from 28,000 nm to 600 nm:

http://www.stsci.edu/jwst/overview/design/

Actually this is an unclear point for me. The main JWST camera, NIRCam,
appears to be almost convenient enough for the gee-wow color photographs,
but not quite. It will be sensitive to 600 nm - but why not extend
that to 360 nm for the sake of public relations? I suppose that they
could make false-color pictures by permuting the spectrum, and maybe
they would look even better than true-color. But NIRCam also only has
two channels, rather than 3. Maybe they are used to faking three
independent colors with two? Or maybe for public relations they plan
to use the other camera, MIRI, which has 12 filters.
--
/\ Greg Kuperberg (UC Davis)
/ \
\ / Visit the Math ArXiv Front at http://front.math.ucdavis.edu/
\/ * All the math that's fit to e-print *

Greg Kuperberg wrote:

Actually this is an unclear point for me. The main JWST camera, NIRCam,
appears to be almost convenient enough for the gee-wow color photographs,
but not quite. It will be sensitive to 600 nm - but why not extend
that to 360 nm for the sake of public relations?

Perhaps the optics may not be good enough for high resolution at 360 nm?

Paul

In article ,
(Greg Kuperberg) wrote:

Granted, the "save the Hubble" camp argues that JWST is an infrared
telescope, not optical. But why should it be? The optical band is
one of the few narrow bands for which the atmosphere is transparent.
(It's opaque to near infrared, incandescent in far infrared, and badly
translucent beyond near ultraviolet.) Astronomers are planning a
ground-based mirror, the Thirty-Meter Telescope, with 25 times the area
of JWST. It should be better than Hubble in the optical range.

But wouldn't that increase the bill for the JWST (if only by requiring
NASA to redefine the scope of what the JWST can do), which in turn would
require NASA to justify extra funding, funding that would presumably be
outside the scope of the Bush plan?

In any case, I notice that the JWST only has a 5-10 year lifespan
(versus 20+ years for Hubble). Since the JWST is presumably not being
designed to be serviced in situ, when it expires that will the end of
the JWST. If astronomers want to keep having a telescope in space they
will need to build and launch another one. Which means that the funding
would need to be found to build and launch that other one.

--
Stephen Souter

http://www.edfac.usyd.edu.au/staff/souters/

In article ,
Stephen Souter wrote:
In article ,
(Greg Kuperberg) wrote:
Astronomers are planning a
ground-based mirror, the Thirty-Meter Telescope, with 25 times the area
of JWST. It should be better than Hubble in the optical range.
But wouldn't that increase the bill for the JWST...

Rather, the TMT telescope should be better than Hubble in the optical
range - albeit only calm days. JWST will be much better than Hubble in
the range that it sees, namely far infrared to yellow light according
to the current spec. I think that part of the point of JWST is that its
shifted frequency range, which is actually broader than that of Hubble on
a logarithmic scale, is more interest to astronomers than Hubble's range.

In any case, I notice that the JWST only has a 5-10 year lifespan
(versus 20+ years for Hubble). Since the JWST is presumably not being
designed to be serviced in situ,...

First of all JWST will be farther away than the moon. Thus, so-called
"service missions" are out the window. Again, the wisdom of the
astronomy community is that LEO sucks for most space telescopes, certainly
for infrared ones.

Second, the "service mission" plan for Hubble turned out to be a
great Faustian bargain for astronomers. It was pursued largely to
the exclusion of launching other space telescopes. Hubble is like an
old computer data center with a 20-year service contract. Sure, they
might get upgrade the memory and even the CPU now and then. Of course
what people really want is completely new computers every few years.
And likewise astronomers want completely new telescopes.

It may be counterintuitive for a lot of people here to think of a
billion-dollar spacecraft as disposable. Well, even on the ground
each giant new particle accelerator makes the old one almost obselete.
And CPU chip factories face a similar fate.
--
/\ Greg Kuperberg (UC Davis)
/ \
\ / Visit the Math ArXiv Front at http://front.math.ucdavis.edu/
\/ * All the math that's fit to e-print *

In article ,
Greg Kuperberg wrote:
First of all JWST will be farther away than the moon. Thus, so-called
"service missions" are out the window...

If you've got a system that could do a servicing visit in lunar orbit,
then with a bit less propulsion and a bit more life support, it could do
one at L2. It's significantly hard but not ridiculous, if the cost is
low enough.

Again, the wisdom of the
astronomy community is that LEO sucks for most space telescopes, certainly
for infrared ones.

If nothing else, most telescopes are better off without Earth filling half
the sky and getting in the way constantly. Whether that's enough to
justify going farther out depends on details, and in particular on whether
there are other reasons as well. Certainly there are for infrared (colder
environment) and for high-energy astronomy (less particle noise from the
Van Allen belts).

...Hubble is like an
old computer data center with a 20-year service contract. Sure, they
might get upgrade the memory and even the CPU now and then. Of course
what people really want is completely new computers every few years.
And likewise astronomers want completely new telescopes.

Keck must be on its last legs by now, to say nothing of ridiculous
antiques like almost everything on Mount Palomar.

Only in space is it accepted that you throw out a hugely expensive
observatory after a few years of use. Anywhere else, it may not be
right at the leading edge any more, but it remains in service.

It may be counterintuitive for a lot of people here to think of a
billion-dollar spacecraft as disposable. Well, even on the ground
each giant new particle accelerator makes the old one almost obselete.

On the contrary. The new ones are built as extensions of the old ones
whenever possible -- e.g., the Tevatron is fed by the original Fermilab
ring (and yes, there have been proposals to do similar things with
Hubble!) -- and even when that's not done, often the older machines are
rebuilt for secondary roles.
--
MOST launched 30 June; science observations running | Henry Spencer
since Oct; first surprises seen; papers pending. |

"Paul F. Dietz" wrote in message ...
Greg Kuperberg wrote:

Actually this is an unclear point for me. The main JWST camera, NIRCam,
appears to be almost convenient enough for the gee-wow color photographs,
but not quite. It will be sensitive to 600 nm - but why not extend
that to 360 nm for the sake of public relations?

Perhaps the optics may not be good enough for high resolution at 360 nm?

JWST's mirror will be actively controlled. In order to
have a decent optical mirror it must be within a small
tolerance of the ideal optical shape. The deviation of
a mirror from the ideal is often quoted as a fraction of
some wavelength (usually representing the average
wavelength used in observations with the instrument).
For a recreational or backyard telescope deviations of
around 1/4 Lambda (wavelength) are usually acceptable.
Lambda/8 would be decent, Lambda/12 or Lambda/16 would
be research grade. For an actively controlled mirror
surface there are huge costs to pay for choosing a
smaller Lambda (going for visual or ultraviolet instead
of IR). First, each actuator has to be all that much
more fine, for near-IR to visible light that can make
a difference of around a factor of 2 or so, which is
no small thing for actuators which are already
insanely precise and expensive. Second, you've got to
increase the actuator density to shore up the "sagging"
in between actuators (though it might be "sagging" too
high or too low). The number of actuators needed
scales roughly with the inverse square of Lambda.
Decrease Lambda by 2 (1,000 nm to 500 nm, for example)
and you multiply the number of actuators needed by
about 4, and they have to be 2x as fine. That drives
up cost and complexity, and also drives up the areal
mass of the mirror. Which is a rather big issue,
sadly, with the launch capabilities we have today.

Especially since JWST is developing this technology
as it goes, it makes a hell of a lot of sense to shoot
for a lower goal first and, maybe, let incremental
improvements and evolution take the technology to
where it might prove useful in visible and UV
wavelengths later on.