Re: Relativistic transit through another solar system

Thomas Womack <twomack@xxxxxxxxxxxxxxxxxxxxxx> writes:

>
> OK, but that still makes the mechanics and the optics of the
> telescopes about about thirty to a hundred times as cumbersome and
> expensive (I think both metrics scale as N^2.5 or so) as equivalent
> resolution in the optical would be. I don't know how much you save
> by needing to have the curves accurate to 200nm rather than to 50nm.

Are you saying that it is much harder to build a telescope to work at
2.2 microns as it is to work at 0.5 microns? This confuses me. No
decent sized telescope that has been built in the last probably 40
years lacks an IR imager or spectrograph. Heck, Gemini south is
optimized for the IR, and was cheaper than Keck because of the smaller
mirror. The near IR is basically the optical as far as mechanical
design and optics are concerned. The instrumentation is slightly more
expensive but everything else is so much cheaper in the near IR than
in the visible, you more than win.

> I got the impression that emission lines from elements tend to be
> unavoidably in the optical (OK, everything including Lyman alpha is in
> the infra-red if you look far enough away, but I'm wearing my
> geologist rather than my astrophysicist hat at the moment); though
> with the geologist hat, I suppose mineral spectra tend to be
> reasonably distinctive in the IR.

The availability of lines is rather dependent on a lot of things. If
you want to take pretty pictures of cold objects, however, you might
want to move your passband to the part of the spectrum they peak at.
A gas giant is not going to be a big emitter of [OII] or OIII. The
optical is too cold for Iron lines and most carbon features are in
absorption. Molecules dominate that near IR and if you want to look
at gas giants, well, that is where it is at.


> I appreciate that optical interferometry needs very clean wavefronts,
> but I don't see that the individual unit telescopes need to be so big
> that it's impractical to get them diffraction-limited (ah, sunlight at
> Jupiter is about magnitude 23 per square arc-millisecond, so at that
> level of resolution there may be trouble, but an arc-millisecond on
> Jupiter from Earth is better than the best spacecraft images).

The dominant cost for interferometry is not in the size of the
telescope. The Keck interferometry program is going to be expanded
by using two additional 1 meter telescopes, but the optical bench to
interfere the images is the main expense, well, that and they need a
working adaptive optics system before they even do the interfering.
All of this pushes development to the near IR.

> On the other hand, I'd be entirely satisfied with an example of
> infra-red interferometric _imaging_ ('behold Io, note that the disc
> comes out circular and these three volcanos are clearly separated' is
> not so bad a target), and I haven't seen interferometric imaging in
> anything shorter than millimetre-wave.

Would you settle for an AO image? Probably not....

No one is doing interferometry with more than a couple of telescopes.
The near IR is not like radio astronomy where you just have delay
lines, a good digital correlator and can do image reconstruction in
software after taking the data. Correlating multiple telescopes with an
actual optical bench is much more difficult.

--
Bradford Holden
"Is that even legal?" CMR on the use of a turn signal.
.

Relevant Pages