Limits of Spectroscopy

What is the faintest "source" that can be spectroscopically analysed
via a telescope for fraunhofer lines and elemental composition?

I know bright galaxies and quasars produce ample quantities of light
for spectroscopy, but surely the multiple stellar make-up of these
objects produces meaningless 'noise' at that level...

Ta.

Abdul Ahad

Ioannis wrote:
? "Abdul Ahad" ?????? ??? ??????
om...

What is the faintest "source" that can be spectroscopically analysed
via a telescope for fraunhofer lines and elemental composition?

I see no reason why it shouldn't be precisely the same magnitude
limit that exists for the larger scopes.

If it can be photographed, it can be spectrographed.

You're going to need a good few photons to be able to remove the
uncertainty as to peaks and troughs in the wavelength profile. In this
recent story:
http://www.astronomynow.com/news/040...t_galaxy.shtml
I imagine they had enough photons to calculate an approximate blackbody
curve to determine the redshift, but I very much doubt they could tell
you what the composition of the galaxy was!
--
Andrew Urquhart
Reply: www.andrewu.co.uk/about/contact/

In sci.astro.amateur, Abdul Ahad wrote:
What is the faintest "source" that can be spectroscopically analysed
via a telescope for fraunhofer lines and elemental composition?

How big is the telescope? What is the throughput and resolution of the
spectrograph, the QE of the detector, and how long are you willing to
integrate on the target? What is the FWHM of the seeing, and how bad
is the light pollution or sky background in your wavelength range?

It's not a simple question.

I know bright galaxies and quasars produce ample quantities of light
for spectroscopy, but surely the multiple stellar make-up of these
objects produces meaningless 'noise' at that level...

In most cases you can't resolve individual stars, so a spectrum of
the galaxy consists of the blended contribution from various types of
stars. That's not "noise" though.

Quasar spectra are quite different from stars, and it is often simple
to distinguish the two sources and subtract out the stellar contribution
to the spectra.

Dave

David Whysong
DWhysong (at) physics (dot) ucsb (dot) edu

In sci.astro.amateur, Ioannis wrote:

"Abdul Ahad"
What is the faintest "source" that can be spectroscopically analysed
via a telescope for fraunhofer lines and elemental composition?

I see no reason why it shouldn't be precisely the same magnitude limit that
exists for the larger scopes.

Spectrographs usually have much less throughput than cameras; there are
losses due to the grating (or grism) and re-imaging optics. Furthermore,
the light is spread over many pixels, each of which has read noise and
dark current.

So you do lose substantial SNR with a spectrograph, compared to imaging.

Dave

David Whysong
DWhysong (at) physics (dot) ucsb (dot) edu

In message , Ioannis
writes

? "Abdul Ahad" ?????? ??? μ???μ?
. com...

What is the faintest "source" that can be spectroscopically analysed
via a telescope for fraunhofer lines and elemental composition?

A devious answer to this is that we can analyse clouds of hydrogen gas
that happen to be in the line of sight between us and a distant quasar
no matter how faint the cloud itself may be. The light from the quasar
allows us to see the composition and redshift of the intervening gas
cloud.

I see no reason why it shouldn't be precisely the same magnitude limit that
exists for the larger scopes.

If it can be photographed, it can be spectrographed.

You have it exactly backwards. Anything that a spectrograph will work on
will be plenty bright enough to photograph. The main reason the biggest
scopes are in demand is to feed enough light into spectrographs.

Imaging a star requires obtaining enough signal to noise on a handful of
pixels. Imaging a spectrum can require a few thousand times more SNR.

Try to make a spectrum with 0.1nm resolution across the visible to look
in detail at Fraunhofer lines and that same amount of starlight is
spread across 4000 pixels with consequent loss of limiting magnitude
(roughly about 9 mags ignoring other secondary factors).

Fortunately you don't always need very high resolution spectra to see
interesting things - some of the wilder active objects have bright
emission lines that stand out even at relatively low dispersion.

There are amateur spectra of some brighter quasars and Seyferts about.

Regards,
--
Martin Brown

(Abdul Ahad) wrote in message . com...

I know bright galaxies and quasars produce ample quantities of light
for spectroscopy, but surely the multiple stellar make-up of these
objects produces meaningless 'noise' at that level...

Not at all! First of all, most of the light from quasars is *not*
from multiple stars, but from the active core itself -- a single
source. Second, it doesn't matter if the light of a galaxy comes
from multiple stars, as long as those stars are all doing the same
thing. When you are measuring red shift, you are looking for certain
spectral lines. Assuming no radial motion to or away from Earth,
all stars in the universe will put those lines in exactly the
same place, so the spectrum of a galaxy would look more or less
like the spectrum of a single star multiplied by a few billion.

Now in fact, galaxies rotate, so that even if the galaxy as a whole
is at rest w.r.t. the Earth, the spectral lines are "smeared" due
to the fact that the stars at one edge are moving towards Earth
and the stars at the other edge away. But when you are looking
at cosmologically significant distances, galaxies as a whole are
moving away from Earth at a large fraction of the speed of light,
which is *far* greater than the rotational speed of a galaxy.

As for other people's comments, barring the ability to measure the
energy of individual photons -- which is *not* currently possible
in the visible spectrum -- of course you need more light to do
spectroscopy than to do simple photography. How much more depends
on how finely you want to resolve those spectral lines.

- Tony Flannders

On Thu, 04 Mar 2004 07:19:13 GMT, David Whysong
wrote:

In sci.astro.amateur, Ioannis wrote:

"Abdul Ahad"
What is the faintest "source" that can be spectroscopically analysed
via a telescope for fraunhofer lines and elemental composition?

I see no reason why it shouldn't be precisely the same magnitude limit that
exists for the larger scopes.

Spectrographs usually have much less throughput than cameras; there are
losses due to the grating (or grism) and re-imaging optics. Furthermore,
the light is spread over many pixels, each of which has read noise and
dark current.

So you do lose substantial SNR with a spectrograph, compared to imaging.

SNR = signal to noise ratio

Dave

David Whysong
DWhysong (at) physics (dot) ucsb (dot) edu

On Thu, 04 Mar 2004 07:16:29 GMT, David Whysong
wrote:

In sci.astro.amateur, Abdul Ahad wrote:
What is the faintest "source" that can be spectroscopically analysed
via a telescope for fraunhofer lines and elemental composition?

How big is the telescope? What is the throughput and resolution of the
spectrograph, the QE of the detector, and how long are you willing to
integrate on the target? What is the FWHM of the seeing, and how bad
is the light pollution or sky background in your wavelength range?


QE= Quantum Efficiency the energy needed to split off measurable
electrons. with perfect efficiency, one photon would split off one
electron. there's inherent inefficiency in this transfer based on
work function and black body properties of the detector material,
amongst other reasons.

FWHM = Full Width at Half Maximum it applies to a well shaped peak
on a graph. its the point halfway up to the maximum of the peak at
the theoretical fattest part of that peak. its the value of the
weighted average of the peak.


It's not a simple question.

I know bright galaxies and quasars produce ample quantities of light
for spectroscopy, but surely the multiple stellar make-up of these
objects produces meaningless 'noise' at that level...

In most cases you can't resolve individual stars, so a spectrum of
the galaxy consists of the blended contribution from various types of
stars. That's not "noise" though.

Quasar spectra are quite different from stars, and it is often simple
to distinguish the two sources and subtract out the stellar contribution
to the spectra.

Dave

David Whysong
DWhysong (at) physics (dot) ucsb (dot) edu

(Abdul Ahad) wrote in message . com...
What is the faintest "source" that can be spectroscopically analysed
via a telescope for fraunhofer lines and elemental composition?

That would depend, if you float in space with a big aperture telescope
and 'stare' at a distant object for several days then the telescope is
capturing quite a lot of light which can be used to learn from.

Like amatuers who use several minute (or more) exposure cameras to
produce the excellent images occasionally posted on their websites.

I know bright galaxies and quasars produce ample quantities of light
for spectroscopy, but surely the multiple stellar make-up of these
objects produces meaningless 'noise' at that level...

I can imagine a rough estimate of the general proportions of materials
in the distant galaxy can be made, though not of individual stars.

(Abdul Ahad) wrote in message . com...
What is the faintest "source" that can be spectroscopically analysed
via a telescope for fraunhofer lines and elemental composition?

I know bright galaxies and quasars produce ample quantities of light
for spectroscopy, but surely the multiple stellar make-up of these
objects produces meaningless 'noise' at that level...

Ta.

Abdul Ahad

The distance to galaxies is determined by spectroscopic analysis.
I think the record is right at 30M; Keck verified primevil galaxies
found in one of the hubble deep fields. The big telescopes are
routinely doing spectro work at M27. This spectro work is concerned
mainly with the Lyman Alpha emissions or other ultraviolet emissions
that have been red shifted down to optical wavelengths. The detection
of the pattern is all that is necessary to determine red shift.
These are typically done on meduim resolution spectragraphs.

High resolution spectro work to determine elemental composition is
being done down to M26. Individual stars in M31 and other rather
near galaxies has been acomplished. These are done on high resolution
(see Echelle) spectrographs. Resolutions in the 0.1 Angstrom range
are routine.

High resolution work to determine shifting spectral content due to
large planets is not at this time photon limited as all the stars
are quite bright reletive to the redshift analysis stuff. The problem
for this analysis is stability of the receiving equiptment {scope
plus spectrograph} and the stability of precision reference line
sources. Resolution in the 20 Hz range allows detection of Jupiter
sized objects to several thousand light years. The stability of
the reference line source must be significantly smaller than
the spectral resolution for years at a time. Several planets have
been confirmed with amateur spectra on 20" sized telescopes.

As to noise: To the spectrograph, if photons are allowed to pass
the 'slit' the spectrograph images the spectra. So if the slit is
set at 1 arc second, the spectrograph creates a spectrum of
everything that passes the slit. Sometimes it becomes aparent
that spectral lines from several sources have been imaged at the
same time. In the early days of spectroscopy (turn of previous
century) many binary stars were determined to be binary by this
technique.