Gratings, Dispersion and magnitudes....

[BillW]

Hi,

I've been getting emails about what is the best grating to use. The answer is there is no easy answer!
It depends on MANY variables but last night I caught the same meteor on two very similar camera systems last night.

Both were Watec 910's with 12mm f0.8 lenses. One had a 600g/mm and the other an 830g/mm. In theory the 830g/mm should give ~1.4 times greater dispersion than the 600g/mm
...and indeed it does!

Graph 1 is the 600g/mm which has a dispersion of ~0.94 nm/pix (note this is not the resolution achieved)



Graph 2 is the 830g/mm which has a dispersion of ~0.66 nm/pix



They have been given identical processing and it is clear the increased dispersion has the consequence of "stretching" the light more so the available light per pixel diminshes and information is lost (look at the zone between the two main lines, 590nm and 777nm). For the 830g/mm to record the same s/n as the 600g/mm the meteor would need to be brighter (another magnitude at least).

High dispersion/greater resolution is the ideal but there is a practical penalty for this viz fewer meteors will be caught.

A graphic (both literal and metaphorical ) illustration of the compromises needed for meteor spectroscopy.

Cheers,
Bill.

[stewartw]

Hi Bill,

Many thanks for posting this - as you say, a picture tells a thousand words and this is a great demonstration of the importanc of the signal:noise ratio.

Working on the assumption that partial spectra or those obtained at a low dispersion angle are of limited use, has anyone ever considered a stacked grating configuration where there is a (for example) 600 g/mm above a 830 g/mm (note, above as opposed to in front)?

Half of the recorded spectra would therefore be on the 600 grating and the other half on the 830 grating thus getting the best of both worlds.

Cheers

William

[BillW]

Hi,

A Polish group tried this with crossed plastic grating sheets. What is does do is allow spectra from almost any direction to be caught with almost equal dispersion.
They got a great result from a Perseid fireball with this technique.
The problem there is that the second dispersion has already undergone a dispersion and so would be fainter, thus it still needs very bright meteors to work.
I'm afraid it a law of diminishing returns.
Cheers,
Bill.

[stewartw]

Hi Bill,

I should perhaps have explained myself better. The proposal would be to NOT overlap the gratings but to have have of the FOV (eg the top) covered with the 600 g/mm grating and the other half (eg the bottom) covered with the 830 g/mm grating.

Cheers

William

[BillW]

I see what you mean. Interesting thought but I'm not sure how practical it would be. What if the meteor only fell in the part of the fov covered by one or other of the gratings. There would be no net gain.
However if it did happen to cross both gratings then it might well work.
I think the key to success was what we were talking about. With multiple stations working with a variety of gratings on overlapping fields then in principle one of the stations will pick up the "best" image for that meteor.

To paraphrase a well used expression from MANY IMC's "we need more data...!" (and observers and cameras and ...)

Cheers,
Bill.

[arp]

Hi Bill,

Thanks for posting this mini-tutorial. As William commented, I can read page after page of information on a topic but it usually takes a clear worked example - with explanatory text - for the penny to drop. :-)

Looking forward to seeing more results from your twin system.

Cheers,

Alex.

[BillW]

Hi,

I got this the other night on the 830g system. This was actually from the "faint" side of the dispersion pattern, that is, not the brighter side that the blaze produces.
To get a signal on the fainter side hints that this was quite a bright one.

It's not brilliant as a spectrum but it gives a better demonstration of what the greater resolution does.



The key issue is the nice group of lines in the green. Apart from the stongest green line which is magnesium the rest are iron and they're reasonably well resolved.
(The strong blue line is ionised magnesium at 448.1nm)
The measured resolution is 0.8nm/pix. This is a bit less than the first example probably due to lower s/n and seeing effects as well as a contribution from optical distortion towards the edge of the fov. In practical terms the full width at half max (FWHM) is 2.3nm for the green magnesium line. It should be remembered though that this "line" is actually three unresolved lines between 516.7nm and 518.3nm.

cheers,
Bill.