A basic science question about emission nebulae --Pelican, North American etc. Anything goes · Tim Hawkes · ... · 12 · 624 · 0

I was wondering about the total spectrum of light that comes from the gas and dust in many of these objects.  They are generally described as 'emission' nebulae because of the characteristic  specific visible lines corresponding to specific electron and spin transitions , HA, OIII, SII etc  from the ionised / excited H, O, S and N  ions that the nebulae comprise.

However does anyone know  - aside from the direct light from stars in the same field -  how much of the total light from these objects is also broadband?  Are they generally  purely emission objects   or does  a fair proportion of their total light also usually originate  from - for example -  scattering  from stars partly obscured behind dust (potentially giving orange hues)  and/or some reflection nebulosity (an obvious example there being the flaming star nebula in Auriga).

I ask because I have been comparing pure NB images with other images (of the Pellican) where I used an OSC camera with a UHC filter.  To my mind - the UHC image possibly provided better contrast of some of the dust lanes in some parts of the image plus of course the overall colour is different - perhaps 'truer'? --and I was wondering whether the extra bandwidth the UH filter encompasses maybe contributes to this?  I suppose the basic question is " Does one stand to miss any significant information --maybe colour or contrast? - by not having any broad band information mixed in with the NB?"

Tim

If someone did a spectrum analysis of light from one of these objects, it would tell you quickly what is being missed by strictly NB. Basically, everything outside the spectral lines of the gases is all the light you're missing doing NB.

If someone did a spectrum analysis of light from one of these objects, it would tell you quickly what is being missed by strictly NB. Basically, everything outside the spectral lines of the gases is all the light you're missing doing NB.

I was looking to see if there is somthing like that in the lit. So e.g. for the crab nebula - obviously very different a sn remnant - the spectrum shows the expected NB lines and a background of now - fading broadband bluish light from the pulsar synchrotron effect. However not yet found anything on more typical HA emission regions driven by UV from new stars. I am guessing the same way as Andrea that there should be broadspectrum emission from scattering, reflection etc but have not seen any studies on it? ***

andrea tasselli:
I think you hit the nail right on the head here Tim. Quite a number of them have some additional scattering/reflection/other-emission-lines contribution to the total light flux that strict NBs are going to miss and, if I can help it, I try to intereleave bradband and dual/tri-band data (shooting OSC, obvs) as much as I can. Or just forgo NB if the optics is fast enough. If I were under much darker skies than I am I'd rarely bother with NB either.

*** Thanks Andrea. I agree and would do the same in clear and dark enough skies. Intuitively the broader filters should provide the truer picture. Id just love to know if there are spectral data on any of this. Tim

You could have a look at this through a spectrograph there are fun DIY options to build out there 

http://www.astrosurf.com/solex/sol-ex-stars-en.html


​​​​​​In this video (https://youtu.be/aHj6wA5GWVM?si=nDAufZPqV6jnEDU7 , in french) at 27:00 you can see the Orion nebula through that spectrograph, around the h-alpha line, with two satellite lines that are actually coming form nitrogen

andrea tasselli:

Tim Hawkes:
*** Thanks Andrea. I agree and would do the same in clear and dark enough skies. Intuitively the broader filters should provide the truer picture. Id just love to know if there are spectral data on any of this. Tim

*I would doubt there is much or anything at all other than emission lines, as continuum signal is usually subtracted and I can't see what useful information can be extracted than wouldn't otherwise available through photometry.  Some interesting spectra here: Nebulae spectra (astrosurf.com)

Obviously nothing beats doing the test yourself, try IC59 & IC63, or even NGC7380.

Thanks so much Andrea for that link.   Yes indeed - the spectra therein make it  clear that  it is impossible to integrate broad underlying peaks and especially with such poor signal to noise  so the spectra really can't be used to calculate what proportion of the total visible light (UHC bands)  from the nebulae is HA + OIII narrow band and how much is broadband

I have tried now two simple photometry experiment/ calculations of my own using some data sets on IC5070, the  Pelican Nebula and on the elephants trunk IC396  region.  I  took 3 sets of data from each using a UHC filter and a  ASI294 camera and either a 7 nm HA filter or a 3.5 nm O3 filter with a ASI294MM  camera.  The UHC ASI294 MC  integration was a CFA x1 drizzled integration image so there was no creation of false colour pixel information via Bayer deconvolution.  The UHC x1 CFA drizzle integration - not corrected in any way - was extracted into its RED, BLUE  and GREEN channels.  All 5 images were aligned and then  exactly the same patches  of each image carefully selected to represent ca 5000 pixel patches of a) a dark region with no nebula in to define as the 'no nebula' background  b)  a first region of high nebula signal with no stars and c)  a second patch of lower signal nebulosity also with no stars.

Then - using the PixInsight Statistics process -  measurements were made of the signal and stddevs in the two homogenous (indicated by small values of Stdev)  patches  (signal - background) and compared across the  the 5 images.

The idea was to quantify how much of the total 'nebula'- specific  signal that is detected  using the broader UHC filter  (i.e  within the two ca. 470-510 nm  and 630-700 IR cut - windows)  i.e.  defined as nebula signal  versus a null ackground region -   can be accounted for as purely  the narrow band signal contribution.

With the following  assumptions..

1) sky quality/ signal was reasonably similar on the different nights on which the data were obtained (this only really affects SNR) -- measured at between 19.6 to 20.6 mags/ sq arcsec
2) H beta peak signal is 1/2.8  x the H alpha line
3) OIII 495.9 nm peak is 1/2.95 the primary line at 500.7 nm  (the 3.5nm O3 filter only detected the primary line) 
4)  Relative overall quantum efficiencies of the B, R and G dye pixels plus UHC filter/ MC camera at 500 nm and at 656 nm and of the OIII and HA filters were about 50% and 60 % , respectively versus ~ 90% for the MM camera plus either cognate NB filters

and after normalising the data for 

gain, exposure time,  filter/ camera efficiency,,  the H beta and 496.4 nm OIII lines only being detectable in the UHC filter but not the  --- etc.

The overall conclusions were ... 

1) the AS1294 MM camera/ narrow band filters appeared to detect ~ 1.5X (27.6/ 16.9 =   1.63;   59.52/ 41.76 =  1.42 ) more  HII nebula-specific red photons than the OSC/ UHC system in unit time at equivalent gain.   These data accord more or less perfectly with the 1.5X difference  differences in overall QE.  Thus more less all of the nebula specific red photons passing through the UHC filter in the 620 to 700 nm window can be accounted for by the single HA line.    Thus no significant contribution to the nebula light in this region from either  broadband light  or other NB light (SII for example )
2) It was a different story with the blue green 470-510 nm band.  Even calculating the H beta based on the HA NB detection and also allowing for the missing OIII band, the NB detection of OIII and HA  was insufficient to account for all of the blue + green nebula specific light passing through the UHC filter .  After allowing for the relative camera QE's the shortfall was about 40% 

So - probably - and based on these initial results the blue/ green nebula-specific light does comprise a significant broad band contribution - likely from the scattered and reflected light from the nearby bright stars.   It seems then that the ideal representation of type II nebula colour would include broadband and not just NB contributions ....However of course the practical difficulty would lie in finding dark enough skies and being able to image for long enough  to get sufficient SNR and particularly using an OSC



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

As a postscript ...I also wonder how much the overlapping colour response of the IMX sensor used in the OSC -  i.e with as much of 10% of NB 656 nm red light being detected as green by green pixels ... inherently means that OSC images will render colour of emission n ebulae differently -- more orangy --than colour images constructed from mono NB and even RGB filters??

andrea tasselli:

Tim Hawkes:
As a postscript ...I also wonder how much the overlapping colour response of the IMX sensor used in the OSC -  i.e with as much of 10% of NB 656 nm red light being detected as green by green pixels ... inherently means that OSC images will render colour of emission n ebulae differently -- more orangy --than colour images constructed from mono NB and even RGB filters??

If you are after color purity then OSCs aren't the place you should be looking into unless a significant amount of fiddling with the channels is carried out to remove cross-band response but then OSCs are there to mimic human's eye response to the visible spectrum so that overlapping it baked in, as it really should. Personally I find some of the responses of color filters too much in terms of primary colors, without nuances.

Yes I agree Andrea...the whole colour thing is so complicated -- I also like the OSC 'look' for that reason  but who knows what 'visual reality' is when it comes to emission nebulae.  I like the idea of trying to get closer to it but it may be a Quixotic aspiration. ..especially since we can hardly really see such deep reds anyway.

Incidentally checking things over - I misread the output of the statistics process in PI and the calculation on the dataset above was  wrong by a factor of 4 -  the result now looks as though in fact the majority of the light detected by the UHC/ OSC system from the nebula can in fact be accounted for as purely the NB contribution  -- so I will have to edit that again