Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Interesting view of Wirtanen in UV Cyanogen band - no rotation evident

  • Please log in to reply
13 replies to this topic

#1 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • topic starter
  • Posts: 7967
  • Joined: 12 Oct 2007

Posted 07 December 2018 - 03:13 AM

Comet 46P/Wirtanen has a known rotation rate of about 8 hours and recently it was observed directly in the UV Cyanogen band around 388nm:

 

http://wirtanen.astr...6P_status.shtml

 

I was hoping to capture some indication of this rotation, first with Sloan filters and recently with a UV CN filter - but it has not worked out.

 

The comet is now moving to the north so northern hemisphere observers will have a chance to image rotation - but I am giving up and will switch to wide field.

 

My Sloan filter images were fairly high res. at 0.4" per pixel and well guided - but I could not see evidence of rotation over 4 hours.

 

So I switched to CN imaging, as used in the recent professional work that captured rotation, but the signal was extremely faint and I was barely able to see the comet at all in each 10m exposure. 

 

The end result of 5x10m exposures is here:

 

get.jpg?insecure

 

and my accompanying text is:

 

This is an interesting but non-impressive view of comet 46P/Wirtanen in the UV passband of the CN, "Cyanogen" emission line around 388nm. I was hoping to capture some detail in the CN emission, but it is very faint and the camera/optics are not optimized for UV sensitivity.

 

This is 5x10m exposures oag guided using MetaGuide shift guiding. Shift guiding was essential here for the 10m exposures since the comet is moving so fast.

 

It is often said that the green color of comets is due to cyanogen - but that is incorrect and the green is due mostly to diatomic carbon, C2, emission. "Cyanogen" or CN is often present, but emits mostly in the UV around 388nm and not in the green.

 

Details can be found here:

https://www.cloudyni...ot-cn-cyanogen/

 

Recently images of rotation in this comet were obtained in the CN band as reported here:

http://wirtanen.astr...6P_status.shtml

 

The rotation period is about 8 hours and can be observed with professional equipment but I was unable to detect it either with Sloan or CN filters.

 

I have imaged other comets with a CN filter as shown here:

https://www.cloudyni...-bicolor-image/

 

but that was with a 300 f/4 dlsr lens - and this image of 46P is with EdgeHD11 at f/7 and 0.4" per pixel.

So - good luck to anyone who might be able to capture this thing spinning.

 

Frank


  • JoRy likes this

#2 Tonk

Tonk

    Cosmos

  • *****
  • Posts: 8794
  • Joined: 19 Aug 2004
  • Loc: Leeds, UK, 54N

Posted 25 January 2019 - 06:23 PM

Whats the FWHM of your CN ( "Cyano not Cyanogen" wink.gif ) filter?

I have recently obtained a 387nm CL 11nm FWHM filter but the British winter has only allowed me one brief imaging session last week. All I can say is the CN signal from Wirtenen is about the same as the C2 signal at 515nm (but not calibrated to account for QE at 387 vs 515 etc.) and its "bright" in relative terms and certainly brighter than C2 in absolute terms as I know my camera QE is lower at 387 than 515nm


Edited by Tonk, 25 January 2019 - 06:38 PM.

  • JoRy likes this

#3 Tonk

Tonk

    Cosmos

  • *****
  • Posts: 8794
  • Joined: 19 Aug 2004
  • Loc: Leeds, UK, 54N

Posted 25 January 2019 - 06:33 PM

So - good luck to anyone who might be able to capture this thing spinning.

Try running your CN images though this processor - http://www.psi.edu/research/cometimen
 

Jorma has done so and created a nice "rotation" animation - (using the same Semrock filter I have just got) - see ...

 

 

http://spaceweatherg...pload_id=150891
 


Edited by Tonk, 25 January 2019 - 07:29 PM.


#4 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • topic starter
  • Posts: 7967
  • Joined: 12 Oct 2007

Posted 26 January 2019 - 06:43 AM

Hi Tonk-

 

It's the same filter I used years ago.  I think it is 10nm bandwidth.  I'm pretty sure it is getting signal from the CN line since my earlier images showed the comet well.

 

I was hoping to see rotation by direct imaging rather than applying a special filter - but I guess that's how the rotation images were created in the first place.  So I may try processing them further at some point.  I have other images of the comet with much more signal using Sloan filters - and I might as well try applying the filter to those also.  But direct imaging of structure is what I was after.

 

I haven't looked at the Semrock filter in detail but I assume it would work better than what I used.  I may get one some time.

 

But at this point it looks like many people will be imaging in the CN and C2 lines when comets come around - so there should be plenty of examples.

 

I still see people saying comets are green because of cyanogen - but overall it is fading out.

 

Frank



#5 Tonk

Tonk

    Cosmos

  • *****
  • Posts: 8794
  • Joined: 19 Aug 2004
  • Loc: Leeds, UK, 54N

Posted 26 January 2019 - 07:58 AM

I was hoping to see rotation by direct imaging rather than applying a special filter


The problem is the features exposing rotation have an extremely low contrast gradient and range. The filters amplify the contrast differences using a radial normalisation process -  for example doing (quoting the site) "division by azimuthal average" or "division by azimuthal median" by converting the input image into polar coordinates (centered on the nucleus) so each horizontal row in the polar representation located at a given radial distance from the nucleus represents different azimuths of the original image. It then finds the average value of all azimuths for a given radial distance, excluding outliers, and divides each azimuthal value by that average. This helps to emphasize azimuthal changes for a given radius that can be hard to see in the original brightness distribution.
  • JoRy likes this

#6 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • topic starter
  • Posts: 7967
  • Joined: 12 Oct 2007

Posted 26 January 2019 - 02:47 PM


The problem is the features exposing rotation have an extremely low contrast gradient and range. The filters amplify the contrast differences using a radial normalisation process -  for example doing (quoting the site) "division by azimuthal average" or "division by azimuthal median" by converting the input image into polar coordinates (centered on the nucleus) so each horizontal row in the polar representation located at a given radial distance from the nucleus represents different azimuths of the original image. It then finds the average value of all azimuths for a given radial distance, excluding outliers, and divides each azimuthal value by that average. This helps to emphasize azimuthal changes for a given radius that can be hard to see in the original brightness distribution.

There have been other comets that showed rotation and spiral structure - including dust patterns in Hale Bopp.  So it can be possible to see it directly and even without a special filter.  For Wirtanen it may be possible to extract some similar evidence of rotation using a morphology filter on other visible passbands or even an L filter - I'm not sure.

 

Applying a special mathematical operation on the image to bring out movement is just something I was hoping to avoid.

 

Frank



#7 JoRy

JoRy

    Lift Off

  • -----
  • Posts: 5
  • Joined: 03 Jul 2018
  • Loc: Finland, Helsinki, 60N

Posted 27 January 2019 - 08:21 AM

I observed 46P/Wirtanen also earlier at 26.12.2018, and made then a gif animation of 36x5min CN 387nm exposures without any special image processing, only dark&bias&flat calibration and basic histogram transformation.

 

Observation can be found from this link https://www.taivaanv...ions/show/79906, where the last image (6) is the 36x5min animation. Sorry partly finnish language in page.

 

Some minor coma/jet changes can perhaps be seen in the animation on the right side of optocenter. Please note that in the animation N is down, E right. Please also ignore the first color image, it has practically no scientific value.

 

Frank mentioned earlier that his system pixel size is 0,4"/px. It sounds quite a high resolution, so perhaps it can be binned for example to bin3 to get 1,2"/px. With binning it's possible to get much better S/N ratio.

 

Jorma


Edited by JoRy, 27 January 2019 - 08:28 AM.

  • Tonk likes this

#8 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • topic starter
  • Posts: 7967
  • Joined: 12 Oct 2007

Posted 27 January 2019 - 04:36 PM



I observed 46P/Wirtanen also earlier at 26.12.2018, and made then a gif animation of 36x5min CN 387nm exposures without any special image processing, only dark&bias&flat calibration and basic histogram transformation.

 

Observation can be found from this link https://www.taivaanv...ions/show/79906, where the last image (6) is the 36x5min animation. Sorry partly finnish language in page.

 

Some minor coma/jet changes can perhaps be seen in the animation on the right side of optocenter. Please note that in the animation N is down, E right. Please also ignore the first color image, it has practically no scientific value.

 

Frank mentioned earlier that his system pixel size is 0,4"/px. It sounds quite a high resolution, so perhaps it can be binned for example to bin3 to get 1,2"/px. With binning it's possible to get much better S/N ratio.

 

Jorma

Hi Jorma-

 

Yes - I had only recently seen your result - and congrats on seeing some rotation.

 

I had also imaged Wirtanen in Sloan filters as shown here:

 

get.jpg?insecure

 

The frustrating thing for me at the time was that I had no scale or exposure info on the Wirtanen images that showed rotation - and I also didn't know that processing had been applied specific to revealing rotation.  So I just went for a high res. close up - hence less signal.  I probably would have been better with a 300 f/4 lens and a much wider field - as I had used with Lovejoy some years ago.

 

I think my field may be too narrow to see anything even if I bin - but I may try at some point.  I just realized your images are at f/2.9??  That is much faster and wider - especially with a larger pixel ccd.  So - yes - I should have used the 300 f/4 lens.

 

I'm also interested to see journal papers on the rotation of Wirtanen during this passage.  It's not clear to me the CN band is needed to see it.

 

I have been trying to image with the CN line since 2007 - first with a Venus UV filter that didn't work well at all:

 

http://www.astrogeek.../cometvz13.html

 

then in 2015 I found an Edmund filter with about the right passband.  It's great if this new filter works even better.  People have been selling "comet" filters for a long time, and even advertising them as CN filters - even though they were just for the Swan bands.  So it's about time something is available.

 

Frank


  • JoRy likes this

#9 Tonk

Tonk

    Cosmos

  • *****
  • Posts: 8794
  • Joined: 19 Aug 2004
  • Loc: Leeds, UK, 54N

Posted 27 January 2019 - 05:13 PM

Except the Semrock isnt made for CN! Its for laser fluoresense microscopy with a specific dye  so by a happy accident works for CN.



#10 JoRy

JoRy

    Lift Off

  • -----
  • Posts: 5
  • Joined: 03 Jul 2018
  • Loc: Finland, Helsinki, 60N

Posted 28 January 2019 - 04:20 PM

Hi Frank,

 

I was thinking the differences of our systems and what are the actual parameters which should take into account when estimating the sensitivity and to get enough 387nm photons to resolve comets CN radical gas jets.

 

The following parameters are perhaps the most important ones:

 

Aperture area
f ratio
CN Filter transmission
CCD Absolute QE at 387nm
CCD pixel size area ratio
Airmass
Comet CN rate ratio
Exposure time, integrated

 

Then I tried to roughly compare these two systems to have an estimation of "photon budget":

 

12" aperture compared to 11", the aperture ratio is 1.00/0.87
f2.9 compared to f7, the ratio is 1.00/0.17
CN filter transmission in Semrock is 0.93, Edmund 0.4(?), then 0.93/0.4
CCD QE at 387nm, QSI690 0.60, ASI1600 0.45(?), then 0.60/0.45
CCD pixel area ratio, QSI690 3.69um bin2, ASI 3.8um bin1, ratio is 1.00/0.27
Airmass, both 1.00/1.00
Comet CN ratio, let 46P/Wirtanen=1.0, so 1.0/1.0
Exposure times similar

 

I compared these two different systems sensitivity with multiplying these ratios, not sure if it's really possible but to get some figures:

12"f2.9: 1.00*1.00*0.93*0.60*1.00*1.00*1.00 = 0.5580
11"f7.0: 0.87*0.17*0.40*0.45*0.27*1.00*1.00 = 0.0071

 

This means 78x sensitivity difference between these systems. I tested equation in excel using bin3(=11.4um) px in ASI1600, then equation gives only 8.6x sensitivity difference. About Edmund CN filter, somewhere was mentioned its transmission is only 0.40 or was it typo?

 

This equation possible have some fundamental errors, but the main parameters affecting to overall sensitivity are perhaps these above.

 

Jorma



#11 Tonk

Tonk

    Cosmos

  • *****
  • Posts: 8794
  • Joined: 19 Aug 2004
  • Loc: Leeds, UK, 54N

Posted 28 January 2019 - 04:30 PM

About Edmund CN filter, somewhere was mentioned its transmission is only 0.40 or was it typo?


Sounds about right. Thats why I decided to avoid the older Edmunds. The newer Edmund high transmission band pass filters only go down to 400nm and the set is small and mostly misses comet spectra lines. Semrock on the other hand go well into the UV and the set of available CL values  is much much larger and hit the important comets spectral lines. There is even a 3nm FWHM at the C2 main line near 515nm but I decided against that one as the available flux through that window  is very low


Edited by Tonk, 28 January 2019 - 04:31 PM.


#12 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • topic starter
  • Posts: 7967
  • Joined: 12 Oct 2007

Posted 28 January 2019 - 05:22 PM

Hi Frank,

 

I was thinking the differences of our systems and what are the actual parameters which should take into account when estimating the sensitivity and to get enough 387nm photons to resolve comets CN radical gas jets.

 

The following parameters are perhaps the most important ones:

 

Aperture area
f ratio
CN Filter transmission
CCD Absolute QE at 387nm
CCD pixel size area ratio
Airmass
Comet CN rate ratio
Exposure time, integrated

 

Then I tried to roughly compare these two systems to have an estimation of "photon budget":

 

12" aperture compared to 11", the aperture ratio is 1.00/0.87
f2.9 compared to f7, the ratio is 1.00/0.17
CN filter transmission in Semrock is 0.93, Edmund 0.4(?), then 0.93/0.4
CCD QE at 387nm, QSI690 0.60, ASI1600 0.45(?), then 0.60/0.45
CCD pixel area ratio, QSI690 3.69um bin2, ASI 3.8um bin1, ratio is 1.00/0.27
Airmass, both 1.00/1.00
Comet CN ratio, let 46P/Wirtanen=1.0, so 1.0/1.0
Exposure times similar

 

I compared these two different systems sensitivity with multiplying these ratios, not sure if it's really possible but to get some figures:

12"f2.9: 1.00*1.00*0.93*0.60*1.00*1.00*1.00 = 0.5580
11"f7.0: 0.87*0.17*0.40*0.45*0.27*1.00*1.00 = 0.0071

 

This means 78x sensitivity difference between these systems. I tested equation in excel using bin3(=11.4um) px in ASI1600, then equation gives only 8.6x sensitivity difference. About Edmund CN filter, somewhere was mentioned its transmission is only 0.40 or was it typo?

 

This equation possible have some fundamental errors, but the main parameters affecting to overall sensitivity are perhaps these above.

 

Jorma

The Edmund filter is just described as > 40% transmission - so 40 is the minimum.  Not clear what it actually is.  So I would view 40 as a pessimistic value for the Edmund - and 93 as an optimistic value for the Semrock.

 

If you compare binned to unbinned with similar pixel size - that is a factor of 4 right there - but I can bin in post-processing so it isn't a hard limit.

 

The response of these cameras and the system transmission are plummeting rapidly below 400nm - so it's hard to say what the actual values are without measuring them. 

 

I just looked at the Semrock transmission spectra and I see they let in IR above around 900nm.  Are you blocking that with an additional filter?

 

So I think the dominant difference is from the f/ratio and binned vs. unbinned - which is a guaranteed factor of  about 23 right there.  And you are capturing a much wider field.

 

I don't know how the system transmission compares - but an SCT definitely has a lot of loss below 400.  If there is no need for high res. detail then I would do this UV imaging with a shorter refractor.

 

As for filters, I already have i', r', g' Sloan filters - and I would be happy to add u' to reach the CN line.  But annoyingly there is a gap between u' and g' filters right where CN lives - so that wouldn't work.  Given the rapid loss of QE and other things below 400, I think it would be OK to have the filter passband extend beyond 387 on the short side - so narrow bandpass may not be critical.  But you would want high transmission at the CN line - and it looks like the u' wouldn't work.

 

Unfortunately the Semrock may be a good enough fit for CN that there is less motivation to make custom filters for comet imaging.

 

So it looks like Semrock is the best we can do - and the main thing to focus on is the camera and optics.  There are many options a wide field imaging situation - but for 0.4" per pixel and high res - that points to a Cassegrain.  So maybe a pure DK or RC with no lenses would be best - and with coatings targeting UV.

 

Frank


Edited by freestar8n, 28 January 2019 - 05:25 PM.

  • JoRy likes this

#13 JoRy

JoRy

    Lift Off

  • -----
  • Posts: 5
  • Joined: 03 Jul 2018
  • Loc: Finland, Helsinki, 60N

Posted 29 January 2019 - 10:31 AM

I use only Semrock 387nm filter without any IR block filter in series. Semrock start to leake in IR end after 950nm and between 950nm-1050nm it's transmission is less than 10%. On this same bandwidth area QSI690 CCD spectral response drops rapidly after 950nm (QE 10%) to being zero at 1050nm (QE 0%).

 

This combined Semrock+QSI red leake response is low and seems that it practically do not affect to final images.

 

Maryland University scanned Semrock filter and results show that it's transmission at 387nm is 93% or even bit more http://wirtanen.astr...lter_test.shtml.

 

You're right, there is perhaps no motivation by any filter manufacturer to make better CN filter having narrower band and good coatings.

 

Jorma



#14 Tonk

Tonk

    Cosmos

  • *****
  • Posts: 8794
  • Joined: 19 Aug 2004
  • Loc: Leeds, UK, 54N

Posted 29 January 2019 - 01:58 PM

Semrock start to leake in IR end after 950nm and between 950nm-1050nm it's transmission is less than 10%.


Also this is a major atmospheric H20 absorption region - so unless you are in space or on top of a really high mountain you should not see any effect from this as the IR here is literally watered down wink.gif


Edited by Tonk, 29 January 2019 - 02:14 PM.



CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics