Centaurus A in UV (B channel)

[madbadgalaxyman (Robert)]

Originally Posted by sjastro

Perhaps Madbadgalaxyman has some comments.

Indeed I do, Steven.
I have many different images of NGC 5128 at many different wavelengths, but it is quite hard to compare them.
Therefore, I am currently in a state of profound cogitation (mentation)(cerebration) regarding your UV image compared to the other imaging data I have on disk.
I shall comment further.

Would you like to have a go at presenting a version of your UV-only image at maximum contrast?

I intend to compare it with, for instance, far-ultraviolet, H-alpha, and U-B colour index, images that I have on file.

Best Regards
Robert

[sjastro]

Would you like to have a go at presenting a version of your UV-only image at maximum contrast?

I intend to compare it with, for instance, far-ultraviolet, H-alpha, and U-B colour index, images that I have on file.

Best Regards
Robert

Robert,

Not sure if this is satisfactory but here it is.

http://members.iinet.net.au/~sjastro...ewcontrast.jpg

Regards

Steven

[madbadgalaxyman (Robert)]

Hi, Steven and other IIS members,

We would all like to know in Steven's UV image of N5128, what exactly is the distribution (both in the image, and also in "real Three-D space") of the hot young blue stars that are associated with the dust lane.

Steven's UV-only image can potentially show us the location of these blue supergiant OB stars......
BUT
in long-wavelength UV images like Steven's, there is still substantial contamination of the image of NGC 5128 by light which comes from the old stellar population of this galaxy which we see as the spheroidal component of this galaxy.

Thus, in Steven's image, the light of the blue knots of young stars is somewhat confused with the light of the old stars in the spheroid of this galaxy.

There are various mathematical approaches for subtracting out the light of the spheroid (old stars) from Steven's image , but another method of isolating the young stars in his image is to compare Steven's UV image with another image that shows only the young blue supergiant stars.

One such image can be found in the 1978 Sky and Telescope, which shows a U-B (colour index) image of NGC 5128 that displays only the youngest stars in this galaxy!
Does Anyone have this article by Dufour and van den Bergh :
1978, Sky & Telescope, Volume 56, page 389

I desperately need a scan of the images in this article, as I only have a very poor scan of the U-B image that isolates the young stars in NGC 5128!!

This article presented a picture processing analysis of NGC 5128; as Dr Sidney van den Bergh is now well into his 80s, and this is "fair dealing for personal research & study", you won't be violating copyright.

Best wishes,
Robert

[sjastro]

.....Steven's UV-only image can potentially show us the location of these blue supergiant OB stars......
BUT
in long-wavelength UV images like Steven's, there is still substantial contamination of the image of NGC 5128 by light which comes from the old stellar population of this galaxy which we see as the spheroidal component of this galaxy......

Thus, in Steven's image, the light of the blue knots of young stars is somewhat confused with the light of the old stars in the spheroid of this galaxy......

From an imaging perspective this makes perfect sense as the original luminance image carries the RGB information from the visual image and is a "source" of the contamination.

With this in mind I have created a synthetic luminance image L=B+30*UV.
Here I have taken my existing B images and merged them with the 10 hr UV exposure where each pixel has been multiplied by 30.

The new luminance image is now in effect acting as a filter and is only allowing blue and UV light through and filtering out other wavelengths to a certain degree. The strongly red field stars are now largely muted and the outer envelope of Centaurus A is no longer visible.
The dust region is much more opaque looking.

The new image is found here.
http://members.iinet.net.au/~sjastro/CentUV_B.jpg

It's no longer a pretty picture but hopefully it might answer some questions.

Regards

Steven

[Terry B]

Originally Posted by sjastro

Amongst the many pitfalls of long wavelength UV imaging is one factor I failed take into consideration, the night sky is "aglow" with single ionized oxygen (OII) emissions. While this has no effect on visual observation or even visible light imaging as the filters cut out the emission, it is a problem for UV imaging as the emission line sits in the middle of the UV pass bandwidth.

It also helps to explain why there is no contrast gain using a UV pass filter.

Clear skies

Steven

Steven
I'm not sure that the OII emisions in the UV are a problem. When I take spectra, the skyglow fron OI at 5577A is very easly seen. Usually the background spectra is subtracted in the processing of the images so you don't see it.
I have attached a relatively raw spectra (only dark subtracted) of a Be star. It has nice hydrogen absorption lines. Theses give an idea of calibration. The image has the UV on the left and IR on the right. My camera is the same as yours ie an ST10XME so should have a very similar spectral response. The response extends from about 3600A.
The slit is most of the height of the image so airglow shows as a vertical line. I have labelled the obvious features. The line to the right of the 5577 airglow is just a defective pixel column.
Of interest is that there is no identifiable airglow lines in the UV. This is only a 5 min exposure but I have done lots of these types of images with long stacks of spectra and have never seen any lines in the UV. There may be emission there but it is not very bright.
Cheers

Terry

[sjastro]

Originally Posted by Terry B

Steven
I'm not sure that the OII emisions in the UV are a problem. When I take spectra, the skyglow fron OI at 5577A is very easly seen. Usually the background spectra is subtracted in the processing of the images so you don't see it.
I have attached a relatively raw spectra (only dark subtracted) of a Be star. It has nice hydrogen absorption lines. Theses give an idea of calibration. The image has the UV on the left and IR on the right. My camera is the same as yours ie an ST10XME so should have a very similar spectral response. The response extends from about 3600A.
The slit is most of the height of the image so airglow shows as a vertical line. I have labelled the obvious features. The line to the right of the 5577 airglow is just a defective pixel column.
Of interest is that there is no identifiable airglow lines in the UV. This is only a 5 min exposure but I have done lots of these types of images with long stacks of spectra and have never seen any lines in the UV. There may be emission there but it is not very bright.
Cheers

Terry

Terry,

I'm surprised you are able to even detect skyglow lines even in the sweet spot region of the ST-10 QE range. I would have thought the slit would act as an effective aperture stop.

I stand corrected on the skyglow issue being caused by the OII emission at 3720A, in fact there is a continuous UV background radiation over the 2500-4000A range caused by O atoms combining to form O2 molecules.

NASA's Nightglow project investigated the near UV background radiation a few years ago.
http://asd.gsfc.nasa.gov/archive/nig...ience_sci.html

Regards

Steven

[Terry B]

Originally Posted by sjastro

Terry,

I'm surprised you are able to even detect skyglow lines even in the sweet spot region of the ST-10 QE range. I would have thought the slit would act as an effective aperture stop.

I stand corrected on the skyglow issue being caused by the OII emission at 3720A, in fact there is a continuous UV background radiation over the 2500-4000A range caused by O atoms combining to form O2 molecules.

NASA's Nightglow project investigated the near UV background radiation a few years ago.
http://asd.gsfc.nasa.gov/archive/nig...ience_sci.html

Regards

Steven

That makes more sense that it is a pretty continuous glow.
The slit doesn't act as an aperture stop. All it is essentailly doing is taking an image of the sky that is a vertical line and then spreading that line out to display the spectrum. A star is on the vertical line so that is displayed as the narrow (vertically) line in the image. All the other light entering the slit from skyglow, light pollution, moonlight etc is also spread out across the image and is visible above and below the star spectra. That image was taken with no moon and I don't have any light pollution so the only extra light on the slit is skyglow. When the moon is full, there is a much more obvious glow visible across the entire spectra.

[Merlin66 (Ken)]

http://www.atoptics.co.uk/highsky/airglow2.htm

This may assist with the determination of airglow bands.
The Na ("Sodium flash") was discussed at some length when a significant Na emission was recently observed in the comet Pan STARRS.

[sjastro]

Thanks Ken and Terry.

The UV background radiation gives me an idea I might be able to improve my UV images by using the UV background radiation noise itself as a flat field.

Regards

Steven

[madbadgalaxyman (Robert)]

Steven (and Ken and Terry B and Rolf)

Thanks for an interesting discussion regarding the problems of UV imaging!

I am glad you blokes are sweating it out, in so far as getting accurate and uncontaminated observations.

I do rely on data like yours - all the time - to do my work/play on the properties of galaxies, so it is good to know that you work on sorting out the complexities of getting good data.

Here, I continue the parallel discussion about the structure and evolution of N5128, but my next post will be in the science forum, as I am now getting heavily into the details about this galaxy........
__________________________

The three-dimensional morphology and evolution of the approximately disk-shaped "Extreme Population I" component of NGC 5128 (young stars + molecular hydrogen gas + dust) that manifests in our images as the dust lane, is a problem that has had a few different solutions in the professional literature. The most recent attempt was probably Quillen et al., 2010, PASA, Vol.27, p.396 (in Proceedings of the Astronomical Soc. of Australia)
Here is the preprint of this paper......the preprint reference is arxiv: 0912.0632 (you have to pay for the final version!)

N5128__the disk__(arxiv 0912.0632)__Quillen +__2010_PASA__27__396.pdf

They have analysed a lot of the multi-wavelength imaging data about this galaxy and they fit a simplified model (or visualization) of the three-dimensional structure of the disk to these observations;
But they give their model in exclusively numerical terms, without a 3-D visualization of the structure that they have come up with. I may try to make a three-dimensional image of their model of the dusty disk galaxy that bisects NGC 5128, but I would have to find some tools to do so; typically, these 'tilted ring' models are usually made and displayed with the tools of radio astronomers.

The following near-infrared image from 2MASS (a composite of J and H and K bands) may be helpful, as it simplifies the structure of the dust lane by only showing the parts with the highest extinction:



Here is the inner disk of NGC 5128, in the central region of this galaxy, as imaged at 8 microns by the Spitzer Space Telescope: (8 microns picks up primarily dust emission, whereas 3 microns detects primarily stellar photospheres)



The disk is not planar, with a complex and warped/bendy structure.....this is a very complicated galaxy!

In the following image :
(1) The left panel shows a map of the two-dimensional distribution of the J-H colour (an infrared colour index) of this galaxy over the central parts of NGC 5128. The parts of the dust lane with the heaviest dust extinction are displayed as the lightest regions on this map. (This infrared colour map greatly resembles a negative of the dust lane. )
(2) The right panel shows a near-infrared (J band) image of the dust lane, displayed as a positive image from the 2MASS survey. (being infrared, it shows only the heaviest dust extinction). On this image are overlaid the contours (isophotes) of the previously displayed 8 micron image from the Spitzer Telescope.



The above two images are from: 2006, ApJ, 645, 1092 (which is freely accessible on the internet)

( I am going to try to overlay this infrared disk with an image of the dust lane in the optical/UV regime, so you can get a better idea of where the infrared disk is located in this galaxy; I think there is a Herschel Space Observatory image that 'ties it all together' )

At present, the disk component that slices through the short axis of this galaxy is an unpopular subject for research; a search of the literature for recent papers about this galaxy using the SAO/NASA Astrophysics Data System reveals that since 2007, as regards NGC5128, professionals have written only one or two papers about the disk component (dust lane) but about 40 (!!!) papers about the orbital structures and kinematics and ages and metallicities of the objects in the spheroidal component, mainly studies of the very many globular clusters and planetary nebulae that belong to this galaxy ( these papers have the still unrealised goal of characterizing the time-evolution of the spheroidal component.....the data archives contain massive data about galaxies, but nobody in the professional community has been smart enough to use it to produce a generally accepted theory of how galaxies evolve!! ).

Here is a section of a still useful (but somewhat out of date) comprehensive overview about NGC 5128, which came out in 1998 (F.P. Israel , Astronomy & Astrophysics Review, 1998, Vol.8, p.237)

N5128_the dusty disk.doc
(this is in MS Word 2000 .doc format)

The rest of this paper can be found at;
http://ned.ipac.caltech.edu/level5/M..._contents.html
________________________________

In the overall context of humankind's attempts to understand how galaxies evolve, this sort of observation and study of an individual case of galaxy merger and accretion can shed a light on how galaxies evolve and build up mass; NGC 5128 has previously undergone at least one major merger (between two large galaxies) in addition to its recent "eating" of the smaller galaxy that produced the dust lane. There is also evidence that the big elliptical has swallowed a few other small galaxies!
In the 'hierarchical model' of galaxy formation, galaxy formation is ongoing in today's universe, as smaller galaxies continue to merge with larger galaxies, while in another recently popular model of galaxy formation, galaxies continue to build up their masses (especially the outermost part of the disk component) by accretion of gas from the intergalactic medium;
we could be seeing both the "merger" model of galaxy evolution and the "gas accretion" model of galaxy evolution happening in NGC 5128, because the outermost portions of this galaxy (the 'ansae" at the two extreme ends of the long axis of this galaxy, as seen in Mr Sidonio's ultra-deep image) show evidence of disky kinematics (orbits in the same plane, rather than in many different planes);
there seems to be a disk forming in the very outer regions, which means that NGC5128 might even evolve into a spiral galaxy!

Here is a map showing the noticeably-elongated Disky distribution of (spectroscopically identified using OIII) planetary nebulae in the outer parts of NGC 5128, overlaid on a 2 by 2 degree DSS image. This image is taken from Peng, Ford, and Freeman (2004, ApJ, 602, 685 )



Is this a giant disk structure that is currently in the process of formation in the outer regions of NGC5128?
(it would require infall of gaseous material for a 'traditional' gas+stars+spiral Disk Structure to form, but many giant elliptical galaxies already have a kinematically cold disk-like component in their outermost regions)
The planetary nebulae have substantial orbital velocities, with line-of-sight velocities that can depart from that of NGC 5128 itself by over 100 km/s.(the rotational component is significant in comparison to the random velocity component)

The current "mild two-component S0 morphology" of NGC 5128 may be evolving towards a true Disk+bulge galaxy (becoming, eventually, a spiral galaxy or an S0 galaxy);
Peng & Ford & Freeman characterise the velocities in the outermost regions as "fast disk-like rotation along the major axis" of this galaxy, together with significant rotation along the minor axis as well. The strong rotation of the outer parts of this galaxy continues out to a radius of at least 50 kpc! ( 163,000 light years)

cheers
bad galaxy man

[SkyViking (Rolf)]

Thanks Robert for that excellent piece of exhaustive information about Centaurus A!
Incidentally I'm working on a new Centaurus A image that will be finished in the next week or so. I quite sure you will find it interesting, and will let you guys know once I'm done processing the data.

[madbadgalaxyman (Robert)]

Originally Posted by SkyViking

Thanks Robert for that excellent piece of exhaustive information about Centaurus A!

I keep on coming back to Cen A and puzzling about it for a few days straight, but this is a particularly complex galaxy.
I have repeated my last post (about its morphology), in the IIS Science Forum, and I also included there an image of the shells in the halo (very-low-contrast features that are signatures of the accretion of smaller galaxies)

I think "exhaustive" is an exaggeration, but I have done my best......there have been so very many papers about this galaxy in the last few years that I don't think even the professional astronomers know where the research on this galaxy is headed.
('exhausted' might be a better word, at least at 1am in the wee hours of the morning...... when I usually finish my regular 5 hours per day of working on various galaxies)

I hope I can come up with a multi-wavelength image overlay to illustrate the relationship of the various features in the dust lane.

Best regards,
Robert Lang

[sjastro]

Robert,

I'll take off my astrophotography hat and replace it with my science hat to try to make sense of the image by going through the literature you have supplied.

Now that the atmospheric UV background has been identified as causing potential havoc I've been able to further improve the synthetic luminance image that filters out most of the older population of stars and emphasises the blue stars in the image.
The role of the UV background can be seen in comparative sum and average stacked images using AIPWinV2 in the attachment.
The attached image is a screenshot, no adjustments to the histograms have been made.

The filtered image can be found here.

http://members.iinet.net.au/~sjastro/CentUV_B.jpg

Details.
10 hr near UV exposure average stacked.
A synthetic L image L=B + 30*UV was produced. B= 0.8 hrs total.

(B+30*UV) L, 0.8 hrs R, 0.8 hrs G, 10 hrs UV
BRC-250, ST-X10ME.

Clear skies

Steven

[madbadgalaxyman (Robert)]

Originally Posted by sjastro

Robert,
I'll take off my astrophotography hat and replace it my science hat to try to make sense of the image by going through the literature you have supplied.
Steven

There is probably not enough in this literature I supplied to make a conclusive identification of what really you are detecting in your UV image.

Here are a few further suggestions for comparison with your image:

(1) Most instructive, for purposes of comparison, might be some of the Herschel Space Observatory images showing what looks like the entire dust lane in emission (infrared emission) rather than in absorption like we usually see it.
The large scale interstellar dust in galaxies is very cold (10-30 kelvin) but it lights up like a flare in Herschel images.
(dark nebulae become bright nebulae!)

The site is: //sci.esa.int/herschel
and put "NGC 5128" in the search box.

(2) The shortest bandpass (filter) of the Spitzer Space Telescope at about 3.6 microns is very useful for Near-infrared images that still enable a "feature for feature" comparison with optical/UV images. These Spitzer 3.6 micrometer images are far superior to the 2MASS (J+H+K) images of galaxies (such as the one I just supplied for Cen A).
These 3.6 micron images are a handy low-extinction comparison with optical images, but they are rarely published as public .jpg files.
They can be got from the NASA infrared archive as FITS files
(usually very large file sizes)

[ The most relevant Spitzer bandpasses for comparison with your optical imagery used the IRAC instrument of Spitzer Telescope to take images at 3.6, 4.5, 5.8, and 8.0 micrometres. ]

I can find the URL of the infrared data archive, if need be;
it is called NASA/IPAC Infrared Science Archive. Not very user friendly!

(3) Two images that accurately show the distribution of very young and very hot (= OB) stars in NGC 5128, and which should provide a much closer analog to your UV image, are the galex FUV image and also the following H-alpha image .
(bear in mind that the dust lane will be much heavier and more extensive in the FUV than in H-alpha)

Here is the GALEX FUV-bandpass image of NGC 5128 displayed first at a linear scale and then at a logarithmic scale; (the FITS data was downloaded from NED)(only young stars are visible in this image!)(no spheroid!!)





Here is the H-alpha image of NGC 5128 from the SINGS hydrogen-alpha survey, in which the H-alpha channel should accurately trace the distribution of OB stars over the face of NGC 5128.
[[ Blue color codes for broadband stellar 'continuum' light, while orange color codes for H-alpha light (only the nebular light of the h-alpha line is displayed, as the stellar continuum was subtracted out before display) ]]



I think that I may have saved you a bit of work with this post, as I have viewed colossal numbers of NGC 5128 images in order to select these ones!!

Robert Lang

P.S. After all this hard work to try to find out what is in your image, done by both of us,
here is one further silly comment;
OMG look at how heavy that dust lane is in the UV!!

[sjastro]

P.S. After all this hard work to try to find out what is in your image, done by both of us,
here is one further silly comment;
OMG look at how heavy that dust lane is in the UV!!

Thanks for your efforts Robert.

If there is one aspect that wasn't a surprise is the heavy dust lane.
I thought this feature would be sufficiently different in detail compared to a visible light image that warranted the effort to image in UV.
The other features are a bonus.

Regards

Steven

[John Hothersall]

Great effort and a very good in depth thread.

John.

[madbadgalaxyman (Robert)]

To compare with Steven's 'long-wavelength' UV, and also to compare with the GALEX satellite FUV-only image which was attached in my previous post (far-ultraviolet essentially isolates only very young and very hot and very luminous stars) ;;

here is another GALEX image which sums the two filters/bandpasses used by Galex (this image displays far-ultraviolet plus near-ultraviolet bands).

Note that far-ultraviolet and near-ultraviolet in the context of GALEX observations have different meanings to the same terms when used elsewhere.

In this display, the galex FUV band displays as blue-white and the galex NUV band displays as yellow.

This image offers very good discrimination between;
- the dust lane
and
- the young hot OB stars (which appear as blue-white)
and
- the old stars of the spheroidal component of NGC 5128, which appear as yellow



(This image was obtained using the excellent GalexView 'virtual telescope' interface for the GALEX images)

One thing to bear in mind is that the GALEX far-ultraviolet band is extremely extremely sensitive to the light of the OB stars found in blue knots within galaxies; a vanishingly faint knot of blue stars which is seen in a visible-light image can light up brightly in the GALEX FUV band.

[sjastro]

Originally Posted by John Hothersall

Great effort and a very good in depth thread.

John.

Thanks John.
It has been a most informative thread.

Originally Posted by madbadgalaxyman

To compare with Steven's 'long-wavelength' UV, and also to compare with the GALEX satellite FUV-only image which was attached in my previous post (far-ultraviolet essentially isolates only very young and very hot and very luminous stars) ;;

here is another GALEX image which sums the two filters/bandpasses used by Galex (this image displays far-ultraviolet plus near-ultraviolet bands).

Note that far-ultraviolet and near-ultraviolet in the context of GALEX observations have different meanings to the same terms when used elsewhere.

In this display, the galex FUV band displays as blue-white and the galex NUV band displays as yellow.

This image offers very good discrimination between;
- the dust lane
and
- the young hot OB stars (which appear as blue-white)
and
- the old stars of the spheroidal component of NGC 5128, which appear as yellow

Attachment 139927

(This image was obtained using the excellent GalexView 'virtual telescope' interface for the GALEX images)

One thing to bear in mind is that the GALEX far-ultraviolet band is extremely extremely sensitive to the light of the OB stars found in blue knots within galaxies; a vanishingly faint knot of blue stars which is seen in a visible-light image can light up brightly in the GALEX FUV band.

A couple of observations Robert.
The transparency of the dust belt increases again as one goes from NUV to MUV and FUV as seen in the Galex image.
Opacity is caused by scattering of photons which is a function of photon wavelength and the particle size of the scattering medium.
NUV wavelengths and particle size has maximized the degree of opacity.
Perhaps it explains the dearth of NUV images due to the lack of belt detail.

The other point is the confusion over terminology.
From Galex.

DETECTORS: Two 65 millimeter diameter, microchannel plate detectors. Far ultraviolet sensitive to light with wavelengths 135 to 175 nanometers. Near ultraviolet sensitive to light with wavelengths 175 to 280 nanometers.

Long wavelength UV (NUV) is classified in the region of 300-400nm.
I'm not sure for the reasons behind the Galex definition but the NUV data in the Galex image isn't really NUV but more like MUV.

Regards

Steven

[madbadgalaxyman (Robert)]

Originally Posted by sjastro

Opacity is caused by scattering of photons which is a function of photon wavelength and the particle size of the scattering medium.

I have a few comments to add about the nature and optical effects of the interstellar grains, but I am going to transfer them to the Science Forum thread that I started on NGC 5128. While it is relevant to understanding what we see in galaxy images at various wavelengths, I feel that a more detailed discussion of extinction belongs more properly in the Science Forum.

Best Regards,
Robert

[sjastro]

Originally Posted by madbadgalaxyman

I have a few comments to add about the nature and optical effects of the interstellar grains, but I am going to transfer them to the Science Forum thread that I started on NGC 5128. While it is relevant to understanding what we see in galaxy images at various wavelengths, I feel that a more detailed discussion of extinction belongs more properly in the Science Forum.

Best Regards,
Robert

I'll wait for your input in the Science forum.

Finally(!) I have been able to produce a UV image which doesn't look like a "stained sample" which I believe indicates what an NUV image of Centaurus A is supposed to deliver, a dust lane that is more optically opaque and an enhancement of the OB associations.

The OB associations are far more conspicuous than in any visual light image.

http://members.iinet.net.au/~sjastro/CentUVnew.html

This has been a real nightmare to successfully process.

Regards

Steven