Carina Dwarf Galaxy: Old Data New Process

[sjastro]

Originally Posted by rustigsmed

interesting stuff Steven, have you tried stacking/processing the ESO data with just the Visual and Blue filters?

Cheers,

Rusty

Thanks Rusty.
Leaving out the U and I data takes a large chunk of data out of the equation.
ESO advised that if I was to combine all the data I should normalize the background for each dataset to minimize the noise.

This is a valuable piece of advice for RGB imaging where a synthetic luminance image is produced. I never even imagined normalizing the background could reduce the noise.

Regards

Steven

[madbadgalaxyman (Robert)]

Originally Posted by sjastro

That's not to say that Rolf's ultra deep sky imaging or even the my own version of "mathematical jiggery pokery" is without scientific merit when applied to activities such as sky surveys.
Steven

In the below (linked) thread, we discussed the feasibility of an amateur H-alpha survey of the Milky Way, given that some individual amateur Ha exposures now show a lot more than some of the short-exposure Ha survey images that were made by professional astronomers.
For instance, VPHAS+, the latest professional H-alpha survey of the MW has the deficiency of:
- short exposures (2 minutes)
- no coverage at high galactic latitude

See:
http://www.iceinspace.com.au/forum/s...+survey&page=2

[Ken Crawford (KenC)]

This is excellent work indeed and it is great to see others pushing the limits in capturing large, faint, extended objects. As owners of our own equipment we can "sit" on an object as long as we want without writing proposals for telescope time (except to our spouses).

And with dark skies and good processing methods these limits can be pushed indeed. The Max Planck pros measured our (R. Jay GaBany and Ken Crawford) star stream data to about 29.3 Mag/sq arc sec with .5 meter systems with about 10 hours of exposure time. They have been used in several AJ papers like this . . . . http://arxiv.org/PS_cache/arxiv/pdf/...003.4860v1.pdf

So I have no doubt that you can exceed this with longer exposure times under dark skies and careful calibrations.

Keep up the great work!!

[sjastro]

Originally Posted by Ken Crawford

This is excellent work indeed and it is great to see others pushing the limits in capturing large, faint, extended objects. As owners of our own equipment we can "sit" on an object as long as we want without writing proposals for telescope time (except to our spouses).

And with dark skies and good processing methods these limits can be pushed indeed. The Max Planck pros measured our (R. Jay GaBany and Ken Crawford) star stream data to about 29.3 Mag/sq arc sec with .5 meter systems with about 10 hours of exposure time. They have been used in several AJ papers like this . . . . http://arxiv.org/PS_cache/arxiv/pdf/...003.4860v1.pdf

So I have no doubt that you can exceed this with longer exposure times under dark skies and careful calibrations.

Keep up the great work!!

Thank you very much Ken.
I have been an admirer of your images over the years.

Regards

Steven

[sjastro]

Olivier Hainaut: ESO wrote

Combining the data over a huge wavelength range UBVI can be counter productive, even if your only goal is to go deep. For instance, the individual U images tend to be much shallower (because the camera is less sensitive in U, and bc the stars are fainter in U), so if you average them with the others without a different weight, you actually increase the noise more than the signal. You could consider to use the following weight for each image:
w = F/t
with t = exposure time (i.e. normalize by exposure time) and
F(U) = 0.2; F(B) = 0.4; F(V) = 1; F(R) = 1.5; F(I) = 0.8

First of all I was mistaken that the ESO image is a total 31 hr exposure, but a much shorter 7.5 hrs.
What is remarkable however is that according to the FITS headers an extraordinary number of subexposures were taken through each filter but only a small number were selected for stacking.

According to the FITS headers:

U filter 132 subexposures taken best 13 selected for combining.
V filter 1192 subexposures taken best 13 taken for combining.
B filter 873 subexposures taken best 13 taken for combining.
I filter 141 exposures taken best 13 taken for combining.

Subexposures varied from 300s -1000s depending on the filter.

I took up ESO's advice to normalize the filter images according to the weight function based on exposure times.

The noise values were calculated on 2 differently processed images.

(1) Filter images not normalized, pixel mapping applied to combined image.
Noise value:- 6.98 X 10^4.
(2) Filter images normalized, pixel mapping applied to combined image.
Noise value:- 6.03 X 10^4.

There was a 14% reduction in noise when using normalized filter images.
I used Pixinsinsight's noise calculation script.

It confirms that noise is increased for stacking individual images that are not normalized.

Regards

Steven

[madbadgalaxyman (Robert)]

Steven,
You must be onto something if even the pros find your image to be excellent.
Your version qualifies as the best image of several that I have seen of the Carina dwarf spheroidal galaxy.

To my eye, Carina does at least look like a galaxy rather than a random field of stars, at least after the "eye+brain system" tries to remove the rich star-field of foreground bright Milky Way stars. (Professional astronomers are always "cleaning" galaxy images, by removing the foreground stars from images, prior to doing photometry, by using various software techniques)
The Carina Dwarf Spheroidal Galaxy itself, in images, looks to me like it is defined by a "cloud" of very faint stars that at least seem to be of comparable brightness to each other.

I will, here, give a general perspective on dSph galaxies, for the benefit of all of the IIS members......

It is hard to resist the impression that a dwarf spheroidal (dSph) galaxy looks like a giant-sized "vastly expanded" globular star cluster,
But dSph galaxies, as a population of galaxies, are:
(1) much much more physically extended, for any given total luminosity, than globular star clusters.
(2) kinematically distinct from globular star clusters.
(3) very Dark Matter dominated (a globular has enough gravity to stop itself from flying apart without the need for any Dark Matter content)
(4) always dominated by old-to-intermediate aged stars, but with diverse star forming histories. Some dSph galaxies have only an 'old' (9-13 billion years old) stellar population , but others of these galaxies have had multiple episodes of star formation, with some of them experiencing low-level star formation even in the present day.
(5) Diverse in their kinematics. For instance, some dSph galaxies have disky (flattened) kinematics and shapes, which tends to contradict the "spheroidal" description.

As there are some newly discovered dSph galaxies that are so poor in stars that the total light of such a galaxy is equivalent to that of a single -1 absolute magnitude star......it seems it is only the need for Dark Matter to hold together these ultra-faint galaxies that distinguishes such a faint galaxy from a star cluster. Not that a -1 absolute magnitude object fits the usual mental picture of a "majestic galaxy"!!

Here are four useful resources about the nearby Dwarf Galaxies in the Local Group of Galaxies and in its nearby environment:

(1) Catalog and description of all known (as of 2012) dwarf galaxies in and around the Local Group ::
http://iopscience.iop.org/1538-3881/144/1/4/article

(2) Igor Karachentsev's Updated Nearby Galaxy Catalog is the "Encyclopaedia Galactica" of nearby galaxies, listing all known galaxies within 11 Mpc ::
(and the vast majority of them are tiny systems of low luminosity)

http://heasarc.gsfc.nasa.gov/W3Brows...eargalcat.html
AND
http://www.sao.ru/lv/lvgdb/introduction.php
(additional information on this website)

(3) Here is a very clear and very concise overview paper about the faintest-known local dwarf galaxies:
http://www.publish.csiro.au/?act=vie...id=AS11023.pdf

(4) Here is a long, technical, but clearly-written, review paper about the dwarf galaxies in the Local Group of galaxies. (This paper is: 2009, ARAA, 47, 371 )
http://www.astro.rug.nl/~etolstoy/tolstoyhilltosi09.pdf

cheers,
Robert

[sjastro]

Originally Posted by madbadgalaxyman

Steven,
You must be onto something if even the pros find your image to be excellent.
Your version qualifies as the best image of several that I have seen of the Carina dwarf spheroidal galaxy.

Thanks Robert.
Hopefully this can lead to further bridge building between amateurs and pros.

To my eye, Carina does at least look like a galaxy rather than a random field of stars, at least after the "eye+brain system" tries to remove the rich star-field of foreground bright Milky Way stars. (Professional astronomers are always "cleaning" galaxy images, by removing the foreground stars from images, prior to doing photometry, by using various software techniques)

After a while I could also trace the dwarf against the background.
Probably viewing a 100% resolution in the original FITS format would improve things.

The Carina Dwarf Spheroidal Galaxy itself, in images, looks to me like it is defined by a "cloud" of very faint stars that at least seem to be of comparable brightness to each other.

I will, here, give a general perspective on dSph galaxies, for the benefit of all of the IIS members......

It is hard to resist the impression that a dwarf spheroidal (dSph) galaxy looks like a giant-sized "vastly expanded" globular star cluster,
But dSph galaxies, as a population of galaxies, are:
(1) much much more physically extended, for any given total luminosity, than globular star clusters.
(2) kinematically distinct from globular star clusters.
(3) very Dark Matter dominated (a globular has enough gravity to stop itself from flying apart without the need for any Dark Matter content)
(4) always dominated by old-to-intermediate aged stars, but with diverse star forming histories. Some dSph galaxies have only an 'old' (9-13 billion years old) stellar population , but others of these galaxies have had multiple episodes of star formation, with some of them experiencing low-level star formation even in the present day.
(5) Diverse in their kinematics. For instance, some dSph galaxies have disky (flattened) kinematics and shapes, which tends to contradict the "spheroidal" description.

As there are some newly discovered dSph galaxies that are so poor in stars that the total light of such a galaxy is equivalent to that of a single -1 absolute magnitude star......it seems it is only the need for Dark Matter to hold together these ultra-faint galaxies that distinguishes such a faint galaxy from a star cluster. Not that a -1 absolute magnitude object fits the usual mental picture of a "majestic galaxy"!!

Here are four useful resources about the nearby Dwarf Galaxies in the Local Group of Galaxies and in its nearby environment:

(1) Catalog and description of all known (as of 2012) dwarf galaxies in and around the Local Group ::
http://iopscience.iop.org/1538-3881/144/1/4/article

(2) Igor Karachentsev's Updated Nearby Galaxy Catalog is the "Encyclopaedia Galactica" of nearby galaxies, listing all known galaxies within 11 Mpc ::
(and the vast majority of them are tiny systems of low luminosity)

http://heasarc.gsfc.nasa.gov/W3Brows...eargalcat.html
AND
http://www.sao.ru/lv/lvgdb/introduction.php
(additional information on this website)

(3) Here is a very clear and very concise overview paper about the faintest-known local dwarf galaxies:
http://www.publish.csiro.au/?act=vie...id=AS11023.pdf

(4) Here is a long, technical, but clearly-written, review paper about the dwarf galaxies in the Local Group of galaxies. (This paper is: 2009, ARAA, 47, 371 )
http://www.astro.rug.nl/~etolstoy/tolstoyhilltosi09.pdf

cheers,
Robert

Thanks for that I'll check out the links.
The Carina Dwarf has been recently in the news as part of the Dark Energy Survey.
http://cosmic-horizons.blogspot.com....uds-decam.html

The link to the paper is at the bottom.

Regards

Steven

[madbadgalaxyman (Robert)]

Originally Posted by sjastro

After a while I could also trace the dwarf against the background. Probably viewing a 100% resolution in the original FITS format would improve things."
Steven

"Yeah, it's definitely there"
(the Carina Dwarf Galaxy, that is).

However, I have just discovered that it has proved difficult even for our "professional cousins" to trace the extent and shape of this ghostly and very extended galaxy, so it should come as no surprise that we are having difficulty defining it! Indeed, there is still some controversy over the morphology of the extremely-extremely-faint outer portions of this galaxy, as is illustrated in the below quick summaries that I have made of two recent papers.

((((
But firstly, a note regarding the method used in these two papers to trace the outline of the Carina Dwarf Galaxy;
A lot of Optical imaging studies of the extremely-low-surface-brightness outermost regions (e.g. halos, very faint extensions, tidal tails) of very nearby galaxies don't try to trace the shape and extent of the integrated or diffuse light of the outermost portions of a galaxy (due to their extreme faintness in diffuse light) ,

but instead they resort to tracing the two-dimensional (on the image) distribution of the outermost stars in a galaxy by finding the locations in a galaxy image of all of those stars which occupy a particular defined part(s) of the Color-Magnitude diagram of the field of a galaxy;
this method allows the finding on an image of many individual stars from a specific part of the Color-Magnitude Diagram, for instance the Red Giant or Main Sequence Turnoff stars that belong to the target galaxy can be isolated and located on the image. The locations of these stars on the image are then used used as 'tracers' of the space distribution of the general stellar population around them......
So by isolating the location of all of those stars within a particular range of magnitude and a particular range of B-V color , in the field of the Carina dwarf galaxy, hopefully most of the stars that are found will belong to the Carina dwarf, rather than being foreground stars of our own Galaxy.

I further note that in very busy Star Fields (such as that of the Carina dwarf Galaxy) where there is a major confusion of the stars belonging to the target galaxy with the foreground stars belonging to the Milky Way, this simple numerical method of isolating or "filtering out" the stars of the target galaxy from the foreground Milky Way stars is going to work a lot better than the other method which tries to subtract out the foreground Milky Way stars from the image by means of statistical techniques. In fact, the technique of making a Color-Magnitude (= Hertzprung-Russell) diagram of the field of one of the satellite galaxies of the Milky Way, and then isolating those stars in the field that belong to the target galaxy (and those stars that don't) because they occupy particular ranges of color index and apparent magnitude, is probably something that amateur astronomers can master.

For instance, here is part of Figure 2 from the second mentioned paper (arxiv 1408.2907). The top panel shows the stars (as multiple points in the graph) in a particular field near to the Carina dwarf galaxy , plotted as multiple points in a Color-Magnitude Diagram (the vertical axis gives the apparent magnitude of each star, and horizontal axis gives the g-r colour index of each star (g minus r is quite similar to B-V, one of the familiar Johnson-Cousins colors). The bottom panel shows how it is possible to discriminate and sort the various stars in the field according to their particular values of magnitude and colour;
for instance, the region labelled MW in the Color-Magnitude diagram is Milky Way stars in the foreground of the Carina dwarf, the region labelled LMC is LMC stars in the foreground of the Carina dwarf, and Old and Intermediate and Young zones are stars of various ages that actually belong to the Carina dwarf...........


(actually, as you can see, this particular field near to the Carina Dwarf Galaxy contains very few stars from the galaxy itself!)
))))

Using this sort of technique, Battaglia et al., in 2012, Astrophysical Journal Letters, 761, L31 , trace the spatial distribution of the stars of Carina dwarf, & they map the extent and shape of this galaxy.
Here is their diagram showing the spatial distribution (displayed as the contours of equal star density) of the stars of the Carina dwarf spheroidal galaxy, clearly showing that this dwarf galaxy extends well beyond its King tidal radius of about 28.8 arcminutes, and also providing some evidence that it has tidal 'tails' or 'extensions' along its Major Axis which could be evidence of tidal disruption:

(the red ellipse is the nominal tidal radius of the Carina Dwarf galaxy)

However, a new analysis of the distribution of the stars of the Carina Dwarf (in this preprint : arxiv 1408.2907)(submitted to MNRAS in 2014), which includes some of the same astronomers among the authors, casts significant doubt on the tidal debris hypothesis, characterizing the evidence for tidal tails as 'tentative'. (The new study is a Very Wide Field study of the star distribution in the field of this galaxy, over an area of 12 square degrees). Apparently, the elongated feature along the major axis of this galaxy becomes significantly less prominent, after contamination of the color-magnitude diagram of the stars in the field (originating from Large Magellanic Cloud stars which are in the foreground of the Carina dwarf) is removed.

(the presence of LMC stars in this field, which is very far from the bright optical body of the LMC, is, in itself, very interesting!!!)


At least it would seem to be beyond question that:
- the isophotes of this galaxy have a somewhat flattened and slightly boxy aspect.
- the isophotes are significantly elongated and extended along the major axis of this galaxy.
- there is some kind of unusual extended component along the major axis, but its nature is not known with any certainty.
-the change in the ellipticity of this galaxy, with progressively increasing radius, is not typical of a "standard spheroidal shape"(for instance, one could argue that the isophotes are 'disky').
- there is a small but significant fraction of the total stellar population outside of the tidal radius of this galaxy. (the second paper estimates a figure of 1.6 percent)

[sjastro]

[/I] For instance, here is part of Figure 2 from the second mentioned paper (arxiv 1408.2907). The top panel shows the stars (as multiple points in the graph) in a particular field near to the Carina dwarf galaxy , plotted as multiple points in a Color-Magnitude Diagram (the vertical axis gives the apparent magnitude of each star, and horizontal axis gives the g-r colour index of each star (g minus r is quite similar to B-V, one of the familiar Johnson-Cousins colors). The bottom panel shows how it is possible to discriminate and sort the various stars in the field according to their particular values of magnitude and colour;
for instance, the region labelled MW in the Color-Magnitude diagram is Milky Way stars in the foreground of the Carina dwarf, the region labelled LMC is LMC stars in the foreground of the Carina dwarf, and Old and Intermediate and Young zones are stars of various ages that actually belong to the Carina dwarf...........
Attachment 171905

(actually, as you can see, this particular field near to the Carina Dwarf Galaxy contains very few stars from the galaxy itself!)
))))

Using this sort of technique, Battaglia et al., in 2012, Astrophysical Journal Letters, 761, L31 , trace the spatial distribution of the stars of Carina dwarf, & they map the extent and shape of this galaxy.
Here is their diagram showing the spatial distribution (displayed as the contours of equal star density) of the stars of the Carina dwarf spheroidal galaxy, clearly showing that this dwarf galaxy extends well beyond its King tidal radius of about 28.8 arcminutes, and also providing some evidence that it has tidal 'tails' or 'extensions' along its Major Axis which could be evidence of tidal disruption:
Attachment 171906
(the red ellipse is the nominal tidal radius of the Carina Dwarf galaxy)

Interesting stuff Robert.
I have never heard of a g-r colour index.
The low and negative values of g-r for the young stars suggests the Carina Dwarf might be a good target for NIR imaging since the redness might be due obscuration from our own galaxy.

Regards

Steven

[SkyViking (Rolf)]

Very impressive Steven. I like how one can usually improve image processing continually and gain new insight into old data. Interesting discussion too for sure. It is true that we as amateurs have an advantage in that we can expose a patch of sky for as long as we want really.
Nice of them to comment on your work like that. I too have found Olivier and Lars to be very approachable and helpful in the past.

Incidentally Lars was my supervisor on an astronomy project back when I studied at Copenhagen Uni in the 90's. We used a Meade LX200 to take images of the double cluster in Perseus from the rooftop of the old Copenhagen Observatory and calculate the age of the clusters. I remember there was no easy image processing back then and all image calibration was done with an excessive number of complicated Unix commands...

[madbadgalaxyman (Robert)]

Originally Posted by sjastro

I have never heard of a g-r colour index.
Steven

Just the g magnitude of an object minus its r magnitude, just as in B-V or V minus R etc.

Color indices using B and V and R and I magnitudes were the traditional ones, developed from photoelectric photometry.
A strongly positive color index usually indicates a red object.

But since the Sloan Digital Sky Survey, the ugriz system has become very popular in the literature;
Here is an example of the transmission curves of the u and g and r and i and z photometric filters ::
http://www1.cadc-ccda.hia-iha.nrc-cn...a/filters.html

The traditional UBVRI magnitudes and filters are briefly compared with the ugriz filterset, in this review paper:
http://www.astrohandbook.com/ch02/be...otosystems.pdf
(This is also a good general reference on various filters used in photometry)

Essentially, a color-magnitude diagram of a cluster or galaxy using B minus V for the colour index is often very similar in appearnce to that using g minus r for the colour index.

To follow the current literature, it is getting ever more necessary to get a good appreciation of u and g and r and i magnitudes ; they are not that different from the standard UBVRI magnitudes.

[Rex]

Very cool, processes you are using Steven, and a very interesting discussion. Thanks for sharing.

[sjastro]

Originally Posted by Rex

Very cool, processes you are using Steven, and a very interesting discussion. Thanks for sharing.

Thanks Rex.

Originally Posted by madbadgalaxyman

Just the g magnitude of an object minus its r magnitude, just as in B-V or V minus R etc.

Color indices using B and V and R and I magnitudes were the traditional ones, developed from photoelectric photometry.
A strongly positive color index usually indicates a red object.

But since the Sloan Digital Sky Survey, the ugriz system has become very popular in the literature;
Here is an example of the transmission curves of the u and g and r and i and z photometric filters ::
http://www1.cadc-ccda.hia-iha.nrc-cn...a/filters.html

The traditional UBVRI magnitudes and filters are briefly compared with the ugriz filterset, in this review paper:
http://www.astrohandbook.com/ch02/be...otosystems.pdf
(This is also a good general reference on various filters used in photometry)

Essentially, a color-magnitude diagram of a cluster or galaxy using B minus V for the colour index is often very similar in appearnce to that using g minus r for the colour index.

To follow the current literature, it is getting ever more necessary to get a good appreciation of u and g and r and i magnitudes ; they are not that different from the standard UBVRI magnitudes.

Thanks Robert.
What I found particularly interesting about the ugriz filter link was the CCD QE performance used on the CFH telescope and how it compares to my ST-10XME.
The CCDs used by the pros absolutely murder the typical amateur CCDs in near ultra violet performance.
I can attest to that as I am currently imaging the SMC in near ultra violet.
For an object that is visible to the naked eye it will take me about 12hrs exposure to get a satisfactory S/N ratio.

Originally Posted by SkyViking

Very impressive Steven. I like how one can usually improve image processing continually and gain new insight into old data. Interesting discussion too for sure. It is true that we as amateurs have an advantage in that we can expose a patch of sky for as long as we want really.
Nice of them to comment on your work like that. I too have found Olivier and Lars to be very approachable and helpful in the past.

Incidentally Lars was my supervisor on an astronomy project back when I studied at Copenhagen Uni in the 90's. We used a Meade LX200 to take images of the double cluster in Perseus from the rooftop of the old Copenhagen Observatory and calculate the age of the clusters. I remember there was no easy image processing back then and all image calibration was done with an excessive number of complicated Unix commands...

A good story Rolf.
In the 1990s I was stumbling around in the dark trying to load my astrofilm onto the reel that went into the developer tank.....

Apart from the feedback from Olivier and Lars what also impressed me was the speed of response. Let's face it, you, me or any other amateur would not exactly rate highly in the order of priority for ESO. It is a credit to each individual to respond, and in a timely manner.

Regards

Steven

[madbadgalaxyman (Robert)]

Originally Posted by SkyViking

Incidentally Lars was my supervisor on an astronomy project back when I studied at Copenhagen Uni in the 90's. We used a Meade LX200 to take images of the double cluster in Perseus from the rooftop of the old Copenhagen Observatory and calculate the age of the clusters. .

Hi Rolf,
did you end up with a Color-Magnitude diagram of the cluster(s)?
And if so, was it accurate enough to superpose some isochrones and then replicate the cluster age that has been determined through very precise photometry?
How did your cluster main-sequence look? (well-defined, or noisy?)

I don't see too many amateurs making Color-Magnitude (e.g. B-V or U-B vs. apparent magnitude) diagrams of clusters.

I thought it was interesting that a fairly simple technique using the C-M diagram of the Carina dwarf galaxy's field was able to quite reliably distinguish the stars belonging to this galaxy from the foreground stars belonging to the MW and the LMC. All that would be needed for amateurs to do this kind of work would be the ability to measure accurate magnitudes, though it would not be easy to do at >20 apparent magnitude.

Cheers, Robert

[SkyViking (Rolf)]

Originally Posted by madbadgalaxyman

Hi Rolf,
did you end up with a Color-Magnitude diagram of the cluster(s)?
And if so, was it accurate enough to superpose some isochrones and then replicate the cluster age that has been determined through very precise photometry?
How did your cluster main-sequence look? (well-defined, or noisy?)

I don't see too many amateurs making Color-Magnitude (e.g. B-V or U-B vs. apparent magnitude) diagrams of clusters.

I thought it was interesting that a fairly simple technique using the C-M diagram of the Carina dwarf galaxy's field was able to quite reliably distinguish the stars belonging to this galaxy from the foreground stars belonging to the MW and the LMC. All that would be needed for amateurs to do this kind of work would be the ability to measure accurate magnitudes, though it would not be easy to do at >20 apparent magnitude.

Cheers, Robert

Hi Robert,

The result from back then was rather miserable as I remember it, I don't think we recorded enough stars to get a descent enough H-R diagram, but is was a good exercise.

It is so much easier these days and I actually have a dedicated gallery with H-R diagrams for a selection of globular clusters here: http://www.rolfolsenastrophotography...de-Diagrams#!/

Cheers,
Rolf