3 orange galaxies, are they interacting?
-
by
Budgieye
moderator
EDIT: I can't do this without scale on the image.
RA..DEC.......................type..redshift.....mag....distance from center.. ID
03h32m10.0s -27d43m33s G 1.010000 24.0V 0.000 GOODS. J033210.04-274333.1
03h32m10.1s -27d43m33s G 1.022000 23.3R 0.008 GAS....... J033210.12-274333.3
03h32m10.1s -27d43m33s G 1.010000 24.9V 0.019 GOODS..J033210.12-274333.3
03h32m10.2s -27d43m34s G 1.020000 23.7R 0.035 COMBO-17 42592I wish there was a scale on the image, it would make identification easier.
target galaxy in middle z=1.01 Lookback time: 7.81 billion years ago.
http://www.wolframalpha.com/input/?i=redshift+z%3D1.01&a=FSelect use cosmological redshift
galaxy to the left, that would be next nearest, with slightly larger RA, same DEC,
Sorry, I can't sort out the coordinates without some scale on the image.
Posted
-
by
Mergie
This looks very similar: orange galaxies that may be interacting.
Image AGZ00087le
Posted
-
by
ramberts
Is it possible some tidal features are too far to be images in Hubble? They seem to be close enough together...
Posted
-
by
JeanTate
in response to ramberts's comment.
Yes, it's possible that any tidal features may be missed ... but not because they're 'too far' away. If these are indeed ellipticals, tidal features would very likely consist almost entirely of stars, and so have a very low surface brightness (making them 'invisible'); tidal features arising from a merger/interaction involving at least one gas-rich spiral will almost certainly contain gas, dust, and star-forming regions, which would make them much more visible.
Near the center of rich clusters, there are many ellipticals; projection makes a lot of them look like they're very close, even though they are quite remote from each other.
One way to tell if there's possibly interaction going on is to look for 'disturbance' ... quantitatively, do the integrated radial intensity profiles deviate from a de Vaucouleurs one, for example? That's something which we citizen scientists can check for ourselves; are you interesting in learning how?
Posted
-
by
ElisabethB
moderator
Or it could just be that with these elliptical galaxies it is just very difficult to see minor interactions? 😉 Remember, without hard data (eg redshift, it is just not possible to determine if these galaxies are really close together or just appear to be. And, here in GZ, we are doing visual classifications ! 😄
Posted
-
by
JeanTate
in response to ElisabethB's comment.
And even with good redshifts, it's possible that they will be similar, making them apparently at 'the same distance' (this is very common for ellipticals near the center of rich clusters). Distinguishing mere overlaps from interactions/mergers in ellipticals near the centers of rich clusters is difficult ...
And, here in GZ, we are doing visual classifications ! 😄
Precisely! 😃 If, in the image presented, you see no signs of interaction/merging - disturbed shapes, dustlanes in ellipticals, 'tidal tails', shells, ... - then you cannot honestly say there appears to be interaction/merging going on.
Posted
-
by
ramberts
Sure I'd be willing to learn, as long as it isn't super complicated graduate/PHD level physics kinda stuff. LMAO!
Posted
-
by
JeanTate
in response to ramberts's comment.
OK, cool! 😃
First, some general words about the principle (or, one principle).
Imagine an image of a galaxy, one that appears perfectly round, and that there's nothing else in the image. Such an image is just an array of pixels, perhaps 424x424. Let's assume it's a monochrome ('black and white') image.
It's fairly straight-forward to find the 'photocenter' of the galaxy, with a ruler marked in pixels (I'll go over how to do this, in some detail, later). If you take a line through that center, from any angle, the pixels it meets, on either side of the center, will have the same brightness (intensity, luminosity, flux ... lots of terms, and the details do matter, but not for now) on each side ... well, the pairs which are the same distance from the center will. Why? Because we assumed a 'perfectly round' galaxy! 😄
Now also because the galaxy is perfectly round, all such lines through the center will be the same, in the sense that pairs of pixels equi-distant on either side of the center, will have the same brightness values. So, if we plot (graph, chart, ... what's your fave verb?) the pixel brightness values as a function of distance from the center, along any line, we will have a 'radial intensity profile'.
Astronomers have spent many decades fitting radial intensity profiles to all manner of galaxies, and - at a very high level ('lies we tell to children') - they found that elliptical galaxies have one profile ('de Vaucouleurs') and the disks of spiral galaxies another ('exponential').
So our task, as citizen scientists, is to derive a radial intensity profile from the image data, and see how much it deviates from either a de Vaucouleurs or an exponential one; the extent to which it does points to some disturbance, something not-quite-right.
(Actually, as you might expect, it's not quite that simple, or straight-forward, but this is a good place to start).
Best of all, none of the above is "super complicated graduate/PHD level physics kinda stuff", and diving into the details doesn't involve any stuff like that either. Full disclosure: I have a BSc in physics ... but it's now over 40 years' old, and I didn't use it directly in any of the careers I had in those 40 years.
Still interested?
Posted
-
by
ramberts
in response to JeanTate's comment.
Sure, physics math is the same thing as finance math right? Haha just kidding! I like learning about stuff so yeah I'm down. 😃
Posted
-
by
JeanTate
in response to ramberts's comment.
Cool! 😃
In this post, dealing with some (obvious) realities ...
Imagine a spiral galaxy as a (very) thin disk, and no bulge, arms, rings, lenses, ... If viewed 'face-on' it would appear as a perfect circle (let's assume); viewed at an angle, not. To deal with this 'projection effect', instead of treating images of galaxies as perfect circles (perfectly symmetrical), treat them as ellipses. We can still derive radial intensity profiles fairly straight-forwardly, by 'correcting for' the elliptical shape (basically using the symmetry in the elliptical shape).
Now elliptical galaxies are not 'flat', and their true shapes are still an active area of research. However, for our work, we can treat them as looking elliptically-shaped, no matter which angle they are viewed from (perhaps they're like an American football/rugby ball, with two axes of (elliptical) symmetry). That means we can treat them just as we treat spirals.
The bulges of spirals introduce a slight complication, but not much: it turns out spirals' bulges are similar to elliptical galaxies, close to perfectly-round ones; their true shape is not much different from that of a ball. So spirals become composite objects, in images, an almost cirular bulge and an ellipse (being the disk, viewed from any angle), with the two having the same center.
The hard-working and clever SDSS people have developed computer code which 'looks at' all galaxy images and 'fits' three different radial intensity profiles (called a 'model') to the data, a 'de Vaucouleurs' one, an 'exponential' one, and a 'PSF' one (I'll get to the last one later). If all real galaxies were either perfect bulgeless spirals or perfect ellipticals, one model would fit perfectly and the other would not. Here's their DR12 page on this. Best of all, from our point of view, all the relevant model parameters - position angle of the fitted ellipse, its axis ratio, log-likelihood of the model fit, ... - are published, for each and every galaxy in SDSS, and you can download them for free!
Next: more real universe stuff, 'PSF' and the Sérsic profile; then GALFIT (etc).
Posted
-
by
JeanTate
Jumping to the end, GALFIT "is a data analysis algorithm that fits 2-D analytic functions to galaxies and point sources directly to digital images." In other words, you can construct a detailed model of your target galaxy (or galaxies), subtract from the original, and the residual will tell you what you missed (or how your galaxy is disturbed or otherwise apparently not normal). It runs on Linux, so if you have a Linux machine, you can do all the fitting and testing you want.
Something I should have covered earlier: the images we get to classify 'dumb down' the data, big time. The cameras used in SDSS, Hubble, etc output digital data, to a depth of much more than 8 bits (8 bits is all you get in JPG images, a mere 256 brightness levels per pixel). And astronomers commonly use a file format called FITS for images. Fortunately, there are quite a few (free) programs you can get which will read FITS files, and allow you 'play' with that data, to visualize it using different assumptions about scaling, etc. A particularly good one - used, by our fellow zooites who are co-authors of the SpaceWarps papers, in their research1 - is DS9, which runs on Windows, Linux, and more (unfortunately its user manual is a bit terse). If you are a bit more adventurous, you could get MONTAGE; I think it is (or at least was) used by the GZ Science Team to create some of the image sets we have been classifying (e.g. one of the Stripe 82 SDSS ones, and perhaps even the current GOODS ones).
Best of all: none of this is particularly complicated, and certainly doesn't require anything more than a good high school education (with sufficient mathematics). If you're good with computers, know a language such as Python, etc - none of which requires a university degree - you'll likely find it all surprisingly easy. 😄
'PSF' and the Sérsic profile will have to wait until tomorrow, sorry. 😦
Still interested? Questions?
1 Julianne Wilcox, Elisabeth Baeten, Christine Macmillan, and Claude Cornen; at least I think they used it; certainly the PI has used it. It's also turned up in Radio Galaxy Zoo, in posts by at least one of the Science Team members
Posted