DECaLS - new images and science
-
by
KWillett
scientist, admin, translator
I thought I'd open a discussion on the new Dark Energy Camera Legacy Survey (DECaLS) images on Galaxy Zoo. Please let us know if you have any questions - we'll hope to get a couple of the DECaLS science team in here as well.
Edited: here's the blog post describing the new images. http://blog.galaxyzoo.org/2015/09/21/new-images-for-galaxy-zoo-part-1-decals/
Posted
-
by
KWillett
scientist, admin, translator
One very neat place to start is the DECaLS Explorer - the extremely smart @dstn, of Carnegie-Mellon University, has created an interface to the DECaLS imaging that works much the same way that the SDSS navigator does. One of the really powerful things about DECaLS is that the way they both detect and characterize the sources in their images is through a sophisticated code called Tractor. You can see the models that it produces for each source by selecting different layers in the upper right of the DECaLS Explorer.
http://legacysurvey.org/viewer
Posted
-
by
KWillett
scientist, admin, translator
I've also added the link to the SDSS viewer back into the Examine page for all DECaLS galaxies. You can now compare the two images of the same galaxy more easily.
Posted
-
by
ElisabethB
moderator
Wow ! Tx Kyle ! 😄
Posted
-
by
williamaskew
Yes very glad to see the sdss link up.
Posted
-
by
JeanTate
in response to wtaskew's comment.
I second that! Thanks a million, Kyle 😄
Posted
-
by
KWillett
scientist, admin, translator
OK - many users have raised interesting questions about the images that are coming up, often with features that are obvious only in one or more bands (remember that the color images you're classifying are a combination of three different filters). Since the GZ Examine tool doesn't work on the current images, at least for the moment, I wanted to show how you can download the raw images yourself.
The raw images are stored in a format known as FITS, which stands for "Flexible Image Transport System". It's been the standard in astronomy for many years, although not used very commonly in other fields. Unlike formats like JPEG or PNG, FITS both preserves the measured values in the image (which we need to make accurate measurements of brightness) and can include metadata about the observation as part of the image.
The most common software for viewing FITS images is called DS9. It's free to download, and there are versions for Windows, Mac, and Linux. There are other ways to view and interact with FITS files, including common packages for IDL and Python.
So - how do you get FITS images for a DECaLS subject? You can do it from the web browser. Let's use AGZ000atp8 as an example. From its page on Talk, I can find the coordinates for this galaxy, which are RA: 318.35963, DEC: 0.97547. Now I'm going to insert those coordinates into the following URL in my web browser:
You can see in the middle of that URL that I've set the RA and dec coordinates for this particular galaxy; you can pick any point on the sky and just substitute those values in. The other two parameters here are pixscale, which is the physical scale of the image in arcsec/pixel, and size, which gives you the number of pixels per side in the square image you're going to download. You can also change either of those values if you want to change the image size or scale. So this URL is going to download a 512x512 pixel FITS image for a galaxy centered at RA: 318.35963, DEC: 0.97547, at a scale of 0.2 arcsec/pixel (which means that the image will be 102 arcsec, or 1.7 arcmin in size).
That URL will download all three individual filters as a single file; DS9 can load it as a "data cube", where each plane of the data corresponds to a different filter. That's useful since you can toggle back and forth between the different filters, or even play it as a movie.
If you just want to download a single filter - say the image in the r-band, for instance, just add it to the end of the URL:
Remember, the three filters you have available are g, r, and z.
This is a bit more work, but will give you the same data that are used to make the color composites. I hope some of you are interested in using this and exploring the data further, especially for objects where you see single-color features that might only come from one band. Let me know if you have questions!
Posted
-
by
JeanTate
in response to KWillett's comment.
Very cool Kyle; thanks for posting this! 😃
In my experience, DS9 is incredibly powerful, and can produce fantastic results. However, its User Manual is, um, underwhelming. Does anyone know of a good tutorial on DS9, one with lots of (worked) examples?
Posted
-
by
KWillett
scientist, admin, translator
(cringes)
Yeah - DS9 was never designed with newbies in mind, and better tutorials would be a fantastic resource. Not something that the astro community has managed to develop, though, to my knowledge.
Here's one that astrobites put out: http://astrobites.org/2011/03/09/how-to-use-sao-ds9-to-examine-astronomical-images/
And this is a good tutorial on IRAF, which has some DS9 stuff in it, but it's lacking in screenshots. http://www.twilightlandscapes.com/IRAFtutorial/
This'd be a good separate thread on Talk, especially if you have specific things you're trying to do or questions.
Posted
-
by
JeanTate
Check out today's GZ blog post, Searching for “tidal debris” in DECaLS images! 😃
It ends: "Please let us know, here or on Talk, what questions you have. Thanks for your help in tackling a new and interesting scientific problem."
I have a few questions, and hope that Sugata Kaviraj will drop by to answer them ...
- How can we tell if these faint features are real, and due to extragalactic tidal debris? The background in all ground-based optical astronomy images is not perfectly 'flat'; there are faint, extended regions that are due to many things other than tidal debris, e.g. scattered light from bright stars/the Moon/planets; airglow; MW 'cirrus'; extended faint MW nebulae.
- In regions where there are lots of galaxies along a line of sight/in a small field, how can we tell which galaxy/galaxies the tidal debris is associated with? An example of a 'false positive' like this (albeit a dust lane rather than debris) is M81 and Arp's Loop (see this APOD, for example)
- In DECaLS images, how to distinguish between tidal debris (mostly stars) and EELR (extended emission line regions; a.k.a. voorwerpjes)?
- Why are tidal debris features longer-lasting signs of a (past) merger than binary nuclei?
- Since, in the local universe, surface brightness is ~distance independent (as observed by optical telescopes), we should be able to spot tidal debris out to, say, z~0.4, right? As long as it is extensive enough ...
- How long is dust expected to survive, in tidal debris? As I understand it, within a galaxy it will be destroyed within ~100 million years or so (though I've forgotten what, mostly, destroys it, UV? sputtering? cosmic rays??)
- Follow on from #6: if tidal debris is seen against a background source (e.g. a distant elliptical), will it seem like a (foreground) dust screen/lane?
Posted
-
by
KWillett
scientist, admin, translator
in response to JeanTate's comment.
Good questions. I'll see if someone on the science team can give detailed answers.
Posted
-
hi Jean,
All excellent questions (and apologies for not able to reply sooner!).
How can we tell if these faint features are real, and due to extragalactic tidal debris? The background in all ground-based optical astronomy images is not perfectly 'flat'; there are faint, extended regions that are due to many things other than tidal debris, e.g. scattered light from bright stars/the Moon/planets; airglow; MW 'cirrus'; extended faint MW nebulae.
This is true. There can be faint features in deep images which are not tidal in origin. However, the features we are after (tidal tails, shells, dust lanes etc.) are typical signatures of mergers (we see these types of features routinely in high resolution simulations of mergers for example) and are (very) unlikely to be produced by anything else. Some examples of these features were given in Kyle's blog post: http://blog.galaxyzoo.org/2015/11/05/searching-for-tidal-debris-in-decals-images/?shareadraft=563a9a6eced7d).
In regions where there are lots of galaxies along a line of sight/in a small field, how can we tell which galaxy/galaxies the tidal debris is associated with? An example of a 'false positive' like this (albeit a dust lane rather than debris) is M81 and Arp's Loop (see this APOD, for example)
Very good question. The short answer is that sometimes we cant. Although if there are two galaxies in close proximity then the tidal debris really belongs to the merging system as a whole so perhaps its not important to associate the tidal debris with any one galaxy.
In DECaLS images, how to distinguish between tidal debris (mostly stars) and EELR (extended emission line regions; a.k.a. voorwerpjes)?
It could indeed be tricky to make this separation. However, tidal debris will appear at all wavelengths whereas the EELR will typically appear in the image that contains the emission line in question. Also tidal debris are much more common.
Why are tidal debris features longer-lasting signs of a (past) merger than binary nuclei?
Generally binary nuclei will coexist with tidal features. However, it is difficult to separate the merging nuclei after a point due to the finite angular resolution of the telescope. Since the tidal features are extended it is easier to pick them up.
Since, in the local universe, surface brightness is ~distance independent (as observed by optical telescopes), we should be able to spot tidal debris out to, say, z~0.4, right? As long as it is extensive enough ...
Correct.
How long is dust expected to survive, in tidal debris? As I understand it, within a galaxy it will be destroyed within ~100 million years or so (though I've forgotten what, mostly, destroys it, UV? sputtering? cosmic rays??)
I am not an expert on dust unfortunately but my understanding is that dust can be destroyed over a few dynamical timescales so a few 0.1 Gyrs. Note however that the tidal debris are formed out of stars. The dust and gas content of these features will generally be quite low. In any case what we see in the images is dominated by stars and not by gas or dust.
Follow on from #6: if tidal debris is seen against a background source (e.g. a distant elliptical), will it seem like a (foreground) dust screen/lane?
Dust lanes and tidal debris are formed out different material. Dust lanes are dominated by gas and dust while the tidal debris are created from the underlying stellar population of the satellite that is being accreted.
Hope this helps! And thanks a lot for classifying these images. The DeCals survey is just the first in a new generation of deep wide surveys that are about to become available, so we are entering a very exciting era of low surface-brightness science.
Sugata
Posted
-
by
sat666leg
good reading..
Posted
-
First red image of probably an asteroid for me, although not in one line with blue and green;
DECaLSQuestion: Why is the red image of asteroids so rare in DECaLS? When looking at SDSS asteroid they always seem to have 3 colors.
EDIT in fact answered by c_cld's asteroid identification example, may even be months in between different bands
Posted
-
Here is another great full volunteer. Thank you KWilett!
Leonie van Vliet
Posted
