New Sloan Digital Sky Survey Galaxies in Galaxy Zoo

The relaunch of Galaxy Zoo doesn’t only include the fantastic new images from the CANDELS survey on Hubble Space Telescope, but also includes over 200,000 new local galaxies from the Sloan Digital Sky Survey. We’ve had a lot of questions about where these galaxies came from and why they weren’t put into earlier versions of Galaxy Zoo, so I thought I’d write a bit about these new images.

The Sloan Digital Sky Survey Project (SDSS) is currently in its 3rd phase (SDSS-III). You can read all about the history of SDSS here, and here, but briefly SDSS-I (2000-2005) and SDSS-II (2005-2008) took images of about a quarter of the sky (which we often refer to as the SDSS Legacy Imaging), and then measured redshifts for almost 1 million galaxies (the “Main Galaxy Sample”, which was the basis of the original Galaxy Zoo and Galaxy Zoo 2 samples; plus the “Luminous Red Galaxy” sample) as well as 120,000 much more distant quasars (very distant galaxies visible only as point source thanks to their actively accreting black holes).

Following the success of this project, the Sloan Digital Sky Survey decided they wanted to do more surveys, and put together a proposal which had four components (BOSS, SEGUE2, MARVELS and APOGEE – see here). To meet the science goals of these projects they realised they would need more sky area to be imaged. This proposal was funded as SDSS-III and started in 2008 (planned to run until 2014).  The first thing this new phase of SDSS did was to take the new imaging. This was done using exactly the same telescope and camera (and methods) as the original SDSS imaging. They imaged an area of sky called the “Southern Galactic cap”. This is part of the sky which is visible from the Northern Hemisphere, but which is out the Southern side of our Galaxy’s disc. It totals about 40% of the size of the original SDSS area, brining the total imaging area up to about 1/3rd of the whole sky. The images in it were publicly released in January 2011 as part of the SDSS Data Release 8 (DR8 – so we sometimes call it the DR8 imaging area).

This illustration shows the wealth of information on scales both small and large available in the SDSS-III’s new image. The picture in the top left shows the SDSS-III view of a small part of the sky, centered on the galaxy Messier 33 (M33). The middle and right top pictures are further zoom-ins on M33.
The figure at the bottom is a map of the whole sky derived from the SDSS-III image. Visible in the map are the clusters and walls of galaxies that are the largest structures in the entire universe. Figure credit: M. Blanton and the SDSS-III collaboration

We have selected galaxies from this area which meet the criteria for being included in the original Galaxy Zoo 2 sample (for the experts – the brightest quarter of those which met Main Galaxy Sample criteria). Unfortunately in this part of the sky there is not systematic redshift survey of  the local galaxies, so we will have to rely on other redshift surveys (the most complete being the 2MASS Redshift Survey) to get redshifts for as many of these galaxies as we can. We still think we’ll get a lot more galaxies and, be able to make large samples of really rare types of objects (like the red spiral or blue ellipticals). Another of our main science justifications for asking you to provide us with these morphologies was the potential for serendipitous discovery. Who knows what you might find in this part of the sky. The Violin Clef Galaxy is in the DR8 imaging area and featured heavily in our science team discussions of if this was a good idea or not.

And interesting things are already being found in just a week of clicks. The new Talk interface is a great additional place for us to discuss the interesting things that can be found in the sky.

For example this great system with tidal tails and a Voorwerpjie:

this weird triangular shaped configuration of satellites:

and an oldie (but a goodie) in the beautiful galaxy pair of NGC 3799 and NGC 3800 (NGC 3799 in the centre, NGC 3800 just off to the upper right):

and just this morning I discovered the discussion of this really unusual looking possible blue elliptical (IC 2540):

There are also rather more artifacts and odd stuff going on in these new images than I think we saw in the SDSS Legacy sample (from GZ1 and GZ2). Remember these are completely new images you are looking at. It really is true that no-one has looked at these in this level of detail (or perhaps ever) before. The original sample had a sanity check at some level, since when GZ1 ran the majority of the sample had already been targeted by SDSS for redshifts (so someone had to plug a fibre into a plate for each galaxy). In this new imaging all that has happened is that a computer algorithm was run to detect likely galaxies and set the scale of the image you see. Sometimes that mistakes stars, satellite trails, or parts of galaxies for galaxies. Always classify the central object in the image, and help us clean up this sample by using the star/artifact button.

And you can enjoy these odd images too. I like this collection of “GZ Pure Art” based on just odd things/artifacts classifier “echo-lily-mai” thought were pretty. :)  If you get confused by anything please join us on Talk, or the Forum where someone will help you identify what it is you’re seeing.

Tags: , , ,

About karenlmasters

Astronomer at Institute of Cosmology and Gravitation, University of Portsmouth. Project Scientist for Galaxy Zoo. Also member of Sloan Digital Sky Survey, Dark Energy Survey and Euclid.

21 responses to “New Sloan Digital Sky Survey Galaxies in Galaxy Zoo”

  1. zutopian says :

    The presented images are all available also in DR7. I think, that it would have been better to present just some of those images, which are only available in DR8. I had asked in a comment to another blog post, why there are some old images in the new GZ. ZK Chris explained in a reply, why it is the case.
    - 1st image: It is one of the galaxies from the Voorwerpje paper!
    - 4th image: It is not IC 2540, but IC 2520! It was presented as an “Object of the Day” on the forum in 2007!

    Here are the links:
    - Blog post:
    - Voorwerpje galaxy, OOTD:
    - IC 2520, OOTD:

  2. karenlmasters says :

    zutopian – thanks for your comments (and the typo correction). A small fraction of older images were also included to ease cross comparison between these morphologies and the old Galaxy Zoo sample. However the vast majority of these images are galaxies entirely new to Galaxy Zoo.

    Ironic I picked such bad examples (and I apologise for not checking more carefully), but it was completely innocently done that I got all old DR7 images! Ooops.

    • Jean Tate says :

      It’s soooo wonderful to have a whole lot of new galaxies to oogle at! :)

      Here’s a suggestion: why not invite a seasoned non-astronomer zooite to join the Galaxy Zoo team? Or perhaps establish some sort of advisory panel? There seem to have been rather a lot of “oops!” things related to the new Galaxy Zoo (see the A thread to talk about Talk – – for more examples), most of which almost certainly would have been avoided by having such a zooite, or panel, aboard.

      In fact, I have a very well suited person in mind: zutopian :D

      Jean Tate

    • zutopian says :

      1st image:
      Why was an image, on which the nucleus isn’t the central object, selected for “reclassification” of the Voorwerpje Galaxy?

      2nd image:
      No comment

      3rd image:
      Why is the lower part of the GZ image missing, though it is available in DR8?

      4th image:
      The nucleus isn’t in the center of the image. I guess, that this image is displayed in the new GZ, because DR7 says STAR, but DR8 says GALAXY.: ZK Chris had mentioned in a reply (another blog post) about such cases. The central object looks like a starforming region.: I am not sure, if we should classify the galaxy or click “STAR”?

  3. Jo Echo (@EchoLilyMai) says :

    Hello, great blog. Thanks for the mention! I can’t really take the credit for this though as there is a pure art thread over on the forum. If you zoom in on certain details on those found images in the SDSS they make great little artworks!

    These are just my pretty artefact finds. I’m glad you like them too. Enjoy :)

  4. zutopian says :

    In the blog post is said.: “Always classify the central object in the image, and help us clean up this sample by using the star/artifact button.”

    The central object in the 1st image seems to be a star! So I would have clicked the “Star” button!
    I guess, that there is also an image, on which the nucleus of the galaxy is in the center, displayed for classification.: Then I would have clicked “Merger” and/or “Disturbed” and “Other” (Voorwerpje) as Odd features.

  5. zutopian says :

    Jean Tate wrote.: “In fact, I have a very well suited person in mind: zutopian.”

    Thank you for your compliment.
    I think, that actually you are the right zooite, who should be involved in the GZ Team. I am impressed, how much knowledge in astronomy you have. Reading a lot of posts by you, I had guessed, that you were studying astronomy! Well,if you need an assistant in the Zoo someday, I would volunteer!
    The other OOTD posters are also very good! I am also impressed of ccld’s an other zooites’ knowledge.

  6. zutopian says :

    “and help us clean up this sample by using the star/artifact button.”

    There are so many!
    Objects from areas like the “scarf area” shouldn’t be displayed!
    OOTD about the “scarf”:

    • Brooke Simmons says :

      There are indeed many artifacts, as Karen said. We’re working on ways to re-filter the data so as to remove as many of the affected images as possible, while retaining all the images that are new and/or deeper than the DR7 images. Your clicks are helping with that — no click is wasted, even if you’re just flagging an artifact!

      Must say, though, that if I were presented with the image above with the Voorwerpje in it, I’d classify the galaxy that passes through the center of the image and then flag on Talk that there might have been some confusion of the galaxy and a foreground star by the SDSS pipeline.

      • Jean Tate says :

        I’ve a suggestion that might help with some of these: add an option to put cross-hairs onto the image.

        Sure, that’d likely be a huge extra development burden, but I’ve encountered several new Zoo images where it was – for me – impossible to decide what the ‘object in the center of the image’ was! E.g. AGZ000538c (bright star? or faint blob?) In these cases, you have to guess … but cross-hairs would help, as would a cut-out that zooms more closely to the ‘object-in-the-center’.

  7. karenlmasters says :

    For the SDSS images, if you follow the Galaxy Zoo Examine link you can zoom in on the Skyserver image. In most cases the central object is pretty obvious, so just do the best you can.

    • Jean Tate says :

      That’s certainly true … though a big downside is that that takes three (or perhaps two) clicks, none of which is where your mouse is (though your hand gets used to doing this, after a dozen or so times), and each of which opens a new window/tab (all of which you need to close, to avoid having your browser eventually crash). And I wonder how many zooites are actually aware of all this?

      • Jean Tate says :

        Oh, and you can only go check on an SDSS object AFTER you’ve finished classifying … at which point discovering that you’ve classified the wrong one becomes moot …

  8. zutopian says :

    karenlmasters says :September 21, 2012 at 9:34 am
    “For the SDSS images, if you follow the Galaxy Zoo Examine link you can zoom in on the Skyserver image. In most cases the central object is pretty obvious, so just do the best you can.”

    As far as I know, there is no link to the Skyserver while classifying. There is a link avaible after classifying. Well, in the URL adress of the new GZ images the Obj-ID (DR8) is given. So one can copy it and look at the image in Skyserver while classifying. So one can make sure, which the central object is. Sometimes it is confusing to look at the GZ images and also the Skyserver images while classifying.: Not concerning the central object. The images have different rotations and the image qualities are different.

    PS: Concerning the different image qualities, you can read, what
    thebamf says :September 12, 2012 at 1:29 pm in a reply to another blog post.:

  9. zutopian says :

    Brooke Simmons says :September 21, 2012 at 9:18 am
    “Must say, though, that if I were presented with the image above with the Voorwerpje in it, I’d classify the galaxy that passes through the center of the image and then flag on Talk that there might have been some confusion of the galaxy and a foreground star by the SDSS pipeline.”

    DR8 says, that the central object, which looks like a star, is a Galaxy.

  10. karenlmasters says :

    We don’t show you images of things DR8 says are stars – everything in the sample has been classed by the pipeline (correctly or incorrectly) as a galaxy. So the clicks identifying stars help disentangle things. We don’t waste any of your clicks.

    Please just do the best you can identifying the central object. You coud use the clicks through to the SDSS to learn from difficult cases and apply that to your future classifications.

  11. zutopian says :

    The new GZ is also announced in a SDSS blog post.:

  12. zutopian says :

    I am confused.:

    klmasters posted in Talk.:
    “We applied the criteria of the difference between PSF magnitude and extended magnitude larger than a certain amount to identify things in DR8 which were not likely to be stars. This was the same as used by DR7 and earlier. I’m not sure why then the DR8 ID is different – perhaps they changed the criteria slightly.”

    ZK Chris had posted as a reply in another blog post.:
    “The second reason is that there seem to be some that are accidentally included as new images in the SDSS DR8 due to changes in how the pipeline decides what is and what isn’t a galaxy.”

    • karenlmasters says :

      It’s probably to do with photometric scatter. Although the DR7 images are (mostly) the same images, they have been completely re processes with new algorithms to improve the subtraction of sky light and make fainter extended emission more obvious. So things which in DR7 just made the star/galaxy cut might now just be on the other side of the cut.

      It’ll just be a tiny fraction of the images.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


Get every new post delivered to your Inbox.

Join 17,568 other followers

%d bloggers like this: