CANDELS: The new data in Galaxy Zoo

This post is the second of a series introducing the new Galaxy Zoo.The first is here, and you should come back in the next few days for more information about our fabulous new site. This post is also part of Citizen Science September at the Zooniverse.

When we look at nearby galaxies, we see several familiar shapes. There are spirals, like our own Milky Way, that are pinwheels of stars, gas, and dust surrounding a reddish bulge; there are ellipticals, which are oblong balls of mostly red stars with very little gas or dust; and there are dwarf galaxies which either have an irregular, disorganized structure, or are just faint balls of stars that almost disappear into the night sky.

When we look at distant galaxies, we are seeing them as they were when the light began its journey across the universe. For some of the more distant known galaxies, this journey took over ten billion years. We are thus seeing these galaxies in their youth. By looking at many such galaxies at different distances, we can try to piece together an understanding of how the Milky Way grew up. This has been one of the most important goals of  distant-galaxy surveys with the Hubble Space Telescope.

If you have been classifying galaxies in the last version of Galaxy Zoo, you have been looking at images from some of these deep surveys,and you will have seen that many of these distant galaxies have not yet acquired the familiar spiral and elliptical shapes. Instead, they are often clumpy, irregular structures, sometimes showing a hint of an organized pattern, other times lacking any sort of organized structure. Sometimes they look like two galaxies colliding and merging together. Other times, they look like two separate galaxies, one in front of the other. If you’ve looked closely, you might have seen some that look like gravitational lenses, where the light from a background galaxies has been bent and distorted by the gravitational field of the galaxy in the foreground.

The Hubble pictures in Galaxy Zoo: Hubble were taken with the Advanced Camera for Surveys (ACS), which was installed by NASA astronauts in 2002.

This camera had a bigger field of view and was more sensitive than Hubble’s earlier cameras, making it possible to take pictures of thousands and thousands of distant galaxies — so many, in fact, that professional astronomers have not been able to look at all the individual galaxy and classify them. That is why they have turned for help to the Galaxy Zoo.

The ACS were taken in visible light. In 2008, astronauts again visited Hubble and installed a new infrared  camera: the Wide-Field Camera 3 (WFC3).

Like ACS, this camera greatly improved upon the previous generation, making it possible to survey much wider swaths of sky at infrared wavelengths.

One of the most ambitious surveys ever undertaken with Hubble is the Cosmic Assembly Near-infrared Deep Extragalactic Legacy Survey (CANDELS), which is in the second year of a 3-year program using WFC3 to obtain detailed infrared images of distant galaxies. You can keep up with news from the survey on the CANDELS blog.

Why are these images important? Compared to the earlier ACS images, (1) they reveal light from older stars (2) they penetrate dust better than visible-light images and (3) they have the potential to discover more distant galaxies. Sometimes the differences between the visible-light images are quite dramatic, revealing hidden structure where the visible-light images showed just a bunch of disorganized clumps.

Now astronomers need your help! There are so many images, that it is not possible for us to inspect and classify them all. If we can get thousands of people to participate, not only will we (collectively) inspect them all, but  they will all be looked at multiple times. For some galaxies, everyone will agree on the shape and structure. For others, people will disagree — which is in itself informative. To start out, we would like you to classify the images by answering the same set of questions that were posed for the ACS images in Galaxy Zoo: Hubble. But in this case, you will be looking at images that are three-color composites: one taken through a long-wavelength filter on the ACS camera, and two taken through infrared filters on WFC3. Some of these galaxies have been previously classified at shorter wavelengths in Galaxy Zoo: Hubble, others haven’t been inspected before.

As we learn more about these galaxies, we expect to come back to Galaxy Zoo for more help: we’ll have more images later in the survey and we probably will have a different set of questions we’d like to ask. Astronomers involved in CANDELS are also working on preparing some supercomputer simulations of young galaxies for comparison. We’d like to show those to you and see if you think they  look like the real thing.

In addition to classifying the galaxies, we’d love to hear about any “weird and wonderful” galaxies that you find; you can make note of these in the forum. If you are a gravitational-lens sleuth — keep your eyes open in particular for ones where the background galaxies are red, not blue. Those could be very distant galaxies indeed!

So, go forth and classify!

Harry Ferguson, CANDELS Co-Principal Investigator
(posted by BorisHaeussler on Harry’s behalf)


12 responses to “CANDELS: The new data in Galaxy Zoo”

  1. Jean Tate says :

    What a fantastic project!!!

    I am a bit puzzled by one thing, however. A three-color, RGB image of a galaxy at z~1 won’t correspond to one of a local (z~0) galaxy, even if they are exactly the same galaxy. Why? Because a) we won’t see most of the fainter detail in the more distant galaxy (due to the fact that the surface brightness will be below the detection threshold), and b) the ‘R’, ‘G’, and ‘B’ wavelengths won’t be the same (an RGB image of the z~1 galaxy would be more like a GBUV, or BUVUV image of the z~0 one). And the differences will be more pronounced as z increases.

    We have seen this even in SDSS images (the original Galaxy Zoo): spiral galaxies at z ~ 0.3 have green arms, not the blue ones we see for those at z~0; all galaxies at z~0.5 are (almost purely) red.

    So, how can classifications of high-z galaxies be compared with ones at low z?

    • Boris Häußler says :

      I think that’s the point why we now use CANDELS data.
      a) CANDELS data is HST data, so better quality regarding both PSF size and image depth (because HST does not have to worry about sky brightness), so you WOULD see the fainter stuff, simply because you’re looking at different (and deeper) data.
      b) CANDELS uses WFC3 data, that means it is infrared data. It’s not entirely true, but the idea is that if you look at IR, you are looking at the same intrinsic (restframe) wavelength that you look at in optical wavelengths at z~0. The CANDELS images are IJH composites, to balance out the redshifting effect as well as possible.

      • Jean Tate says :

        Thanks Boris! 🙂

        It’s taken me FOREVER to find the info, but I think I’ve finally got it.

        “I” is actually the same “i” as in the 2X8 collage in the blog entry; i.e. the F775W filter, and is NOT one of the 18 (?) on WFC3!

        “J” is F125W, and “H” is F160W; both are IR filters available on WFC3.

        Assuming the effective central wavelengths are coded into the filters’ names, and using the data on the SDSS Filters webpage, the redshifts (z) at which the various central wavelengths will ‘move over’ are as follows (assuming SDSS g -> B, r -> G, and i -> R, for z = 0):
        z=0.23 (G -> R)
        0.3 (B -> G)
        0.6 (B -> R)

        That’s pretty close to what I wrote earlier.

        For CANDELS, I get these:
        z=0.64 (B -> I)
        1.0 (G -> J)
        1.1 (R -> H)

        z=1.62 (B -> J; G -> H)
        2.2 (B -> H).

        So a CANDELS IJH image will be – very roughly – similar to a restframe (z=0) SDSS gri one, for redshifts somewhere between ~0.7 to 1.1. Did I get it right?

        And a CANDELS IJH image will be similar to a z ~0.3 SDSS gri one for z ~1.6.

        Finally, the most distant objects – the approximate equivalents of an SDSS z ~0.6 – will be a CANDELS z ~2.2.

        If that’s (more or less) right, I’ll see if I can do the surface brightness calculations …

        Oh, and this takes no account of where the 400nm break falls, nor what the filter (etc) transmission characteristics are (especially near the redshifted 400 nm break) …

  2. Jean Tate says :

    Today’s Object of the Day, in the Galaxy Zoo forum, features this blog entry. It’s called “Spirals, Forever Spirals”. Here is the URL:

  3. Tom zolotor (@FreeTheSoulss) says :

    Often the fussy galaxies in Hubble CANDELS can be called FHB galaxies and tag as #FHB in TALK. 🙂

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 )

Facebook photo

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

Connecting to %s

%d bloggers like this: