Dust in the Zoo – chapters opening, continuing, and closing

Anna Manning and I are back at Kitt Peak, using the 3.5m WIYN telescope for
more observations of overlapping-galaxy systems from the Galaxy Zoo sample.
This trip started with an unexpected dust encounter. Indulging my fascination with some of the technological excesses of the Cold War, I dragged Anna (and my mother-in-law as well) to Tucson’s Pima Air and Space Museum. I particularly wanted to see their newly-restored B-36 aircraft, one of only 4 of these vintage giants left. The wind had been high already, but really whipped up and caught us in a dust storm (with added rain so it was like tiny mud droplets stinging the skin). Anna pointed out the irony, especially since I had announced on Twitter that “dust will be revealed, in detail”. Maybe next time it is I who should be more detailed.

As I recently announced on the forum, the sample gathering for overlapping galaxy systems from the classic Zoo (1 and 2, both using the Sloan survey) is now complete. Keeping some statistics on systems reported on the forum, and on how many were reported for the first time, shows that we had reached a point of diminishing returns before the switch to Hubble Zoo. Our working list now contains about 2187 pairs (there may still be a few double entries to weed out before giving an exact value). We’re well into preparation of this catalog for publication, after refining it to remove pairs with evidence of interaction, or where the background galaxy is so small or distant as to backlight only a tiny region of the foreground galaxy. (We’ll retain these in a complete electronic version with appropriate codes).

The catalog includes some special groups, usable because we have so many to select from. There are pairs with such large redshift differences that they cannot be interacting with each other, and guaranteed to be the kind of distant superpositions we would like to use to measure dust absorption; there is a pending Hubble proposal to do “snapshot” images of some of these. There are pairs of edge-on galaxies crossing at sharp angles (X-pairs), which show us how think their dust layers are. There are pairs with elliptical or S0 background galaxies, ideally suited to measuring foreground dust because they are smooth and symmetric. These differ in which questions they can be used to address, and in just what measurements we need to derive the effects of dust. Here are some examples, from data we took in the previous observing sessions at the WIYN telescope. As so often happens, 10% of the work is in deriving the answer, and 90% in working out the error range on that answer. For these, the error is mostly due to departures from symmetry in the two galaxies, which we estimate using the segments of the galaxies that are not overlapped. Each graph shows the fraction of light transmitted (in the blue B and deep-red I bands) as a function of radius within the foreground galaxy.

Arp 198 is what we have started calling a Q-type pair, where the background galaxy is seen edge-on and almost radial to the foreground one. When we first used this technique on small numbers of galaxies in the 1990s, I did not yet appreciate how useful such pairs can be. The small amount of foreground galaxy backlit for examination is balanced by the fact that we only need to interpolate across a small wedge in angle to estimate its light in front of the background galaxy, so we get unusually accurate measurements of the changes in dust absorption between spiral arms and interarm regions. In favorable cases, we can go beyond these averaged radial dust profiles and map the dust absorption in two dimensions – that is, make a full map, spiral arms and all, of the backlit part of a galaxy.

Our observations now are especially targeted at a subgroup which was selected for entirely different reasons. The ultraviolet extinction in galaxies due to dust is well known only in certain cases such as the Andromeda galaxy or starbursts, and increasingly important as Hubble data show us the high-redshft Universe in light which started out in that spectral band. In a part of this project funded by NASA, we are using survey data from the GALEX satellite observatory to do parallel analyses in the ultraviolet. New ground-based images will provide not only the visual-light anchor measurements, but help us to interpret the GALEX data whose resolution is not nearly so fine, so we need additional data to know which parts of the image are arm or interarm, and which parts we can use for assessing galaxy symmetry. Accordingly, our target list is strongly weighted toward galaxy that are bright in the UV or that are in fields where GALEX has obtained long exposures to give sensitive images. Because elliptical galaxies are so faint in the ultraviolet, we cannot use them as background sources. This means we have to use spiral/spiral pairs, and put up with larger errors due to departures from symmetry than with elliptical background sources. This means we will have to trade exposure length for a larger sample size to get mean properties. Watch the forum and blog for updates during this observing run.

The ultraviolet data are a stepping stone that will help with a major comparison which we can only now start, after having had it in mind for a couple of years – dust properties in the local versus early Universe, which observationally means SDSS versus Hubble. Some overlaps in the Hubble fields are already identified, thanks to the massive efforts put into using the worlds largest optical telescopes to measure spectra of thousand sf galaxies there. Where we see two quite different redshifts in one spectrum, we know that we are dealing with an overlap. But there remain many galaxies with resolved structure and no spectral data, or where the galaxy cores are far enough apart not to appear in one spectrum for automated distinction. One again, putting together this sample will be an ideal project for Zooites, reported in the forum (in this thread). The goal is to see how much difference we find in dust properties with cosmic time for similar galaxies, which then factors in to how we estimate the galaxies’ history of star formation, overall mass of interstellar matter, and changes in stellar population. (People doing supernova surveys are also interested, since this could yield corrections they would like to apply to measured supernova properties in distant galaxies).

There has been a lot going on behind the scenes with the dust project. I’ll leave off for now with a particularly interesting tidbit. Serendipitous observations in a Hubble survey turned up one spiral galaxy with a dusty disk extending even beyond the apparent end of the stellar disk. We know that these cannot be common or Zooites would have seen them easily, but it is very interesting to know how common they are and start to find out what makes them different. The Galaxy Zoo sample has produced what look like two more examples of such extended dusty disks. One we already observed from WIYN, and it was only on inspecting those results closely that it became clear that there is a reddened region centered on the foreground galaxy but well beyond its spiral pattern:

The other, also known as UGC 4008, is ordinary-looking in the SDSS images, but its companion galaxy to the north shows an odd region of missing light on the southeastern side. This looks a bit like dust in front of it. Long-exposure imaging with the SARA 1m telescope confirms that UGC 4008 has a very dim disk just reaching out to this area, which you can perhaps convince yourself of with the SDSS data shown just below. We plan to look at UGC 4008 more closely this week. Stay tuned for more updates…