Well, That Was Fast!

By Julianne Dalcanton | September 18, 2009 12:14 pm

Last week saw the first public release of data from the refurbished Hubble Space Telescope, and its new imaging camera (Wide Field Camera 3, or “WFC3”).

Over the past week, four papers have shown up on astro-ph using new WFC3 data of the Hubble Ultra Deep Field (see the prescient comment by Brian Mingus in the original blog post):

Bouwens et al
Oesch et al
Bunker et al
McLure et al

All of these papers are based upon data released on September 9th, from a large “Treasury” program to extend the wavelength coverage of the Hubble Deep Field. The first two papers were produced by the team that actually proposed the observations, and the second two were from groups that were sitting around eagerly waiting for the first group’s data to be publicly released1.

All of the papers deal with the statistics and properties of extremely high redshift (i.e. distant and young) galaxies. The dominant technique for finding high redshift galaxies has been looking for “drop out” galaxies. These are galaxies that have essentially zero flux in blue filters, due to absorption from intergalactic gaseous Hydrogen, and significant flux in all redder filters (at wavelengths that are largely unaffected by the same gas). Higher redshift galaxies “drop out” of progressively redder filters, because the rest frame (un-redshifted) wavelength at which the gas absorbs the galaxy’s light appears redshifted to longer and longer (redder) wavelengths for highly redshifted galaxies. This technique was pioneered by Guhathakurta, Tyson, & Majewski in 1990, and put on the map as a technique for galaxy selection by Chuck Steidel throughout the 90’s.

"Drop-outs" from the Hubble Ultra Deep Field

These early works looked for “U-band dropouts”, which turn out to be star forming galaxies at a redshift of 3 (about 2.5 billion years after the big bang). Subsequently, people had the bright idea to just keep pushing the drop out technique to redder wavelengths, to look for ever more distant galaxies. However, this comes at a cost, since more distant galaxies tend to be much fainter, so you need to work harder to have statistically significant detections in the red filters, and strong contraints on the absence of detections in bluer filters. The new WFC3 data pushes this technique out of the optical and into the infrared, selecting galaxies at redshifts from 6 to 9, when the universe was 0.5-1 billion years old. (Note: in the picture above, each successive column shows an image taken at redder and redder wavelengths. The likely distant galaxies are those that show up in the rightmost three columns but none of the leftmost columns)

The new papers all find that at these early times, the star formation rate of the universe is on the rise. This isn’t too surprising, given that you’re getting so close to the beginning of the universe — early on, structure hasn’t really had much time to form, so naturally you should find that fewer galaxies have yet had time to go about their business.

Evolution of the star formation history of the universe

All in all, these are nice results doing just what the new data was designed to do.

1 In Jackson Hole, I once saw a bald eagle sitting high in a tree above a stream. There was an osprey sitting lower down the same tree. The river guide said the eagle waits for the osprey to catch a fish, and then just steals the fish from the osprey.

2 After posting this, I found a nice write-up by Ron Cowan here, as well as discussion of the result from the always lovely Peter Coles here. There’s also a cute discussion of the stress involved in such publications over at andxyl’s place.

  • Sili

    Re footnote: Ouch! Why not link to the PAMELA brouhahah, too?

  • http://cosmicvariance.com/julianne Julianne

    The footnote is actually not meant to be particularly judgmental. Both are effective strategies for getting fish.

    Moreover, the analogy is imperfect in that in Nature, either the osprey or the eagle gets the fish, and the other one goes hungry. In science, both groups get to make a tasty meal from the same data. Both groups will get citations. And by competing, both will wind up shaking down possible shortcomings in the other’s work. The cost is perhaps some lack of thoroughness when you’re rushing to eat your fish before someone else does, and a higher stress level when you’re all amped up about whether Group B is going to get their paper onto arXiv a whole two days before yours.

    Astronomy is much more open about such things than Physics, since there has been a strong recent history of data being made public for all to work on. The community has largely decided that the benefits above outweigh the costs. Thus, there is nothing unexpected or improper about the eagle-group.

    It’s still a funny Nature story, though!

  • http://lablemming.blogspot.com/ Lab Lemming

    By capitalizing Nature, are you referring to the journal?

  • http://blog.denniswilliamson.us Dennis

    “Moreover, the analogy is imperfect in that in Nature, either the osprey or the eagle gets the fish, and the other one goes hungry.”

    The analogy is better than that. The other does not necessarily go hungry and die. He could go hungry and get better at his technique (or uses a different one).

  • http://cosmicvariance.com/julianne Julianne

    But sometimes you really want to eat that particular fish. And besides, you need a snack after all that hunting.

  • http://telescoper.wordpress.com Peter Coles


  • Julianne


  • eric gisse

    Regarding Hubble and the WFC series of cameras…

    Why can’t we build a handful of Hubbles and incrementally upgrade the hardware and launch them? I’d love to see how much science we could get done if there was a half dozen Hubble-class telescopes in orbit, of increasing utility.

    If one goes tits up, it isn’t the end of the world. No need to spend billions of dollars and have heroics to save one instrument with six others up there. I refuse to believe nobody has thought of this.

    Is the Hubble really so big that we couldn’t toss a new one into a Delta 5 or, god forbid, an Arianne or whatever Russia’s workhorse lifter is?

  • http://andyxl.wordpress.com Andy Lawrence

    Was that really Peter C calling someone else a flirt ?

  • Bill

    The HST is 26,000 lbs and cost several $billion to build. The short answer to eric gisse is, “no bucks, no Buck Rogers”.


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Cosmic Variance

Random samplings from a universe of ideas.

See More

Collapse bottom bar