24
Feb 2012

How Science Works: Is GJ1214b a Waterworld?

POSTED BY: S. DOMAGAL-GOLDMAN
 

Many of you will have heard of the “waterworld” (IMO, much more interesting than this waterworld) that was confirmed with a publication earlier this week by Berta et al. What’s cool about this story to me as a scientist is what a beautiful example it is of the scientific method in action. This is far from the first publication on this planet: a quick google scholar search for “GJ1214b” in the title returns 11 results. The planet first appears in references in this Nature article, announcing its discovery as a planet with 6.55 the mass of the Earth and 2.68 times Earth’s radius. Putting these two values together, you can get at the density of the planet, and doing that allowed the authors to draw a hypothesis about its composition: “We find that the planetary mass and radius are consistent with a composition of primarily water enshrouded by a hydrogen–helium envelope that is only 0.05% of the mass of the planet.” A loose re-wording of this might be “Based on the density of this planet, we hypothesize it is a waterworld.” Because this planet transits its parent star – and because it is fairly close to our Solar System – their hypothesis on the planet’s composition was capable of being tested with follow-up observations.

The community entered the game at this point, providing alternative hyptheses and models showing how they could be tested. First, Rogers and Seager published a comprehensive suite of geophysical models, the results of which suggested three possibilities for the composition of GJ1214b: 1) a hydrogen-helium atmosphere overlaying an rocky, icy core; 2) a hydrogen atmosphere overlaying a rocky planet or 3) a water-rich atmosphere overlaying a water-rich interior. Then, Miller-Ricci and Fortney (now Kempton and Fortney) published atmospheric models and corresponding model spectra (roughly, the amount of different “colors” leaving the planet). The importance of these models is they demonstrated how observations could be used to test the atmospheric composition. Specifically:

Space-based observations revealing a flat spectrum for GJ 1214b wouldtherefore rule out the presence of a cloud-free hydrogen-rich atmosphere—although H/He atmospheres with haze or cloudlayers that extend high into the atmosphere may not be excluded.However, spectra taken at a precision of at least several parts in 105 would be necessary to discriminate between atmospheres composed of H2O, CO2, a mix of these two gasses, or something else entirely. Detection of water features between 1 and 3 μmand the CO2 features at 4.3 and 15 μm may prove to be the mostuseful in making this distinction.

Do you see what’s happening here? A set of observations led to a hypothesis that the planet had a water-rich atmosphere. A second set of researchers further presented alternative hypothesis of a hydrogen-/helium-rich atmosphere. A third set of researchers then stepped up with independent models of observations, and put together instructions for discriminating between the two hypotheses…. which made it time to go out and get more data.

There were a series of observations, the most recent set of which are still in press but available here. Despite impressive efforts (by Jacob Bean in particular), the data collected had largely been inconclusive or contradictory…. until the paper this week was published. That paper was written by both the original discovery/hypothesis team and the team that published the atmospheric models that allowed for the hypothesis test. It leveraged data from the Hubble Space Telescope in a new wavelength range to remove the ambiguity from prior observations, and knock the hypothesis test out of the park. Take a look at the last figure in their paper, which shows spectral information (again, think “color distribution”). Specifically, it shows spectral models (blue and red lines) and observational data (grey and black shapes):

(and models, too!)

The blue line is a model spectrum for a “water-rich” atmosphere and the red line is a model spectrum for a “hydrogen-/helium- rich” atmosphere. The grey shapes represent prior observations, most of which were more consistent with the blue line… but some of which were consistent with the red line or with neither. Enter the black circles, which are the observations from Hubble. Simply put, these are in wonderful agreement with the “flat” spectrum of a water-dominated atmosphere, and in strong disagreement with the “solar” atmosphere dominated by hydrogen and helium.

I love this story. Observations led to multiple hypotheses, test for them were developed using modeling tools, and follow-up observations were made until one hypothesis was strongly favored over the others. This is how it’s supposed to work, folks. Sometimes science is messy and meandering, but when hypotheses beget predictions beget further observations that confirm and/or deny the hypotheses… the power of the method is clear.

What’s more, the hypothesis is a provocative one, our ability to test it is awesome, and this whole exercise is a preview of yet more awesomeness to come. We are able to confirm the atmospheric composition of a planet that is light years away from us. This has been done before, but never for a planet as small (in size) and cool (in temperature) as GJ1214b. Consider this a preview of things to come. And when JWST unfurls its mirrors, it will be able to do this for even smaller, cooler planets. This will allow us to do this for (relatively) nearby potentially habitable planets, and start to confirm their habitability. And that will be awesome, in both senses of the word.

5 Comments  
Share
 
 
S. Domagal-Goldman Posted by
S. Domagal-Goldman
read more
 
 
  • http://people.ku.edu/~dimitra Dimitra Atri

    How robust are these models? I am not an expert but I guess one has to make a lot of assumptions and stretch them beyond prescription in some cases?

  • Emily

    Great post Shawn! I just couldn’t stop thinking: “decreasing the suck, increasing the awesome!”

    • http://www.paleblueblog.org S. Domagal-Goldman

      :-)

  • http://www.paleblueblog.org S. Domagal-Goldman

    Dimitra, there’s uncertainty in the models, but not nearly to the scale of the discrepancies between the blue and red lines at the wavelengths Hubble observed at. Although, to be totally honest no one has quantified those uncertainties. I’m proposing to start building such a capability in an upcoming proposal. We’ll see if it flies…

  • http://people.ku.edu/~dimitra Dimitra Atri

    Good luck!

Pale Blue.blog
 
RSS Twitter Facebook
paleblue.you posts
Characterizing the “Missing Era” of Prebiotic Evolution
by Sara Imari Walker
Posted on: 13, Apr 2012
Water flows on the surface of Mars.
by Noah Hammond
Posted on: 11, Apr 2012
Found the Higgs… now find the aliens!
by Adi Chopra
Posted on: 4, Apr 2012
View more
post on paleblue.you
  • Astrobiology Magazine Top Story
    PaleBlue.you