Monthly Archives: September 2013

exoplanets.org is hiring!

I need to hire a computer / IT person for long-term, very part time work (~10 hrs/month) to manage exoplanets.org.


The backend was all written by the very capable Onsi Fakhouri, who built the site as a graduate student and maintained it at cut-rate prices for years as a labor of love.  Alas, his job and family commitments no longer allow him to keep this hobby, and we are grateful for his service.  So, a new era dawns, and we need a new Webmaster.

And there’s lots to do!  We have some minor bug requests (error bars don’t work across units on the cvs output, line thicknesses can be hard to change in Chrome, etc.) and lots of new features to implement (upper limits!  Better merging of Kepler and EOD planets!)

Here are the specs for his replacement:

Webmaster needed for long-term, part-time work to upgrade and maintain the backend for educational/research website at www.exoplanets.org.  Job may also include additional Web design for this site.  The successful applicant will work as a consultant using their own computer and bill hourly on a monthly basis. 
The ideal applicant will have, at a minimum:
 
– Moderate experience with Ruby.  
– Extensive experience with large complex Javascript codebases (jQuery in particular).
– Basic experience with version control (GIT).
– Experience provisioning servers and deploying web applications to servers (Google sites in particular). 
– Good visual and UI design sense
– Familiarity with visual display of scientific information (error bars, log scales, upper and lower limits, units of measure, etc.)
– Flexibility to devote around 10 hrs/month to this project, mostly at their own pace but with occasional need for timely work on short notice
– Ability to commit to the project for at least 2 years, preferably longer
– Initiative to implement new features to improve site functionality
Interested applicants should contact me, Jason Wright, at astrowright@gmail.com with a resume that highlights the above job requirements, lists references I can contact, and (if possible) examples of their UI and Web design sense (UI elegance is one of our site’s primary advantages!).  
I hope fans of this blog will pass this ad along to people who would be good candidates, and encourage them to apply!

Resonances, Inclinations, and True Masses for Exoplanets

When we discover planets with the radial velocity method, we generally just get five pieces of information:

  1. The orbital period
  2. The minimum mass of the planet (technically, the mass function of the binary system, but we usually know the mass of the star pretty well so this is equivalent).
  3. The orbital eccentricity of the planet
  4. An orientation of the point of closest approach with respect to the line of sight (the argument of periastron)
  5. The orbital phase of the planet during the observations

When two planets are discovered, however, a whole world of additional information opens up to us.  Xianyu Tan of Hong Kong University, working with his thesis adviser Man Hoi Lee, has analyzed a classic 2-planet system, HD 82943.  This has always been one of my favorite systems.  It was announced back in 2000 by the Geneva team, but not published until 2004 as part of a big 16-pack of exoplantary systems.  It is a pair of planets in a 2:1 mean motion resonance, with orbital periods 220 days and 440 days;  the planets are at least 2 times the mass of Jupiter, each.
The mean motion resonance means that the planets are perturbing each others’ orbits so much that the five pieces of information we get are actually changing quickly enough for us to detect.   As a result, one can determine the true masses of the planets involved. 
Determining whether a system is in true resonance (that is, that the orbital periods, on average, are dynamically locked in a near-integer ratio) requires demonstrating that certain resonant arguments librate (oscillate) around a fixed value, as opposed to circulating (varying by an arbitrarily large amount).   It requires a lot of radial velocity information to do this, because the effects are quite subtle, but demonstrating that it is happening can reveal a lot about the origin of the system.
Screen Shot 2013-09-28 at 1.39.55 PM.png
Detecting these interactions also allows one to put constraints on the true masses of the objects in the system, and thus determine the inclination of both of the planets in the system.  The only remaining unknown orbital parameter is then the longitude of the ascending node of the two planets, which does not contain any physically interesting information unless one can actually image the system (which we cannot).
Detecting these interactions also allows one to rule out more exotic explanations for the radial velocities we see in this system, including the possibility of 1:1 resonance (two planets sharing the same orbit) or a 1:2:4 “Laplace resonance” (like the one claimed for GJ 876).  
Not many planetary systems have all of this information worked out for them, so it’s a big advance for us to be able to add HD 82943 to the list.  It turns out that both planets are almost five times the mass of Jupiter!  Monsters in a dance it takes centuries to see.
Congratulations, Xianyu, on a very nice piece of work!  The paper is now accepted to the Astrophysical Journal.

Incidentally, there were some issues getting all of the co-authors properly listed on this paper.  Representative Frank Wolf (R-VA) put an amendment into effect in 2011 to NASA’s funding that bars any cooperation with China.  There is no exception for purely scientific research, with no military or space application.  Since some of the American co-authors have NASA funding that could have been used to do work on this project, we had to go out of our way to make sure that we didn’t do any work on this project on NASA’s dime after the amendment went into effect.  We managed it fine and didn’t bend any rules, but it seems an absurd thing for us to have had to worry about — one would think America would be pleased to have such great science come out of NASA’s investment into astronomy!  Instead, we had to carefully avoid that outcome.

Hangin’ with Freeman

I just finished an invigorating Google Hangout with Freeman Dyson, Jill Tarter, Matt Povich, and Franck Marchis.

How it went down is Matt was visiting the SETI Institute to discuss science and collaboration regarding our WISE search for Kardashev Civilizations.  Franck, whom I know because we overlapped at Berkeley back in the day, was excited to hear about the project.  Since he helps organize these SETI chats / hangouts, he suggested we do one together, with Jill.  We of course accepted.
Then he said that he was going to invite Freeman Dyson.
For those who don’t know, Prof. Dyson is one of the most celebrated and brilliant scientists/mathematicians in the world.  Franck was inviting him because of a foundational 1960 paper he wrote on the possibility of detecting alien civilizations through their waste heat, but he is known in many other fields for even more seminal contributions.  Donald Richards, a statistician here at Penn State, asked me if the “Freeman Dyson” of “Dyson spheres” could really be the same “Dyson” he knew for fundamental work in statistics (indeed he is).  I saw him talk at Berkeley about his “garbage bag theory” of the origin of the cell (and life).
Now, I had intended to ask to give a talk at Princeton after our search was mature and I had really figured out what we were doing.  While there, I imagined, I would get an audience with Prof. Dyson, armed with plenty of data in hand.  Instead, I was going to meet him right away!
Before agreeing to the hangout, Prof. Dyson had some questions about our search, and so I sent him a blurb about it.
But he had more questions, so I sent him our proposal to the Templeton Foundation.
But he had more questions, so I sent him a very rough draft of our first paper (50 preprint pages and counting, although some of those were hastily composed after he requested it!).
That, he said, answered “most” of his questions, and so the Hangout was on.  *Whew!*
And it went great!  I started out nervous, but as I got talking I got excited about the project, much to Franck’s amusement.
Here it is, recorded for posterity:

Make Award Winning Posters

Back in March my graduate advisee Sharon Wang won first prize in the 2013 Penn State Graduate Research Exhibition.

Last week, my postdoc Ming Zhao won third place in the 2013 Penn State Postdoc Research Exhibition with this poster:

Screen Shot 2013-09-17 at 4.32.55 PM.png

He credited Sharon’s winning entry with inspiring his poster design.
I asked Ming to write up the secret to poster success.  Here it is (with some edits by me):
======
There are a lot of good materials online teaching us how to do this. I strongly recommend this online tutorial: http://www.makesigns.com/tutorials/scientific-poster-parts.aspx
For those who do not have time to go through them, below is a short summary. It’s far from perfect, but is perhaps a good start.
  • First things first, know your audience. This determines how much technical details or basic introduction your should include in your poster. A poster for general audience from many different disciplines will need a good amount of background information and should avoid the use of jargons, whereas a poster for a specialized meeting (like the ExoClimes, or Kepler Science meetings, for instance) needs very minimal amount of general introduction, and can include commonly used jargons.
  • A poster is a graphical representation of your abstract. People usually only spend 3-5 minutes on a poster, unless they are particularly interested (then they will likely talk to you for more details). So the goal of a poster is to make it easy to scan over quickly and leave one or two take-away points for readers to remember, or to open up a dialogue for those who are interested. Your poster is a visual aid for your oral pitch.
  • Keep it simple and clear. Include only what is absolutely necessary. Use bulleted lists whenever possible. Use fewer words (<250-300) rather than a lot of them.  Less is definitely better!
  • A poster usually consists of the following parts:
  • A short title that clearly describe the main point/topic of the poster (Do not use all UPPER CASE type for titles).
  • List of authors and affiliations.
  • Introduction (What is it about? Why is it important? What are the objectives?).
  • Observations/Methodology/Data Analysis (How do you do it?).
  • Results (What do you find?)
  • Conclusion/Summary/Future prospects
  • References
  • Acknowledgements
  • An abstract is not necessary. I used to include abstracts to my posters. Usually, it is just a copy of the abstract I submitted to the conference. However, I now realize that since a poster is supposed to be an illustrated abstract, with the goal of making it easy to scan over quickly, it is kind of redundant to include a real abstract. It certainly has lots of text to read, which is what you want to avoid for posters.
  • Use short sentences as section titles if you can — this is a tip I learned recently from Dr. Julia Kregenow’s teaching class at Penn State. A short sentence can briefly summarize the major point of the section, thus is often more descriptive and effective than simple section headings like “Introduction”, “Results”, or “Method”.  This tip is also very useful for powerpoint presentations.
  • Make the flow of a poster simple, intuitive, and easy to follow. Use sufficient white space around text and sections to ensure a clear layout and flow.
  • Use sans serif fonts (Helvetica, Arial, Calibri, etc. are acceptable — font snobs will have better, harder to find recommendations) Use fonts big enough for people to see easily at a distance of 1 – 1.5m (40-60in).
  • These font sizes are a good start:
  • Title: 85-90 pt
  • Authors: 56pt
  • Section headings: 56 – 60pt
  • Body text: 28 – 32pt
  • Captions: 24 – 28pt
  • Headings and other text having the same level of importance should be the same font size.
  • Use bold face or italics or combinations to emphasize words and phrases.
  • Used left-aligned, ragged-right text. Using fully justified text will create large gaps between some words and make it difficult to read, especially for captions or text in a narrow text box.
  • Use colors to engage readers. But no more than 3-4 colors. Too many colors make it look chaotic and distractive, but having no color makes it boring and not attractive . Also remember that a lot of astronomers are red/green color blind. So choose colors carefully so that they actually help the poster presentation.
  • Choose good contrast between background and text. Avoid using bright background colors and low contrast text, like blue on red, dark on blue, dark on red, etc. A general rule is to use dark type on light color background and keep the contrast high.
  • Stay away from backgrounds that are too busy or too colorful, since it makes text hard to read. Often times a solid tone, or a very simple gradient is a better option. A white background seems simple but works very well too.
  • A picture is worth a thousand words. Graphs and figures are a great way to convey information effectively and attractively. Make simple figures. Label them well. And use big clear fonts. Don’t make readers stand on their head to read your data.
  • Use high resolution figures. PDF and EPS figures usually come out great on posters. Again, consider the color blind when possible.
  • Finally, choose a good software for poster making. Adobe Illustrator is a great tool to make fancy posters. It manipulates PDF or EPS figures easily. But it may not be available everywhere and the learning curve could be a bit steep. Powerpoint is a good and widely available tool, and it is very easy to use.

[Edit:

More resources at AstroBetter

Astrowright group members Jason Curtis and Kimberly Cartier both won AAS Chambliss awards for their posters following these guidelines. You may also find Kimberly’s “meta-poster” on the topic useful, as well as her more general science presentation poster here.

And now Jacob Luhn has also won a competition using a variant on this design:

…and Jacob won again in Grindelwald at EPRV4:

]