## Wednesday, January 30, 2013

### Cartoon Owen

Owen and I had dinner with Jorge Cham and his lovely family this weekend. We also saw his studio where the PhD magic happens. He sketched a portrait of Owen on his computer screen/canvas and Owen colored it with his cool stylus. The result is pretty uncanny!

## Monday, January 28, 2013

### Microhistory made (I can tweet!)

Can a brotha get some followers? :)

## Sunday, January 27, 2013

### Annika Peter: More Facts

To continue my exploration of gender parity in astronomy, I have called on my friend and fellow astronomer Annika Peter to guest blog for me. Annika and I have had several illuminating discussions over coffee about academia in general and women in science in particular. Here's the second in a series of posts from Annika. The first can be found here. The data contained herein has informed a number of articles that I'll post soon.

Women scientists in academia and family structure:  A number of studies indicate that, at the faculty level, a large proportion of women physicists and astronomers are partnered with other academic scientists (especially other physicists!).  The exact numbers are hard to come by---a lot of the time, all physical scientists are lumped together in studies, even though there are hints that there are major differences across fields (with Stanford’s Clayman Institute for Gender Research finding that physicists have the most “endogamous” marriage habits).  I have found only one survey specific to physicists, and it is not especially recent (1998 to be specific).  Moreover, I have not yet been able to find cohort studies that examine family structures at a variety of career stages, or studies of the reasons why both men and women leave the academic track.  I am also interested to see if there is greater or lesser selection pressure on dual academic couples.  In my experience, a high percentage of women in physics and astronomy are coupled with other academic scientists at just about all career stages, but I would like to see some cold, hard numbers on this.

Nevertheless, if you lump all physical science faculty together, one finds large differences in the partnering and child-rearing patterns of men and women.  A study referenced in this article (PDF document) shows that women scientists are more than three times more likely than their male counterparts to be married to someone who also holds a science PhD.  Women science professors are far more likely than their male counterparts to be single or have no children.  When women science professors are partnered, they are far more likely than men to have a “two-body problem”.  The subject of the two-body problem is near and dear to my heart (as it is to many women scientists and the men or women they are partnered with), and I will be devoting some future blog space to this subject.

Women scientists still do far more housework than men, even in dual-academic-career households, as noted by Schiebinger & Gilmartin.  The difference in time devoted to household work is significant, especially compared with their additional finding that men and women in dual career couples have statistically almost indistinguishable distributions of hours worked (median of about 55 hours, the width of the distribution is 11 hours).  Thus, women scientists have a significantly higher work+housework time commitment than men scientists.  In addition, the authors of the study find that the average scientist spends 19 hours per week on housework.  There are a lot of illuminating figures in that paper, and I recommend you check it out.  Note that this study only considers work related to inanimate objects; child-related tasks are not included.

Efforts at top-tier PhD-granting universities to hire and retain women faculty: About a decade ago, some top-tier universities starting realizing that they were not hiring or retaining very many women faculty in their science and engineering departments.  MIT is particularly famous (or infamous) for having had so few women on its faculty as recently as the late 1990’s.  This realization is an indication that the AIP’s finding of a lower proportion of women at PhD granting institutions than colleges and universities as a whole is in part because they simply were not hiring and retaining women at rate one would have expected based on PhD completion rates.  Since then, these universities have undergone self-studies to identify concrete steps they can take to improve the retention of women faculty, and have implemented a number of changes.  Here are links to a few of those universities’ reports and findings:
-MIT:  1999 report; 2011 progress report (
-Princeton: 2003 report (
-The NSF funded ADVANCE (Increasing the Participation and Advancement of Women in Academic Science and Engineering Careers) programs at a suite of universities. Information on UC Irvine’s ADVANCE Program can be found here: http://advance.uci.edu/.  Information about the NSF ADVANCE grant program can be found here: http://www.nsf.gov/funding/pgm_summ.jsp?pims_id=5383

## Saturday, January 26, 2013

### Jesus, take the Wii!

Take it from my hen! An absolutely beautiful karaoke fail, brought to you via Lori's gchat status:

## Friday, January 25, 2013

### Blast from the past

I just found my first drivers license photo. Lookit all that hair!

## Thursday, January 24, 2013

### Humans are awesome

Humans can do anything. Well, except pick good music to accompany their Humans Are Awesome videos. The visuals are amazing, though! Thanks, Bri.

## Wednesday, January 23, 2013

### 100 years of Markov chains!

From Bekki: Did you know that today is the 100th anniversary of Markov chains?

http://computefest.seas.harvard.edu/markov

## Tuesday, January 22, 2013

Have you heard about the Up-goer Five contest? The challenge is to explain your research using only the 1000 most commonly used words. Here's what my research group is up to this week. Also check out Prof. Wright's research statement.

The space team put a far-looking light cup in space to get light from stars. The light cup looks for star light to get smaller when a world goes in front of a star. We make sure that the light-small that is seen is really from a world and not from a star around a star. We also see how big the star is and what it is made out of and how hot it is. To do these things we use big light-cups on the ground. We then see how long it takes for the world to go around the star, how far the world is from the star, and how big the world is. Sometimes the world is smaller than our world, and that makes us very happy to share what we found with people on this world. It would make us especially happy if the world was both small and warm enough for water. We hope that the light cup works for four more years.

### Owen's haircut

After using my old Oster clippers for about two years too long, I decided to bite the bullet and buy a new set. I went with the Wahl Lithium-Ion Cordless.I have absolutely no idea why I waited so long to get a real set of clippers. These things handle like a dream. They're light, ergonomic, quiet and going cordless is the only way to go.

Plus, they cut like a hot knife through butter. Owen's hair grows in 3-4 separate swirl patterns. Following his lines with my old clippers was like trying to trace a van Gogh painting with a dried-out ballpoint pen on a chain. But the Wahl's just zip right on through, against the grain or from the side, it doesn't matter. Haircut night used to take 10-15 minutes per head. Tonight, it only took 5-7 minutes per head.

Here's Owen's before and after shot:
 Before: Momma's baby

## Monday, January 21, 2013

### The cost of science

Over the past few years U.S. astrophysics has hit hard times financially. Everyone is feeling the pain as they submit ever more grant applications each year, only to get more rejection letters citing the lack of funding provided by Congress.

Science is an expensive enterprise. It takes money to pay student, postdoc, engineer and professor salaries. Telescopes require a lot of funding each year to maintain. For example, a night on the Keck 10-meter telescope costs $100,000, or about \$167 per minute. And that's a bargain compared to ALMA, LSST and JWST.

So let's put the tremendous national cost of astronomy on an easy-to-understand scale. Take the Iraq War. The total direct costs of the war from 2003 to 2010 were about a trillion dollars, or 390 million dollars per day.

The combined NASA Astrophysics and NSF AST astrophysics budgets are 1 billion dollars a year. That's a lot of money. Tax payers spend a billion bucks a year on astrophysics. Think about it, that's 2.5 days of war. PER YEAR!

Astronomy for a year costs 2.5 days in a war of choice. Jeez!

We're lucky congress is so generous to fund such a luxurious enterprise as understanding the mysteries of the universe. In a more ideal world, we'd swap astrophysics for two and a half days invading a foreign land for...um... weapons of mass... er...to get rid of terrorists... hmmm... Screw it, WE WON! (Oswalt 2006)

## Sunday, January 20, 2013

### Work-Life Balance Through Working Efficiently (Part 3)

 Figure 1: Awesome figure stolen from Life Without Biases

Avoid the land of diminishing returns

People, I can't stress this enough: avoid the land of diminishing returns, also known since ancient times as the Realm of Wind and Sadness. This is a cold, barren wasteland where weary souls wander listlessly, staring at their computer screens murmuring, "I can get 2-3% improvement. Just a bit more. Almost there…"

Meanwhile, in the Land of Productivity there are scientists who have identified the key physics of their problem, sloughed off the unimportant second- and third-order effects, and are happily shepherding their Nth first-author publication to ApJ (where N is large). Look at the contented smiles on their faces as they use the time they save on their project to hang out with friends on a Friday night, or spend dinner time chatting with their spouse and kids.

Roughly speaking, 20% of your effort goes into completing 80% of the project. Squeezing that last 20% requires the other 80% effort. Keep in mind that these numbers aren't exact, but getting that extra significant figure on the percentages would require a lot more of my time and it wouldn't change my conclusion. Namely, those who can identify the "knee" in the effort-production curve for their projects will be able to produce more science with less time, and that science will be produced in a timely manner so as to keep the conversation moving forward within their subfield.

Want an opportunity to broaden your skill set or expand into a new subfield? Collaborate! Worried that you're about to get scooped by someone else? Ask them to join forces, or coordinate the release of your papers. Need to generate a strong third reference letter? Guess what you need to do.
 Figure 2: Rad shirt. Image from CentralDesktop
However, be careful when selecting your collaborators. Ask around first, or spend some time chatting with the potential workmate at a conference or meeting. When deciding whether I want to collaborate with someone I think about
• Does this person share my scientific approach (my vision) and can they work at my pace? You don't want someone who is going to run out way far ahead of you, and you also don't want to be held back.
• Do we get along on a personal level?
• Do we communicate well? Do they respond to my emails in a timely manner (days rather than months)? When things get tense, does this person give me the benefit of the doubt and understand that I would not intentionally slight them, or do things go all nonlinear and escalate into hurt feelings and/or arguments?
• Will we both gain from this collaborative relationship? Like any relationship, you want it to be balanced in terms of effort invested and benefits gained.
Personally, my favorite kinds of collaborations involve a theorist learning more about how observing is done and data are analyzed, and an observer learning how to think like a theorist. My collaboration with Bekki Dawson falls into this category.

Collaborating allows you to expand your "footprint" and increase your capacity for productivity (increase $\dot{S}$). Collaborating with someone at a later career stage (postdoc while you are a student) will also allow you to pick up a third/fourth strong reference letter.

Learn to delegate

Once you have a broad set of collaborators, and/or once you have built up a research team as a PI or professor, you must learn to transition from lone-wolf scientist to team manager. Building a large collaborative effort will only increase $\dot{S}$ faster than $R$ if you can learn to lean on the people in that team and trust them to do things right.

This means that you must be careful in selecting team mates (see above) and you must trust them to do the job correctly without your micromanaging. This last part is very hard for many people, myself included. But the more I trust my team to get their jobs done, the more time I find available for both my work and my life.

By trusting people you also show them that they are valued. And people who feel valued do valuable work for you. I'd say that this is one of the keys to the success of the Exolab. I bring in the best and the brightest, provide them with the resources they need to work effectively, and I work hard to ensure that they feel valued. As a result, they produce for me in a big way

via Phil:

## Wednesday, January 16, 2013

### Training Day

 http://xkcd.com/1156/

This reminds me of the story of an MIT prank from back in the day. The story goes that MIT students spent a Summer going to Harvard's football stadium wearing striped shirts, blowing whistles and spreading bird seed.

## Tuesday, January 15, 2013

### Work-Life Balance Through Working Efficiently (Part 2)

In a previous post I presented a simple relationship between academic happiness and happiness, number of responsibilities and impact of the stuff you do. In this post I continue with some of the consequences of that simple relationship.
 Figure 3: A turtle. With a rocket on its back. Lookit that turtle go!
Learn to work smarter rather than harder

One of the best recommendation letters I've read stated that the applicant wasn't just smart, but she worked smart. The recommender went on to describe how the applicant was good at identifying projects that yielded huge impact (large $\eta$) relative to the amount of work invested. On the flip side, I know of a lot of extremely smart individuals who are stuck in postdoc positions because they focus too much on topics that very few people really care about. Don't get me wrong, you should pursue topics that move you. But if you are looking for a job, you have to offer something in return. Working on projects that don't help other scientists or that don't advance the field is not offering much in return for employment, which increases stress and decreases happiness.

Further, having projects with small $\eta$ reduces your cross-section for interaction with other scientists, which in turn reduces $\dot{S}$ because you can't build productive collaborations that help you get stuff done. A reduced scientific cross section also shields you from good luck because your opportunities to shine at the right moment are decreased.

ABW: Always be writing

I often get questions about how I manage to have time to write blog posts when I also have to write papers, proposals, recommendations, etc. One key to my writing success is that I live in a steady state of writing. But here I'm using "writing" as not just the act of putting words on the page, but the process of developing an idea, composing things mentally, and giving myself the space I need to take the idea from inception to final product.
 Figure 4: A scene from one of the best movies/plays ever featuring a wonderful cameo by Alec Baldwin. Always be clo...um....writing!
The first thing I do is I make use of dead time in my day to think about what I'm writing at the moment. Walking to work is great for this, as are long flights, waiting for "next" on the basketball court, walking from one place to another on campus. I've learned to make use of snippets of time, no matter how large or small, to compose things in my mind. Then, when I get to a computer I dump everything I've been thinking about onto the page with no regard to order, grammar, spelling, etc.

The second thing I do is once I have everything spilled onto the page (usually in Google Docs), I designate 30-minute writing blocks into my day. No longer, no shorter: 30 minutes exactly. During those half-hour periods I write from minute-one to minute-last. Writing in these sessions might be composing a section, writing a single paragraph, polishing something previously written, outlining at my blackboard, or just standing in the middle of my office with my eyes closed and headphones on envisioning the final product. The latter activity is really important for me because if I don't keep the problem in front of me, it tends to artificially grow more and more scary in my brain, which in turn initiates procrastination. However, if I separate the task at hand into a discrete piece, independent of all the other things I need to accomplish, I can spend 30 minutes doing highly efficient writing.

Now here's the magical key: once the 30 minute session is up, I stop. I walk right away from the Google Doc and go onto the next thing. If I go over time by 30 minutes, I start wearing myself out, which reduces the chance I'll want to get back to it tomorrow. And if I don't get to it tomorrow, then I'm hosed, because my train of thought is interrupted and I have to waste energy in the phase transition back into writing.

But if I string together 4 30-minute sessions on a writing project in a single week, I can look back and be proud of my steady stream of progress. The paper/proposal/chapter that I'm working on looks SO much better than it did on Monday, and the stress I felt on Sunday with that task hovering over my head, all big and scary looking, begins to dissipate harmlessly.

Finally, working in the steady state of ABW, I get way more practice than the average astronomer, which allows me to position myself and my group better for grants, speaker slots, prizes, fellowships, etc. The more I write, the better I become, the easier it is to write, and then I write more. It's a nice cycle to be in. You just need the initial investment of discipline.

For more on this, check out this book.

----------------

Believe it or not, this whole blog post took only 2 thirty-minute sessions this weekend, and one was spent during an NFL playoff game (on TV, not at Gillette Stadium). Thus, $\dot{S}$ was pretty large, and R was constant because I've already decided to spend part of my time maintaining this blog.

Now the question to you, dear reader, is $\eta$ for this post large? Please sound off in the comments, or send me an email, or comment on the Facebooks. I met many of you at the AAS meeting, and you are all amazing people with good things to add to the conversation. Stop lurking and jump into the fray. Let's use this blog as a forum for how to change our field into what we want it to be.

## Monday, January 14, 2013

### Work-Life Balance Through Working Efficiently (Part 1)

Let's face it, if you're a professional scientist the number of responsibilities you have will not decrease with time. As a postdoc, I looked back wistfully at the halcyon days of grad study, particularly the time I spent as an Nth-year. I had so few responsibilities! Now as a prof, I look back at those glorious postdoc years when I had so little I had to do. And if you want nostalgia, how about that first year as a prof when I didn't even have to teach?!

To look at it another way, I think of my ability to do more at each stage of my career as being like a telescopic camping cup, like the one shown in Figure 1. At each stage you don't think you can take on anything more, yet at the next stage you find a way to pull up the next segment and increase your capacity.
 Figure 1: A telescopic camping cup. Right: your first year of grad school. Left: As you asymptotically approach tenure.
But what allows us to pull up an additional segment? If the number of responsibilities is increasing, then it must be our ability to do more with the same amount of time; we must somehow become more efficient. This should be obvious. However, it's not the full story. Getting a large number of things done is important, sure. But what if those things are not all that valuable once they are done?

Allow H to be one's happiness, $\dot{S}$ the amount of stuff one can get done per unit time, R the number of responsibilities one has (measured in units of stuff), and $\eta$ the average scientific or career value of each unit of stuff accomplished, then we can use the well-known relationship

$H \sim \eta (\frac{\dot{S}}{R})$

(the $\sim$ symbol means "scales as" and is kinda like an equal sign). Note the similarity between the form of academic happiness and the one for productivity, P

$P \sim \eta_I \dot{A}$

Where $\dot{A}$ is one's publication rate (measured in ApJs) and $\eta_I$ is the scientific impact of each paper. It's not just your publication rate, but also the impact of your work. But I digress...

 Figure 2: A spherical approximation of the astronomy community. This post is written under the assumption that you want to be a happy astronomer (yellow sphere) rather than an unhappy one (blue spheres).
The relation for H demonstrates several important things:

Learn when and how to say no

First, taking on too many responsibilities is detrimental to your happiness. You don't want to be so over-subscribed that you can't get anything done and you have no time to think carefully about your research. Thus, you have to learn to say no, when appropriate. But be careful that by saying no you aren't passing up things with large $\eta$. You may not have time to fly out and give a talk at Harvard. But the value of doing so likely outweighs the downside of taking on a big increase in R.

It's important to realize how hard saying "no" can be. You'll get very few emails that say something like, "Hey, I was wondering if you'd like to take on a huge amount of responsibilities with little value for your career." No, the emails will be deceptively flattering and enticing, involving things like compliments about your expertise in some field, the value you'd bring to the committee/ panel/ proposal, how really famous people have done this in the past, etc. The only way you'll learn when to say no is if you practice saying no.

This is especially true at the professor level. Junior faculty of the world, learn this important phrase: "I'm sorry, but at this stage of my career I must focus on getting tenure. I must therefore decline your very kind offer to [do something that won't help me get tenure and will bring me very little other direct benefit]." Just be sure to fill in the brackets appropriately after you paste this sentence into your email message.

That said, keep in mind that as a young researcher you will be applying for grants, if not as a postdoc then for sure as a professor. And when you submit a grant proposal, you want knowledgeable, thoughtful, friendly people on your review panel. The same goes for the referees of your papers. You don't want every potential committee member that meets these criteria to turn down the invitation to fly out to D.C. to sit on your review panel, or every young person in your subfield to turn down the referee job for your paper. So evaluating whether or not to increment R, remember that you are a part of the community and your active participation is needed to keep the field moving forward. Thus, it's not trivial to know when to say no, and when to bite the bullet to put in your service (In Part 2 I'll talk about how you can make use of the time during the flight to D.C. to make good progress on that paper you need to write.)

I think a good rule of thumb is to referee one paper for every one that you submit (I actually referee ~2 for every one of my first-authors), and you should sit on a review panel once for every N you submit. To be honest, I'm still figuring this out for myself ($N \approx 3$ for me), and I'm making up for my paucity of in-person review panel attendance for remote (online) reviews. I suppose the key is to be conscious of the need to balance service and keeping R comparable to $\dot{S}$.

Stay tuned for Part 2, coming to a blog near you!

----------------

Believe it or not, this multi-part blog post took only 2 thirty-minute sessions this weekend, and one was spent during an NFL playoff game (on TV, not at Gillette Stadium). Thus, $\dot{S}$ was pretty large, and R was constant because I've already decided to spend part of my time maintaining this blog.

Now the question to you, dear reader, is $\eta$ for this post large? Please sound off in the comments, or send me an email, or comment on the Facebooks. I met many of you at the AAS meeting, and you are all amazing people with good things to add to the conversation. Stop lurking and jump into the fray. Let's use this blog as a forum for how to change our field into what we want it to be.

## Sunday, January 13, 2013

### So cold!

Fun video sequence of Los Angeles people overreacting to the "cold" weather we've had lately (h/t Julie).

## Friday, January 11, 2013

### The Gandolf flow chart

Bri is on fire today.

### Robot apocalypse update

While you work at your desk, they are getting smarter, better, faster. Sure, they may move like me after a AAS party, for now. But soon, we will be working in their iron mines. (h/t Bri)

## Thursday, January 3, 2013

### Alt-J

alt-J on your keyboard is the ∆ symbol. On your computer, it's good music:

### Billions and Billions of planets

PASADENA, Calif.—Look up at the night sky and you'll see stars, sure. But you're also seeing planets—billions and billions of them. At least.

That's the conclusion of a new study by astronomers at the California Institute of Technology (Caltech) that provides yet more evidence that planetary systems are the cosmic norm. The team made their estimate while analyzing planets orbiting a star called Kepler-32—planets that are representative, they say, of the vast majority in the galaxy and thus serve as a perfect case study for understanding how most planets form.
"There's at least 100 billion planets in the galaxy—just our galaxy," says John Johnson, assistant professor of planetary astronomy at Caltech and coauthor of the study, which was recently accepted for publication in the Astrophysical Journal. "That's mind-boggling."

"It's a staggering number, if you think about it," adds Jonathan Swift, a postdoc at Caltech and lead author of the paper. "Basically there's one of these planets per star."

The planetary system in question, which was detected by the Kepler space telescope, contains five planets. The existence of two of those planets have already been confirmed by other astronomers. The Caltech team confirmed the remaining three, then analyzed the five-planet system and compared it to other systems found by the Kepler mission.

The planets orbit a star that is an M dwarf—a type that accounts for about three-quarters of all stars in the Milky Way. The five planets, which are similar in size to Earth and orbit close to their star, are also typical of the class of planets that the telescope has discovered orbiting other M dwarfs, Swift says. Therefore, the majority of planets in the galaxy probably have characteristics comparable to those of the five planets.

While this particular system may not be unique, what does set it apart is its coincidental orientation: the orbits of the planets lie in a plane that's positioned such that Kepler views the system edge-on. Due to this rare orientation, each planet blocks Kepler -32's starlight as it passes between the star and the Kepler telescope.

By analyzing changes in the star's brightness, the astronomers were able to determine the planets' characteristics, such as their sizes and orbital periods. This orientation therefore provides an opportunity to study the system in great detail—and because the planets represent the vast majority of planets that are thought to populate the galaxy, the team says, the system also can help astronomers better understand planet formation in general.

"I usually try not to call things 'Rosetta stones,' but this is as close to a Rosetta stone as anything I've seen," Johnson says. "It's like unlocking a language that we're trying to understand—the language of planet formation."

One of the fundamental questions regarding the origin of planets is how many of them there are. Like the Caltech group, other teams of astronomers have estimated that there is roughly one planet per star, but this is the first time researchers have made such an estimate by studying M-dwarf systems, the most numerous population of planets known.

To do that calculation, the Caltech team determined the probability that an M-dwarf system would provide Kepler-32's edge-on orientation. Combining that probability with the number of planetary systems Kepler is able to detect, the astronomers calculated that there is, on average, one planet for every one of the approximately 100 billion stars in the galaxy. But their analysis only considers planets that are in close orbits around M dwarfs—not the outer planets of an M-dwarf system, or those orbiting other kinds of stars. As a result, they say, their estimate is conservative. In fact, says Swift, a more accurate estimate that includes data from other analyses could lead to an average of two planets per star.

M-dwarf systems like Kepler-32's are quite different from our own solar system. For one, M dwarfs are cooler and much smaller than the sun. Kepler-32, for example, has half the mass of the sun and half its radius. The radii of its five planets range from 0.8 to 2.7 times that of Earth, and those planets orbit extremely close to their star. The whole system fits within just over a tenth of an astronomical unit (the average distance between Earth and the sun)—a distance that is about a third of the radius of Mercury's orbit around the sun. The fact that M-dwarf systems vastly outnumber other kinds of systems carries a profound implication, according to Johnson, which is that our solar system is extremely rare. "It's just a weirdo," he says.

The fact that the planets in M-dwarf systems are so close to their stars doesn't necessarily mean that they're fiery, hellish worlds unsuitable for life, the astronomers say. Indeed, because M dwarfs are small and cool, their temperate zone—also known as the "habitable zone," the region where liquid water might exist—is also further inward. Even though only the outermost of Kepler-32's five planets lies in its temperate zone, many other M dwarf systems have more planets that sit right in their temperate zones.

As for how the Kepler-32 system formed, no one knows yet. But the team says its analysis places constraints on possible mechanisms. For example, the results suggest that the planets all formed farther away from the star than they are now, and migrated inward over time.

Like all planets, the ones around Kepler-32 formed from a proto-planetary disk—a disk of dust and gas that clumped up into planets around the star. The astronomers estimated that the mass of the disk within the region of the five planets was about as much as that of three Jupiters. But other studies of proto-planetary disks have shown that three Jupiter masses can't be squeezed into such a tiny area so close to a star, suggesting to the Caltech team that the planets around Kepler-32 initially formed farther out.

Another line of evidence relates to the fact that M dwarfs shine brighter and hotter when they are young, when planets would be forming. Kepler-32 would have been too hot for dust—a key planet-building ingredient—to even exist in such close proximity to the star. Previously, other astronomers had determined that the third and fourth planets from the star are not very dense, meaning that they are likely made of volatile compounds such as carbon dioxide, methane, or other ices and gases, the Caltech team says. However, those volatile compounds could not have existed in the hotter zones close to the star.

Finally, the Caltech astronomers discovered that three of the planets have orbits that are related to one another in a very specific way. One planet's orbital period lasts twice as long as another's, and the third planet's lasts three times as long as the latter's. Planets don't fall into this kind of arrangement immediately upon forming, Johnson says. Instead, the planets must have started their orbits farther away from the star before moving inward over time and settling into their current configuration.
"You look in detail at the architecture of this very special planetary system, and you're forced into saying these planets formed farther out and moved in," Johnson explains.

The implications of a galaxy chock full of planets are far-reaching, the researchers say. "It's really fundamental from an origins standpoint," says Swift, who notes that because M dwarfs shine mainly in infrared light, the stars are invisible to the naked eye. "Kepler has enabled us to look up at the sky and know that there are more planets out there than stars we can see."

The paper can be found on the arXiv here:

http://arxiv.org/abs/1301.0023

In addition to Swift and Johnson, the other authors on the Astrophysical Journal paper are Caltech graduate students Timothy Morton and Benjamin Montet; Caltech postdoc Philip Muirhead; former Caltech postdoc Justin Crepp of the University of Notre Dame; and Caltech alumnus Daniel Fabrycky (BS '03) of the University of Chicago. The title of the paper is, "Characterizing the cool KOIS IV: Kepler-32 as a prototype for the formation of compact planetary systems throughout the galaxy." In addition to using Kepler, the astronomers made observations at the W. M. Keck Observatory and with the Robo-AO system at Palomar Observatory. Support for all of the telescopes was provided by the W. M. Keck Foundation, NASA, Caltech, the Inter-University Centre for Astronomy and Astrophysics, the National Science Foundation, the Mt. Cuba Astronomical Foundation, and Samuel Oschin.

## Wednesday, January 2, 2013

### The Caltech ExoLab at AAS

I'm very proud of my group at Caltech: The ExoLab. My goal upon arriving at Caltech as a professor was to set up a diverse research team working on a diverse collection of projects spanning theory, observation, and instrumentation. Thanks to the outstanding students and postdocs working with me, we have attained that goal over the past three years. Go team!

We've also managed to be quite productive over the past year, as evidenced by the large number of talks and posters we'll present at the upcoming American Astronomical Society meeting in Long Beach next week. My postdocs, grad students and undergrads will be presenting every day of the meeting. Be sure to to stop by and chat with us, and our collaborators, about our work focused on the detection and characterization of exoplanets and the stars they orbit.

149.06. Minerva: A Dedicated Observatory for the Detection of Small Planets in the Solar Neighborhood
Kristina Hogstrom; John A. Johnson; Jason Wright; Nate McCrady; Jonathan Swift; Philip Muirhead; Michael Bottom; Peter Plavchan; Ming Zhao; Reed L. Riddle

149.07. Optimizing Doppler Surveys for Planet Yield
Michael Bottom; Philip Muirhead; John A. Johnson; Cullen Blake

149.08. Improving Radial Velocity Precision for Faint Star Extra-Solar Planet Surveys
Andrew Vanderburg; John A. Johnson; Philip Muirhead

149.10. Ultra-Precise Radial Velocimetry with Lock-In Amplified Externally Dispersed Interferometry
Rebecca M. Jensen-Clem; Philip Muirhead; Gautam Vasisht; James K. Wallace; John A. Johnson

158.09. Measuring the Distribution of Active M Dwarfs in the Galaxy
J. Sebastian Pineda; Andrew A. West; John J. Bochanski; Adam J. Burgasser

109.06. Precision Near-Infrared Radial Velocity Instrumentation and Exoplanet Survey
Peter Plavchan; Guillem Anglada-Escude; Russel J. White; Charles A. Beichman; Carolyn Brinkworth; Michael P. Fitzgerald; Ian S. McLean; John A. Johnson; Peter Gao; Cassy Davison; Michael Bottom; David Ciardi; James K. Wallace; Bertrand Mennesson; Kaspar von Braun; Gautam Vasisht; Lisa A. Prato; Stephen R. Kane; Angelle M. Tanner

252.12. A Serendipitous Doppler Survey of B-type Stars at Keck with HIRES
Juliette Becker; John A. Johnson; Tim Morton

36.01. Hot on the Trail of Warm Planets Orbiting Cool M Dwarfs
John A. Johnson

334.02D. Enabling the Kepler Exoplanet Census
Tim Morton

334.04. Characterizing the Cool KOIs: Sub-Earth-Sized Planet Candidates Around Mid M Dwarfs
Philip Muirhead; Juliette Becker; Andrew Vanderburg; John A. Johnson; Bàrbara Rojas-Ayala; Kevin R. Covey; Katherine Hamren; Everett Schlawin; James P. Lloyd

334.06. Robotic Transit Follow-up: Adaptive Optics Imaging of Thousands of Stars
Nicholas M. Law; Tim Morton; Christoph Baranec; Reed L. Riddle; Shriharsh P. Tendulkar; John A. Johnson; Khann Bui; Mahesh Burse; Pravin Chordia; H. Das; Richard Dekany; Shrinivas R. Kulkarni; Sujit Punnadi; A. N. Ramaprakash

343.18. Retired A Stars and Their Companions: The Latest Discoveries
Marta Bryan; John A. Johnson; Andrew Howard

343.23. Model-Independent Stellar and Planetary Masses From Multi-Transiting Exoplanetary Systems
Benjamin Montet; John A. Johnson

407.02D. Constraining Planetary Migration Mechanisms with Highly Eccentric Hot Jupiter Progenitors
Rebekah I. Dawson; John A. Johnson; Ruth Murray-Clay; Tim Morton; Justin R. Crepp; Daniel C. Fabrycky; Andrew Howard

407.04. Kepler-32 and the Formation of Planets Around Kepler's M Dwarfs
Jonathan Swift; John A. Johnson; Tim Morton; Justin R. Crepp; Benjamin Montet; Daniel C. Fabrycky; Philip Muirhead