Mike Brown's Planets

The fall term always gets a little overwhelming, as classes get into session and lectures need to be written, problem sets graded, exams created. I have an amazingly long backlog of things about which I want to write at this point but which I have not yet had the time to even get started. To top things off, my life appears to be changing forever. Most of these pieces get written on weekend afternoons while Lilah is napping. But the days of napping appear to be coming to a close.I understand intellectually that this is likely, after all, few 10 year olds nap, but I had never really stopped to think about the effect on my life. It’s not all bad; being able to pondering going out to do something with Lilah in the afternoon could be quite fun! But it will definitely gobble up my quiet afternoon writing time. But, today, after a late Halloween night and a no-doubt sugar-induced-early-morning wakeup, Lilah is currently snoozing away and I am going to now type as quickly as possible. Ready? Go! (Halloween? Yes, I started writing this almost a month ago, giving a perfect demonstration of the point I am trying to make.)

Back at the end of August I asked everyone to review my paper on Titan fog, and, to my surprise, many people took the task extremely seriously. The paper was discussed in classes and in on-line forums and was stared at by many eyes. If you recall back in August one of the reasons for attempting this open review was the fear that having only a single official reviewer leads to a huge random factor as to whether or not you will get anything useful out of the process. In this case, I have to say that the official review was pretty difficult to decipher. The reviewer commented on a few typos, complained about the location of the references, and said that the paper was generally incomprehensible.

Incomprehensible? Now, I will admit to having written papers that are incomprehensible before (how about this one; I can barely understand it myself 20 years later), but I actually thought that the paper was pretty clear. What’s more, of the many comments I had gotten from outside the official review process, no one had quite said “incomprehensible.” So what was going on here?

I reread the paper several dozens of times, and reread all of the comments that I had gotten, and realized, I think, the source of the problem. I think I was much too terse in my explanation of what I had actually done. Sure, I discussed fog and its discovery in gory detail. But I perhaps did not do a great job of describing how I really sorted through all of the data to find fog. It’s a pretty crucial step. If you don’t provide enough details in your paper that someone else could come after you and reproduce precisely what you did, you have failed an important point of having a paper in the first place.

One reason for describing all of this poorly was that the real process was actually quite different from the way I attempted to describe it in the paper. The real process consisted of this: I was looking at a bunch of pictures of Titan and said “Whoa; what the heck is that?” That turned out to be fog. I suspect that many discoveries are made that way, but if you read scientific papers you will rarely learn that fact. If you read my paper, you will find something like “Fog is very important so one fine day we decided to go look for fog on Titan. And we found it.”

OK, partially this description is true. After the first few times of accidentally seeing the fog we did, one fine day, systematically search through the entire data set. But that description was all pretty muddled.

The solution was a nearly complete rewrite of the paper. Had I just gotten the official review I would have fixed the typos and reworded a few things here and there and wondered what the heck the reviewer was talking about, but with the strength of the large number of comments, I could really tell what people were seeing and reading and I could make it significantly better. At least I think it is. But don’t take my word for it. Remind yourself of the first version, here. And now go read the new version here as it is about to appears in this week's Astrophysical Journal Letters. You still will not read the new version and realize that the real way we found fog is that we stumbled on it accidentally, but you will at least, I think, have a better idea of precisely what we did and how we did it. Want to go find fog yourself? I think the roadmap is now significantly more clear.

My conclusion from this experiment? I can’t tell you whether this system will always lead to such dramatic improvements in the quality of a paper, but in this particular case there is no doubt that when you read the two versions of the paper and you note any improvements almost all of those improvements came from the open review, rather than the official review. All of the comments that were sent to me were incorporated in one way or another. And for that, I would like to say a hardy THANK YOU to everyone.

But wait, there’s more!

Fresh on the heels of the Titan fog paper, I have submitted a paper to the Astronomical Journal called “The size, density, and formation of the Orcus-Vanth system in the Kuiper belt.”

This paper, I will admit, is less accessible than the paper about fog on Titan, yet, still, would you give it a read? It’s been posted online for a week and one reader already pointed out a rather stupid math error (thanks Alan Martin) of the sort that creeps into papers when you work on them one hour a week for 3 months (the error is still there, until we fix it in the next round of reviews, so feel free to go track it down and marvel at how stupid I sometimes can be).

Normally I would spend a few pages here telling you what the paper is about but, conveniently, I did that last spring, when we were searching for an appropriate name for the moon of Orcus. Go back and reread the post about coming up with names for the moon of Orcus, where I talk about the strange characteristics of what we now call Vanth. And, with a bit of continued Lilah napping over the weeks to come, stay tuned for thoughts about searching for the real Planet X, why I hate the 5 dwarf planets, and strategies for Lilah-weekend-nap-inducement.

And look! Lilah is done with her nap, and ready to start in on last night’s candy. Back to the sugar frenzy. (and with that, Lilah was awake, and we were off, and now it is a month later and Lilah is settled into a post-Thanksgiving nap and I finally have a spare moment to finish and post. Classes end next week for the year, so I look forward to a bit more time for reflection soon. Stay tuned.)

Look in your newspaper this Saturday, and you may see a paragraph about Saturn’s moon Titan and a giant storm that moved across the surface last May and what that means. With luck they’ll even print it with a tiny little picture of Titan to catch your eye. Your response, if you have one, will likely be “huh.” It’s OK. I’m not offended. It’s hard to distill the richness of a full scientific paper into a paragraph. And it’s even harder, still, to distill the richness of a decade of scientific inquiry into a short scientific paper. But if you’re curious about what that little paragraph means, and how it came to be in your newspaper, and what we’ve been doing for the past decade, read on. It’s a long story, but that’s somewhat of the point.

I became interested in Titan ten years ago, almost as a matter of convenience. It was an excellent solar system target for the then-new technique of Adaptive Optics, which attempts to undo some of the effects of the smearing of starlight caused by Earth’s atmosphere. Titan was a great target because it is just small enough to be completely smeared by the atmosphere, but big enough that, if you could unsmear it, you would still have a nice view. Just as importantly, no one had ever had a nice view of the surface of Titan before because the satellite it covered in a thick layer of smog which mostly doesn’t let light penetrate to the surface. When the Voyager spacecrafts flew by, they took pictures of Titan which look like a big orange billiard ball. I should have said, though, that visible light doesn’t penetrate to the surface. On the earth, red light penetrates smog better than blue light (hence the nice red sunsets on a smoggy day in Los Angeles). The same happens on Titan. Red penetrates better than blue, but infrared penetrates better still. In fact, if you go far enough into the infrared, you can take a picture of Titan and almost not notice any smog there at all. Conveniently, the new technique of Adaptive Optics works best in the infrared. Hence Titan became a natural target to try out the new techniques on. Antonin Bouchez, then a relatively new graduate student at Caltech, signed on to do this project as part of his Ph.D. thesis.

Our first goals were to obtain maps of the then-almost-totally-unknown surface of Titan. And what a strange looking surface it turned out to be! We speculated endlessly about what all of those dark and bright spots on the surface might be (for the most part it is fair to say that we – and everyone else – had no idea whatsoever until we got better images from Cassini a few years later). And then, in late 2001, we found a cloud sitting at the south pole of Titan.

A cloud!

It doesn’t sound like such a big deal, except that it had long been predicted that Titan was incapable of having clouds. Occasionally there was speculation that clouds of methane might be present, but that, if so, they would be tightly confined to the equator. And yet there it indisputably was: a cloud at the south pole.

Antonin and I were so astounded by this that we put Sarah Horst, then an undergraduate at Caltech, at work looking through a tiny 14-inch telescope on the roof of the astronomy building at Caltech. We had developed some special telescope filters which would – we hoped – be capable of penetrating the haze deck and seeing if Titan got a little brighter due to a cloud or two. We wouldn’t be able to tell anything else, but that would be enough to go back to the giant Keck telescope and say “Look now! There will be a cloud!”

It worked. Just a month after our first cloud detection Sarah saw something that looked just like what we expected a cloud to look like. We called people at the Keck telescope and begged them to snap a picture, and there it was. A much bigger splotch, still near the south pole.

I’m an astronomer, not a meteorologist. I had to spend six months learning about how clouds worked, trying to understand precisely why people thought they wouldn’t occur on Titan, and figuring out what was wrong. On a long summertime flight across the country where we continuously skirted afternoon thunderstorms, it all came together: no one had ever previously bothered to consider the effect of Titan’s surface heating. Like Arizona on a summer afternoon, Titan’s surface can heat up and eventually drive convective clouds over it. On Titan, though, it doesn’t happen in the afternoon. It happens in the summertime, when the south pole spends something like 10 years in continuous sunlight.

It was a compelling story, and, I think true. But, even better, it made some fairly clear predictions. The clouds were at the south pole when we discovered them only because it was very close to southern summer solstice. Titan (and Saturn) takes 30 years to go around the sun, so its seasons are quite long. But if you had the patience to watch, you should see the clouds move from the south pole to the north pole over the next 15 years before coming back 15 years later.

Antonin eventually got his Ph.D. and moved on to take a job working with the technical team continuing the development of Adaptive Optics at the Keck Observatory. It was the perfect place to be. Whenever there was a spare moment or two at the telescope, he would swing it over towards Titan and snap a picture. The clouds were nonstop. Sometimes there were just a few tiny specks, but occasionally there would be a huge outburst. It was a thrilling show to watch.

Emily Schaller entered graduate school at Caltech at just about that time, and she decided to do her thesis on watching and understanding these developing clouds on Titan. The first year was exciting, indeed. She saw a monster cloud system cover the south pole of Titan and remain for more than a month (disappearing just as one of the first close Cassini flybys went in to take pictures; Cassini saw a few wispy little clouds but missed almost all of the action). Henry Roe, a recently graduated Ph.D. from the University of California at Berkeley who had been using the Adaptive Optics on the Gemini telescope to study Titan, moved down to Caltech to work with us, and the odd discoveries about the clouds poured in. They appeared to finally move north from the pole; they appeared tied to one spot at 40 degrees south latitude for a while; they untied themselves; bright clouds in one spot seemed to foretell bright clouds in another. It was clear that we were amateurs here. We enlisted the help of Tapio Schneider, a professor of environmental engineering at Caltech and one of the world’s experts on atmospheric circulations, to help us make sense of what was going on. Things were finally falling into place.

In one final piece of exceedingly clever astronomy, Emily Schaller replaced our clunky nightly observations with a 14-inch Celestron, originally begun by Sarah Horst, with a sleek set of nightly observations from NASA’s Infrared Telescope Facility on top of Mauna Kea. The IRTF would take a quick spectrum of Titan every night possible, and Emily could quickly look at the rainbow of infrared light to tell precisely how many clouds were there. And when they looked good, she could tell Henry Roe, who would get the Gemini telescope to examine them.

And then the clouds stopped.

For years and years Emily would look at her data in the morning and walk across the hall to my office to mournfully say “no clouds again last night.” Seeing no clouds is scientifically interesting, and she dutifully wrote papers and indeed an entire chapter of her Ph.D. thesis demonstrating and trying to explain this years-long lack of clouds. But, really, I understood. Explaining a lack of something is not nearly as satisfying as actually getting to see something happen. As her advisor, I would have been happy to fly to Titan to perform a little cloud-seeding, but no one had yet figured out exactly what chemicals or incantations might do the trick.

On April 14^th last year, Emily walked across the Caltech campus to finally turn in her thesis. Then she did what she did most mornings: she walked to her office, downloaded the data from the night before, and checked to see if Titan had clouds. That morning, I suspect, she came close to falling out of her chair. She was likely exhausted from those final stretches of thesis writing, and I am sure that the first time she plotted her data she did what I always do when I see something astounding: she assumed she had made a mistake. She probably re-downloaded the data, double-checked the coordinates, and shook herself a little more awake. But it was no mistake. Titan suddenly had the largest cloud system seen in years. She likes to say it was Titan throwing a graduation party for her. But I know better: I think Titan likes to hide its secrets as long as possible, and knew it was finally safe to let go.

The scientific paper that Emily wrote along with Henry, Tapio, and I that appears in Nature describes the big cloud outburst and its scientific implications. And the implications are pretty fascinating. This big cloud outburst – the biggest ever seen – began in the tropics of Titan, where it has been speculated that clouds, if they ever form, should be weak wispy things. The tropics are where, of course, the Huygens spacecraft that landed on Titan took dramatic images of things that look like stream beds and shorelines and carved channels. How could those be at the equator if there are never clouds and never rain? People asked.

This discovery doesn’t actually answer that question, because we don’t know why there was a huge outburst of clouds in the tropics of Titan. But it does perhaps answer that lingering question: How could those be at the equator if there is never rain? Because there is rain.

Now, however, I am going to allow myself to speculate a bit more than we were comfortable speculating in the scientific paper. I am going to ask: Why? Why were there clouds in the tropics? Why did they appear suddenly at one spot? What is going on?

What I think is going on (again, I warn you, rampant speculation follows…) is that Titan occasionally burps methane, and I think Emily found one of the burps. For many years scientists have wondered where all of the methane in Titan’s atmosphere comes from, and, I think, here is the answer. The surface occasionally releases methane. Call it what you want. Methane geysers? Cryovolcanoe? Titanian cows? Whatever happens, the methane gets injected into the atmosphere and, at that location, instantly forms a huge methane cloud. Massive rainout ensues downwind. The stream channels, the shorelines, and everything else in the otherwise desert-seeming regions are carved in massive storms.

Evidence? Evidence? Where’s the evidence? You scream. Fair enough. I am giving you a snapshot of how science is done, and, at this point, this is the hypothesis stage. Or hunch. Or speculation. This hunch is the type that then guides what we go off to try to observe next. What will we see? Will the spot Emily found burp again? That would be pretty striking confirmation. Will other spots blow? (I should mention that we do indeed think we saw a different spot burp a few years ago).

At this point we have observed Titan well for about 7 years, from the winter southern solstice until the northern spring equinox, which actually just occurred last week, the terrestrial equivalent of late December to late March. What will the rest of the year reveal? We’re still watching, waiting. Maybe in 23 years, when we’ve finally seen an entire season, we’ll call it a day.