Chris Lintott’s Universe

I’m sitting at the back of the second press conference of the day, desperately swapping between laptops to run the Astronomy Cast Live feed, catch up with press releases from the speakers and to write this.

Currently, we’re listening to the announcement of the smallest extrasolar planet to date, a 3 Earth Mass planet orbiting a brown dwarf. 3 Earth masses is small enough to all but convince me that we must be talking about a rocky or an icy planet, not a gaseous system, and it orbits its host at about the distance Venus is from the Sun. Of course, as its parent star is much fainter and smaller than the Sun, it will be freezing and there’s no hope of liquid water on the surface.

Phil, whose head is visible in the live feed at times, has all the details.

One issue that came up in the questions, and which has intrigued me for a while, is whether there is any real division between large planets and small brown dwarfs. This is something the International Astronomical Union are going to have to wrestle with as when they demoted Pluto, the declined to define what a planet is beyond our solar system.

The best supported theory of brown dwarf formation claims that these small stars are just the detritus left over from the formation of larger stars. If this is correct then there should be a theoretical minimum mass of about 5-10 times the mass of Jupiter, but as smaller and smaller brown dwarfs are found this limit is being approached.

In one of the other papers presented at the exoplanet session, Michael Liu of the University of Hawai’i talked about weighing brown dwarfs by watching binary pairs orbit each other. This isn’t easy - you need adaptive optics to give you a resolution equivalent to being able to read a magazine from a mile away, you need pointing accuracy equivalent to hitting the bullseye on a dart board in Hong Kong from an oche here in St Louis, and, perhaps most difficult of all, you need the funding and the motivation to study the slow-moving stars for years on end.

Having done all that for a pair of pairs, the results are just about perfect; our theoretical predictions are good, but not that good. The masses are supposed to depend on the energy output of the star, and on its temperature. One star was cooler than expected, and the other warmer. While this isn’t a big deal - no theories are going to be overturned, and no Nobel prizes won - it’s this kind of detailed work that’s essential in making sure our observations and understanding are as good as they can be.

The masses of the stars were well above the theoretical limit; the smallest were about 30 times Jupiter’s mass. More careful work with this technique and study of other pairs might bring us closer to the limit, though.

Astronomy can sometimes feel like an arms race between observers and theorists. Both groups are often convinced that they’re completely right and - at least over a drink at the end of the day - take great pride in being ahead of the game. The latest battleground is in the field of extrasolar planets, and results from this lunchtime’s press conference will have cheered both sides.

Rory Barnes of the University of Arizona scored an early goal for the theorists, proudly announcing the first successful prediction of a planet since Neptune was discovered more than 160 years ago. His theory suggests that systems with multiple planets are often dynamically unstable, meaning that a small change in their orbits would lead to chaos. He also suggested that all such positions tend to have planets - solar systems tend to be ‘packed’. The system he studied, HD 74156, was already known to have two planets, and computer simulations indicated the presence of a gap. Sure enough, a couple of months ago, a planet was found in the gap just as Barnes predicted. Sara Seager, who gave an excellent talk on the field yesterday, congratulated him, and pointed out that this field littered with predictions which turn out to be wrong - but not this one.

Any theorists who were looking forward to the full time whistle should have listened to Carl Melis’ talk just a few minutes later. Looking through an old data set (from the IRAS satellite which went into orbit in 1983). His team have found two very confusing stars. They look like a type of young star called classical T Tauri stars, which have dusty disks, jets and lots of infrared light (see image above). So do the two stars Melis was studying, but there isn’t much lithium in them, which indicates that they are old. (Stars `burn’ lithium so its concentration should be highest when they’re young). So these are old stars pretending to be young, and possibly undergoing a second wave of planet formation. No-one predicted this, and I’m looking forward to seeing what the reaction might be and what explanations people will come up with.

I’ve said again and again that the most exciting area of observational astronomy at the moment is the search for extrasolar planets. It’s incredible to think that in not much more than a decade we’ve moved from finding the first planet in a solar system other than our own to having several hundred in the catalogues. It’s fitting that the first science talk of the meeting, by James Kasting of Penn State, dealt with exactly that. While he talked about lots of the exciting results, it also proved an excellent reminder of just how hard it is to get science done sometimes. With the honourable exception of Kepler, due for launch next year (which he mentioned) and the French COROT mission which I don’t think he did, most of the planned planet seeking missions are either under review or postponed, both in Europe and the US.

Meanwhile, the brave planet hunters soldier on, looking for the first rocky planets in the habitable zone of their star (or, as I prefer, the Goldilocks zone - not too hot, not too cold). Some thought we’d got there with the discovery of Gliese 581c last year, but new results Kasting talked about show that’s it’s too warm, receiving 30% more light from its parent star than Venus does. Time to keep looking…

Update : More detailed lecture blogging here.

Excellent news! I’ve written before about SuperWASP’s planet search, and I’m pleased to say that they’ve just announced their second round of successes. Three new planets are announced today, detected by the small dip in their parent stars’ brightness as the planet passes in front of it. WASP-3 was discovered by the first set of cameras, on the island of La Palma in the Canaries, but WASP-4 and WASP-5 are the first discoveries from the new station in South Africa. These are fairly typical ‘hot Jupiters’, large planets extremely close to their parent stars. As I’ve said before, WASP is based on a brilliantly simple idea and is an extremely ambitious, but low budget, project and I’m really happy it’s continuing to get new results.

I think that’s what they were trying to say in their press release, here. Unfortunately, an early version claimed that The WASP project is the most ambiguous project in the world designed to discover large planets. I blame the spell checker, but whether ambitious or ambiguous, or both, let’s hope there are lots more WASPy planets to come.

I’ve mentioned before just how difficult it is to find extrasolar planets. I’ve also made little secret of the fact that I’m a huge fan of the SuperWASP experiment which adds lenses to professional standard CCD cameras, essentially getting rid of the telescope in order to monitor a huge area of sky for faint dips in the light coming from a star which might indicate the passage of a planet in front of it. The technique works well - one of the leading teams announced the discovery of the largest known planet this way yesterday. (Largest - not most massive. That had me confused for a bit).

I know have a new appreciation for quite how difficult this is. Scrolling through the list of new papers, I saw that Clarkson et al. had published the first tranche of superWASP data, and the numbers in the abstract made me stop dead. They had worked hard to obtain 141,895 lightcurves (each of them tracing the brightness of a single star over time). Buried within this set were 2688 - that’s less than 2% - which have transit-like features. They then looked at those few thousand light curves by eye, and end up with 44 (0.03% of the original sample) which were worth following up further.

What next? Job done? Nope! Another 24 are removed to leave only events which are statistically significant, leaving 20. The next task is to search through everything that’s known about those 20 stars, a process which ruled most of them out. Finally, after a huge amount of work, we’re left with 0.003% of the original sample - 4 stars which might, just might, have planets orbiting them. Wow!

NASA’s Kepler mission is one of the furthest advanced of all of the missions designed to search for extrasolar planets, and is due for launch next year. I’ve just listened to a NASA science press briefing, which included lots of good stuff from right across planetary science and astrophysics. (The highlight was probably details of a Cassini flyby of Enceladus due for next March.)

Most of it was an overview of things to come, but there was one point of interest. In order to avoid cost overruns, Kepler’s main science mission has been reduced from 4 years to 3 and a half years; this will mean the same number of targets will be studied, but the best guess is that there will be an 11% decrease in planet detections. Obviously if the mission is successful then there will probably be an extended lifetime, but the pressure is now on.

The launch of COROT is excellent news, but the work is only just beginning for the team behind what will be the first of several planet-hunting missions. The hard bit is extracting the tiny signal which represents the transit of a Earth across the face of a star from the huge mass of data. COROT’s chosen tactic is to stick to one field for 150 days and then move on, and obviously being in space will also help with removing many sources of interference that affect ground-based searches. I hope they’re successful, but we’ve had an excellent warning as to how hard this is.

I was really pleased we featured SuperWASP on the Sky at Night - it seemed an excellent example of a rapidly put together, cheap, experiment based on a simple but brilliant idea (in this case, using camera lenses and top-quality cameras to survey the whole sky). At the time (2004, if memory serves) I expected it to produce a plethora of planet discoveries, but two years on and we’re only just reporting on the first two. I don’t mean to disparage their efforts, but it’s certainly surprised me how difficult this is. Let’s see how COROT gets on.

(More on this in my state of the Universe report coming up before New Year).

One of my favourite stories out of all of those we’ve covered on the Sky at Night in the last three years was that of SuperWASP. WASP stands for the Wide Angle Planet Search, and the idea is to scan as much of the sky as possible, as quickly as possible, in order to catch the extremely faint blink of a planet passing in front of its parent star. There is no telescope - the two WASP installations use top of the range cameras and camera lenses (some of which, rumour has it, were obtained via ebay!). I reported on the story during my trip to the Canaries in 2004, and had almost given up hope of hearing of detections. The sheer amount of signal processing required to extract anything from the flood of data is immense, but this week brought the first two WASP planets, and I suspect there will be many, many more to come. An excellent result from a simple idea, brilliantly executed.

My only complaint? Why did they have to announce it less than a day after we’d filmed this months News Notes?