Freshly back from Nanjing, I've started Honours in physics and astronomy, and as part of this I'm over in Hawai'i working with my co-supervisor, Frantz Martinache, at Subaru Observatory on the Big Island. This is the third science-related trip for this degree, which is something I'm deeply thankful about. This blog began as a bit of an essay on some experiences I've had in the last week over here in Hawai'i, so it's quite long and I do go into the science a bit - but if you're interested, read on!
I’ve been in Hawai’i for almost a week now. For someone whose life is urban and aspirations urbane, it’s been a bit of a shock – the place is rugged and full of life and utterly outside human control. It’s a flat 26-28 degrees all year round in Hilo, and rains intermittently and without any warning, at any point in the day, almost every day. There are cockroaches as big as your thumb, moths as big as your hand, scurrying beetles, hordes of ants, whining mosquitoes, little green geckos (introduced – they kill the native yellow ones all the time, and the poor little things scream horribly). At a population of 40,000, it’s the second biggest in Hawai’i but still manages to feel tiny, and it has that small-town vibe too: nobody locks their doors, people wave at strangers in the street and any accidental acquaintance leads to a cheerful conversation. They call it the “Aloha spirit”, from the Hawaiian word for “love” – more in a sense of agape than eros, a real surfer-dude Jack Johnson vibe. There’s another note to it: Hilo has a cost of living comparable to elsewhere in the States, but a comparatively low median income. People are, predominantly, quite poor. Most of the tourism is confined to Kona, on the other side of the island (Hilo was Kamehameha’s main port; but Kona where his palace was), which I’m told is very beautiful; as a result, aside from the small tourist strip along the waterfront in “Downtown” Hilo, nothing’s open late, shops are Wal-mart and food is McDonald’s and Taco Bell, and there’s an air of desperation in some neighbourhoods.
The mission objective here is to work with Frantz Martinache at Subaru Observatory; I’ve also had the pleasure of meeting his colleague Olivier Guyon. (There is a very strong French contingent at Subaru!). I didn’t come alone – I’m here with Peter Tuthill’s PhD student Paul Stewart. I’ll write more elsewhere about the work we’ve been doing – at the moment, we think we’ve discovered several new low-mass binaries in archival Hubble Space Telescope data, and I’m currently trying to calculate the significance of our findings. I’ve also been reading papers on star formation, planet formation, brown dwarf atmospheres and interferometry to get ready for my literature review. I’ll leave a proper discussion of this method of interferometry to another time. For the moment, what I’ve done here is wrestle with LaTeX and with my simulation code, and wrestle with my fresh Ubuntu installation, which is a bit of a hassle. (Beautiful compared to Windows though!)
Yesterday must be one of the most exciting days of my life. At work, I got into reading Doug Lin’s masterful reviews of planet formation. Formation of stellar systems had previously seemed a dry subject to me; I can’t say that the brief dabblings we had in first-year Astronomy or in the Berkeley Stellar Physics course really gave me a sense of how exciting the field is. It’s not just about exoplanets – what you see in a radial velocity search or a transit search is a bewildering mixture of strange systems, with hot Jupiters and heavy terrestrial “super-earths”, with coronagraphy you see far-flung gas giants; if what we observe is a representative sample then the Earth, as many have argued, must be incredibly special. But it’s like observing the diversity of life today, and not having fossils; or looking at Europe and saying it’s funny how French and Italian and Spanish are so alike, but not having any account of the Roman Empire. In this way, observing the precursors and alternate outcomes of the formation of planetary systems really tells you the physics of what’s happening. We tend to find a disk of protoplanetary gas and dust, from which planets and asteroids form. The remnant of this in our Solar System is the faint zodiacal light, which can be seen after sunset, from dark enough places, tracing out the ecliptic in a faint glow of reflected sunlight. In younger systems, however, it is dense and warm and significantly contributes to the infrared brightness! We can observe this in many ways: millimetre arrays like ALMA can image the dust in the disk, while lower frequency interferometers like MERLIN can use masers – microwave lasers that naturally form in clouds of gas in space, in water, methanol and other small molecules – to probe the dynamics of the gas motion. (I had a brief but unsuccessful project in first year, as it happens, looking for methanol masers with Lisa Harvey-Smith, who’s now gone on to greater things as the CSIRO project scientist for the SKA!). Optical astronomy is catching up, and a great deal of our attention in high resolution imaging is devoted to observing just these sorts of objects, with the hope of catching planet formation in flagrante! With these observations in hand, and detailed theoretical models, we see explanations for hot Jupiters in terms of type II planetary migration, whereby the planet clearing a lane through gas loses orbital energy and slowly spirals in towards the star, and the observed mass gap between low-mass terrestrials and high-mass gas giants in terms of runaway accretion. Moreover we also see that the circumstances where a solar system like our own might form are special, but not amazingly special: you would expect a great many systems to form their first high-mass core near the ice line, where volatiles like water and carbon dioxide freeze out, giving you a lot more solid matter to play with. If the gas density is right, and the orbital interactions play out right, your gas giant might not migrate in a way that’s too destructive, and will tend to help lower-mass cores form and accrete to smaller gas giants at the pile-up outside its own cleared orbit. The overabundance of similar-mass planetary cores (roughly the size of Mars) is cleared down to near the minimum that can fit stably in the system, as they eject one another and clear their orbits; the leftover debris should damp down the eccentricity of the orbits, as these asteroid-sized objects are swept up on slingshot orbits by the planet that shoot them out into the solar system and rob it of the excess momentum that boosts the orbital eccentricity. So really, our solar system is rather like the kind of solar system we predict when few of the orbital interactions are too extreme and you have a fairly low-mass, moderate metallicity disk from which to form.
Reading the scientific literature on this is like seeing decades of Eureka moments of theory, and decades of baffling observations, finally coming together in a synthesis of amazing import: we are coming gradually to a unifying model of how planets came to be, where they arise, what sort of worlds there are and what sort of worlds we might yet find. Every year we develop better and better methods for detecting planets, and we’re already starting to probe their atmospheres; it is a real possibility that in the near future, if there are any nearby Earth-like planets, we will be able to find signs of life in their spectra. For instance, the dominant feature of the spectra of Earth’s continents, seen from space, is the “red edge” – the cutoff in the very near infrared below which energy chlorophyll cannot absorb light, and becomes reflective. Chlorophyll is just one molecule, but it’s expected, given general thermodynamic arguments, that any molecule with the same function will have a red edge in its spectrum, though not necessarily in the same place. Furthermore, by studing the apparent phases of exoplanets as they orbit their star – like phases of the moon – and measuring the light we observe as a function of this phase, we can work out something of the spatial distribution of brightness around the planet. This has already been used to make a weather-map for a hot Jupiter! Every day, almost, brings a new discovery in the planetary sciences. We are living at the dawn of a Copernican revolution in our understanding of our place in the universe.
I’m not quite in that field yet, though I hope to be at some point. We want to directly image planets – but they’re very dim! It’s been said that it’s like looking for a firefly next to a searchlight. We therefore want to develop high-contrast imaging techniques to tease out the smallest signal from the dazzling glare of the host star. We’re starting to get there – a few tentative images of faint dots around other stars have already been found. But the starting point for these techniques has to be in looking for less extreme contrasts and seeing the promise to extend down to fainter companions. That is why, at present, we’re focusing on brown dwarfs – objects tens or hundreds of times the mass of Jupiter, but much less than a tenth the mass of our Sun, maybe only a few percent. There are other scientific reasons to look at these objects, other than to simply test our astrometric techniques. We believe these should form more or less like stars, rather than like planets, though there’s considerable confusion about how star formation differs at very low masses. In particular, there’s a wonderful relation – the Salpeter function – which tells you how often stars form at different masses (fewer at high mass, many more at low mass, with a scale-invariant expression), and while there are many explanations (it’s very easy to justify such a simple function in terms of simple axioms!) none are universally accepted. Interestingly, at low masses the Salpeter function rolls over and you get few stars at very low masses, which is only partly explained and subject to considerable uncertainties. The atmospheres of these objects aren’t like those of stars – they’re predominantly not ionized, and while hot and luminous, they have clouds of stable molecules (especially methane and water) which give them very rich spectra. We can model what these spectra should be, but it’s very difficult, and we can’t estimate the mass of these brown dwarfs well as a result, which makes it hard to get interesting scientific information out about how they form and so forth. As a result, it’s quite important to find resolved binaries, so that we can observe their orbits and therefore determine their masses, and thereby calibrate all our other observations of brown dwarfs. The HST has been put to use in this regard, but the techniques used were only a bit more sophisticated than blowing up the images and looking really, really carefully for anything that might be a binary! Our technique, which extends the mathematical ideas of aperture masking to a post-processing technique on a full pupil (so long as the wavefront errors aren’t too bad), is much more sophisticated and has let us find two new objects so far, which we think are brown dwarfs, and which are close enough in orbit that over a couple of years we could easily determine their masses by repeated observations. (Remember Kepler’s third law: a closer orbit is completed faster, so that while most of the existing catalogue have many years or even decades-long orbits, these are only a couple of years, which is much more easier to do on a scientist’s busy schedule!). This would be tremendously exciting, which is why I’m currently engaged in dotting the i’s and crossing the t’s on this paper to make sure we’re absolutely faithful to the data and don’t make the horrible mistake of publishing a false positive. That’s a nightmare scenario for me, I don’t even want to think about it.
Anyway, one of the great luminaries of planet-finding optical technology is my supervisor’s colleague Olivier Guyon. I don’t think I’ve ever met a man so hard-working in my entire life – he has jobs at JPL, Arizona, NASA Ames, and Subaru, and is perpetually flying between Pasadena, Tucson, Mountain View and Hilo, sometimes on a daily basis. While his most famous work is in coronagraphy – techniques for blocking out the light of a star so you can see the faint objects around it – he’s got some interesting side projects, like one I helped out with last night. Mauna Loa and Mauna Kea are the two main mountains on Hawai’i, the Big Island: Mauna Loa means “Long Mountain” (it is indeed the largest mountain on earth by volume), and Mauna Kea “White Mountain” (because of the snowy cap), whose peaks lie above 4000m. Mauna Kea is an extinct volcano, while Mauna Loa is still occasionally active. (Interestingly, there is a sound change between the Polynesian dialects of New Zealand and Hawai’i: Maori /t/ is cognate to /k/, for instance in their word for Tahiti “Kahiki”, and /r/ to /l/, As a result, “Aotearoa”, long white cloud, contains the cognates to Kea and Loa!) All the major observatories are on Mauna Kea (e.g. Keck, Subaru, Gemini North, UKIRT), and not just because it’s taller – because you don’t want to invest a hundred million dollars in building something at the summit of an active volcano! However this very fact makes Mauna Loa ideal for geophysical observatories, and for smaller astronomical instruments that would benefit from the geophysical observatories’ infrastructure but aren’t worried long term about the danger of an eruption (major eruptions occur infrequently, decades apart).
There’s a really neat robotic observatory up Mauna Loa which scans the sky looking for variability in young stellar objects, so as to see how they accrete gas and to observe the dynamics of the disk around them. It does this basically by staring and watching how their brightness varies! Olivier’s got a side project with this group using two ordinary Canon DSLRs on a good equatorial mount, with a laptop in a box running the whole thing remotely using software he wrote with Frantz. The idea is that you can, with extremely sophisticated software, get photometry accurate to about a percent, with robust, cheap devices. This means that anyone can make one, once Olivier and Frantz release their software, and these can be distributed not only as an educational outreach programme, but as a serious way to discover transits! Covering the whole sky, densely and with overlapping fields, allows you to cross-correlate different data feeds and dig out the signal of an exoplanet transiting its host star: like detecting the Transit of Venus just by noting the drop in the Sun’s brightness! In this way it might be possible to significantly increase our transit survey capability with very little outlay in cost, and get the public and amateur astronomers involved in the search for extrasolar planets. Equatorial mounts are notoriously fiddly, and he found that it had come a hundred pixels or so off true polar alignment – which meant a drive up the mountain with an assistant. So last night he picked me up about 11:00 and drove us up to Mauna Loa Observatory. You go up out of Hilo to the Saddle Road, whose peak (on the saddle between Mauna Kea and Mauna Loa) isn’t quite as high as Mt Kosciuszko. You then take a sharp turn onto a dirt track that goes up to Mauna Loa. The drive is an experience – no streetlights, mostly no asphalt, with the track littered here and there with volcanic a’a from the surrounding slopes: treacherously loose grey gravel that forms out of viscous turbulent lava flow. Near the top of the saddle road or a little onto the Mauna Loa track you punch through the cloud layer – Hilo being perpetually covered with drizzly low clouds – and then the majesty is overwhelming. This vast mass of lava reaches up into the most beautiful sky I have ever seen. With no moon, it was like taking long exposure photos with the naked eye, you can pick out the faintest details and the finest structure in the most glorious Milky Way. It took me several minutes to find any constellations I knew – Scorpius, Ursa Major, Cassiopeia being the easiest – because they were swamped in an overwhelming sea of stars. Over the course of the night, I saw no less than four meteors, maybe even six, all in the south, some of them very bright indeed – the others, though, might easily have been lost in the glare of a city, but were crystal clear out here. We soon set about aligning the equatorial mount. Olivier’s method of alignment was as follows: take a long exposure photo while rotating the camera around the equatorial mount at a constant rate, so as to see on the sky where the equatorial mount’s axis projected. Then, take a succession of long exposure photos keeping it steady, and carefully inspect the resulting images to look for the star trails, from the rotation of the earth itself. You want these two axes aligned! So after each photo, you use the star trails to pick where the celestial pole is and try and adjust the mount to get these coordinates to agree. This took about an hour, and was exactly the sort of geometrically elegant method which works best in astronomy.
Mauna Loa Observatory is primarily geophysical, and they use lidar and spectroscopy to probe atmospheric chemistry to a sensitivity that allows them to pick out distant volcanic eruptions, and even smog blowing across the ocean from Asia. Most important, however, is that this was the main observatory that provided evidence for the steep rise in carbon dioxide concentration in the atmosphere. I had the great pleasure to go inside and have a look at the original IR spectrometer that produced the hockey-stick graph. Yes, that one, the Al Gore one, the one that mobilised the whole environmental movement around the climate change problem – yes, that graph – this instrument took the data. It was a beautiful quaint old thing like the one we had at school, with the readout transcribed as an analogue line onto graph paper. Up on Mauna Loa I felt, as I never had before, an overwhelming sense of the Earth as a system; seeing the Galaxy stretching overhead, the meteorites, the ocean, the US military bases down on the coast, and the troubles of our industry. I was acutely aware of how fragile and small our planet is, the one habitable rock in the only solar system we yet have. And yet how much smaller still we humans are! We are so closely connected to every scale of the universe and yet we somehow pretend that the only relevant scale is personal, that nothing we do can change something as big as the Earth, the world wouldn’t be that unfair; that nothing from outside could affect us; that we can just enjoy a passive, static world to our heart’s content. The trouble with this is that the world is not static and the part of it on which we depend is so sensitive to change, so much under our influence and yet out of our control. The slopes of this volcano, the a’a on which I stood, were only decades old. A great many people I know were older than this bit of earth. The Earth is dynamical, and indeed, not autonomous but tightly bound up in the universe as a whole. Sure, we can pump carbon dioxide into the atmosphere – but it will change things, and when it goes too far, there’s nothing to help us cope with the consequences. It’s not so much an issue of sustainability or environmentalism here, it’s an issue with a fundamental misunderstanding of how the world works: Nature is red in tooth and claw, and if we ignore change, rather than accept its reality and try to understand it, in the long term we will face ruin.