Centauri Dreams

This week offers two interesting papers about the TRAPPIST-1 planets, one from Hubble data looking at the question of hydrogen in potential planetary atmospheres, the other drawing on data from the European Southern Observatory’s Paranal facility as well as the Spitzer and Kepler space-based instruments. We’ll look at the Hubble work this morning and move on to the second paper tomorrow. Both offer meaty stuff to dig into, for we’re beginning to characterize these seven planets, which form a unique laboratory for the study of red dwarf systems.

Published in Nature Astronomy, the Hubble results screen four of the TRAPPIST-1 planets — d, e, f and g — to study their potential atmospheres in the infrared, using Hubble’s Wide Field Camera 3 in data collected from December 2016 to January 2017. The data allow us to rule out a cloud-free hydrogen-rich atmosphere on three of these worlds, while TRAPPIST-1g needs further observation before a hydrogen atmosphere can be conclusively excluded.

Image: These spectra show the chemical makeup of the atmospheres of four Earth-size planets orbiting within or near the habitable zone of the nearby star TRAPPIST-1. The habitable zone is a region at a distance from the star where liquid water, the key to life as we know it, could exist on the planets’ surfaces. To obtain the spectra, astronomers used the Hubble Space Telescope to collect light from TRAPPIST-1 that passed through the exoplanets’ atmospheres as the alien worlds crossed the face of the star. Credit: NASA, ESA, and Z. Levy (STScI).

Pay particular attention to the purple curves in the above image. These show the signature we would expect to see from gases like water and methane, which would be found if any of these planets had a hydrogen-dominated atmosphere like Neptune’s. The spectroscopic signature should be strong in the near-infrared. The Hubble results are indicated by the green crosses, clearly showing no evidence of such an extended atmosphere for TRAPPIST-1 d, f and e.

Julien de Wit (Massachusetts Institute of Technology), lead author on the paper, explains the significance of the finding:

“The presence of puffy, hydrogen-dominated atmospheres would have indicated that these planets are more likely gaseous worlds like Neptune. The lack of hydrogen in their atmospheres further supports theories about the planets being terrestrial in nature. This discovery is an important step towards determining if the planets might harbour liquid water on their surfaces, which could enable them to support living organisms.”

The work proceeded through transmission spectroscopy, in which some of the light of the star passes through the planetary atmosphere and leaves a distinctive trace in the star’s spectrum. The beauty of the TRAPPIST-1 system is that we have so many transits to work with. Moreover, all seven of the planets orbit their star much closer than Mercury is to the Sun, so we have transits occurring frequently, and the possibility of liquid water on some planetary surfaces. We’re also dealing with a planetary system that’s a relatively nearby 40 light years away.

Image: The graphic at the top shows a model spectrum containing the signatures of gases that the astronomers would expect to see if the exoplanets’ atmospheres were puffy and dominated by primordial hydrogen from the distant worlds’ formation. The Hubble observations, however, revealed that the planets do not have hydrogen-dominated atmospheres. The flatter spectrum shown in the lower illustration indicates that Hubble did not spot any traces of water or methane, which are abundant in hydrogen-rich atmospheres. The researchers concluded that the atmospheres are composed of heavier elements residing at much lower altitudes than could be measured by the Hubble observations. Credit: NASA, ESA and Z. Levy (STScI).

If heavier gases like carbon dioxide, methane, water and oxygen are atmospheric constituents in this system, the James Webb Space Telescope may well be able to find them. What the Hubble work achieves is to take one possibility off the table before JWST goes to work, assuming the latter is successfully deployed in 2019.

The TRAPPIST-1 planets may have had hydrogen atmospheres when first formed, assuming they formed further away from the parent star and migrated into their present positions. The primordial hydrogen would then have been lost as the planets moved close to the star, allowing the formation of secondary atmospheres. Our own Solar System’s rocky planets evidently formed in hotter and drier regions much closer, on a relative basis, to the Sun.

Hannah Wakeford (STScI), one of the scientists involved with this work, adds:

“There are no analogs in our solar system for these planets. One of the things researchers are finding is that many of the more common exoplanets don’t have analogs in our solar system.”

The paper is de Wit et al., “Atmospheric reconnaissance of the habitable-zone Earth-sized planets orbiting TRAPPIST-1,” Nature Astronomy 5 February 2018 (abstract). Tomorrow we’ll look at work just published in Astronomy and Astrophysics on new constraints on the mass, density and composition of the seven planets around TRAPPIST-1.

I was pleased to be a guest on David Livingston’s The Space Show last week. David’s questions are always well chosen, as were those of the listeners who participated in the show, and we spoke broadly about the interstellar effort and what it will take to eventually get human technologies to the stars. The show is now available in David’s archives.

I suspect that if David and I had spoken a couple of days later, the topic would have gotten around to gravitational microlensing, and specifically, the news about planets in other galaxies. On the surface, the story seems sensational. In our own galaxy, we can use radial velocity and transit studies on stars, but here our working distances are constrained by our method. The original Kepler field of view in Cygnus, Lyra and Draco, for example, contained stars ranging from 600 to 3000 light years out — get beyond 3000 light years and transits are not detectable.

Image: The Sun is about 25,000 light years from the center of the galaxy, about half the distance from the center to the edge. The blue cone shows the region of the Milky Way that Kepler explored for planets. Kepler looked along a spiral arm of our galaxy. The distance to most of the stars for which Earth-size planets can be detected by Kepler is from 600 to 3,000 light years. Less than 1% of these stars in the region are closer than 600 light years. Stars farther than 3,000 light years are too faint for Kepler to observe the transits needed to detect Earth-size planets. Credit: NASA/JPL-Caltech/R. Hurt (SSC).

Gravitational microlensing, in which a star moves in front of a more distant star so that light from the background object is distorted by the foreground star’s gravitational field, can turn up distant planets within our own galaxy. In fact, it’s quite a useful tool because it is not limited by line of sight — no planet needs to transit — and is not dependent on the planet’s distance from its star.

Microlensing has allowed us to find planets thousands of light years away, near the center of the Milky Way. We see the pattern of the microlensing event temporarily disrupted by a spike of brightness as the planet around the closer star causes its own gravitational disruption.

But how do we hope to find planets billions of light years away? At the University of Oklahoma, Xinyu Dai and Eduardo Guerras have tackled the question using data from the Chandra X-Ray Observatory, working with microlensing models calculated at the university’s Supercomputing Center for Education and Research. Their work revolved around the microlensing properties of a supermassive black hole at the center of quasar RX J1131-1231. The background quasar, about 6 billion light years away, is what is being lensed by the foreground galaxy, which is 3.8 billion light years out.

We are dealing with what is known as a quasar-galaxy strong lensing system, one in which a background quasar is being gravitationally lensed by a foreground galaxy. The result is that multiple images of the quasar form, as seen in the image below. The light from the background quasar crosses different locations in the foreground galaxy, and is lensed as well by nearby stars in the lens galaxy, an effect called quasar microlensing. The latter is a useful tool, for astronomers have used it to study accretion disks around supermassive black holes at the center of quasars. It can also provide information about the lensing galaxy itself.

Image: The gravitational lens RX J1131-1231 galaxy with the lens galaxy at the center and four lensed background quasar images. Credit: University of Oklahoma.

Let’s turn to the paper, where the relevance of this to extragalactic planets emerges:

As we probe smaller and smaller emission regions of the accretion disk close to the event horizon of the SMBH, the gravitational fields of planets in the lensing galaxy start to contribute to the overall gravitational lensing effect, providing us with an opportunity to probe planets in extragalactic galaxies…

The authors analyze the Einstein ring created by lensing effects to show that emissions close to the Schwarzschild radius of the central supermassive black hole of the source quasar will be affected by planets in the lensing galaxy. Thus we have a way of identifying a population of planets in another galaxy, though not planets orbiting a central star. For the paper goes on:

We have shown that quasar microlensing can probe planets, especially the unbound ones, in extragalactic galaxies, by studying the microlensing behavior of emission very close to the inner most stable circular orbit of the super-massive black hole of the source quasar. For bound planets, they contribute little to the overall magnification pattern in this study.

These planets would be, in other words, so-called ‘rogue’ planets not associated with any star, for bound planets of the kind we are used to studying with radial velocity and transit methods would be below the threshold of detection — they would not change the magnification patterns being observed.

The numbers on these rogue planets are impressive: Roughly 2000 objects per main sequence star in sizes ranging between the Moon and Jupiter — 200 of these per main sequence star would be in the Mars to Jupiter range.

These numbers, Dai and Guerras note, are consistent with theoretical studies showing a large population of unbound planets in our own galaxy. Stanford’s Louis Strigari and colleagues, for example, have found that there may be up to 10⁵ compact objects per main sequence star in the Milky Way, using evidence from microlensing as well as direct imaging. Our galaxy may, in other words, be well populated with such objects, most of these relatively small but some larger than Jupiter. See Island Hopping to the Stars for more on Strigari’s work.

What to make of all this? Before this work, the only evidence for an extragalactic planet was the microlensing event PA-99-N2, detected in 1999, and consistent with a star in the disk of M31, the Andromeda galaxy, lensing a background red giant. A planet of 6 times Jupiter’s mass is one explanation for the lensing profile, but there is no way to confirm the possible planet.

Now we have not the detection of individual planets but a hypothesis that lensing data of a galaxy 3.8 billion light years away can be explained by the presence of a population of unbound planets and other compact objects much smaller than planets. The idea that there would be planets in other galaxies is hardly unusual, given their numbers in our own Milky Way. But it’s an exciting thought that we can now begin to study extragalactic exoplanets, even if we’re extremely early in the process and there is much to be learned. As the paper notes:

It is possible that a population of distant but bound planets (Sumi et al. 2011) can contribute to a significant fraction of the planet population, which we defer to future investigations. Because of the much larger Einstein ring size for extragalactic microlensing, we expect that two models, the unbound and the distant but bound planets, can be better distinguished in the extragalactic regime.

The paper is Xinyu Dai & Eduardo Guerras, “Probing Planets in Extragalactic Galaxies Using Quasar Microlensing,” accepted at Astrophysical Journal Letters (preprint). And if you’re interested in the PA-99-N2 event, one source is Ingrosso et al., “Detection of Exoplanets in M31 with Pixel-Lensing: The Event Pa-99-N2 Case,” Twelfth Marcel Grossmann Meeting: on Recent Developments in Theoretical and Experimental General Relativity. p. 2191 (preprint).