Tag Archives: direct imaging

Meeting the Neighbours

One of six mirrors for the Giant Magellan Telescope

One of six mirrors for the Giant Magellan Telescope

The proximity of a planet to our own solar system will be critically important in the near-future when we begin to characterise the atmospheres and assess the potential habitability of extrasolar planets through direct imaging spectroscopy. Additionally, the nearest extrasolar planets will likely be the first targets of interstellar probes launched by humanity to investigate these worlds up close. To this end, there is considerable interest in discovering planetary companions to nearby sun-like stars. I’ve compiled a brief table below to show the nearest sun-like (FGK) stars to us. I’ve omitted the sea of red dwarfs (as well as a couple A-type stars: Sirius and Altair) interspersed between them in the interests of brievity.

Nearest Sun-like (FGK) Star Systems
D (pc) Spectral Type(s)
Sol 0.0 G2V
Alpha Centauri 1.3 G2V + K1V + M5V
Epsilon Eridani 3.2 K2V
Procyon 3.5 F5IV + DQZ
61 Cygni 3.5 K5V + K7V
Epsilon Indi 3.6 K5V + T1V + T6V
Tau Ceti 3.6 G8V
Groombridge 1618 4.9 K7V
40 Eridani 4.9 K1V + DA + M4V

This interest in our sun’s nearest neighbors can hardly be said to be confined to the scientific community. How many of these star names do you recognise, even if simply from works of fiction?

Until this year, there has been evidence for a giant planet around ε Eri (though the existence of this planet has been questioned recently). But over the last couple of months, we have seen extraordinary announcements of planetary companions around both α Cen and τ Cet.

Firstly, and perhaps most importantly, the discovery of a planet around α Cen is an important milestone in our attempts to understand our place in the Universe. The nearest star system to our own has been found to have a planet – only twenty years after it was not known for sure that extrasolar planets existed at all. But for the details (which are likely not news to the reader): The planet orbits the secondary star in the system in a 3 day orbit, exposing the planet to temperatures that completely throw the question of habitability out the window. But we now know that the system was conductive to planet formation billions of years ago, and we have learned that planets often come in groups. There may yet be more planets around α Cen B, and perhaps even planets around α Cen A as well.

But from the perspective of technological progress, the planetary companion reported around \alphaCen B has a minimum mass that is roughly equal to that of Earth. This planet isn’t some super-Earth or a mini-Neptune: It’s almost certainly a terrestrial planet, and it is the lowest-mass planet detected thusfar from Doppler spectroscopy.

With an increasing number of planet candidates in their star’s habitable zones, and an increasing number of planets discovered nearby, we are approaching a time where there will be a known sample of small planets in the habitable zones of the nearest stars. These planets will likely be our best targets for a search for biosignatures in their atmospheres. Large ground-based telescopes with aperture sizes on the order of 25 – 30 metres, such as the Thirty Metre Telescope and the Giant Magellan Telescope will likely be the key to characterising the atmospheres of these planets.

Thirty Metre Telescope (Concept image)

Thirty Metre Telescope (Concept image)

There are about 125 main sequence stars within 8 pc. Depending on what the abundance and properties of planets in general are in the solar neighbourhood, there may well be a hundred or so small planets that will be available to characterisation by the next generation of ground-based telescopes. You can estimate the underlying distribution of planets (with help from the Kepler mission results) and calculate their angular separation from their stars and their brightness contrast. This can give you an idea of what planets exist in our celestial backyard that are available to us for direct study. Looking in the infrared will make this all easier as the contrast is greater between the two (still extreme, but not as difficult as visible light). A study was recently posted to arXiv about this specifically.

It is notable that the recentlty discovered low-mass planet candidate α Cen Bb presents a contrast of 10-7 and is eminently observable in this baseline scenario (assuming a radius of 1.1 RE) … Despite the worse contrast achieved on this fainter star, the massive planets GJ 876 b and c can be directly characterised if Ag (Rp / RE)2 > ~19 and ~6, respectively (Ag > ~0.02) … The currently known planets GJ 139 c and d can also be characterised of Ag(Rp / RE)2 > ~0.9 and ~1.8 respectively (i.e. Ag > ~0.4), and the planet candidates τ Cet b, c, and d could be characterised for Ag (Rp / RE)2 > ~0.08, ~0.4, and ~1.2 respectively (Ag > ~0.04, ~0.14, and ~0.35). Theoretical efforts to model all these planet’ near-infrared reflectance are clearly warranted.

With nearby planets at α Cen, GJ 876, τ Cet and 82 Eri (GJ 139) being potentially available to direct atmospheric characterisation in the next decade, we are approaching an exciting time where we will have some idea of what these planets are like. This understanding will transcend the basic understanding we have been able to gather from transit light curves and radial velocity.

Still, it will be some time until these planets are known as well as the planets of the Solar System. Our understanding of them will likely be comparable to our current understanding of Pluto. We know there are surface features, we know there are ices of various compositions, but beyond that, Pluto is very much a mystery. Unlike Pluto, a world where in three years time we will have close-up images of thanks to the New Horizons spacecraft, we are not likely to see close-up images of extrasolar planets for the foreseeable future, at least certainly not within our lifetime. Our generation, and likely the next several, will be forced to be content with getting only hints of the nature of the surface conditions on planets outside our solar system. But those hints may be very revealing. They can imply the presence of interesting chemistry and perhaps even biological activity. The limited nature of the understanding of these worlds that we will see unfold in our lifetime is therefore no reason to be pessimistic or to despair. There is much to learn even in the near future, and we should be thankful to be alive in an exciting time where we are beginning to discover the nearest planetary systems to our own and, in the near future, characterise their atmospheres.


Great Mysteries Revealed

Giant squid finally caught on video

Giant squid finally caught on video

It was recently announced that Japanese scientists have managed to finally capture video footage of a giant squid. We’ve known for some time that they exist, but until now, they have never been seen like this in their native environment.

Previously discussed detection methods for extrasolar planets leave much to be desired. Doppler spectroscopy tells you only a minimum mass for the planet and its orbital period. Transit photometry tells you the size and orbital period of the planet, and is typically not capable of confirming the detected object as a planet (since the mass is unknown, a brown dwarf or low-mass star could masquerade as a planet). Gravitational microlensing tells you the planet’s mass but only a projected distance from the star, and usually nothing about the orbit. Astrometry tells you the planet’s true mass and orbit in three dimensions, but otherwise permits you no more information about the planet than Doppler spectroscopy. Combining these techniques will allow one to tease out more information, but it’s still awfully indirect. We know the planets are out there, but it would be nice to actually see them.

The ultimate detection method of the future that will provide the most information is direct imaging. It’s very straightforward, all it involves is a large telescope with a decent imaging contrast ratio. It sounds easy but there are daunting challenges involved due to the proximity of planets to their stars and the brightness of their host stars.

The first directly imaged exoplanet

The first directly imaged exoplanet

As of the time of this writing, only a couple dozen or so planets have been imaged with 8 – 10 metre class telescopes as well as HST. For the most part, they were all imaged in the infrared, they are all very massive as far as planets go, and they are all widely separated from their stars – typically hundreds of AU.

The greatest problem is that stars are very bright, and planets are comparatively very dim. In visible light, there’s a brightness contrast between the Sun and Jupiter of a factor of a billion, but depending on the wavelength and system age and planet mass, the brightness contrast can be as low as ten thousand. This problem is made worse by the diffraction of the starlight across the focal plane of the telescope produces a large amount of “noise,” with the star’s light spread out over a greater area of the image. A method of removing this excess light from a star is a requirement for directly imaging its planets.

One method of doing so is to use a coronograph. This is effectively an object in the telescope to block the light from the star, allowing one to see “around” the star. While you might therefore expect there to be dark “hole” in the image where the star is, the diffraction of light around the coronograph still produces a brighter, noisy area where the star would be. Imperfections in the coronograph will result in extra noise, and it is not clear that perfect coronographs are achievable. Since coronographs (even perfect ones) only attenuate the coherent part of the light’s wavefront (the shape of the waves of light). Imperfections in the wavefront (called aberrations) can leak through the coronograph to produce a residual noise in the form of a speckled halo around the star. Methods exist to correct this, such as wavefront correction with deformable mirrors or to calibrate images for speckles. It’s not perfect, but it certainly removes a lot of the excess starlight.

Four planets at HR 8799

Four planets at HR 8799

Some caution is necessary when discovering planets through direct imaging. A supposed planet next to a star could turn out to be a background star. Monitoring of the planet candidate over some time will be needed to determine if it is bound to the star. Since stars and their associated planets move through space together, there is a certain motion in the sky that the planet and the star will follow if they are bound, whereas the two will have vastly divergent motions in the sky if they are not.

The advantages of directly imaging extrasolar planets is beyond having pictures – it permits direct spectroscopy of the planet’s light. Now while it is possible to gather information about the atmosphere of a transiting planet at a great distance from the star, the fraction of planets at large distances from their star that will transit are quite low. Direct imaging opens up access to all planets with a sufficient brightness and angular separation from the star.

I suspect in the future that this will be the most prevalent way of truly characterizing the planets in the solar neighborhood. Statistically speaking, few nearby Earths will transit, so we will require direct imaging to test their atmospheres for the presence of bio-markers that may be indicative of life.