No-one should have any doubt about what is one of the most exciting and challenging projects in modern astronomy of the early 21^st century. It is the race to find Earth-mass planets lying in their habitable zones. It is easy to understand why this topic has captured the imagination of astronomers, the media and the general public. First, because there exists a widely held belief amongst astronomers that nothing in the universe is unique. Wherever we look, the same types of objects are generally found in many places, although some objects are certainly rare.

Secondly, it is because we recognize that the Earth provides a special environment that has favoured the development of life. That special environment includes having a rocky surface, a protective atmosphere, a temperature suitable for liquid water and hence favourable for life forms. In addition, our planet is at the right distance of one astronomical unit from a solar type star to receive the required amount of radiation in the form of heat and light. Certainly quite a few conditions have to be satisfied, but solar-type stars are ubiquitous in our Galaxy and it is now clear that Jupiter-mass planets are quite common too. Probably Earth-mass planets are also common, but detecting them at present is right on the limits of current technology.

A huge race between many research teams is now going on to find the first truly Earth-mass planet in its habitable zone. As it happens, three different techniques of finding Earth-mass planets have it in their grasp to deliver this amazing prize within the next year or two. It is extraordinary that, in spite of huge technological challenges, all three look like coming to fruition at about the same time.

The three techniques are the highly successful radial-velocity method, whereby we detect a planet by the tiny reflex wobble imparted to the parent star, by using the periodic Doppler shifts of lines in the stellar spectrum. This was the technique used for 51 Pegasi in 1995 by Michel Mayor and Didier Queloz and by many others for several hundred stars over the last decade and a half. At present the lowest mass planet found is a 2.0 Earth-mass object (lower mass limit) orbiting the M dwarf star Gliese 581 in an orbit of just 0.03 AU, so it is much closer than the habitable zone. Another planet orbiting the same star is 7 Earth masses at 0.22 AU, so it may be in the habitable zone.

Secondly, the transit method, first employed by David Charbonneau and others for HD209458 in 2000, detects a planet by the periodic drop in the star’s light level as the planet transits across its face. The Kepler satellite, now in orbit (since March 2009), is designed to detect these tiny drops of order 10^-4 in light level caused by Earth-sized planets transiting in front of solar-type stars. In addition, the French Corot satellite, launched in December 2006, also can find planets by the transit method, and indeed seven have been found, mainly Juipter-sized objects. One however has a mass of 4.8 Earth masses.

Finally gravitational microlensing has the capability of finding Earth-mass planets if they are orbiting lens stars. The method has the advantage that it is good at detecting Earth-mass planets in the habitable zone, provided a near perfect alignment of source and lens stars occurs. In addition, only for microlensing can an instantaneous snapshot be taken revealing both planet position (projected onto the sky plane) and planet mass, whereas the Doppler and transit methods require one or ideally several orbits to be completed before a planet can be confirmed. Already planets with masses of 3.3 and 5.0 Earth masses have been found orbiting low mass stars at respectively 2.1 and 0.62 AU (projected distance from parent star).

At Mt John Observatory in New Zealand we have already made a huge investment in the microlensing method with the Japanese-NZ MOA project. MOA astronomers have contributed to the discovery of all nine extrasolar planets found by microlensing since the first was found in 2003.

Now we have started a new programme to search for Earth-mass planets orbiting our nearest star, alpha Centauri – in this case using the Hercules spectrograph and the Doppler method. The programme is a joint one with Stuart Barnes (Anglo-Australian Observatory) and Mike Endl (University of Texas, Austin). The problem is that an Earth-mass planet in the habitable zone imparts a velocity wobble of only about 10 cm/s to either alpha Cen A or its companion alpha Cen B – actually about 10 per cent more wobble for star B because it is 20 per cent less massive.

We are encouraged to undertake these key observations for several reasons. First, theoretical studies by Javiera Guedes et al. in 2008 showed that Earth-mass planets are likely to have formed in the alpha Centauri system. In their simulations, planets of mass 1 to 2 Earth masses always form in the habitable zone around alpha Cen B, no matter what the initial conditions. What’s more, Paul Holman and Matt Wiegert found that stable orbits are possible in this binary provided they are within about 3 A.U. of either star. The orbits are almost certain to be coplanar with the binary star orbit, which has a semi-major axis of 23 A.U. What is more, the binary orbit is tilted at 79 degrees to the line of sight, so any putative planetary orbits are likely to be at that same favourable angle for detecting Doppler shifts (if the angle is small, the orbits would be close to face on and no Doppler shifts are then detectable).

The McLellan 1-m telescope at Mt John is a small telescope with a powerful spectrograph. It is being used for the alpha Centauri project. (Photo by Fraser Gunn).

The Hercules spectrograph at Mt John is a fibre-fed vacuum échelle spectrograph which, with an iodine cell in the light path, can deliver about 2.5 m/s precision in its velocities, and possibly even better in the near future. Our modelling shows that some 30,000 spectra of either alpha Cen A or B over three years can rather readily detect an Earth-mass planet in its habitable zone, about 1.2 AU from alpha Cen A or 0.75 AU from alpha Cen B. We also know from past observations that alpha Centauri does not contain any planets as massive as Jupiter. Such planets would easily have been detectable by the Doppler method during the last decade if they existed.

Fortunately the exposure times are short for very high signal-to-noise spectra – about a minute for star A and a few minutes for B. We already have acquired several thousand spectra of each target and the aim is to intensify the campaign over the next few years. There is a strong optimism that finding an Earth-mass planet orbiting either of our two nearest stars is a realistic target within our grasp.

alpha and beta Centauri at lower culmination from Tekapo, still 15 degrees above the southern horizon (photo by Fraser Gunn)

One other fortunate circumstance makes this the ideal programme for Mt John Observatory. Being the world’s southernmost optical observatory (at 44ºS), we are able to observe alpha Centauri for 12 months of the year, even in November and December. The star is circumpolar and passes the southern horizon at lower culmination at an altitude of some 15 degrees, when it is still readily observable. No other observatory can see alpha Centauri all year, and a periodic gap in the data every year can be disastrous when trying to detect periodic signals which may well have around a one-year period. We therefore plan to press ahead with the alpha Centauri campaign during 2010 and 2011, with the hope of making the historic discovery of an Earth-like analogue orbiting our nearest star, at just 4.3 light years distance.

Possibly in the present century an unmanned probe could be sent to alpha Centauri and explore any planet that might be detected there. The journey is likely to take at least half a century, but it is not an impossible dream, even though far beyond the realm of current technology. That’s all the more reason to search for planets now by the Doppler method, as a prelude to any later space mission.

(written 15 October 2009)

This entry was posted on Thursday, October 22nd, 2009 at 12:00 am and is filed under Extrasolar planets, MJUO. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.