Explainer: How to find an exoplanet (part 2)

The search for planets around other star systems has been going on for decades and the results are now coming in thick and fast. This series looks at some of the techniques astronomers use to find and learn more about these exoplanets. You can read part 1 in the series here.

Over the past two decades, we have surged into the exoplanet era. It was just 21 years ago that the discovery of the first planet orbiting a sun-like star was announced.

Now, we know of 3,268 planets beyond the solar system, and new discoveries are announced every week.

But how do we find these distant worlds? We’ve already described the two most successful methods used to date, known as the radial velocity and transit techniques.

Between them, they have resulted in the discover of some 3,170 planets, a whopping 97% of the total known. The other 3% were found by a variety of different methods.

Direct imaging

Only a handful of the planets found to date were discovered directly. Seeing the tiny speck of light that is a planet against the contrast of its host star is incredibly challenging.

The planets directly imaged to date typically orbit far from their host stars, and are both massive and young.

The combination of their mass and youth means they remain hot, still radiating away the heat generated by their formation, which makes them brighter and easier to see.

In coming years, such discoveries should become more plentiful. The next generation of instruments on the world’s largest telescopes feature tools specifically designed to detect the faint glimmer of light from distant worlds peeking out of the glare of the host stars - so watch this space!

Timing variations

If a star does something periodically, then the effect of orbiting planets on that behaviour can be measured, and the planets inferred from those variations.

This is actually what we’re doing with both the radial velocity and transit methods, but there are more subtle periodic behaviours that can reveal the presence of a planet.

Take, for example, pulsars. Tiny objects, known as neutron stars, the size of a city but more massive than the sun, these are remnants of the deaths of once massive stars.

And they spin incredibly fast, beaming radiation to space from their magnetic poles. If those beams cross the Earth, then we see pulses, more regular than clockwork, many times per second.

Were a planet to orbit a pulsar (once thought unlikely, given that they are the products of supernovae), then the pulsar would rock back and forth. Sometimes farther from Earth, and sometimes closer. The pulses themselves would shift, arriving early when the pulsar is closer than normal, and late when it is more distant.

Amazingly, a few such planets have been found, including one pulsar with three tiny planets, and another two with a single planet each. All spotted by the minuscule wobble they induce on their host star’s pulses.

Other planets have been claimed based on the timing of eclipses between components of binary star systems. The idea here is the same; the stars are sometimes a little closer or more distant, causing the eclipses to arrive early and then late.

To date, several such planets have been claimed, although only eight have stood up to further scrutiny.

Transit timing variations

In some cases, the behaviour of a known planet can reveal the presence of its unseen siblings. Such is the case with the transit timing method. Here, we have systems with multiple planets, where at least one has been observed to transit its host star.

If such a system has planets on orbits that are tilted with respect to one another, they may not all transit their host. How, then, can we find them?

If we measure the timing of the first planet’s transits with sufficient precision, the influence of the other planets on its orbit can be revealed.

Those planets will perturb the orbit of the first, tweaking it slightly. The result? The transits will arrive slightly early, or slightly late, breaking the perfect symmetry of its orbital period.

Reverse engineering these timing variations to disentangle the orbits of the other planets is a challenging task. For this reason, just 15 planets have been discovered this way to date. But with the wealth of data available from the Kepler satellite, more are sure to follow.

Gravitational microlensing

One of the predictions made by Einstein’s theory of General Relativity is that mass can bend the path of light, warping space-time to create a gravitational lens.

On the largest of scales, this results in spectacular, kaleidoscopic views of distant galaxies, their light distorted, warped, and fragmented by foreground galaxies.

But on a smaller scale, this gives us another tool to search for hidden planets: gravitational microlensing.

Within our galaxy, the stars are all in motion. Should a foreground star happen to pass close to our line of sight to one that is more distant, its mass can act as a lens.

Over a period of a few months, as the foreground star passes by, the background star will brighten and then fade in a well-understood fashion.

If a planet orbits the foreground star, then it too can act as a lens. Since the planet is smaller, its effect will be less pronounced, a minor tweak in the overall variation.

The odds of any given star being involved in a microlensing event is tiny. Despite this, 36 planets have been found this way, albeit all far too distant to be followed up using other techniques.

Astrometric observations

As a planet and a star orbit their common centre of mass, they wobble. This is core to the radial velocity technique, which measures the component of that wobble along our line of sight.

But what if you had a telescope sufficiently accurate that you could actually see the star wobbling around in the night sky?

That technique – the detection of unseen companions through astrometry – is actually well established.

As early as 1844, Friedrich Bessel predicted that Sirius, the brightest star in the night sky, must have an unseen companion, causing it to wobble in its path through space.

Almost 20 years later, that companion (the first known white dwarf star) was found. It is an object the mass of the sun but the size of the Earth, but far outshone by its full-sized sibling.

To find faint stars by this method is one thing, but to find planets is quite another. To date, just a single planet has been found this way, the evocatively named DENIS-P J082303.1-491201 b. But the astrometric method is definitely one for the future.

The future

Where the radial velocity method once ruled the roost when it came to exoplanet discoveries, the future is likely the domain of transit and astrometric surveys.

Next year, NASA is scheduled to launch TESS, the transiting exoplanet survey satellite. The successor to Kepler, TESS will search for transiting planets around stars across the whole night sky.

It will doubtless lead to a great windfall of exciting new worlds. And unlike Kepler, TESS will target bright stars. This will allow astronomers to follow up its discoveries, and to learn a great deal more about the planets it finds.

And then we have Gaia.

The Gaia space observatory, launched by the European Space Agency in December 2013, is designed to measure the positions and motion of stars with unprecedented precision.

An astonishing tool, it will revolutionise our understanding of our galaxy. It will carry out astrometric observations of a billion stars, around 1% of all stars in the Milky Way.

As well as creating a three-dimensional map of the stars in our galaxy, it may also prove an incredible planet-finding tool. It has the potential to find thousands, if not tens of thousands of new planets.

In a galaxy with more than a hundred billion stars, most of which we now think will host planets, one thing is certain. We have barely scratched the surface, and there are many exciting discoveries still to come!

See also:

Explainer: How to find an exoplanet (part 1)