The Search For Earth-Like Planets

The Kepler Space Telescope's Disoveries (so far)

When the Kepler space telescope was launched, hopes were high that it could answer a key question posed by scientist Frank Drake when he was trying to work out a formula to determine if there was other intelligent life in the Universe: how many Earth-like planets are there out there? Are they common, or rare? Until Kepler, all we had to go on was what we knew of our own solar system: not much, basically. Now, NASA’s space telescope Kepler is giving us some hard data about exoplanets (planets outside our own solar system).

The mission was originally scheduled to end this year (2012), but it has been granted funding for another four years, which is really good news as the longer it searches, the better! Over the last 3½ years it has found some 2300 “planet candidates” (where a star’s light has been seen to dim at least three times in a regular pattern - as if something, maybe a planet, is passing in front of it regularly). The candidates are checked with regular telescopes to see what may be there - so far, around 100 have been confirmed as definite planets. If Kepler can observe these stars for longer, it can find planets with longer orbital periods - to detect a planet with a period of 1 year (like the Earth) for example, Kepler needs to look at the star for at least three years so a definite pattern of regular dimming, once a year in this case, can be spotted in the data. The good thing about Kepler is that it is able to watch 150,000 stars simultaneously and from the data so far, we know that planets in general are extremely common.

So far, the figures are suggesting that at least 1 in 3 stars has at least one planet, including binary and multiple-star systems. The planet candidates so far look like this. It is worth showing figures for the candidates, as from the work done so far, it is clear that the vast majority of these will turn out to be planets in fact and not just errors or ‘noise’.


Radius compared with the Earth’s

Number Found

<1.25 x Earth’s radius (Earth-like)


1.25 - 2 x Earth’s radius (possibly livable)


2 - 6 x Earth’s radius


6 - 15 x Earth’s radius (Jupiter-like planets)


>15 x Earth’s radius (super-Jupiters)



The smallest planet-candidate so far appears to be about the size of Mars, about half the radius of the Earth. As the search goes on, it is expected that many more roughly Earth-size planets will be found.

Kepler 22b

This was the first planet Kepler found that was orbiting in the habitable zone around its star (habitable zone - neither too hot nor too cold for liquid water to exist on a planet’s surface). Furthermore, the star it orbits is a G-type star, like our own Sun. 22b has an orbital period (year) of 290 days. So far, so good. The bad news, from a habitability point of view, is that its radius is 2.4 times that of the Earth - so it is a bit on the big size and gravity would be far too strong for humans to live on it. Bacteria and the like could probably survive, of course, as long as it has suitable conditions. We don’t know the surface conditions or even whether it has a suitable atmosphere yet - further observations would be necessary to try to see reflected or refracted light from it, something Kepler is not designed to be able to do. Such as it is, the evidence from its velocity suggests the planet is gassy rather than rocky like the Earth, or there is a tiny possibility that it has an extremely deep ocean instead. We also don’t know yet if the entire orbit lies within the habitable zone, or just a part of it. Assuming it is, we can guess that if it has little or no atmosphere, its average surface temperature will be around -11C. If it has an Earth-like atmosphere, it will be around 22C. If it has a runaway greenhouse effect atmosphere like Venus, the temperature will be around 460C.

Kepler 16b and Kepler 47b and 47c

Before Kepler, it was not known whether binary or multiple star systems would have planets. Now we know that they do. In fact, about 70 candidate planets are “circumstellar” planets in binary systems, that is, the planet candidates appear to be orbiting one of the stars in the system.

Kepler 16b is different: it is a “circumbinary” planet - it orbits both stars of its system. It is closer in to the two stars than theory had suggested was possible, so it is back to the drawing board for the scientists, I suppose. It orbits near the outer edge of its system’s habitable zone, and appears to be a gas giant with a radius of 8.45 times that of the Earth.

Kepler 47b and 47c are both circumbinary planets. 47b has an orbital period of 49.5 days and a radius of 3 times that of the Earth. 47c has an orbital period of 303 days and a radius of 4.6 times that of the Earth. However, 47c is within its system’s habitable zone, so it is not impossible that it could have habitable moons, although no moons have been discovered by Kepler at all, yet, although they are expected to be common, given how many there are in our own solar system. The priority at the moment is to find the planets first.

Future Missions

Well, Kepler is finding the planets, but it is not able to tell us much beyond their broad physical characteristics. By observing the orbits carefully, and by observing the wobbling of their stars as they orbit them, especially in multi-planet systems, we can estimate the planet’s masses fairly accurately, and their density. We can guess at their surface temperatures, given one of the three atmospheric scenarios as mentioned with Kepler 22b above.

A key future mission will be to detect what kinds of atmospheres the planets have so we can detect water, carbon dioxide and ozone (easier to detect than oxygen, but it will only be present if oxygen is present). This can be done by looking at the spectrum of reflected light from them and seeing which substances are absorbing or reflecting their star’s light. It should also be possible to work out whether they are water worlds, deserts, or even jungle planets, as long as they are not completely covered in reflective clouds. We will also be able to determine the day length on relatively cloud-free worlds by regular variations in their brightness as they rotate.

Another mission, now that we know that planets are common, will be to try to examine nearby stars for planets, where observation may be easier.