Could we detect an earth-like planet (e.g. similar distance from star, and similar size) with current planet finding techniques? If an earth-like planet were orbiting our nearest star, excluding the sun :-), would we have detected it by now?

It seems current techniques tend to find large planets orbiting close to their star, or for direct observation, find large planets further away from their star. Likewise, we have found many very large planets very close to their star (e.g. hot Jupiters), and some rocky planets somewhat larger than earth close to stars, but not very many (if any) that are very similar to earth (e.g. distance from star and size). It seems the current techniques available do not readily detect earth-like planets though, so it may be that there really are many earth-like planets, but we haven't seen them yet. Here is an article on current planet detecting techniques.


  • $\begingroup$ The main method used, as far as I'm aware, is brightness change detection from the star being transitted by the planet. So for a planet in an earth-like orbit, it would transit roughly once per earth year. So it'd take a few years possibly to confirm the brightness dip. Further, because earth-like planet would be smaller than a Jupiter sized planet, the brightness dip would be much smaller. $\endgroup$ Mar 13, 2015 at 15:58
  • $\begingroup$ astronomy.stackexchange.com/questions/687/…, but there was no distance limit $\endgroup$
    – Mithoron
    Mar 14, 2015 at 15:25

2 Answers 2


The short answer is yes, the Kepler mission has used transits to detect nearly "Earthlike" planets, one of the classic examples is Kepler-22b is only about 2x the radius of Earth and has a year that is 290 (Earth) days. Kepler has detected many planets smaller than Earth and closer to the star. With more observing time Kepler could discover similarly sized or smaller planets at comparable orbits, so technologically transits can already detect "Earthlike" planets. Practically, with Kepler's lost reaction wheel this has been difficult, but more Earth sized planet detections are likely from the K2 Kepler reboot.

Microlensing is more sensitive to planets further from stars, a next-generation space telescope like the WFIRST space telescope, which might launch in the 2020s, expects to be able to detect planets smaller than Earth at slightly larger distances from stars.

Taking a photograph or spectra, or directly imaging, a small planet far from its parent star is much more difficult, light from the Sun is around 10 billion times brighter than the sunlight the Earth reflects (a $10^{10}$ contrast), the sensitivity of current ground based instruments (like GPI) and the Hubble space telescope are closer to $10^7$. WFIRST project will be be able to detect giant planets and may be able to detect Earthlike planets but the technology is not quite there yet.

Because of instrument stability and stellar variability, radial velocity instruments are not yet sensitive enough to pick out the incredibly small wobbles caused by a small planet in an Earthlike orbit. Future instruments, like the 100 Earths Project may get to this sensitivity in the moderately near future.

  • 1
    $\begingroup$ There is no confirmation that Kepler 22b is earth-like. We know that it is a lot bigger than the Earth (2.4 Earth radii). The mass is almost unknown (<124 Earth masses) and it is in a closer orbit (0.85au) to its (slightly cooler) star. As stated by Borucki et al. in the discovery paper - "there is no evidence that Kepler-22b is a rocky planet". I have no doubt that Earth-like planets will eventually be confirmed, but this isn't one (yet). $\endgroup$
    – ProfRob
    Mar 14, 2015 at 12:29
  • $\begingroup$ good point that 22b may not be rocky, the questioner asked about planets with similar sizes and orbits, 2x in radius arguably qualifies as similar. A stricter definition of Earthlike would certainly change the answer, we are far from characterizing a rocky planet with Earthlike atmospheric constituents. $\endgroup$
    – E. Douglas
    Mar 15, 2015 at 19:17

The short answer is no; we cannot quite detect earth-like planets around Sun-like stars with orbital periods of 1 year.

The two main planetary detection techniques are transit photometry and the radial velocity variation technique. Direct imaging of earth-like planets at 1 au from the host star is utterly impossible with current technology - the problem is not the sensitivity, it is the contrast achievable at small angular separations.

The first demands high precision photometry (a transit by an "earth" across a "sun" produces a lightcurve dip of about 0.01%). This kind of precision has been achieved (by space-based observatories), but they have not observed stars for sufficiently long to build up the requisite number of transits (you need at least a few) to confirm a detection at periods of 365 days The Kepler primary mission ceased after about 4 years, meaning it will be tricky to dredge out convincing Earth-like transit signals at periods of 1 year (but not impossible) - and even then you need to perform some sort of follow up to prove it is a planetary mass object, rather than some false positive, and actually estimate the mass to show that it is a rocky planet.

Which brings us to the doppler radial velocity technique. The motion of the Earth-Sun system results in the Sun executing a 1-year orbit around the common centre-of-mass, with an amplitude of about 9 cm/s. This is about a factor of 5-10 smaller than the best precision that is available at any telescope in the word right now.

So - although none have been confirmed yet (there are candidates in the Kepler data), that does not mean that Earth-like planets are uncommon. Indeed most sensible extrapolations of the frequency of Earth-sized planets found at closer orbital distances suggests that they could be quite common (e.g. $\sim 25$% Petigura et al. 2013)


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .