News blog

Earth is less likely than you think

EarthEver since the astronomer Frank Drake penned a rough equation in 1961 for the likelihood of finding an alien civilization, astronomers have been struggling to quantify the chances of encountering extraterrestrial life. For the most part, it’s been a guessing game (Drake himself suggested there might be about two advanced civilizations in the Milky Way at any given time).

People are still guessing all these years later, but now they’re guessing with data. A paper that appeared Tuesday on the physics preprint server analyses the chances of finding other earths in a set of 150,000 stars studied by the Kepler mission. It finds that the odds are quite a bit lower than anyone thought – about 1 in 100.

The number comes from an analysis by Joseph Catanzarite and Michael Shao at Caltech who are analysing the chances of finding “Earth analog planets” out there (earth analogs in this case refers to planets with orbits between .95 and 1.37 that of earth and with planetary radii between 0.8 and 2 that of our home world). It’s based on the latest Kepler data release, which came out in February.

Kepler works by looking for tiny eclipses of planets as they pass in front of their home stars. For a confirmed detection, the satellite needs to see at least three eclipses, and since Kepler only launched in 2009, it’s unlikely that it would have spotted any planets exactly like Earth. Even if it had, it would take at least another year or two to verify it was there.

Rather than wait, Catanzarite and Shao look at the candidate planets Kepler did find and extrapolated the number of earth-like planets the satellite is likely to find. It’s still guesswork: the team uses a ‘power law’ that works well for small planets but not for gas giants in the Kepler data set, and they also have to estimate the number of ‘false detections’ the satellite might be seeing.

But unlike Drake, who pulled numbers out of the air, this is an informed guess. And the results are surprising: just 1.4%-2.7%, of sun-like stars will have an earth analogue, the authors guess. Because the star and planet have to be perfectly aligned for Kepler to see them, the satellite will spot just 12 or so earths.

How does that compare to other guesses? Well Drake naively assumed that every planetary system would have 2 habitable planets, far more than this latest estimate. More recent guesses by the authors and others have suggested that perhaps 20% of sun-like stars would have habitable earths.

Based on this new, lower number, the authors suggest that plans for future planet hunting missions may need to be revised. In particular, it may mean that more work will have to go into spotting a lot of good earth candidates before we develop a very powerful telescope cable of determining whether the planets are, in fact, home to life.

Image: NASA


  1. Report this comment

    Michael said:

    I don’t like the response to their calculated findings. We need to remember that these are in fact ASTRONOMICAL numbers we’re dealing with. So 1 in 100 of every 150,000 stars may have an Earth-like planet. And how many hundreds of billions of stars are in our galaxy alone? And how many hundreds of billions of galaxies are there in existence? Just a thought!

  2. Report this comment


    Encouraging to me, as any chance is better than no chance, and 1/100 is OK, and there are undoubtedly more to be found in the 150,000 as more time and discoveries emerge. Fascinating stuff!

  3. Report this comment

    Nathan said:

    Good points. 1/100 is 1%. 1% of all sun-like stars still suggests a universe awash in earth-like planets.

  4. Report this comment

    James T. Dwyer said:

    I find no mention of any stabilizing large moon being necessary to produce an Earth-like planet. I suspect that would significantly reduce the probabilities of finding another planet suitable for complex life.

    If the objective is to determine our probability of finding another planet that has produced intelligent life, an estimate of evolutionary development time should also be considered: it took 4.5 billion years for Earth to produce modern humans, and another ~200,000 years for us to produce extraplanetary communications technology. We’ve been at that now for a couple of hundred years – further reducing probabilities…

Comments are closed.