What are the odds of finding Earth 2.0? (Synopsis)

"The diversity of the phenomena of nature is so vast and the treasures hidden in the heavens so rich precisely in order that the human mind shall never be lacking in fresh nourishment." -Johannes Kepler, and the adopted saying of the Kepler mission

The latest haul from NASA's Kepler mission indicates that, in it sample of some 150,000 stars, there are over 2,000 confirmed exoplanets, with approximately 40% of them rocky worlds. If we extrapolate this to our entire galaxy, we have about 60 billion habitable zone planets in our galaxy alone.

The 21 Kepler planets discovered in the habitable zones of their stars, no larger than twice the Earth's diameter. Image credit: NASA Ames/N. Batalha and W. Stenzel. The 21 Kepler planets discovered in the habitable zones of their stars, no larger than twice the Earth's diameter. Image credit: NASA Ames/N. Batalha and W. Stenzel.

But there's a big difference between habitable zone and capable of hosting humans as they are right now. The astronomy indicates that there are so many opportunities for life -- and intelligent life -- to arise, but how many of these chances are actually borne to fruition?

Artist’s depiction of the worlds found by Kepler thus far. Image credit: NASA/W. Stenzel. Artist’s depiction of the worlds found by Kepler thus far. Image credit: NASA/W. Stenzel.

Time to take stock. Here's how close we are to finding the first possible Earth 2.0.


More like this

“If I can't make it through one door, I'll go through another door - or I'll make a door. Something terrific will come no matter how dark the present.” -Rabindranath Tagore Time continues marching on here at Starts With A Bang, just as it does everywhere. My Patreon supporters have stepped up their…
"How vast those Orbs must be, and how inconsiderable this Earth, the Theatre upon which all our mighty Designs, all our Navigations, and all our Wars are transacted, is when compared to them." -Christiaan Huygens With a field-of-view encompassing 150,000 stars, NASA’s Kepler mission delivered an…
“I’m sure the universe is full of intelligent life. It’s just been too intelligent to come here.” -Arthur C. Clarke For the past three years, Kepler has been looking at 150,000 stars, searching for planetary transits. Image credit: NASA. The science haul has been huge, but mostly larger planets…
"Day after day, day after day, We stuck, nor breath nor motion; As idle as a painted ship Upon a painted ocean. Water, water, every where, And all the boards did shrink; Water, water, every where, Nor any drop to drink." -Samuel Taylor Coleridge, Rime of the Ancient Mariner Despite the discovery of…

This is a test.
Find at least two things wrong with the following:

“According to research by Adam Frank and Woody Allen, if humanity isn’t a rarity in the Universe, that means that the probability of those three big steps — life takes off on a world, life evolves into complex organisms, and one such lifeform becomes technologically advanced — must be at least 2.5 × 10^-22, and if we’re not a rarity in the galaxy it must be at least 1.7 × 10^-11. But it could be much, much higher, particularly if we don’t require a technologically advanced civilization.”

By See Noevo (not verified) on 10 May 2016 #permalink


Here’s a more fundamental probability question:
What are the odds that, for the various universal constants (e.g. the strong, electromagnetic, weak and gravitational forces), each would be at its very precise and *necessary* setting (i.e. *necessary* for our lives and our universe) by chance?

Off the top of my head, I’d say it’s one in infinity to the Nth power (where N is the number of universal constants).

But you like those odds.

For you’re a gambler with enormous, other-worldly faith, aren’t you?

By See Noevo (not verified) on 10 May 2016 #permalink

Off the top of my head, I’d say it’s one in infinity to the Nth power (where N is the number of universal constants).

Okay SN, since you can calculate this, you must be able to answer these questions, because they are necessary for any odds estimation:
1. What is the available range of values for c, and how do you know that range?
2. Within this range, are c values quantized or not? If quantized, how many exist in that range? And how do you know this?
3. Are all values in the range equiprobable or do they follow some probability distribution? How do you know this?

If you can't answer any of these questions, then what's coming off the top of your head is just pure, unadulterated bullflop.

A better question is "What is the probability sn really believes infinity to a power" makes sense? I'd say that answer is 1.

Oh, G-d, it's so desperate that it's circled back around to the already-failed "fine tuning" "argument."

From the Forbes article:

And where, perhaps with only minor adaptations, we could survive and thrive on the surface.

This seems a big step to me. AIUI our atmosphere is a result of the particular evolutionary pathway that life took on Earth. Specifically, we have an oxygen rich atmosphere with few sulfides, carbonates, nitrides, etc. because (we think) the earliest forms of life ate the latter and pumped out oxygen as a waste product. If Gould is right and rewinding the tape of life couldn't be expected to lead to a similar end-state, then there is very little reason to believe that a different rocky planet in a habitable zone with life would develop atmospheric conditions similar to ours. Gould's point is that we wouldn't even expect the same atmosphere on a planet that started out with biota and conditions exactly like Earth's, let alone planets that start out significantly differently.

Fortunately, this is just a riff on the "we can only look under the lamppost' problem astronomy has already had to deal with, and it has done so successfully in the past. Just as we can only see planets aligned to us etc., the fact that looking for signatures like free atmospheric oxygen only allows us to identify planets with similar life forms - and not life that may be radically different - simply means we'll have to look at more systems before expecting to see anything.

But the bigger point is that there is probably still a big factor to consider between 'life-supporting rocky planet with standard temperature and pressure' and 'humans can live on the surface with minor assistance'.

Of course if you're paranoid about alien invasion, you can take comfort from the reverse problem too - I would expect it to be highly unlikely that any alien would find our aerobic, extremely oxidizing atmosphere pleasant or even survivable.

See Noevo, God Bless ya brotha. Ya keep pluggin away.Missing the mark here but hitting it there.

@All other 's who are either atheist or closet atheist.

And what are the ODDs?
Hell, can the universe really handle the truth?

By Ragtag Media (not verified) on 11 May 2016 #permalink

A big concern, and something I didn't see addressed in the article was tidal locking. Even if you do get a planet the right size, made from the right stuff, and the correct distance from the star to support liquid water on the surface, if it is tidally locked it is not going to be Earth 2.0. Although TESS (Transiting Exoplanet Survey Satellite), the successor to Kepler, is going to open up the entire sky rather than just Kepler's small patch, it is geared specifically to check nearby dwarf stars. I fear every habital zone rocky world it finds is going to be tidally locked due to the proximity of the habital zone to the star.

@3: a VERY superficial analysis of our own system would seem to reject your conclusion. We have three rocky planets in/close to our habitable zone, and only 33% of them (one of the three) is close to being tidally locked.

So, are you saying our solar system is highly unusual in that Earth and Mars aren't tidally locked? If so, what makes you think that?

@Ragtag Media #7

And what are the ODDs?

We have yet to determine any other possible way to set up a universe than exactly how ours is set up, and have so far only conclusive proof of this universe. Of course the odds could change with new information but as it stands today the odds are 1:1.

@eric #9

So, are you saying our solar system is highly unusual in that Earth and Mars aren’t tidally locked? If so, what makes you think that?

The habitable zone extends outwards with increased star luminosity. The likelihood of tidal locking decreases with distance. With M Class Dwarf stars the habitable zone lies close enough to the star that planets in that zone are likely tidally locked. In a bigger star like ours the habitable zone lies further out where the planets are less likely to be tidally locked.

That seems fair. So, dwarf stars are probably not the best candidates if we're searching for Earth-like life. But they are also the most common type of star. So we are somewhat special in that our planet doesn't orbit the most common star type, though that is not particularly metaphysically newsworthy. Follow-on question: if data and extrapolation lead us to think ~40% of stars have rocky worlds, is that 40% divided between all star types equally or is it distributed unevenly? If dwarfs are underrepresented in that 40%, your conclusion that tidal locking is a problem for most or many earth 2-candidate planets would not necessarily be true.

Have a question about some "back-of-the-napkin" statistics.

If of the 150.000 sampled stars, we found 2.000 exoplanets, of which 40% are rocky... that would mean (to round it).. 1000 rocky planets for 150.000 stars, or in other words.. a little less then 1% of the numbers of stars. How can then we have 60 billion estimated habitable worlds in our galaxy, if our galaxy has 100 billion stars. By math above, we should less then 1 billion rocky worlds.. of which some would be in habitable zone. In other words.. in millions.. but not in billions. One order of magnitude lower.

I'm either missing something important, or there's a mistake somewhere in numbers.

By Sinisa Lazarek (not verified) on 11 May 2016 #permalink

@eric #12

if data and extrapolation lead us to think ~40% of stars have rocky worlds, is that 40% divided between all star types equally or is it distributed unevenly?

We don't have enough data from larger stars yet to make a determination. The exoplanetary survey we have so far is dominated by short orbital period worlds simply because of the testing methodology. That is not to say that are definitely more common in the universe, but the short orbital period worlds fill the minimum requirements for detection earlier so we find them first.

The majority of rocky worlds we've found orbit dwarf stars. The data is biased, but I haven't read any good analysis/estimate on how biased. Maybe Ethan has.

@Sinisa Lazarek #13

I’m either missing something important, or there’s a mistake somewhere in numbers.

You're missing something important. Kepler is finding exoplanets via the transit method which means the orbital plane of the solar system has to be inline enough for Kepler's cameras to see the planet cross the face of the star. Your back-of-the-napkin calculations left out the odds of that alignment.

The term Earth 2.0 is misleading, since a planet can have very similar physical characteristics, and yet end up being unsuitable for complex life (simple, single-celled life is quite another matter).
For details, please see David Waltham's
"Lucky Planet: Why Earth is Exceptional - and What that Means for Life in the Universe"

By Birger Johansson (not verified) on 11 May 2016 #permalink

@ Denier

my calculation is based on numbers provided by the article, which states that by extrapolating those numbers we get 60 billion. Either that claim is wrong/incomplete, or we have more then 100 billion stars in the Milky Way.. about 10 times more :)

By Sinisa Lazarek (not verified) on 12 May 2016 #permalink

@Sinisa Lazarek #17

The claim is not wrong or incomplete, and I strongly suspect that you didn't read the article or my response citing your error.

from the article:

...But that doesn’t mean that only 1%-2% of stars have planets around them; the likelihood of having a good planetary alignment with our line-of-sight is very low, and furthermore, we can only detect planets with orbital periods that are less than Kepler’s observing time

also from the article:

About 80% of star systems are expected to have planets around them

The vast majority of planets are three times the size of Earth or smaller, not gas giant worlds

Of the 150000 sampled stars Kepler found 2000 planets. That is not the same as the being an aggregate total of 2000 planets in orbit around the 150000 targeted stars. A Kepler detection means there are planets AND the orbital plane is aligned correctly. If there is no detection then it could mean either there are not planets -OR- the orbital plane is not edge-on to us.


I read your resposne, that's why I replied. But no, I didn't read the article, nor plan visiting Forbes as long as they have their malware spreading campaign on.

But thank you for copying snipets. If I am reading this correctly, it means that in fact Keppler's data IS NOT what gives the estimate, and it's not Kepler's data that is being exptrapolated to give 60 billion. Thus the original claim that if we extrapolate Kepler we get 60 billion is misleading. That's all I wanted to know. Thanx.

By Sinisa Lazarek (not verified) on 12 May 2016 #permalink

@Sinisa Lazarek #19

You're still not grasping your error. The Kepler data IS what gives the estimate, and Ethan's math is not wrong in the way you think. I do have a quibble with it but I am going to try this one last time.

Imagine you had a sphere (not a circle) a little bigger than the orbit of Neptune around our star with our sun in the middle. Now imagine you have 1000 Kepler telescopes. Each Kepler telescope is deployed to a random spot on that sphere with its camera facing in towards the sun and kept them there for 20 years.

After 20 years of data gathering, statistically speaking only 11 of the 1000 Kepler telescopes would detect Mercury, and even one of those is a very questionable detection. The other 989 telescopes detect no planets at all.

Of the 11 telescopes that detect Mercury, only 6-7 of them also detect Venus, and only 4-5 of those detect Earth. Of the 1000 Kepler telescopes gazing inward, only 3 detect Mars. Jupiter, the biggest planet in our solar system, gets seen maybe by one Kepler telescope and even that would be lucky.

Being that we really only need to find Mercury to conclude that our star has a planet, lets stick with that. 11 Telescopes out of 1000 means there is roughly a 1% chance that a random Kepler telescope will detect a planet in orbit around our star.

Being that we have only 1 Kepler telescope, and it found 2000-3000 planets, that probably means there were 200000-300000 planets and it was in the correct position to detect only 1% of them. That is what you are missing. Take your numbers and multiply them by 100.

That would give you a number higher than Ethan's which brings me to my quibble about his data. He gives numbers of planets not solar systems. Some of those discovered exoplanets are in multi-planet systems. That is why your number is higher than his.

If you want to understand why only 11 of 1000 telescopes would detect Mercury there is a good write up here:


Look at the section titled 'Geometric Probability'.

denier said "A big concern, and something I didn’t see addressed in the article was tidal locking."

I dunno if you remember the Perimeter Institute 'new kind of ' journal innovation which they called The Journal of Brief Ideas. Basically the upshot is 200 words maximum could be submitted by anyone that wanted to...threshold standards were collapsed so to give pretty much anything legible a good chance of acceptance. The rationale was along the lines of high standards could be exchanged for brevity (200 words max) because it then only takes a second to scan and snap-judge for plausible new usefulness.

Anyway I think it was probably a silly idea that fizzled out, but at the time I did my bit and my submission, which I think I used some other pseudonym probably keeping 'Chris', was the same good point that you are making there.
My 200 words worth, focused on the habital zone of red dwarfs which are mathematically doomed to tidal locking to the star, making them probably nonviable. The suggestion was to look for co-orbiting similar size binary planets....like the Earth and Moon, but both similar sized so more like two Earths. It would be rare but that had to be balanced by far greater analytical reach we'd enjoy as to the theoretical and properties of such systems, both abstract theoretical, and empirical / practical. The centrepiece reason to be bothering was the same insight as yours, in that two co-orbiting similar size planets would tidally lock to oneanother, thus preventing tidal locking to the star. Obviously that would have to be worked through and verified - but by the same coin that's eminently doable.

Other thoughts to fill out the 200 words included that co-orbiting similar size planets, also enjoy larger habitable since each respective planet only has to be in the habitable for a portion of the co-orbiting cycle.

Another though was that the effect of binary planetary systems on the wobble of a star, could feasibly be similar to, and mistaken for, super-Jupiter's at crazy 4 day period distances from the star. This is because two rocky planets orbiting each-other much further out in the habitable zone, could generate something like a 4 day wobble due to coming much closer to the start at one stage of the cycle while much further away at another. Could possibly emulate a Jupiter up close and nasty to the star (i.e. 4 day, say, co-orbiting periods of the planets, but as a system taking a year or whatever to complete a full orbit of the star.

It's probably a load of shit, but they asked for it... init.

By Chris Mannering (not verified) on 14 May 2016 #permalink

@Chris Mannering #21: Interesting analysis! The only point that I have an obvious argument against is your very last one -- "that the effect of binary planetary systems on the wobble of a star, could feasibly be similar to, and mistaken for, super-Jupiter’s at crazy 4 day period distances from the star."

I don't think this is the case. One of the really nice things about radial-velocity observations (as compared to Kepler's transit-only search) is that you can get a really detailed waveform for the stars motion. Both Keplerian and non-Keplerian motion (such as your hypothetical large double-planet system) can be modeled with very high accuracy. Those motions can be extracted from the radial velocity curve quite well.

While a double-planet system might be able to emulate, for example, the magnitude of the peak velocity due to a "hot Jupiter", the system could not emulate the full velocity curve. Rather, it would show the curve for a long-period planet (with a mass equal to the double planet system), with a much lower magnitude (due to the distance) short-period oscillation on top. That low magnitude oscillation would not look like a hot Jupiter.

By Michael Kelsey (not verified) on 14 May 2016 #permalink

eric @3, moreover, the question is like asking what the chance of 10^26 molecules of H2O forming in exactly the given fractal pattern and ignoring that the pattern is the result of the hole the water is sitting in.

The Kalam "argument" is by far the dumbest one available.

What's the chance of an omnipitent being existing? It's infinitely less likely than the "fine tuning" of the constants.

@ Denier #20

Ahh. understand now what you mean. I missed that the sample Kepler can even in principle sample is one order of magnitude smaller then the rest of possible configurations.
Thanx. My bad.

By Sinisa Lazarek (not verified) on 16 May 2016 #permalink