Can red dwarf stars have Earth-like planets with life?
This is an important question, at least in the long run, because 80% of the stars in the Milky Way are red dwarfs, even though none are visible to the naked eye. 20 of the 30 nearest stars are red dwarfs! It would be nice to know if they can have planets with life.
Also, red dwarf stars live a long time! They’re small—and the smaller a star is, the longer it lives. Calculations show that a red dwarf one-tenth the mass of our Sun should last for 10 trillion years!
So if life is possible on planets orbiting red dwarf stars—or if life could get there—we could someday have very, very old civilizations. That idea excites me. Imagine what a galactic civilization spanning the 80 billion red dwarfs in our galaxy could do in 10 trillion years!
(No: you can’t imagine it. You don’t have time to think of all the amazing things they could do.)
Proxima Centauri
Let’s start close to home. Proxima Centauri, the nearest star to the Sun, is a red dwarf. If we ever explore interstellar space, we may stop by this star. So, it’s worth knowing a bit about it.
We don’t know if it has planets. But it could be part of a triple star system! The closest neighboring stars, Alpha Centauri A and B, orbit each other every 80 years. One is a bit bigger than the Sun, the other a bit smaller. They orbit in a fairly eccentric ellipse. At their closest, their distance is like the distance from Saturn to the Sun. At their farthest, it’s more like the distance from Pluto to the Sun.
Proxima Centauri is fairly far from both: a quarter of a light year away. That’s about 350 times the distance from Pluto to the Sun! We’re not even sure Proxima Centauri is gravitationally bound to the other stars. If it is, its orbital period could easily exceed 500,000 years.
If Proxima Centauri had an Earth-like planet, there’s a bit of a problem: it’s a flare star.
You see, convection stirs up this star’s whole interior, unlike the Sun. Convection of charged plasma makes strong magnetic fields. Magnetic fields get tied in knots, and the energy gets released through enormous flares! They can become as large as the star itself, and get so hot that they radiate lots of X-rays.
This could be bad for life on nearby planets… especially since an Earth-like planet would have to be very close. You see, Proxima Centauri is very faint: just 0.17% the brightness of our Sun!
In fact many red dwarfs are flare stars, for the same reasons. Proxima Centauri is actually fairly tame as red dwarfs go, because it’s 4.9 billion years old. Younger ones are more lively, with bigger flares.
Proxima Centauri is just 4.24 light-years away. If explore interstellar space it may be a good place to visit. It’s actually getting closer: it’ll come within about 3 light-years of us in roughly 27,000 years, and then drift by. We should take advantage of this and go visit it soon, like in a few centuries!
Gliese 667 Cc
Gliese 667C is a red dwarf just 1.4% as bright as our Sun. Unremarkable: such stars are a dime a dozen. But it’s famous, because we know it has at least two planets, one of which is quite Earth-like!
This planet, called Gliese 667 Cc, is one of the most Earth-like ones we know today. But it’s weirdly different from our home in many ways. Its mass is 3.8 times that of Earth. It should be a bit warmer than Earth—but dimly lit as seen by our eyes, since most of the light it gets is in the infrared.
Being close to its dim red dwarf star, its year is just 28 Earth days long. But there’s something even cooler about this planet. You can see it in the NASA artist’s depiction above. The red dwarf Gliese 667C is part of a triple star system!
The largest star in this system, Gliese 667 A, is three-quarters the mass of our Sun, but only 12% as bright. It’s an orange dwarf, intermediate between a red dwarf and our Sun, which is considered a yellow dwarf.
The second largest, Gliese 667 B, is also an orange dwarf, only 5% as bright as our sun.
These two orbit each other every 42 years. The red dwarf Gliese 667 C is considerably farther away, orbiting this pair.
What could the planet Gliese 667 Cc be like?
Tidally locked planets
Since a planet needs to be close to a red dwarf to be warm enough for liquid water, such planets are likely to be be tidally locked, with one side facing their sun all the time.
For a long time, this made scientists believe the day side of such a planet would be hot and dry, with all the water locked in ice on the night side, as shown above. People call this a water-trapped world. Perhaps not so good for life!
But a new paper argues that other kinds of worlds are likely too!
In a thin ice waterworld, an ocean covers most of the planet. It’s covered with ice on the night side, maybe 10 meters thick. The day side has open ocean. Ice melts near the edge of the ice, pours into the ocean on the day side… while on the night side, water freezes onto the bottom of the ice layer.
In an ice sheet-ocean world, there’s a big ocean on the day side and a big continent on the night side. As in the water-trapped world, a lot of ice forms on the night side, up to a kilometer thick. But if there’s enough geothermal heat, and enough water, not all the water gets frozen on the night side: enough melts to form an ocean on the day side.
Needless to say, these new scenarios are exciting because they could be more conducive to life!
Read more here:
• Jun Yang, Yonggang Liu, Yongyun Hu and Dorian S. Abbot, Water trapping on tidally locked terrestrial planets requires special conditions.
Abstract: Surface liquid water is essential for standard planetary habitability. Calculations of atmospheric circulation on tidally locked planets around M stars suggest that this peculiar orbital configuration lends itself to the trapping of large amounts of water in kilometers-thick ice on the night side, potentially removing all liquid water from the day side where photosynthesis is possible. We study this problem using a global climate model including coupled atmosphere, ocean, land, and sea-ice components as well as a continental ice sheet model driven by the climate model output.
For a waterworld we find that surface winds transport sea ice toward the day side and the ocean carries heat toward the night side. As a result, night-side sea ice remains about 10 meters thick and night-side water trapping is insignificant. If a planet has large continents on its night side, they can grow ice sheets about a kilometer thick if the geothermal heat flux is similar to Earth’s or smaller. Planets with a water complement similar to Earth’s would therefore experience a large decrease in sea level when plate tectonics drives their continents onto the night side, but would not experience complete day-side dessication. Only planets with a geothermal heat flux lower than Earth’s, much of their surface covered by continents, and a surface water reservoir about 10% of Earth’s would be susceptible to complete water trapping.
From a technical viewpoint, what’s fun about this new paper is that it uses detailed climate models that have been radically hacked to deal with a red dwarf star. Paraphrasing:
We perform climate simulations with the Community Climate System Model version 3.0 (CCSM3) which was originally developed by the National Center for Atmospheric Research to study the climate of Earth. The model contains four coupled components: atmosphere, ocean, sea ice, and land. The atmosphere component calculates atmospheric circulation and parameterizes sub-grid processes such as convection, precipitation, clouds, and boundary- layer mixing. The ocean component computes ocean circulation using the hydrostatic and Boussinesq approximations. The sea-ice component predicts ice fraction, ice thickness, ice velocity, and energy exchanges between the ice and the atmosphere/ ocean. The land component calculates surface temperature, soil water content, and evaporation.
We modify CCSM3 to simulate the climate of habitable planets around M stars following Rosenbloom et al., Liu et al., and Hu & Yang. The stellar spectrum we use is a blackbody with an effective temperature of 3400 K. We employ planetary parameters typical of a super-Earth: a radius of 1.5 R⊕, gravity of 1.38 g⊕, and an orbital period of 37 Earth-days. The orbital period of habitable zone planets around M stars is roughly 10–100 days. We set the insolation to 866 watts per square meter and both the obliquity and eccentricity to zero. The atmospheric surface pressure is 1.0 bar, including N2, H2O, and 355 parts per million CO2.
And so on. Way cool! They consider a variety of different kinds of continents and oceans… including one where they’re just like those here on Earth—just because the data for that is easy to get!
Here’s a question I don’t know the answer to. To what extent can models like Community Climate System Model version 3.0 be tweaked to handle different planets? And what are the main things we should worry about: ways Earth-like planets can be different enough to seriously throw off the models?
We live in exciting times, where just as we’re making huge progress trying to understand the Earth’s climate in time to make wise decisions, we’re discovering hundreds of new planets with their own very different climates.
Azimuth 
Great article, mate. Fascinating information which gets the imagination run wildly trying to guess what would life be like in such system. It got me thinking, if life existed and evolved to an intelligent life, if that life survived its own destruction and the natural disasters that can occur in such place, I think they would be a resilient species. Which means that they could potentially conquer the entire galaxy.
Star wars/star trek sort of thing.
Thanks for sharing….
MKT
Thanks for sharing this John – it’s a fascinating thought that there could be very old civilizations on these worlds. Your article inspired me to surf a bit, and I found this speculation, that creatures like tardigrades could conceivably live in space and possibly on other planets: http://www.mn.uio.no/astro/english/research/news-and-events/news/astronews-2012-02-17.html
Thanks! Of course there will only be very old civilizations on these worlds when the universe becomes very old.
Our civilization is pathetically juvenile: fire for 1.4 million years (that’s almost respectable), calendars and paintings for 35 thousand, but cities for only 9 thousand years.
If civilizations need metal—your art certainly does—then they couldn’t arise until the first metal-rich Population I stars were formed from the material spewed out by earlier metal-poor stars. This happened about 4.8 billion years ago. Our Solar System is not much younger, about 4.55 billion years old. If life takes the same time to reach our point on every solar system (a silly assumption, but just for argument’s sake), we can expect some to be 0.25 billion years ahead of us. That’s 250 million years.
So, we could easily expect some civilizations 250 million old in our galaxy. The main puzzle is why we haven’t seen them.
But what excites me these days is the future. We may destroy ourselves, but with 100 billion stars in the galaxy some civilizations will survive and spread. Our Sun will boil Earth’s oceans in just 1.1 billion years, so G-type stars like ours are only good places to get started. But there are 80 billion red dwarfs to settle around, which will last for trillions of years! We could get really deep, mature and complex cultures.
There’s no puzzle, despite what a lot of people say.
Suppose natural processes would produce life that might turn into a complex expanding civilization, with some density (measured in civilization-starting events per year per cubic lightyear), which in space is roughly evenly distributed (a decent assumption, since we can smooth out on a scale larger than galaxies, since expansion can be fast), but in time has a peak far in the future. (We know the peak is not at time zero due to the metal issue you mentioned. We can guess it’s far in the future since life here seems to have reached its current situation after a series of lucky accidents (Robin Hanson has an interesting paper about that).)
These events either produce a failed or static civilization (we’ll ignore those for now), or an expanding one. If expanding, it either purposefully hides and doesn’t interfere with natural processes in its territory (so we won’t see it — we’ll ignore those for now too), or it doesn’t (preventing the otherwise-natural process of new civilizations starting, in its territory). So looking just at the expanding civilizations that don’t hide, they form a series of “future cones” in spacetime, whose apexes (starting points) are distributed according to the time-varying density I mentioned.
Pick a random starting point of such a civilization (or any other point in spacetime not covered by one of those cones, e.g. the current point for us if it turns out we’re in a failing or static civilization). (Or pick a point covered by exactly one of those cones, i.e. the current point for us if it turns out we’re in a successfully expanding and not hiding civilization.)
From any such point, it almost certainly appears that there are no expanding civilizations (except perhaps one’s own). (The non-expanding ones are not very visible, even if they’re not hiding.)
Or to put it another way — using the above picture of cones again, each cone’s portion that’s not covered by other cones has a relatively thin surface layer of points in spacetime from which that cone’s civilization was visible from that point, but had not yet overrun that point. So we are saying “any point which is not overrun is probably also not in that layer of points from which a potentially-overrunning-civilization is visible”.
The above argument has a few more assumptions which I haven’t listed, mostly about the likely actual timescales of the processes involved, but AFAIK they’re all reasonable. For example, known historical events that affected our becoming civilized (like the asteroid that wiped out the dinosaurs) had random time components of hundreds of millions of years, so we know the distribution of start-times is not sharper than that. In fact, we can assume we’re in the “early tail” of it, since otherwise we’d have been overrun. It also assumes that our potentially-expanding civilization is “typical”, which is reasonable without evidence to the contrary.)
(The above argument places no limit on density of civilizations which could have overrun us if they chose, but are “hiding” instead.)
Note that this is an application of the “anthropic principle”. That is a principle which can never explain an observed fact, but can tell you that you ought not to be surprised about an observed fact which you might otherwise think was very surprising. (E.g. that you happen to be located on a planet’s surface, even though that property is very rare among all points in space. Or that that planet has a big moon. Etc.)
Bruce wrote:
Well, there is a puzzle… but I think your answer to the puzzle is right. Thanks for explaining it so clearly.
If I understand correctly, you seem to be saying that “Do not be surprised that we have not observed any alien civilizations, because if we have observed an alien civilization, you would not be asking this question”. I am fine with such an answer if somebody is surprised that we are located on a planet’s surface, because a random point in space would most likely be deadly to any for of life. But observing an alien civilization does not necessarily mean an end to the observer civilization. We could observe or even interact with an alien civilization and still go on with our lives.
replying to temur:
> you seem to be saying that “Do not be surprised that we have not observed any alien civilizations, because if we have observed an alien civilization, you would not be asking this question”.
I hope that’s not what I’m saying! For one thing, I’m distinguishing “detecting aliens” from “being overrun by aliens”; for another, I’m making and using some arguments about plausible time and spatial scales, distributions, and “spacetime geometries” of both civilization-starts-expanding events and the expansion itself (versus our ability to detect it). So let me try to be clearer, and also to make explicit whatever assumptions I left implicit (or missing) the first time.
(Also, perhaps the term “overrun” is too loaded. I’m not saying it’s impossible they’d be interfering but not bad — I’m just ignoring that issue, since I think we all agree they’re not visibly interfering with us at all. But I don’t know a shorter term for “a place they’ve reached in the sense that whatever happens there is able to be heavily influenced by them”.)
What I’m arguing: that (given our existence) there is at least one reasonably likely-seeming scenario (described below) that results in the situation we observe, namely that we (residents of this solar system) seem to be the only civilization we can see. Therefore, the fact that we don’t observe other civilizations can’t be considered surprising. (This is a very weak point, which makes it easier to argue.)
Assumptions (remember that for my argument to hold, I don’t need these to be proven, only to be reasonably likely, i.e. “not surprising if true”):
1: events of the type “a civilization starts expanding to other stars” are distributed in time in a way that has a random component of at least a few hundred million years, but is otherwise unknown; and in space in a way similar to stars (when smoothed on spatial scales of a few million lightyears or more).
2: we might be close (in space and time) to an event like that, since it’s possible we’ll soon start expanding that way.
3: for a significant fraction of expanding civilizations, the civilization is likely to interfere (detectably to other civilizations in those places) with most places it reaches (on timescales of millions of years after the expansion starts), since the only way it could refrain from that would be to enforce strict rules for all time on all its expanding elements — since the first ones to reach new places would determine the overall nature of how they interfered, and it’s possible to expand faster if you don’t refrain from interfering with the places you reach (e.g. by using up lots of their resources). (I’m not sure if my argument actually requires this point. Exercise for the reader: did I use it anywhere, below?)
4: it’s hard to detect any civilization from farther away than a few million lightyears, unless it’s expanding rapidly in a highly interfering way (e.g. turning much available matter into light to help clear the way for the expansion), but in that case you only detect it when it’s almost reached you (relative to the time elapsed since it started expanding), since its light signal expands at the speed of light, only somewhat faster than the civilization itself. (Whether “civilized” is a good term for aliens behaving that way is questionable, since the light they made would destroy much of what it reached before they got there, but the choice of term is not relevant to the argument.)
So, if some civilization is expanding rapidly and making lots of light, you might detect it from far away, but the volume of space from which you can detect it is not much greater than the volume it’s already overrun. And if it’s not doing that, you won’t be able to detect it from far away. (Note that the relevant spatial scale is related to the assumed timescale in (1), i.e. it’s hundreds of millions of lightyears.)
Given these assumptions, we look at the likely geometry of how spacetime is filled with “expansion events”. Each of those generates a “future cone” of places it has overrun, and a larger region of places from which it’s detectable (since some sufficiently powerful signal it’s emitting has reached them). (If you want to visualize only space, not spacetime, then these regions are nested spheres of about the same radius, except when they are small (just starting to expand).)
Point (4) is saying that these regions (overrun vs. detected-from points) have comparable volumes, on the relevant scales, except whey they are small. That is, for any given expanding civilization (unless we’re close to the start of its expansion), if your relative position in time and space is random, if you’re able to detect it then you’ve probably been overrun by it. This is not about what it is likely to do to you (aside from being detected or “getting here” respectively), only about relative volume in space of two different kinds of physical consequence of its expansion.
So given all this, what does the universe look like? First it’s mostly empty. Then there is a small density of places from which an expanding civilization can be detected. (Since “the fraction of places like that, as a function of time” can’t change very rapidly from near-0 compared to hundreds of millions of years, by our assumptions. Note that I’m not saying *when* that time arises, only that it exists and doesn’t have very short duration.) That density increases over time, and eventually reaches almost 1 (when so much time went by that expanding civilizations mostly hit each other).
Following the wave of ability to detect civilizations from given points, comes a wave of their overrunning given points. So the density of such points is smaller but also eventually reaches almost 1.
Now where are we in that picture? At a random point which is distributed much like the points at which expansions start. (Since it’s reasonably possible that we’re within a short time of starting to expand like that.)
(To visualize it in spacetime, “upside down”, think of a random union of very pointy mountaintops, like volcanic cones. We’re near the tip of one, which is not buried inside another one.)
So picking a random point at which an expansion started (and not already overrun by someone else), how likely is it to be a point from which we could already detect some other civilization? (In the mountain range picture, how likely is the tip of one mountain, picked randomly among visible tips, to be very close to the sloping side of some other (higher) mountain, either in the sense that looking at its top means looking up almost 45 degrees from horizontal, or that we’re within a few million light years of it in absolute distance?)
Not likely, because:
– except at distance scales shorter than a few million light years, and at a time at which many visible start-events (mountain tips) occur, the volume of space from which you can detect others is similar to the volume they’ve overrun;
– but if that small distance scale is relevant (since the start-event-density is high at this time), it means that we’re probably within a few hundred million light years of the expansion-start-points of several expanding civilizations, but if the density of expansion-starts was that high now, then due to the known time-smoothing in (1) it has been high for at least a few hundred million years, so some of them have probably overrun us by now.
I’m sure this could be made clearer, especially in the logical organization of the last few points, but I already spent more time than I have available for it right now, so I’ll stop here and ask whether this has actually made my overall point clearer.
Thanks Bruce. I like your way of thinking about this. Couple of Qs-
1) For this model to work, it seems that static civilizations would have to be a *lot* less intrinsically visible than the overrunning subset of expanding civilizations. Is this plausible?
2) Your “relatively thin surface layer” wouldn’t be very thin unless the overrun frontier expanded at something approaching lightspeed, right? If that is right, this might suggest that the overrun *mechanism* is typically informational rather than physical.
Bruce, I see your latest clarification may have shed light on one or both of my Qs. So you can ignore them for now.
Now I remember why assumption (3) is important: it rules out a scenario being likely where we see some civilizations, but are not overrun by any, because (in this scenario) it’s the norm for civilizations to waste (or use for beacons) lots of energy for a very long time (thus being visible from far away), but to expand much slower than lightspeed (thus overrunning only a small fraction of the volume they’re visible from). Point (3) says “maybe some would do this, but (plausibly) not almost all”; even one of them (starting at a similar time and distance from us as the others we see), expanding rapidly, is enough to overrun us. (And the others can’t stop it since they don’t get here in time.)
(Arch — Thanks. I think I covered some but not all of your points, but I don’t have time now to reply — maybe later today.)
(Arch, this replies to only one of your points.)
By “informational vs physical expansion” were you distinguishing whether we (or some similar aliens) send physical things that can think and act, vs pure information signals (whose decoding and use is entirely up to the recipient)? Or were you assuming we send physical things, and distinguishing whether they are our biological descendents vs. some kind of robots with AI or uploaded minds? I guess the latter, since I don’t imagine pure information can “overrun” (though I suppose someone could disagree and cite religious conversion or computer viruses, as at least some SF novels have done).
For the argument I’m making, if we or someone sends something that can think and act, it makes no difference whether it’s descendents or robots. But it does matter whether we can send at high speed something physical that can act, not just a signal (since I’m not considering “overrunning by pure physically-passive information”). So your comment shows me another assumption I forgot to make explicit: that we can send a physical actor quickly. That is:
Assumption 4: I think it’s plausible to assume that beings much smarter than us, mastering nanotech, etc — that is, much farther than us on the trajectory of progress in technology that we have at least some chance of staying on for awhile yet — can send robots (small, but the lucky ones able to act as seeds using what they find there) to other stars and galaxies at near lightspeed (at least, say, 90%, which is not very high in terms of magnitude of relativistic effects). Whether they are likely to decide to do that or not doesn’t matter for this argument, as long as there’s some chance they decide to, since making this decision is exactly what I mean by saying “the civilization starts expanding”. (My point (2) assumes that for us, there is a significant chance we’ll decide to do this, or make something which does.)
I have a hard time taking the opposite position seriously, given that even with our current level of understanding there have been plausible proposals for how to do that, but I understand that many people do. So for them, this overall argument might not work — if they are optimistic about our ability to detect civilizations from extreme distances (e.g. by expanding slowly in a very controlled way, to super-galactic sizes, and building networks of telescopes and quantum computers which act as one big telescope of effective diameter equal to the entire range of our expansion, which can thus resolve millimeter-sized features at a distance a thousand times its diameter), but at the same time pessimistic about the chance of expanding quickly (so they think some absolute physical/technology limit or necessary social effect prevents a sustained expansion from ever being faster than, say, 30% of lightspeed), then they could conclude that a scenario I “ruled out” above — lots of slow expansions but no fast ones — is overwhelmingly likely, which could lead to a random point near an expansion-start being able to detect other civilizations (perhaps many of them) but be far from having been overrun.
(Even then, to detect them, one would have to build much better telescopes than we’ve done so far, unless the maximum expansion speed is more like “galaxy diameter / random timing of start-events” == 0.1% of lightspeed (and the start-event density is high enough). To be fair, I guess that must describe the thinking of actual early SETI proponents, so it’s not fair to dismiss it entirely — only to say “this puzzle does have an easy solution if you’re willing to take certain things as plausible which lots of people now think are very likely, like assumptions (1-4)”.)
Arch — you also said
> 1) For this model to work, it seems that static civilizations would have to be a *lot* less intrinsically visible than the overrunning subset of expanding civilizations. Is this plausible?
Yes (provided there are not a *lot* more of them), because they’d be so much smaller. E.g. if we just let the big ones expand for the same duration as the assumed minimum scale of randomness in our own start time, 100 million years or so, at a significant fraction of lightspeed, and assume their density is such that none of them reached us yet, and that the small (unexpanding) (yet highly visible, ie not dead) ones are no more than a thousand times as dense as the big ones, then the closest small ones tend to be at least 10 million light years away. So they are 10 times closer but (let’s say they expanded to our-galaxy-size and then stopped) a million times smaller in visible area from here, thus intrinsically 10,000 times less visible (even if they’re emitting as much energy as the expanding ones, and even if we don’t count volume, only visible area, as helping to make an expanding one detectable by emitting energy).
I do see a way I could be wrong about that — if an expanding one is trying very hard to hide *during the expansion*, even though it’s going to interfere enough once it gets here to count as *overrunning* us, but if static ones are not trying to hide (in fact if they are trying to signal us), then this could even out their relative visibility.
I admit that’s possible, but assumption (3) makes me claim it’s not very likely to be a universal law applying to most expanding civilizations, so it doesn’t nullify the overall claim that “the puzzle of missing aliens has an easy solution”.
Bruce,
Thanks for fleshing your ideas out. As your avg typing speed seems greater than my peak reading speed I must be selective in my response.
Re: your 2/17 8:53 pm msg, by an “informational” overrun mechanism I did mean one based purely on message-sending (by e.g. getting someone – anyone – at the receiving end to do/build/think something that would trigger their civilization’s demise). Admittedly this presents challenges and I haven’t thought it through.
A physical expansion at near c seems challenging, too. I suppose the easiest way to do this is with small self reproducing automata. But unless they could harvest material at speed from the interstellar medium in order to build their progeny, it seems the SRAs would have to stop occasionally to reproduce, slowing their avg expansion rate. I think there have been analyses of this, but don’t recall the upshot.
arch1 wrote:
Or if you want to be more optimistic, something that would trigger their civilization’s becoming an ally and joining the expansion. (Something a bit like this happened in Carl Sagan’s Contact; I think the basic idea (maybe with demise more common than alliance) has happened in other novels but I don’t know/recall them.)
A physical expansion at a moderate fraction of c (say 90%) has been explored often, both in serious SF and serious analyses of possible real methods. I have not studied details (and can’t give references) but I’d be very surprised if it was beyond the power of very advanced tech. My general impression is (in terms of what can be imagined with any confidence now) 10% of c is thought to be vastly easier than 90% which is vastly easier than 99%. But personally this is just an interest, not even a hobby, so I don’t invest any time in studying this seriously — also partly because any such study (by anyone) will be way obsolete by the time it’s close to being practical. (Accordingly, I admit everything I say about it is “opinion” (somewhere between semi-educated guess and general bias) rather than “strong claim”. Except for a claim like “something is plausible enough that if true it would not be too surprising”, which is weak enough to make even under those conditions.)
Bruce wrote:
As you know, if you have the ability to keep accelerating, you can keep asymptoting up to c. Then, I believe, the main difficulty becomes dealing with collisions with small objects. Quoting a website whose accuracy I haven’t checked:
The nonrelativistic calculation is pretty close to right at 0.5 c; none of this even gets into the relativistic effects that become important as you get to 0.9 c or more; I leave these as an exercise for the reader who has time for fun calculations!
Dust grains are one good reason to use lots of small “ships” (like a millimeter or less in diameter), not traveling right next to each other (so when one explodes it doesn’t harm the others), rather than a few big ones. (Credit: a short story by, IIRC, Marc Stiegler.)
I imagine that outside of galaxies, where you would spend most of your time if you’re trying to go far, there are a lot fewer dust grains, anyway.
Yes, I suspect there’s almost no dust in intergalactic space, it would have to be formed by stars and diffuse out there.
Here are some other rather obvious points:
You don’t get anywhere much faster at 0.99 c than at 0.9 c, so there’s not much point in going at highly relativistic speeds unless 1) you want to sneak up on someone, reaching them shortly after they first see you or 2) you want to take advantage of the relativistic time dilation to age more slowly en route.
I believe any advanced civilization will have ways of ‘putting the crew in suspended animation’ during long flights, e.g. simply turning them off. So, 2) doesn’t seem like a real issue, unless we let some idiot at NASA launch a manned interstellar mission before we’re ready to do a good job of it—sort of like the stupid Mars mission idea multiplied by 600,000.
On the other hand, it seems like 1) is a real issue for civilizations that are trying to expand and colonize the Galaxy before anyone else can stop them, or for civilizations engaged in internal wars.
So, if we wait long enough, we should expect to be overrun by a wave of probes that arrive shortly after we see them… unless we set off such a wave of probes ourselves.
Which reminds me: on 17 January 2012 I wrote this on G+:
I tend to look at these issues from an evolutionary persepective. I come to similar conclusions to Bruce Smith, if anything stronger. I think there are some general lessons we can learn from evolution on Earth, which will apply to any organisms that disperse and replicate in space.
Whenever evolution faces a ‘choice’ – some organisms can make a living this way, others that way – evolution explores both paths. One or both may prove dead ends later, but everything that can happen now does happen now. I don’t believe that intelligence or politics or culture can withstand Darwinian selection for long. So the answer to a question like “will it be humans or genetically modified humans or cyborgs or self-replicating machines?” is yes to all of them, plus many things we don’t have names for yet. Similarly for “hide or show?”, “static or expanding?”, “nm or um or mm or m or km or Mm?” and so on. Likewise, if most organisms descended from a particular planet disappear into another universe, the ones left will still colonise this universe. In more scientific language, any attempt at channelling evolution down one path will not be an evolutionarily stable strategy.
Another lesson is that when a species acquires a key innovation (http://en.wikipedia.org/wiki/Key_innovation) a rapid diversification follows. The ability to disperse and replicate in space is a key innovation to beat all others. The diversification is going to be staggering.
A third lesson is that all but the largest organisms will have predators, and all but the smallest will have parasites. This may not seem relevant to the current discussion but I think it is, because it adds to the general messiness and uncontrollability of everything. (I think of computer viruses as a harbinger of this phenomenon.)
I can think of two major differences between evolution so far and evolution in space. The first is that the relative importance of dispersal and replication changes, at least until much colonisation has taken place. On Earth, there is a struggle for existence. In space it will be, or at least begin as, a race. The other big difference is that many organisms will design their own progeny. I don’t know what effect that will have, except making things even more complicated.
John said: “You don’t get anywhere much faster at 0.99 c than at 0.9 c, so there’s not much point in going at highly relativistic speeds unless 1) you want to sneak up on someone, reaching them shortly after they first see you or 2) you want to take advantage of the relativistic time dilation to age more slowly en route.”
If you are packing your kids off to a galaxy 10 Mly away, and you can send them at 0.91c while your neighbours can only manage 0.9c, your kids arrive in a probably nearly empty galaxy over 100k years earlier: the galaxy is theirs. All organisms want to give their kids the best possible start in life, and what could be better than that? Going faster and faster is not something that some weird aliens or machines with weird motivations might choose to do. It is something that a population of organisms which disperses and replicates in space cannot resist doing.
Sadly, I don’t think red dwarfs have much of a future.
Thanks Bruce, Graham and John; much to chew on and great refs.
BTW, on reflection I’m decreasingly concerned about the various overrun scenarios. Marauders and their proxies are famously susceptible to honor challenges. We just lure them to the Los Angeles area for a winner-take-all intellectual showdown at which (leveraging James Randi’s card-forcing expertise) we get them to select John, Terry Tao, and Sean Carroll as their random opposition. (OK, plus maybe a 4th person coming off the bench to cover sports trivia and British 70’s punk bands.) The poor buggers will never know what hit them.
Graham wrote:
I think they have a great future, starting around 100 billion years from now when all other stars will have burnt out. They will last 10 trillion years. I can’t believe that such useful sources of energy, lasting for such a long time, will fail to be exploited! Your comment on how every niche gets filled is relevant here.
I’m really disappointed by the short-term thinking of some comments here.
I never said that red dwarfs were important in the short term.
Okay, I stated point 1) too narrowly. I probably should have said that the speed advantage is only worth the extra nuisance in invasions, wars, and other speed competitions.
(The situation you describe sounds like sending your kids off to Europe—but they’re really invading another, most likely already occupied, galaxy. There will be competition between them and other invaders from our galaxy, and also competition with civilizations in that other galaxy.)
Graham: that’s a very good and clear explanation of some of the issues, including some new ones (to this discussion anyway).
If there is indeed a competitive race, I fear it might use up excessive resources and cause excessive destruction, though I think this depends on technical details of what’s possible that it’s very hard to forsee. In the worst case, the best strategy to spread rapidly could be to convert most local matter to light to help propel you, and to help push interstellar matter out of your way which would otherwise be a hazard to high speed travel (also making it easier to eat, by making its velocity closer to yours before you reach it). We can only hope that either that’s not true, or there’s not a race (at least started by us — we can’t avoid some aliens expanding that way) (avoiding a race might be possible if we have a good enough unified rule-enforcement mechanism — the rules need not be onerous in their ultimate effect), or that such races only destroy the center of each independent civilization’s expanding bubble, since when they’ve gone far enough that they predict they might run into alien neighbors any eon now (and are no longer close to the co-species neighbors they were competing with), they ought to want to conserve some resources for longer-term use, since the expansion phase won’t last forever (they will hit similarly powerful aliens and be stopped, and at that point it’s in their interest to ally with those aliens, or at least not go to war with them).
(I am optimistic that there is *some* way to avoid a destructive race, in the sense that a significant fraction of expanding civilizations avoid one (so they can expand at a leisurely 90% or 99% c instead, and not disturb what they expand through very much until they want to). All of Graham’s arguments that a race is “natural” are correct, so avoiding it is not easy. I’m afraid my main reason for optimism is just because I want to avoid it so much (i.e. wishful thinking) — I’m a fan of nature and don’t want to destroy all of it during the expansion. But since a lot of people feel that way, and the expanders will be very intelligent, maybe what’s necessary for that will prove possible.)
John: I think you are missing a few of the points here. I don’t know for sure what Graham meant about red dwarf stars not having much of a future, but maybe it was that they were likely to be “eaten” by something impatient to use their matter as energy much faster than is happening naturally. (Whether this is actually a likely fate, even in a race, depends on the technical details I can’t forsee which I alluded to above.)
As for a newly reached galaxy (in a race to expand) being “likely to be occupied”, the argument I gave earlier predicts that’s very *unlikely* (except by competitors in the same race, i.e. from the same origin point) until we’ve gone a very long way (past at least billions (by volume) of unoccupied galaxies, perhaps arbitrarily more).
Bruce is right that I think red dwarfs will be eaten long before 100 billion years goes by, though not necessarily by ‘racers’. The following is meant to expand his last paragraph.
Earlier, Bruce mentioned a “minimum scale of randomness in our own start time, 100 million years or so”. I think that’s OK as a minimum, but a realistic estimate would be bigger. To keep things simple, suppose the rate at which galaxies produce organisms capable of expanding was zero until 1bn years ago, then a constant r per bn years. That might roughly match a presumed need for metal plus some time to build up oxygen levels plus some time to evolve suitable organisms. I believe that nearly all such expansion-capable planets will produce *some* organisms which *do* expand at 0.5c or more, and in a way we couldn’t help but notice if they got here.
The Virgo Supercluster (http://en.wikipedia.org/wiki/Virgo_Supercluster) contains a lot of galaxies (I can’t tell from Wikipedia, maybe 2500?) within about 70Mly of us. The planets here have had at least .8bn years to overrun us, given my assumptions, and they haven’t. So r <= 1 expansion-capable civilisations per 2000 galaxies per bn years. I think you could put a lower bound on r by considering galaxies further away. That's why I think any galaxy is likely to be "nearly empty". There may be many planets with life which cannot leave its home planet, but of course that life cannot fill up much of a galaxy.
I thought I should mention a specific reason to be optimistic. Here on Earth we have a lot of “civilization”, a phenomenon in which, under certain conditions, most strangers care about each other’s welfare (or act that way, which is good enough). Just looking at the animal world and thinking about evolution, I think you would never predict this or even conceive of it. It seems to me it’s an artificial construct (relative to the animals-are-normal, humans-are-new-and-different point of view), which requires intelligence (and maybe other conditions) to be created, but *is* created and maintained here, fairly reliably, when the situation is stable enough. So we can hope that even higher intelligence might lead to even better forms of civilization or something like it, to help us escape from otherwise-likely-seeming wasteful conflicts, even with strangers, maybe even with intelligent aliens we are only predicting but haven’t yet met.
(In the context of expanding into the universe, I think even our current level of civilization, if maintainable, would let us create and protect “nature preserves”, including most planets with native life if those turn out to be rare among current-human-habitable planets.)
Note that I only said we can “hope” that some kind of “higher-level civilization” is possible, not we can “predict” that. Personally I think it could go both ways, and we ought to be trying hard to work towards the “higher civilization” way if we can figure out how. (Please don’t misread this as a claim that any specific known form of that, or claimed form, is good enough, or even more good than bad.)
Bruce, another thought just hit me in connection w/ your near-lightspeed-overrun scenario. Yes, the scenario, if true, does make it unlikely that we would see anyone a long time before being overrun; but unless overrunning-expansion start-points (SPs) are extremely rare (say, <1 per visible-universe volume), doesn't your scenario *also* make it extremely likely that we would have already been overrun? If so, the fact that we have *not* been overrun would seem to suggest that for your scenario to be plausible, its SPs must in fact be extremely rare.
Do you buy this so far?
Bruce wrote:
Indeed, in my initial comments about civilizations using red dwarf stars as a resource lasting 10 trillion years so they could think about interesting things, I’d imagined that they’d have tired by then—as I have—of the biological imperative to propagate themselves as much as possible at all costs. I was imagining something like Europe to the 20th power, where there’s a kind of sensible restraint built into the system, after the youthful follies are over.
But of course there’s no way to be sure: as Graham said, everything that can be tried will be tried, and ‘sensible restraint’ will only spread if and when that’s a strategy that does well against all competitors.
In reality what happens in 20 years, much less a million, billion, or trillion, is extremely unpredictable. My only reason for thinking about the far future is that it’s interesting and it gives me a way to locate my life in a context that eases the pain of mortality.
I don’t find my predictions depressing, perhaps because I am used to them. Or maybe I’m a short-termist. After the first billion years it all seems a bit unreal to me. And can life ever rival black holes as a destroyer of resources? Perhaps it should try!
I think that, more likely than not, there is nothing out there in the whole Virgo Supercluster that you could have a conversation with right now. And that seems very sad and lonely, as well as a terrible waste of starlight (or stars). I prefer to imagine it colonised, even by organisms which are “overpaid, overfed, oversexed and over here”. I do not think that evolution simply selects for profligacy, its more complex than that. Better to say it selects for a very wide range of frugal to profligate lifestyles. I think there’ll be lots of niches for those who want a quiet life. But a mixture of frugal and profligate looks profligate overall, and that’s what we’ll see first.
arch1 asked (I’m paraphrasing): “doesn’t your argument imply that expansion start-points (SPs) are extremely rare (say, <1 per visible-universe volume)?"
It's important to think of that volume as in spacetime, not space, since as you get farther away you see the results of less and less time since the universe's beginning, and since the rate of start-point creation per unit volume might be lower for earlier ages.
But given that — yes, that's the main point. The puzzle is "why don't we see them"; the argument includes "if they were there (in our past-cone), they'd probably have overrun us by now (unless they're hiding)"; and one conclusion is that a plausible explanation for our not seeing them is simply that they're rare enough that we shouldn't expect to, so far. (That's not a mystery in itself, since we have no evidence whatsoever, practical or theoretical, that such start-points are not rarer than any specific density.)
I guess that's a simpler way of saying it than I actually said. What I said looked at the ratio of spacetime volumes in which we observe they didn't start and then overrun us, vs. in which we observe they didn't start and then become visible to us. But your (and Graham's) way of putting the argument is simpler and ultimately equivalent.
Hi, John! Another important question, beyond the issue of life on red dwarfs is the following: could human beings live on Earth-like planets in these kind of stellar systems? I am thinking about “eye problems”, radiation problems, and so on. Life on red dwarfs will be likely very different from that here on Earth. I am wondering if someastrobiologists have written about the adaption of human species or fauna to that red sun environtment or the kind of life we could find on it…
By the time we’re able to get ourselves to another solar system, we’ll be able to refashion our bodies in ways that can handle life in many environments—even hard vacuum. Biotechnology, robotics and artificial intelligence are advancing much faster than our ability to send large payloads multi-light-year distances. We should not, in the long term, remain attached to the current human form.
That’s true, but then why do you think red dwarfs are important at all? (As settlement targets, that is — as opposed to as places to speculate about finding native life.) By the time we can do those advanced things, all that should matter to us about where to live is what places have access to resources and energy and sufficient physical stability, plus other considerations we can’t predict now (e.g. social, political, aesthetic). Unless those other considerations are very special, “near an unstable star” seems like a relatively bad place to live.
(Though I guess you might say, “they’re unstable in ways that won’t bother us then, since we’ll be well-shielded, and stable in ways that do matter, since if we can prevent ourselves and anyone else from eating them or attacking us for all that time, they’ll last as energy sources”. But then, so would a bunch of cold comets we were slowly running through more efficient fusion engines while hiding in the dark spaces between galactic clusters.)
Bruce wrote:
Right. Red dwarf stars calm down as they age. The younger ones have annoying flares, but that dies down, and then they’ll be putting out energy in a reliable way for trillions of years, long after larger stars have died out. That’s why they could be good long-term locations for life.
In the next few billion years they may not be so important, but I’m indulging in some long-term planning.
Wikipedia says:
All this is separate, of course, from whether red dwarfs have planets that can develop their own life. That’s mainly interesting to me because 80% of stars are red dwarfs.
That is interesting.
I am thinking that the circumstellar habitable zone depends on the definition of life; for example, for me, an object with consciousness is life (a conscious robot, or a conscious computer like a conscious sterile mule).
If I were using this definition, then almost each planet near a star is habitable (there is a stable form of energy and mineral content), with each atmosphere, with each size (so that the probability to detect life could be high); so why look for sign of life as human ones (signature of hydrocarbon in the atmospheres), or to search Earth-like planets, if there may be alternative types of life?
We don’t know where life arises, so it makes some sense to pay special attention to places that resemble the one place where we know it has arisen.
However, this should not prevent us from looking at other places.
I don’t think all planets are created equal: some have complex mixtures of chemicals in liquid form, while others are basically rocks under vacuum. I would be quite surprised to find life on Mercury, since I can’t imagine how it would get started or flourish there. I would be less surprised if we found life on Jupiter or Europa, since I can imagine roughly how it would work.
I unclearly said that if there are old civilizations, that can build conscious objects, then there is not a habitable limitation; for example mars exploration rover, or philae, are probes without consciousness, but it can be possible in some time (with different technologies) to use artificial intelligence technology for exploration and extraction of minerals; then the probability to observe civilizations is greater of Drake equation estimation.
So that, for example, a human Mars colonization may be possible, but Mars is not in the habitable zone.
Okay, now I understand: almost any planet receiving sufficient sunlight should be a habitat where conscious beings can live, for example if they were built elsewhere and designed for that purpose.
This indeed is my hope: that in the long run life will spread across the Galaxy. And though they should be much smarter than us, and able to figure out how to survive for a very long time, I was still happy to learn that 80% of the stars in our Galaxy will produce easily used power for trillions of years: this makes their job rather easy… at least for a while.
So unless you don’t use any (gen)tech upgrades, humans with light sensitive eyes may eventually have an easier life on Gliese 667Cc ? I haven’t found any scientific investigation for this, but I think that at least to some extent having more light sensitive eyes may have eventually been an advantage for northern civilizations (just imagine hunting in northern nights). And based on another rather unscientific observation about people in my vicinity those light sensitive eyed seem to have on average more often blue or green, i.e. lighter eye colour. Again I haven’t found any scientific study for this, so this correlation might be wrong, but it seems also a reoccurring theme in the internet.
So slightly exaggerating it seems you would eventually either need to hurry up with Gliese 667 Cc settlement or shift more towards Gateca -like methods.
This comment was also thought as a referring to your comment:
I don’t know the latest on this story from June 2014:
In astronomy metallicity refers to proportion of all baryonic stuff other than Hydrogen &. Helium, not just metals in a chemical sense of the word. Therefore both water and organic compounds are high on metals.
Moreover, the majority of Earth like planets is much older than Earth itself, several percent being at least twice as old.
Therefore you may want to consider the possibility of abiogenesis being an extremely unlikely process even under the right circumstances.
Please note we do not have proper theoretical understanding of this particular transition, certainly not one that would make quantitative predictions possible. The only thing we do know for sure is existence of evolvable replicators is a precondition to Darwinian evolution, and there is strong indication that even their initial Kolmogorov complexity should be quite high.
As for empirical evidence, all we know is the conditional probability of abiogenesis, provided we ponder about this very question. Its value is exactly 1. But it tells us next to nothing about its intrinsic probability, other than it is a possible event, that is, its probability should be greater than zero. But, as we know only too well, positive numbers can be arbitrarily small.
Considering the universality and ragtag nature of the genetic code, we can be quite sure there was a unique event leading to its emergence, freezing it in thereafter.
So, if spontaneous abiogenesis were likely on Earth like planets, we would have a problem. It does not really matter, if propagation of advanced life forms can approach the speed of light or not, entire ecosystems can’t hide from observation anyway.
By now we can see at least this much. Universal replicators are possible, there is nothing in physics that would prohibit their existence. These creatures would not need anything else to proliferate than free energy and raw materials, mostly carbon. First of all, they would not need planetary surfaces to survive and replicate. It’s even better if eventually it is figured out how to fuse 3 4He nuclei into 12C.
For Darwinian evolution it does not matter, if the source of variability is random mutations or an ongoing process of intelligent design and re-design, selection still prefers the fittest, that is, the one with more offspring.
Therefore hiding is not an evolutionally stable strategy, in a pristine environment replicators can increase their numbers exponentially until they come to dominate it. In a universe filled with stars that means using starlight as a source of free energy and using as much as possible. Until no stars are visible any more, presumably overshadowed by Dyson swarms.
Even at a slow speed of propagation, entire galaxies or local clusters can be transformed into this state in a short time, relative to the age of the universe. That means there should be infrared galaxies with as much or more power flux output as ordinary ones, but at a much lower temperature, that is, with orders of magnitude higher rate of entropy production. One can’t possibly miss such a beast or a bunch of them with our current space based infrared telescopes.
Icarus Volume 151, Issue 2, June 2001, Pages 307–313
doi: 10.1006/icar.2001.6607
An Estimate of the Age Distribution of Terrestrial Planets in the Universe: Quantifying Metallicity as a Selection Effect
Charles H. Lineweaver
If one has a growing organism (let’s interprete here “producing offspring” e.g. as growth of a swarm/cell organism, like something similar to some bacteria or trichinella collection) then if this organism is within an environment like a human body then further growth may make the human body use e.g. penicillin to kill that organism, i.e. “dominate”. So given the circumstances it may be more “fit” for that organism to have stop growing and it would be eventually even more “fitter” when being flexible enough to e.g. help out within the gut flora.
The human body is not a pristine environment, whereas as space mostly is. But yes, things can get very complicated when different species interact: Eg:
http://news.sciencemag.org/2013/03/scienceshot-why-bacteria-commit-suicide
Interesting research as it seems. Maybe I oversaw something but your science mag article seems to have no link to the original article. So in particular I couldn’t understand how the researchers proved that it was suicide and not “murder”.
Never heard that strange song before. Is this a british folk song? Who sang that song?
In particular the sentence is not in the
lyrics of that song by the Manic Street Preachers.
By the way – a lot if not most suicides are not painless.
Re the song “Suicide is Painless”: it’s not exactly by the Manic Street Preachers. I know it as the theme song to the old series M.A.S.H. from the 1970’s, which is probably the way most people came to learn of the song. You can hear an old recording by Johnny Mandel (also with lyrics) here: https://www.youtube.com/watch?v=2-BtquTKw78.
I am sorry. I somehow got the impression that Jonny Mandel and Mike Altman were part of the Manic Street Preachers, which if I understood you correctly is/was not the case.
Anyways the second part of Traci Watson’s sentence:
may indicate that he or she was having another “suicide is painless”-song in mind which is neither by Jonny Mandel, Mike Altman or the Manic Street Preachers and it would be interesting to know what kind of song this is and who wrote that song.
I don’t think the author meant to refer to a song in the second part of that sentence. If he or she had written, “…, the old song is right: ‘Suicide is painless’. As well as a blessing to one’s neighbors,” it might have been clearer.
nad said: “it may be more “fit” for that organism to stop growing”…
yes, and this can apply even more to an expanding cilivization, if it’s intelligently controlled (governed) and *imagines* that it’s part of a larger system containing perhaps yet-more-powerful civilizations. (Or in a finite place whose resources would run out if it didn’t stop growing.)
For the universe setting (as opposed to Earth), the slowness of communication makes that intelligent governance much harder — it pretty much has to be agreed on, with a perfectly reliable but not overly restrictive enforcement mechanism already running, before the expansion gets underway at all.
[…] How suitable could planets orbiting red dwarf stars be for supporting life? […]