Certain facts about M-class red dwarf stars are vexing for authors and readers of SF. Not to mention reviewers. I am vexed.
First fact: they’re economical. Because they are low mass, you can make a lot more of them from a given amount of matter than you can make of mid-K to mid-F class stars1). Also, they last a long long time, even by galactic standards. Someone or something must have been frugal, because the vast majority of stars are red dwarfs. This proportion will only increase once the stelliferous era draws to an end in the near future (by galactic standards).
What’s so bad about most of the galaxy being composed of long-lived stars? Well, I am happy you asked…
A lot of science fiction authors simply ignore red dwarfs, if only because simple math suggests that the odds of an Earthlike world being in the habitable zone of a red dwarf must be pretty slim. After all, the Sun is fairly bright as stars go and it only has three potentially habitable worlds in the Goldilocks zone—two of which are dead as Dillinger. A back of the envelope calculation suggests that if the Solar System is any guide, most Earthlike worlds in red dwarf systems would be too close or too far away.
But the universe does not necessarily conform to reasonable expectations. A surprising number of red dwarfs have potentially habitable worlds in their tiny Goldilocks zones (as we now know, having learned to detect extrasolar planets). Proxima Centauri, for example, has one, despite the fact that its habitable zone is roughly the width of a piece of paper. Gliese 581 has one, too. TRAPPIST-1 has three and what’s up with that? It seems any quasi-realistic setting will have not just a surfeit of red dwarfs, but a surfeit of habitable worlds orbiting them. Sorry—potentially habitable. Let me explain.
Red dwarfs are roughly as bright as a 40-watt bulb. For a world to be close enough to a red dwarf to be potentially habitable, they have to be close enough that, like our Moon with the Earth, they would be tidelocked (technically, what is called 1:1 spin-orbit resonance.). One side will perpetually face their primary and one perpetually face away2].
That’s fine for the Moon. We have an interesting view because most of the mysterious someone/something’s SF/X budget was spent on the Lunar nearside. Nothing of importance is lost because we never see the farside.
A planet, however… one side will be bathed in continual sunlight, while the other lies in Stygian darkness. In olden days, some feared this would lead all the volatiles like water and oxygen to precipitate out on the night side. Recent models suggest even a modest atmosphere would prevent that from happening. Unfortunately, proximity to the star means exposure to solar flares: goodbye atmosphere. Well, maybe.
This presents hard SF authors with the annoying possibility that the Milky Way is replete with worlds that would be habitable if only they weren’t in orbit around a red dwarf. This seems wasteful. But until our telescopes get good enough to say whether or not the potentially habitable worlds of red dwarfs are actually habitable or if they are radiation-soaked airless rocks, there are some dodges SF authors can use to handwave habitability3.
The first and easiest is to simply ignore issues like spin-orbit resonance and flares and assume habitable worlds of red dwarfs are pretty much like Earth, except that the light is a bit redder4. Example: In Rogue Queen, de Camp’s Ormazd orbits the dim star Lalande 21185, but it seems to be astonishingly Earthlike. That’s the boring solution.
The second easiest solution is to accept that there’s a vast distance between “habitable in the sense that some form of life-as-we-know it could survive there” and “habitable in the sense that humans could survive there without sophisticated technology.” Life does persist in places where humans would quickly perish, after all. So one can embrace the implications of 1:1 spin-orbit resonance and the occasional flare, and see what story ideas fall out of it. Stephen Baxter’s Proxima, for example, is set on a hypothetical planet of Proxima Centauri, one not especially friendly to humans. (To add to the misery, the world is being settled in a manner seemingly calculated to maximize human unhappiness—as one might expect from a Baxter novel.)
The third solution is to imagine a way in which tidelocking has been avoided or mitigated. Larry Niven’s Draco’s Tavern series, for example, features a race of aliens who evolved on a double planet orbiting a red dwarf; the two worlds are tide-locked to each other and not their star. The series notes that such double worlds are not exactly common, but…in a galaxy of 400 billion stars, even a small fraction is a large absolute number. The Chirps have found lots of worlds like their home. Now, given the existence of the Earth-Moon and Pluto-Charon systems, double worlds might seem like a reasonable hand-wave. On the minus side, the forces operating on such worlds may destabilize the double planets in a geologically short time. Note that no world closer to the Sun than the Earth has a natural satellite….
There are other solutions.
The Mercury Solution: spin-orbit resonance doesn’t have to be 1:1. Mercury, for example, is close enough to the Sun to be tidelocked, but for various reasons, it spins three times for every two passages around the Sun. This means it does not have a permanent day side and permanent night side. It also means that Mercury’s Solar day (the interval between noon to noon) is about twice as long as its year. Oddly, although Mercury’s 3:2 spin-orbit resonance was discovered half a century ago, I cannot think of many SF authors who were inspired to imagine worlds with 3:2 tidal locks elsewhere in the galaxy. The closest example that comes to mind is the homeworld of the alien Betans in Poul Anderson’s Avatar. Beta orbits its K3 star in about 3000 hours and is tide-locked into a 2000 hour day. Not exactly what I had in mind. If you know of a better example, please provide it in comments!
Venus offers still another solution. Venus revolves around the Sun in about 225 days. It revolves around its axis once every 243 days. It revolves backwards, because apparently Venus is the Ginger Rogers5 of the Solar System. Why there is that slight mismatch is an interesting question. The important thing is that there is one: If Venus’ spin and orbit can be slightly out of phase, so could the spin and orbit of a world orbiting Ross 128. At least until the astronomers show us otherwise. Again, Poul Anderson provides an example: in “The Three-Cornered Wheel” the planet Ivanhoe orbits a red sun, but has a day sixty hours long.
Buy the Book
Static Ruin
Note that when rotation and revolution almost but don’t quite match, solar days can be counterintuitively long. If, say, a hypothetical world orbited Proxima in 16 hours and rotated on its axis in 15 hours, 50 minutes, it would take about 1485 hours (over 60 Earth days) for Proxima to return to the same point in its world’s sky. Assuming I did not mess up the math. On the plus side, that gives inhabitants more time to get out of the ocean’s way (greater tidal forces, Bay of Fundy tides).
Again, I cannot think of a novel featuring a world with a long day, orbiting a red dwarf, but Dave Duncan’s West of January features a world, Vernier, where a near match between revolution and rotation has given it a day two Earth centuries long.
If one is a pessimist and assumes that naturally occurring habitable worlds around red dwarfs are vanishingly rare, there’s still hope. The key word there is “naturally.” What is a dead world—tidelocked to its star and scoured clean of air and water by flares—but a supreme challenge for your dedicated terraformer? Begin building shades in orbit, import the volatiles that almost certainly exist in the system6, put some hardy lifeforms to work and voila! In just ten thousand years you might have an anoxic Precambrian world!
It is a small investment of time, given that planets can be habitable for billions of years. Pity humans don’t think in those scales.
1: Wait, do I need to explain this bit? The Morgan-Keenan system rates stars from hottest to coolest thusly: O, B, A, F, G, K and M. This is easily remembered with the mnemonic obafgkm, which (as I explained to my fellow Scrabble players) is a resinous wine made from the flesh of certain cacti found in the Yukon. Or it will be, once I introduce cacti to the Yukon and convince people to start making a resinous wine from it.
I won’t get into luminosity classes except to say: if your home planet is orbiting anything that isn’t a class V main sequence star, you’re either in command of some impressive technology or very, very screwed.
2: More or less. The effects of other bodies in the system can make worlds wobble a bit, which is why pre-space-age Earth-bound observers could map more than half of the surface of the Moon.
3: One feature every habitable world will have (so obvious that authors need not mention it) is a powerful magnetic field. That should provide some protection against the charged particles in flares, although it won’t help with the x-rays.
4: Human eyes wouldn’t notice the spectrum shift, but a great many SF authors are convinced that it would be like living under a red lightbulb.
5: Ginger Rogers did everything Fred Astaire did, except backwards and in high heels.
6: It turns out water, which is made from the most common element in the universe and the third most common element in the universe, is itself pretty common, contrary to what certain television franchises would have us believe.
In the words of Wikipedia editor TexasAndroid, prolific book reviewer and perennial Darwin Award nominee James Davis Nicoll is of “questionable notability.” His work has appeared in Publishers Weekly and Romantic Times as well as on his own websites, James Nicoll Reviews and Young People Read Old SFF (where he is assisted by editor Karen Lofstrom and web person Adrienne L. Travis). He is surprisingly flammable.
Another dodge I thought of after sending the final draft of this off: planet orbits red dwarf outside the red dwarf’s Goldilocks zone, red dwarf + planet orbit K or G star within the K or G star’s habitable zone. Planet’s ecosystem is mostly powered by the brighter star.
Interesting article, thank you.
I wrote a novella (“The Warrior Within”, published by Tor.com) that takes place on a planet tidally-locked to an M or K class star. I rather wish I’d read your article before I designed my world. I got — from my own researches — that atmospheric mixing could even out the temperature extremes and possibly make a world marginally habitable. I missed the fact that solar flares would likely strip off the atmosphere. Oops. Still, my planet is extensively geo-engineered, so I may be able to handwave my way out of that difficulty.
I wonder how much protection a strong magnetosphere would offer to such a planet?
I don’t remember seeing papers on flare-driven atmosphere loss for terrestrial worlds of red dwarfs until about the time Proxima b was noticed.
Even if the atmosphere survives, red dwarfs’ habit of huge flares might make visiting such worlds a bit problematic, as crews of orbit to orbit ships often prefer not to be baked like a potato by a flood of x-rays. That could be useful for authors who want to have a world isolated in some way. Sure, it’s the only planet with (valuable biological product here) but getting there and back requires military grade shield or luck. Or military grade shielding _and_ luck. Be a good refuge world for people who don’t want visitors.
Come to think of it, there’s another potential atmosphere-killer: thanks to that whole mass-luminosity deal, potentially habitable worlds orbit their stars at higher velocities than the Earth does the sun. I make the orbital velocity of Proxima b more than 100 km/s. Earth’s is more like 30 km/s. Impacts happen at much higher speeds than on Earth. Kinetic energy per unit mass scales with the square of the velocity….
Alternatively, the planet could be tidally locked, with the sunward side ruled by technology and the dark side ruled by magic, with a narrow, perhaps shadowy, band of twilight around the middle.
We, in some strange power’s employ, ferment a resinous wine.
We don’t know enough to assume that flare activity would inevitably strip a planet’s atmosphere in a red dwarf habitable zone. Flares are most common while a red dwarf is relatively young, and once they subside, a planet’s volatiles could be replenished by comet impacts or outgassing. There are a lot of uncertainties.
Besides, even if habitable planets are less common around red dwarfs, red dwarfs are far more common than any other type of star, so that improves the odds.
‘World-Building’ by Stephen Gillett (ed. Ben Bova) is a non-fiction book that covers flare-driven atmospheric loss, tidal locks (and resonant locks), and a bunch of other physical characteristics to guide the reader to understand how to create a spot where critters that reach for the stars could evolve, or at least live. M-class stars are addressed in some detail. The science is a bit dated but not that hard to update.
So that’s why Syfy’s Krypton is tide-locked (and why those atmospheric domes are so important)!
@7/Tracy: I have World-Building, but there’s been quite a lot of research on red dwarf planets and habitability since the book came out, a lot of which disproves long-held assumptions (like the above-mentioned work showing that a moderate atmosphere would keep the dark side from completely freezing). The Centauri Dreams blog is a good place to catch up on the research, if you search the archives:
https://www.centauri-dreams.org/
Wasn’t Krypton’s star a red giant? (googles) Oh, DC has gone back and forth on that. Well, red anything will have issues of some kind or another, although I’d rather deal with tide locking and flares than impending planetary nebulas or supernova close enough the neutrino dose alone is lethal.
So I got about a third of the way through before I clicked we weren’t talking about Lister, Rimmer, the Cat et al, who are technically millions of years old.
I must admit, I was quite looking forward to seeing how SF would solve the problem of them all getting too old for their parts.
I remember a story in Analog, sometime in the past 10-15 years, that had a planet in what was probably a resonant lock. Though the exact ratio did not make enough impression on me for me to remember. Anyhow, the planet was populated by two permanently separated groups of people, one on the “sunrise” twilight side and one on the “sunset” side, each following the twilight at the rate of a few miles a day (I think). And following a plant that circled the globe. The solar “day” length was a few hundred Earth days, as I recall. Or maybe thousand.
Good article, and much food for thought!
@6/CLB: But the length of the ‘relatively young’ phase is, well, relative, so in absolute terms the flare-y phase of a M dwarf is longer than that of a more Solar-like star, and far harsher on the planets involved since the habitable zone is so much closer in. Plenty of opportunities for cometary replenishment work to be undone by the next round of flares…
Yet that does present some interesting storytelling possibilities. A world where life—and perhaps one or more civilizations—has arisen and been wiped out multiple times? A civilization on such a world that has learned about the fates of its predecessors, and now must figure out what to do as the result of observations indicating the cycle is about to start again?
@14/Ian: As I said, though, red dwarfs are considerably more common than other types of stars, so even if it’s rare for the conditions to align just right for habitability, it’s still going to happen a fair number of times.
Anyway, I think I saw a paper not long ago suggesting that a planetary atmosphere might be more robust against flares than has been assumed. I don’t remember the specifics, though. Again, there’s a lot we don’t know, and a lot of our long-held assumptions about red dwarf planets have been challenged or debunked by recent theoretical work.
@15: Agreed, in general. I’m objecting to your claims of ignorance when what we are confident about provides plenty of material upon which to build stories. For example, flares are localized, they come in a range of energies, and their occurrence could be influenced by tides or hot spots raised by planets in the system; those three give plenty of opportunity to envision combinations of orbital resonances and dumb luck that allow for planets that have survived billions of years without encountering an atmosphere-destroying flare. Now, as a recent paper suggested, targeting M dwarfs is not currently an efficient use of limited SETI telescope time, but that doesn’t mean that there is an urgent need to jump away from current science in order to invent some measure of plausibility for the purposes of fiction.
Besides, I’m guessing most stories involving such planets will probably involve humans, and thus probably some sort of FTL drive, so the bar for ‘unlikely event’ will have already been set quite low. :-)
I’ve forgotten – Piper’s Four-Day Planet is four days to the year. What kind of star was it around?
@@@@@`17,
According to Project Gutenberg:
Fenris is the second planet of a G4 star, six hundred and fifty light-years to the Galactic southwest of the Sol System. Everything else equal, it should have been pretty much Terra type; closer to a cooler primary and getting about the same amount of radiation. At least, that’s what the book says. I was born on Fenris, and have never been off it in the seventeen years since.
Everything else, however, is not equal. The Fenris year is a trifle shorter than the Terran year we use for Atomic Era dating, eight thousand and a few odd Galactic Standard hours. In that time, Fenris makes almost exactly four axial rotations. This means that on one side the sun is continuously in the sky for a thousand hours, pouring down unceasing heat, while the other side is in shadow. You sleep eight hours, and when you get up and go outside—in an insulated vehicle, or an extreme-environment suit—you find that the shadows have moved only an inch or so, and it’s that much hotter. Finally, the sun crawls down to the horizon and hangs there for a few days—periods of twenty-four G.S. hours—and then slides slowly out of sight. Then, for about a hundred hours, there is a beautiful unfading sunset, and it’s really pleasant outdoors. Then it gets darker and colder until, just before sunrise, it gets almost cold enough to freeze CO2. Then the sun comes up, and we begin all over again.
You are picking up the impression, I trust, that as planets go, Fenris is nobody’s bargain. It isn’t a real hell-planet, and spacemen haven’t made a swear word out of its name, as they have with the name of fluorine-atmosphere Nifflheim, but even the Reverend Hiram Zilker, the Orthodox-Monophysite preacher, admits that it’s one of those planets the Creator must have gotten a trifle absent-minded with.
Did I overlook consideration of living underground, in which case several factors won’t matter – including length of the day, I suppose. Or under the sea. Tricky for humans, yes. Aquaman manages…
@@@@@9/ChristopherLBennett Thanks for the link.
There are quite a few papers about the topic of the habitability of both tidally-locked planets and those orbiting red dwarfs; one place to look is arxiv.org, for example https://arxiv.org/vc/astro-ph/papers/0609/0609799v1.pdf, https://arxiv.org/ftp/arxiv/papers/1006/1006.0022.pdf, and https://arxiv.org/ftp/arxiv/papers/1702/1702.06936.pdf, the last for the TRAPPIST-1 system.
Dan Blum@5: Well played, sir!
> mnemonic obafgkm
James, James.
For ordinary mortals, this is “Oh be a fine girl kiss me”
http://www.star.ucl.ac.uk/~pac/obafgkmrns.html
Mnemonics for the
Harvard Spectral Classification Scheme
–created by Annie Jump Cannon!
For those who are curious as to the origins of the admittedly bizarre sequence OBAFGKM: the original classification of stellar spectra was done by the Harvard group (mostly women, including Annie Jump Cannon, Antonia Maury, and Williamina Fleming) based on the strength of absorption lines of hydrogen. Stars with the strongest hydrogen lines were classified as “A”, the next strongest were “B”, etc. But it was realized that this didn’t work for *other* absorption lines, which seemed to jump around discontinuously from one stage to the next.
Annie Jump Cannon revised this system into something more logical, partly by reclassifying some stars (which is how we lost C, D, E, H, etc.), and mainly by re-ordering them into a sequence where all the spectral lines changed strength continuously: OBAFGKMRNS. (In this sequence, the hydrogen-line strength starts weak, then gets stronger, then gets weaker again as you move on past A.) Other astronomers and physicists then figured out that this sequence corresponded to one of decreasing temperature.
(The original version of the mnemonic was apparently “Oh, Be A Fine Girl, Kiss Me Right Now, Smack”; in the late 1980s, when I learned it, it had evolved into “Oh, Be A Fine Girl/Guy, Kiss Me [Right Now, Sweetie]”.)
Later on the R, N, and S stars were mostly re-classified as M, though they still show up sometimes, and some brown dwarfs are classified as “T”.
Good timing on this:
https://www.baen.com/small-stars
https://www.tor.com/2017/11/28/did-we-all-write-a-book-about-space-elevators-why-unfortunate-coincidences-happen-in-science-fiction/
Now, they also do a piece on 40 Eridani or the Problem of Proxima, that would be weird.
I think there is a typo in footnote 1, last paragraph. Class V Main stars were not ever mentioned.
@28: The footnote is slightly ambiguous but not incorrect: luminosity Class V and the Main Sequence are effectively interchangeable, both terms identifying those stars that are in the primary hydrogen-burning phase of their lifetimes. M dwarfs pretty much remain at class V their entire lives, so a habitable planet that somehow managed to form around one could host a very long-lived civilization.
If you actually watch their films, Fred was dancing backwards about as much as she was. And he was co-creator of the choreography in most cases.
One of Poul Anderson’s David Falkayn stories takes place on a tidally locked planet over a red dwarf. The sun-side is very dry and heavily desert, with the zone near the terminator much lusher and with a day-night cycle due to libration of the planet. It’s also an important plot point.
The discussion about planetary atmospheres being stripped by solar flares and/or solar wind has reminded me of the remarkable difference between Mars and Venus. Both worlds are unprotected by magnetic fields and Venus is much closer to the sun and thus subject to much stronger charged particle bombardment and yet Venus’s atmosphere is over TEN THOUSAND times denser than Mars’. I have not seen any explanation for this in popularized science articles or TV shows. The best theories that I have been able to come up with involve the greater mass of Venus. First that the larger core of Venus retained it’s spin longer and only lost it’s magnetic field recently. Second that the greater gravity of Venus was able to retain oxygen molecules from water vapor disassociated by solar ultraviolet even though the hydrogen was lost to space. The free oxygen combined with anything available adding to the CO2 already present helping to thicken the air of Venus. Can anyone tell me if either of my guesses is correct?
@31/DoctorZarkov: As I recall my college astronomy, the key difference between Venus and Mars is that Venus is highly volcanically active, so that gases continuously erupt from the interior of the planet and keep the atmosphere very dense despite loss to space, while Mars is volcanically dead, so that its atmosphere losses to space are not replenished. Earth is in the sweet spot where we have enough volcanic activity to replenish the atmosphere but not so much that we get a runaway greenhouse effect.
Humanity: “Hold my beer!”
DoctorZarkov @@@@@ 31:
My impression is that this is something which really isn’t very well understood. How much of this is due to the initial formation (maybe Mars started off with fewer volatiles and thus less of an atmosphere), early Solar System evolution (effects of massive impacts on young planets; fainter Sun but more UV radiation and stronger solar wind), evolution in the first few hundred million years versus gradual processes operating over billions of years…
Certainly the difference in mass has to play an important role: it’s simply easier for ions, atoms, and molecules to escape from Mars (escape velocity = 5.0 km/s) than from either Earth or Venus (escape velocities = 11.2 and 10.4 km/s, respectively).
One additional factor may be that Venus’s atmosphere resists erosion due to its very thickness, which enables it to it to have a more robust ionosphere and thus better shielding of the lower atmosphere from the solar wind. It turns out that the interaction of the solar wind and the ionosphere creates localized magnetic shielding (an “induced magnetosphere”) and diverts the solar wind around a planet, even if the planet doesn’t have its own, internally generated magnetic field like Earth does. Venus being closer to the Sun might, ironically, help: Venus experiences a higher UV flux, which ionizes more of the upper atmosphere, and thus Venus ends up with more of an ionosphere than it would were it at the distance of Mars.
There has also been speculation that Mars might have lost significant amounts of atmosphere early on due to massive impacts early in the Solar System’s history. Really massive impacts would have blown some of Mars’s atmosphere into space. Such impacts would have happened to Earth and Venus, too, but their higher surface gravities would have meant less atmospheric loss from each impact.
ChristopherLBennett @@@@@ 32:
I don’t think ongoing volcanism is very relevant, for several reasons:
1. I’m not sure how well volcanism on Earth would replenish Earth’s current atmosphere. Volcanic outgassing on Earth is about 60% water vapor and about 10-40% CO2, with almost no nitrogen. But our atmosphere is 78% nitrogen and 21% oxygen. (Over geological time scales, volcanism replenishes the CO2 content, yes. But probably not the rest of the atmosphere.)
2. It’s not at all clear that Venus has volcanism on the same scale as Earth, particularly over long geological periods. Some volcanism, yes — but 10 or 100 times that of the Earth?
3. It’s also not clear how you’d replenish Venus’s atmosphere with volcanism, because where would the CO2 come from? On Earth, CO2 in the atmosphere is drawn down and deposited on the ocean floor in the form of carbonate rocks, via water-based weathering processes (and organic production via photosynthesis). When the ocean floor goes into the mantle at subduction zones, the heat and pressure produce CO2 from the carbonate rocks, which can then accompany magma that moves up and erupts in volcanoes. But Venus doesn’t have a water-based CO2 weathering cycle (no water!), and it doesn’t have plate tectonics, either. So there’s probably no way of getting CO2 out of the atmosphere and into the mantle, and thus maybe not a lot of CO2 coming out of Venus’s volcanoes.
I remember reading that, counter-intuitively, Earth-like planets around M and K stars will have greater surface UV fluxes than similar planets around F or G stars, as the optical thickness of the planet’s ozone layer increases faster than the star’s UV output.
I’ll try to hunt up the reference and add it in a follow-on post.
I think the jury is still out regarding the origin and maintenance of the thick atmosphere on Venus, but recent data from the MAVEN spacecraft seems to suggest how Mars lost its atmosphere. Because of its small mass, it became geologically dead—and thus lost any internal process to generate a protective magnetic field—more than 3 billion years ago when the Sun was more active. Combined with its position on the outer edge of the habitable zone, Mars thus entered a feedback loop that caused volatiles to freeze out or sputter away into interplanetary space. There is a good writeup in the July 2018 issue of Sky & Telescope (I cannot find a non-paywalled link, unfortunately).
@36/swampyankee: While the amount of ozone will certainly affect atmospheric conditions for individual planets, the effect you describe has nothing to do with ozone but rather with the way the shape of a blackbody spectrum changes with temperature. As the surface temperature of a star drops, not only does the peak of the spectrum shift to redder wavelengths but the power distribution also flattens over the range from near-UV to mid-infrared; hence, for any given total flux over that wavelength range (which is really the primary factor used to identify the habitable zone), the relative contribution of UV will be higher for cooler stars.
Ian @@@@@ 37:
the way the shape of a blackbody spectrum changes with temperature…. the relative contribution of UV will be higher for cooler stars.
No, that’s the opposite of how blackbodies work. Cooler temperatures mean less contribution (in both absolute and relative senses) from shorter wavelengths (such as UV).
The reason M stars can have relatively higher UV fluxes is due to stars not being perfect blackbodies. Instead, you get flux — including emission lines — coming from the active chromosphere, in addition to the mostly blackbody flux from the coolor photosphere. (The Sun has a chromosphere, too, but it’s not as dominant as the chromospheres of M stars.)
Thanks for all the feedback. While I have your attention perhaps you all can help with a question from an old science fiction story. In Poul Anderson’s novelette A Sun Invisible the plot hinges on the idea that “everybody” knows that blue giant stars don’t have planetary systems. In our solar system most of the angular momentum resides in the planets rather than the sun. Assuming that this is true for all systems, can the spin of stars be measured and plotted against their mass thereby determining if a family of planets has syphoned away a sun’s angular momentum? In all the reports of exo-planet discoveries I don’t recall any involving blue giant stars. Perhaps it is just my faulty memory but why would astronomers, searching for livable planets, bother to glance at O,B or A stars. These giants are too short lived to allow time for the evolution of life forms.
@38/PeterErwin: Having one degree in physics and another in astronomy, I am fully aware of how blackbodies work and how stellar atmospheres deviate from them. :-) And while the deviations you mention are certainly true—absorption and spots in a star’s atmosphere would tend to block energy at visible and IR wavelengths without affecting UV and X-ray flux—they are second-order effects.
I think the problem is confusion regarding relative flux. Relative to what? Now, certainly, cooler blackbodies emit less UV or X-ray flux as a fraction of their own total flux than warmer blackbodies. But the flux intensity received at some other location is not solely determined by the object’s temperature: a G dwarf may emit more total UV than an M dwarf, but that is primarily a function of their relative masses (and thus luminosities), not their temperatures; M-supergiant Betelgeuse emits a lot more UV than the Sun despite having a much cooler surface temperature.
A key thing to remember is that the total flux energy over the biologically-critical set of wavelengths—roughly 300-2500 nm—needs to be approximately the same as that found here on Earth in order for a planet to be considered habitable. Yet the shape of the blackbody curve is not only asymmetric but varies with temperature; hence, given two blackbody curves of different temperatures normalized to the same total area in the specified range of wavelengths, the cooler cureve will actually be above the warmer one at the UV end (with the crossover point being somewhere in the visible range given the temperatures of F/G/K/M stars). That is the sense in which K and M stars emit ‘relatively’ more flux in the UV than F and G stars: for a given total energy budget over the UV-to-IR range, the cooler stars put a bit more into the UV end. K/M stars may produce less UV, but a habitable planet needs to orbit more closely and the inverse-square law (over)compensates.
FWIW, I believe UV-flux variations for G vs. M stars are orders of magnitude less important than differences in flaring activity, but some extra sunscreen when visiting beaches around M dwarfs is probably advisable.
@39/DoctorZarkof: Planet searches around OBA stars are mainly avoided for observational reasons: such stars are too bright to make planets directly visible via reflected light, and they are too massive for ‘wobbles’ due to planetary orbits to be visible. Only a direct transit from an edge-on orbit is likey to be observable.
As for rotations, the speed is known for some stars from analysis of absorption-line widths. But we still don’t know enough about the details of either stellar or planet formation to translate such values into meaningful estimates of how much angular momentum may have beeen ‘stolen’ by a planetary system.
Ian, I’ll admit I’m a bit puzzled as to what point you’re trying to make. Some of what you say is correct (e.g., “certainly, cooler blackbodies emit less UV or X-ray flux as a fraction of their own total flux than warmer blackbodies”) and then you go on to say things that directly contradict that.
I think the problem is confusion regarding relative flux. Relative to what?
You made it fairly clear in your previous comment: UV flux relative to the total flux from the star a planet was orbiting: “… for any given total flux over that wavelength range (which is really the primary factor used to identify the habitable zone), the relative contribution of UV will be higher for cooler stars.”
And that statement is wrong.
… the total flux energy over the biologically-critical set of wavelengths—roughly 300-2500 nm—needs to be approximately the same as that found here on Earth in order for a planet to be considered habitable.
The standard (astronomical) definition of “habitable” is “permits liquid water on the surface”, and this depends on the total energy received by the planet from the star, not just the 300-2500 nm window (and the planet’s albedo, and details of its atmosphere and greenhouse effect, and so on). One can certainly argue that this is insufficient (and some people are suggesting that “temperate” be used in place of “habitable”), but your definition is not a standard one.
But, OK, let’s use your 300-2500 nm range as the “optical/near-IR”, and 100-300 nm as the UV. If I integrate the blackbody function from 300 to 2500 nm and call that the “optical-near-IR flux”, and then integrate the blackbody function from 100 to 300 nm and call that the “UV flux”, then this is what you get for different blackbody temperatures (calculations done in Python, using the astropy.modeling.blackbody functions and the scipy.integrate.quad function)
T = 10,000 K: UV/optical-near-IR = 0.49; UV/UV+optical-near-IR = 0.33
T = 5800 K (e.g., Sun): UV/optical-near-IR = 0.035; UV/UV+optical-near-IR = 0.034
T = 3000 K (e.g., M5 star): UV/optical-near-IR = 0.00010; UV/UV+optical-near-IR = same
T = 2500 K (e.g., M8 star): UV/optical-near-IR = 7.8e-6; UV/UV+optical-near-IR = same
The relative contribution of UV flux *decreases* dramatically as you go to cooler temperatures.
@42: While the presence of liquid water is currently the sine qua non of habitability, it is merely a minimum starting point. The point I was trying to make is that the total energy flux isn’t the only factor that will affect habitability (if more broadly defined), the spectral distribution will have effects too (e.g. total flux enough to melt water may not be sufficient if the spectral distribution does not contain enough blue flux to activate some critical biosphere processes). But hold that thought for a moment…
Yep, I concede that I misremembered the detailed behavior of of the blackbody curve at the short-wavelength end (the slope on the UV side actually changes less quickly with temperature than I recall). After I made some more detailed calculations, scaling cooler spectra upward so that they matched the Sun’s flux at 300 nm (scale factors ranging from approximately 36 for a K5 star to more than 10^5 for M8), I see to my dismay that the cooler stars’ integrated UV/X-ray/gamma-ray spectra in these scenarios end up falling just a bit short of matching that generated by the Sun. Yet the ratios end up more in the factors-of-2 rather than orders-of-magnitude range, meaning that adding additional absorption in the red/IR range can likely be sufficient to skew the UV/visible+IR ratios towards values higher than that of the Sun. However, I don’t believe that such absorption effects would be sufficient on their own, rather it is their interaction with the rescaling effects that can result in some M stars’ habitable zones receiving slightly enhanced UV relative to what is typical on Earth.
…now, the other thing that occurred to me while thinking about spectral energy ratios is that it is almost certainly not a coincidence that the Sun’s spectrum peaks right in the middle of what we consider the ‘visible’ spectrum (and also midway between the near-UV and near-IR regions utilized by some plants and animals). The OP correctly stated that the human eye-brain would automatically compensate for a spectral distribution peaking more towards the red, but what other effects might that spectral shift cause with regards to other aspects of biology? Would human immigrants on a world circling an M dwarf need a supply chain for vitamin D supplements? Could a person become disoriented when looking at a landscape in cool-hued light while simultaneously being warmed by the additional IR emission? What crops might thrive or fail in that type of light? Would beings who evolved on such a planet have more heat-tolerant skin, perhaps making them excellent smiths, firefighters, or dragon tamers? Would such beings shy away from blue LED lights to avoid sunburn? Possibilities abound for authors who want to explore the interactions between light frequencies and biochemistry!
Ian @@@@@ 43:
… scaling cooler spectra upward so that they matched the Sun’s flux at 300 nm
But why are you doing that? What physical justification could there be? (It’s almost as if you are implicitly assuming that only 300 nm photons actually cause heating of planets…)
All other things being equal, a planet orbiting a blackbody with an M5-star temperature and receiving enough flux to be as habitable as the Earth (in the liquid-water sense) will be receiving the same total flux (ergs or Joules per square meter) as the Earth does. That’s what radiative energy balance requires. But this flux will be distributed differently than is true for the Earth: it will be skewed redder. As I showed, a T=2500 K blackbody produces essentially no UV flux. It’s simply impossible to “skew the UV/visible+IR ratios towards values higher than that of the Sun”.
In practice, once you consider non-blackbody effects, M stars often do often have more UV flux than you would expect from the equivalent blackbody; this is coming from the chromosphere and the corona. (I’m not sure why you apparently think this is associated with “absorption”; the chromosphere and corona are optically thin.) But this is by factors of a few, so that the UV/optical+IR ratio goes from something like 0.0001 to 0.0002 or 0.001, which is still nowhere near the ratio for a T ~ 6000 K blackbody (like the Sun). (See, for example, this paper by Stelzer et al., especially their Fig.7, which estimates for nearby M stars what fraction of the observed GALEX NUV [near-UV] flux comes from sources other than the photosphere.)
@44/Peter: “But why are you doing that? What physical justification could there be?”
I can’t speak for Ian, but don’t you answer this question as soon as you say “All other things being equal?” As I read it, Ian’s point was not to generalize to a universal situation, but to talk specifically about the spectral distribution of a red star relative to the Sun. And the best way to show that is to set the star’s flux equal to the Sun’s for the purposes of the comparison, so that it’s clearer how their proportions differ from each other.
ChristopherLBennett @@@@@ 45:
No, my remark about “all other things being equal” was meant to apply to planetary properties, such as albedo, greenhouse effect, and so forth, which might affect how close to a star you can position a planet so as to have liquid water (potentially) on its surface. This has nothing to do with what the distribution of light coming from the star is.
For the purpose of talking about the spectral distributions of stellar light, setting a star’s flux equal to the Sun’s is completely unnecessary. If 10% of a blackbody’s emission is in the UV, then scaling the total luminosity up or down won’t change that at all. (Imagine you want to talk about how large someone’s head is relative to their body — e.g., the length of their head is X% of the length of their body. Do you multiply their total length by some arbitrary constant first? Of course not — that has no effect on the relative sizes of head and body.)
@45/CLB: You spoke for me quite well there, thanks! :-) But I can offer a couple additional justifications for my reasoning.
A significant fraction of the Earth’s current temperature arises from visible light that is absorbed by water/land/vegetation, re-emitted in the IR, and trapped by water vapor, carbon dioxide, and other greenhouse gases. Those same gases will also block IR from the Sun from reaching the ground. So, shifting the spectral peak while maintaining constant bolometric luminosity might result in some cooling. Hence, if the goal is to maintain Earthlike conditions for comparison purposes, to first approximation the insolation should be scaled to keep the visible-light flux approximately constant.
But perhaps more importantly, the topic of this discussion is not really astrophysics but sci-fi, with the particular goal of finding ways to continue using planets around K & M stars as settings. Since past stories historically involved people in Earth-like conditions, and presumably future stories will continue to do so, the goal is thus to identify the conditions under which the inhabitants of such a planet could experience insolation matching as closely as possible that of the Sun on the Earth. In that context, any physically plausible assumptions that can help minimize the gap between the baseline/typical properties of a star as measured at a given planet and those of the Sun as measured at Earth are not merely justifiable but arguably sort of the point of the exercise.
Ian @@@@@ 47:
Those same gases will also block IR from the Sun from reaching the ground. So, shifting the spectral peak while maintaining constant bolometric luminosity might result in some cooling.
When the gases “block IR from the Sun”, they generally do so by absorbing the IR and thus absorbing the energy. This helps heat the surface (because heating the air means more IR radiation from the air, some of which goes downward, and also due to heat conduction from the lowest layers to the ground). In effect, if the atmosphere absorbs IR strongly, then the albedo for the planet is lower for those wavelengths.
Shifting the spectral distributinon of incoming flux, while keeping the total constant, could have interesting effects — but to figure them out, I think you’d really have to look into the details of how the albedo varies as a function of wavelength, which will depend on the composition of the atmosphere (including clouds) and the composition of the surface. For example, the albedo of snow and ice tends to be highest in the blue and gets lower to longer wavelengths, and is quite low in the near-IR. So snow and ice reflect blue light more efficiently than red or IR light. Shifting the dominant fraction of incoming radiation to longer wavelengths might thus mean that the effective albedo of a planet with significant amounts of snow or ice goes down, meaning a larger fraction of the radiation is absorbed, resulting in heating.
@48: Yes, I’m aware of the relevant physics involved. But you are illustrating the very problem I am trying to avoid: bogging down the model with complexities dependent upon poorly constrained details that may not even be relevant to the topic at hand. The particular surface and atmospheric compositions of a planet may cause it to be a net absorber or reflector of IR, but a randomly chosen planet could fall anywhere within that range. We do know that visible-spectrum light is an important energy source for Earth, so it’s best to keep the flux in that range constant when developing a baseline for comparison; only afterwards is it appropriate to start adding in more details, especially if they involve factors that may be specific to a particular planet.
We’re talking about toy models intended to clarify thinking and sketch out the range of possibilities, not provide a detailed list of viable options from which authors need to choose.
@48,
of course, the “other properties” includes surface pressure. I happen to think “liquid water” is a poor measure of habitability, as it covers quite a broad temperature range, up to about 647 K, at sufficient pressure.
Kepler-13b is a superjovian around an A0 star, so it’s not out of the question for stars as early as early A at least.
(“as early as early” is not a typo, i just don’t know how to recast the sentence)
Jovian exoplanet KELT-9b’s star may be a late B type.
@50/swampyankee: Well, that depends on whether you’re defining “habitability” in the human-centric sense of “Could I build a house there?” or in the more scientifically objective sense of “Could any form of biology exist there?” Liquid water is important to exobiology because it’s the best (and only confirmed) solvent for supporting self-replicating organic molecules. There are other hypothetical solvents that might support more exotic forms of life, like ammonia or methane, but those forms of life would require rarer elements and their solvents might not support biology as well as water does (e.g. life in lakes is less likely to thrive in a solvent that sinks when it freezes instead of floating on top of the lake). So a place that has liquid water at any temperature and pressure is therefore a candidate for supporting life, even if it’s far too hostile an environment for humans.
@CLB: Perhaps exploring the meaning of ‘habitability’ is in fact the best way to ‘rescue’ non-G stellar types for SF purposes. Human-friendly conditions probably require Earthlike planets around Sunlike stars, while cooler stars will perhaps favor different types of biology. Truly some speculative fiction required!
@Del: The best part about habitable planets around A or B stars, compared to the long-lived cooler types, is that you probably won’t need to negotiate with any pesky locals for land rights, nor worry about any invisible threats to your advance scouts driving in their fighting machines. But such locales would only be short-term rentals, and even PeterErwin can’t dispute that a steady supply of SPF eleventy-billion sunscreen is probably a must in such systems. ;-)