Architect of Worlds – Some Initial Results
Over the past couple of weeks, I’ve been applying the star-system and planetary-system design sequences from the Architect of Worlds draft to generate planetary systems for nearby stars. In a number of cases, this has involved tweaking the parameters of the model to fit strong exoplanet candidates that we already know are there. In other cases, it also involved tweaking the parameters to avoid creating exoplanets that we would reasonably have detected by now, if they were there.
So far, the model has held up surprisingly well. I’ve had to make a few adjustments to make the results more plausible, and to allow for some of the real-world cases. That’s to be expected, but by and large I haven’t had to do any major redesign.
Let me summarize some of the results thus far.
Alpha Centauri
We know of two strong exoplanet candidates for this trinary star system – one close-in planet for each of Alpha Centauri B and Proxima Centauri. I had no difficulty at all fitting either of these candidates to the model. The result was five planets for Alpha Centauri A, nine for Alpha Centauri B, and twelve planets for Proxima. The tight or loose packing of planetary orbits makes a big difference, and of course Proxima has no close companion to block off the outer system for planetary formation. Thus the little red-dwarf companion gets the most extensive family of planets.
The mass of the known candidate for the B-component suggested a lighter protoplanetary disk than usual, but this still yielded two Earth-likes in the habitable zone, generously defined. One of these is the best candidate I’ve generated so far for a habitable world. The mass of Proxima’s known companion suggests a denser than usual protoplanetary disk, so Proxima ended up with a few gas giants up to Saturn size, at fairly wide orbital radii.
Barnard’s Star
Low-mass star, very low metallicity, not much material with which to build massive planets. The system ends up with nine planets, all packed well within one astronomical unit of the star, all but the last of them little Mars-likes. This turns out to be fairly typical of red-dwarf stars, given the assumptions of the model.
Wolf 359
Very small and dim red dwarf, although the metallicity is high and that might help form planets. We end up with ten planets this time, several of them cold super-Earths. The innermost planet is in the habitable zone, but is too small to retain much atmosphere.
Lalande 21185
Another small red dwarf. Interesting in that we have a strong exoplanet candidate here, a super-Earth very close in. The problem is that we don’t seem to see any more heavy planets or super-Jovians further out.
This star caused me to make the first adjustment to the model: I added a rule that permitted a massive, volatiles-heavy “failed core” to appear close to the star in rare cases. This makes sense, given that some of our known exoplanets are both massive and not very dense, suggesting that they’re not rocky “terrestrial” planets, but something rich in water and other light compounds instead. If a gas giant can migrate inward, perhaps a smaller planet can form out past the snow line and barrel in close to the star as well.
Placing the exoplanet candidate as a “failed core” rather than a “terrestrial planet” permitted me to keep the assumed density of the protoplanetary disk to a reasonable value. The rest of the planets turned out to be quite small close in, leading to a few modest-sized gas giants on the outskirts, safely under our current detection level. Ten planets in all, two of them within the generous habitable zone, both of those probably too low-mass to be truly Earth-like.
Luyten 726-8
Two very low-mass red dwarf stars, co-orbiting at a close distance that probably forbids either from having many planets. The model gave me two planets for the A component, one for the B component, all cold “failed cores.”
Sirius
A very hostile star system. The bright A component ended up with nine planets, all packed close in, a mix of gas giants up to sub-Jovian size and a few rocky super-Earths. All of the planets are far too hot for human comfort, the coolest of them running a blackbody temperature over 400 K.
I didn’t bother to generate planets for the white dwarf B component – I need to work out rules for applying the model to white dwarf stars, and in any case there’s no possibility of an Earth-like world in such a planetary system anyway.
Ross 154 and Ross 248
These two red dwarf systems each turned out to be barren, with four and five planets respectively. I’m not going to report on any more red-dwarf systems, as they’re all going to be similarly uninteresting.
In fact, my research tells me that the probability of any red dwarf giving rise to an Earth-like world is going to be very low. Any world that’s warm enough will almost certainly be tide-locked, with all the problems that implies. Not to mention that red dwarf stars put out most of their radiation in the infrared range. That means a world that’s warm enough to live on is likely to get so little visible-light insolation that photosynthesis is going to be problematic.
Epsilon Eridani
This was an interesting case – we have at least one strong exoplanet candidate here, a super-Jovian, with a strong indication of a dense asteroid belt just inside that candidate’s orbit. None of this was a problem for my model. The mass of the known gas giant, together with the known density of the current debris disks, suggested a high-density protoplanetary nebula. I was able to generate the rest of the planetary system to match.
I ended up with nine planets, one of them a super-Earth squarely in the middle of the habitable zone. The star system is quite young, well under a billion years old, so that planet is almost certainly a heavily oceanic “pre-garden” world, lacking complex life or a human-breathable atmosphere. Still, maybe a terraforming candidate? Meanwhile, the asteroid belt is in place and I would be comfortable marking that as a rich resource zone. This looks like a star system that people would come to visit, even if there isn’t an Earth-analogue there.
61 Cygni
No surprises here. I got thirteen planets for the A-component, due to tight packing of planetary orbits, and only seven for the B-component. Nothing so massive as to be detectable from Earth, so there’s no sign of Mesklin, alas. Each star ended up with a planet in the habitable zone, but both were too low-mass to be good Earth-like candidates.
Procyon
As expected, this system ended up like Sirius A, but not quite so extreme. Eight planets, all of them rocky worlds, no “failed cores” or gas giants in the mix. The outermost might almost be cool enough for habitability, but it’s far too small, so it’s more like a baked-dry Mars than anything else.
Epsilon Indi
This star gave me a little trouble, and caused me to tweak the model again slightly. The problem is that we have an exoplanet candidate here, but it’s simultaneously very massive (about 2.5 Jupiter masses) and very distant from the primary.
As written when I got to this point, my model permitted massive gas giants to form, but that would tend to require a dense protoplanetary disk, which would in turn force the gas giant to form close in and migrate closer. A gas giant forming much further out would suggest a lighter disk, which would render the planet less massive. A paradox. I solved the problem by adding another size category for gas giants – in rare cases, even a fairly light disk can now give rise to a super-Jupiter. Which makes sense, as this isn’t the only massive gas giant we’ve detected in the cold outer reaches of a star system.
Final result was nine planets, the innermost of which wouldn’t be a bad Earth-like candidate, except that the system is fairly young. Probably another “pre-garden” world here.
Tau Ceti
We have several strong exoplanet candidates here, all of them super-Earths fairly close in to the star. I computed the most likely disk density and proceeded, adding Mars-sized mini-worlds to fill in two gaps left by the known exoplanets. Final result was nine planets, none of them further out than about 6 AU, which gives us room for this star system’s apparently very wide and rich Kuiper belt.
Two of the known super-Earths are close to the habitable zone, but one of them is probably too hot, and the other one is probably only habitable if it’s running a very aggressive greenhouse effect. Probably interesting places to visit, but not somewhere anyone would want to live.
Summing Up
Twenty-seven stars so far, in twenty star systems, and so far I’ve only generated one strong candidate for Earth-like conditions (Alpha Centauri B-V). I’ve also concluded (or, rather, verified for myself) that red dwarf stars are very unlikely to give us Earth-likes. Depending on one’s assumptions, this may mean that such stars are best ignored when building a star map for fictional purposes.
I’m going to continue with this, probably saving myself a bunch of time by skipping over most or all of the M-class stars. Meanwhile, this is enough for me to start building a new version of my solar-neighborhood map. Stay tuned.
5 thoughts on “Architect of Worlds – Some Initial Results”
“In fact, my research tells me that the probability of any red dwarf giving rise to an Earth-like world is going to be very low. Any world that’s warm enough will almost certainly be tide-locked, with all the problems that implies. Not to mention that red dwarf stars put out most of their radiation in the infrared range. That means a world that’s warm enough to live on is likely to get so little visible-light insolation that photosynthesis is going to be problematic.”
I think your assessment might have been overly pessimistic. While red dwarf stars certainly pose problems for the evolution of complex life, a couple of recent papers argue that these problems do not pose as big a challenge as previously thought.
Simulations of the climate of tidally locked planets suggest that an atmosphere of only 10 % the mass of Earth’s atmosphere transports enough heat to the nightside of the planet to prevent atmospheric collapse, and that tidally locked planets can have liquid surface water under a variety of atmospheric conditions. Wandel and Gale give a good overview of some of the findings in “The Bio-habitable Zone and atmospheric properties for Planets of Red Dwarfs” (2019).
And in “Photosynthesis on habitable planets around low-mass stars” (2019), Lingam and Loeb argue that the buildup of atmospheric oxygen through oxygenic photosynthesis is possible on planets around a red dwarf star with a mass of 0.21 solar masses or bigger, and that these planets might therefore be capable of sustaining complex, multi-cellular life.
In regards to storytelling, I think the idea of complex life evolving on a tidally locked planet around a red dwarf star is fascinating. It would be nice to have that as a possibility in the worldbuilding system, even as a low-probability event.
Interesting. Thanks for those references! It’s been a couple of years since I last looked at that question specifically, but now that I’m actively working on the next piece of the design sequence, I’m going to want to review the current literature.
I think, as you suggest, the big question is going to be whether a red dwarf’s planet can retain a substantial atmosphere, which in turn will depend on whether a tide-locked world can have a strong enough magnetosphere to shield its atmosphere from the primary’s flares and stellar wind. I seem to recall that there’s been some recent work on that point too.
In any case, if I remember correctly that post was at a time when I was doing a quick-and-dirty application of the earlier pieces of the sequence on solar-neighborhood data. Now that I’m building the next steps in the sequence, I’ll want to consider even the rare cases.
Thanks for your interest in the work :-).
“… which in turn will depend on whether a tide-locked world can have a strong enough magnetosphere to shield its atmosphere from the primary’s flares and stellar wind.”
There is a paper by Zuluaga et al. titled “Evolution of magnetic protection in potentially habitable terrestrial planets” (2012) where they argue that even slow-rotating tidally locked planets can produce a substantial magnetosphere, as long as they have an iron core and an Earth-like silicate mantle. The magnetic field is generated by convection in the interior of the planet.
One major problem that remains is that red dwarf stars are extremely active early in their lifespan. Any planet that is close enough to be in the habitable zone is going to be bombarded by flares early on, to the point where an Earth-sized planet would probably lose its entire atmosphere to atmospheric erosion.
However, this doesn’t necessarily spell doom, since:
a) terrestrial planets that are located in the outer system, where the early bombardment isn’t as strong, could migrate inwards and into the habitable zone later
b) a planet in the habitable zone that lost its atmosphere early on could accrete a secondary atmosphere through outgassing later
c) a big “super Earth” planet in the habitable zone has a chance of retaining a substantial part of its atmosphere despite the early bombardment
My personal favourite is, once again, a planet with an iron core and silicate mantle similar to Earth’s. After the red dwarf’s early high-activity phase is over, such a planet could acquire a secondary atmosphere though volcanism, and it would have a magnetic field that protects the atmosphere from erosion during that later phase when the host star is less active.
In terms of worldbuilding rules, maybe this could be modelled through a minimum metallicity of 1 (same as Sol), so that enough iron is available for a terrestrial planet to have a large iron core?
In any case, thank you for your interesting work. I remember how my first thought when I found it was: “This is great! Reminds me of the worldbuilding system from GURPS Space, except that it’s even more complex.” Whereupon I discovered that you are the man who designed the GURPS Space system …
I think I’ve seen results similar to the Zuluaga paper you mention. I think what’s going to end up in the design sequence will depend on a world’s density (higher density indicates a large metallic core), its mass (larger mass means the core will stay fluid much longer), and possibly the degree of tidal interaction (a world that gets “flexed” a lot will tend to keep generating internal heat sufficient to keep the core fluid). I think that should be enough to catch cases like Mercury, and at least some of the common “in a spin-orbit resonance with a red dwarf” cases.
Incidentally, I find I need to apologize – I was just reviewing the status of the old Sharrukin’s Archive site in preparation for dismantling it entirely, and I found a cache of email that apparently hadn’t been forwarded to me properly, including at least one message from you. Glad you were able to get in touch eventually by way of the blog :-).
No worries. I assumed you had read my e-mail, but were simply too busy to reply. Which would be perfectly alright, since we all have real-life obligations.
I’ll be keeping an eye on the “Architect of Worlds” project, and also on “Human Destiny”, which looks like a fabulous RPG setting in the making.