[Space] Climate & habitability of tide-locked planets

[Agemegos]

There has been a long debate in planetary science about whether planets that are in 1:1 spin:orbit resonances with their stars (colloquially, "tide-locked") might have conditions on their surfaces habitable to humans.

Consideration of the temperature of different parts of the surface as though they were roughly in radiative equilibrium gave us a model in which the sunny side would be unbearably hot (and "therefore" dry), the dark side as cold as a moon of Neptune, and the twilight zone swept by perpetual gales of ice-cold, bone-dry winds from the dark side to the lit. It seemed obvious that all the water on the lit side and in the twilight zones would evaporate, pass into darkness in high-level winds, freeze out, and form an ice cap on the dark side, where it would be so cold that the ice would be in effect a rock-forming mineral. A little later it became fashionable to realise that the cold of the dark side would be deep enough that carbon dioxide would sublimate to dry ice, that nitrogen and oxygen would liquefy and even freeze. This was called "atmospheric collapse", and we supposed that tide-locked planets would be effectively airless.

Then in 1997 Joshi, Haberle, and Reynolds¹ used a general circulation model (supercomputer model of the atmosphere) to simulate the atmospheres of synchronously rotating terrestrial planets orbiting M dwarfs. This model included the advection of sensible heat in the global circulation of the winds, and it delivered the bombshell conclusion that even with an atmosphere of only 200 millibars (⅕ of an atmosphere) of CO₂, and even with a dry surface, the transport of heat from the lit side to the dark side would be enough to raise the temperature of the dark side to above the temperature at which CO₂ sublimates to a solid. Since that temperature is well above the boiling point of nitrogen and oxygen it follow that atmospheric collapse is not to be expected in the Goldilocks zone² unless the atmosphere is far too tenuous³ to support respiration anyway.

Joshi et al used models that did not include water vapour, and gave us the idea of tide-locked planets in which warm air flowed from the lit side at altitude, cooled and sank on the night side, forming cold air that blew at ground level to the lit side. We, or at least I, did not consider that though a tide-locked planet does not rotate with respect to the direction of its star, it does actually⁴ rotate, and there are Coriolis effects on its atmospheric circulation. And when layfolk informally added water to this model our conclusions were dominated by an impression of warm air flowing across the terminator at altitude, cooling in the darkness with the effect that its humidity would rain or snow out, sinking, and then flowing back to the lit side as a cold, dry, gale. It seemed that the sunlit side of a tide-locked planet would become desiccated, most of the planet's water accumulating as a super-Antarctic ice-cap on the dark side. The availability of water on the lit side would depend on the return of water into the light by the flow of glaciers into the twilight, and would be best in the twilit band around the terminator. We (or at least I) did not think that the convergence zone at the subsolar point would be rainy for the same reason that the tropical convergence is, or thought that it would be too hot for human habitation⁵.

That was the approximate state of the art in 2006, when Zeigler & Cambias⁶ wrote Space for GURPS 4th edition. Tide-locked worlds in GURPS Space emerge from the determination of rotation rate in Step 30 of the advanced world generation sequence, specifically, on p.117. That is before the habitability and population of the world are determined in steps 32 (p. 121) and 35 (p. 122). But the effects of tidal locking on surface conditions are not calculated until the section on "Special Cases" on p. 125. That means that the habitability rating is calculated on the basis of the average surface temperature, i.e. the surface temperature in the twilight zone near the terminator. i.e in GURPS Space 4th edition the human population of a tide-locked planet is assumed to settle in the twilight zone. Following the state of the art about 2006, the rules for tide-locked worlds on p.125 apply adjustments to the day face and night face temperatures that reflect the transportation of sensible heat in the circulation of the atmosphere and therefore depend on the density of the atmosphere. Very dense atmospheres keep the night face very nearly as warm as the day face and effectually prevent any freezing out of either water or atmosphere; very thin atmospheres freeze out, cause the oceans to freeze out on the dark side, and allow the day face to heat up by 20% (that nearly doubles outgoing thermal radiation) while the night face is cooled to only 10% of equilibrium temperature. Thin to dense atmospheres produce various intermediate cases, with a degree of atmospheric transport of heat producing smaller contrasts of temperatures, reductions of hydrographic percentage by partial freezing-out of water, and less or no reduction of atmospheric pressure by freezing-out.

The result is that the star system and planet generation procedures in GURPS Space 4th edition put the human populations on those tide-locked planets where it is the twilight zone that has an equable temperature, and not on those where the dark face or the subsolar region has an equable temperature.

Since that time there has been further progress in modelling surface conditions on the synchronously-rotating planets of K and M stars. In 2010 Merlis & Schneider⁷ examined models in which the planet was acknowledged to be rotating and in which the surface was modelled as a "slab ocean" (which provides as much water as will evaporate, but that does not transport heat). They found dramatic differences in the patterns of winds and rainfall depending on whether the planet was orbiting and rotating quickly or slowly⁸, but that in either case that the temperature of the dark side was much more uniform and warmer that I had expected. Where Cambias & Zeigler [2006] suggested that the dark side of a tidally-locked Earth should be about 230 K, Merlis & Schneider's models showed the dark side of a tidally-locked Earth to be a huge patch of about 250 K (-23 C), which is not as cold as Antartica at night/winter⁹. That is warm enough that substantial glacial flow can be expected, especially with basal warming of the glaciers from geothermal flux. Where Cambias & Zeigler [2006] suggested that the sunlit side of a tidally-locked Earth should be about 325 K, Merlis & Schneider's models showed the subsolar region of a tidally-locked Earth to be little more than 300 K.

Merlis & Schneider's models also show little precipitation of rain or snow on the dark side. The details are substantially different in the fast-rotating and slow-rotating cases (see Merlis & Schneider [2010], figure 2), but generally the subsolar region is very rainy (kind of like the tropics, but perhaps more so) and is surrounded by a broad region in which potential evaporation exceeds precipitation (kind of like the Sahara, Arabian, Thar, Namib, Atacama, and Australian deserts in Earth's horse latitudes), surrounded by a huge area on the dark face and in the twilight zone where the net of precipitation minus potential evaporation is mostly positive but small (the significant exception is in the fast-rotating case, east of the subsolar point).

But wait! There's more!

In 2013 Hu and Yang¹⁰ published a paper reporting the results of modelling synchronously rotating planets using a model that included ocean currents and the transportation of heat in them. They assumed a uniform ocean as deep as the average of Earth's ocean, nowhere interrupted by continents nor oceanic shallows, and point out that this leads to stronger and more symmetrical effects than are to be expected on any habitability candidate. That being acknowledged, their models suggest that the transportation of heat from the sunlit side to the dark side of a synchronously rotating Earth would dramatically reduce the difference in temperatures between them — to less than 50 K. Moreover, the sea ice would be thin — 3–5 metres at most, allowing free currents below it and no ice-cap. A tidally locked planet with an equable temperature and plenty of water might even have ice-free oceans on its dark side.


So it all seems to me as though the adjustments to temperatures and hydrographic percentage for tidally-locked planets that are provided in GURPS Space 4th edition, at least inasfar as they are applied to planets with substantial atmospheres and water that are in the Goldilocks zone, are probably too large.

Since permanent human settlement requires (a) photosynthetically active illumination, (b) an excess of precipitation over potential evaporation, and (c) an average temperature between 273 K and 303 K it seems to me that the habitation candidates among synchronously rotating planets are not those that are about as warm as Earth (settled in the twilight zone), but those that are on average a little cooler than Earth, which will be settled in the subsolar region where it rains and where the light is brighter.

[Agemegos]

¹ Joshi, M. M., Haberle, R. M., and Reynolds, R. T., 1997 Simulations of the Atmospheres of Synchronously Rotating Terrestrial Planets Orbiting M Dwarfs: Conditions for Atmospheric Collapse and the Implications for Habitability.

² When the entire planet is colder than what Joshi et al considered, naturally the dark side is colder than they concluded. CO₂ does freeze out on ice moons.

³ The modern consensus seems to be that 10 mbar of CO₂ will not freeze out with Earth-like solar heating, but that 3 mbar is unstable, and may.

⁴ In any inertial frame of reference.

⁵ Humans do not form permanent settlements where the average temperature is above 30 C or below 0 C. The reasons are probably agricultural, and would also rule out habitation where the photosynthetic photon flux density was too low to support the growth of crops.

⁶ Zeigler, J.F. and Cambias, J.L., 2006 Space. Steve Jackson Games, Austin TX.

⁷ Merlis, T.M. and Schneider, T., 2010 Atmospheric dynamics of Earth-like tidally locked aquaplanets.

⁸ I consider it unfortunate that the cases they published involved (a) one rotation per 24 hours, which is a lot faster than we expect for any habitability candidate, and (b) one rotation per 8 640 hours, which is a good deal slower than we expect for any habitability candidate.

⁹ The reason seems to be that Antarctica is isolated from tropical air by the Ferrel cell and polar easterlies, features produced by Coriolis effects. Since the anti-subsolar point in the tidally-locked Earth is not a rotational pole is it not similarly isolated from the global circulation, and there is no pool of cold air trapped on the dark face.

¹⁰ Hu Y and Yang J, 2013 Role of ocean heat transport in climates of tidally locked exoplanets around M dwarf stars.

[AlexanderHowl]

Tide locked worlds are probably unlikely to be habitable by humans outside of the thin twilight zone (this is why orbital resonant worlds are better). The near side will always be covered by clouds due to evaporation and the near side will always dark, so photosynthesis is impossible without technological assistance. The difference in dayside and nightside will also not help.

Now, cooler worlds with tidelocking could potentially allow for more habitable area, but the nightside would be absolutely frozen in the case. With an average atmosphere, the dayside is +12% while the nightside is -20%, meaning that a 240K world would be 271K on dayside and 192K nightside. One problem though is that carbon dioxide snows at the nigh side temperature, so the biosphere would collapse, as there would be no carbon dioxide in the atmosphere to allow for photosynthesis.

[Agemegos]

Originally Posted by AlexanderHowl

Tide locked worlds are probably unlikely to be habitable by humans outside of the thin twilight zone

Really? I just finished explaining why that belief is outdated.

Now, cooler worlds with tidelocking could potentially allow for more habitable area, but the nightside would be absolutely frozen in the case. With an average atmosphere, the dayside is +20% while the nightside is -20%,

Not according to the scientific papers I cited in the OP.

meaning that a 240K world

Earth's average surface temperature is about 288 K.

would be 288K on dayside and 192K nightside.

Merlis & Schneider (op. cit.) used a global circulation model and found that a synchronously rotating Earth would have a temperature a little over 300 K in the middle of the daylit side and about 250 K over a huge stretch of the dark side. And I told you about that in the message you are replying to.

One problem though is that carbon dioxide snows at the nigh side temperature, so the biosphere would collapse, as there would be no carbon dioxide in the atmosphere to allow for photosynthesis.

Dude! Did you even read my post?

The belief that CO₂ will snow out on the dark side was debunked by Joshi et al in 1997. I explained that and both cited and linked my source in the very post that you are replying to.

[Prince Charon]

I have to wonder how many people reply to a thread after only reading the title, and at most skimming the opening post. This is far from the first time I've seen someone appear to have done that.

[Daigoro]

Originally Posted by Agemegos

Since permanent human settlement requires (a) photosynthetically active illumination, (b) an excess of precipitation over potential evaporation, and (c) an average temperature between 273 K and 303 K it seems to me that the habitation candidates among synchronously rotating planets are not those that are about as warm as Earth (settled in the twilight zone), but those that are on average a little cooler than Earth, which will be settled in the subsolar region where it rains and where the light is brighter.

Does that mean that there'd be a gradient of habitable worlds, with a habitable band located anywhere from the subsolar point to the twilight zone, depending on their average temperature (and hydro load, etc.)? Or is there absolutely going to be an arid band around the subsolar pole?

[Daigoro]

Originally Posted by AlexanderHowl

Tide locked worlds are probably unlikely to be habitable by humans outside of the thin twilight zone (this is why orbital resonant worlds are better). The near side will always be covered by clouds due to evaporation ...

The people of the British Isles will be surprised to hear that cloud cover prevents habitation.

[Crystalline_Entity]

Originally Posted by Daigoro

The people of the British Isles will be surprised to hear that cloud cover prevents habitation.

Indeed, it's a little cloudy here today, and I'm still alive :)

I think part of the issue with things like GURPS: Space is that our understanding of the science behind the way planets form and interact with their environment is increasing rapidly as the initial point was so low. Back when I got my first sci-fi RPG in the 1990s no extrasolar planets had been discovered, so the underlying assumption was that every planetary system was like ours, not allowing for things like hot Jupiters. I'm sure I remember reading last year that observations suggest there could be Neptune-size exomoon, which is a common trope in sci-fi but there was no evidence for in reality until now.

I've not had time to read the linked journal articles in the OP, but it's a very interesting topic to me - it'd be a very different alien feel than your typical Earth-like world, and thus an interesting planet type for worldbuilders to consider or characters to visit in an RPG.

[Agemegos]

Originally Posted by Daigoro

Does that mean that there'd be a gradient of habitable worlds, with a habitable band located anywhere from the subsolar point to the twilight zone, depending on their average temperature (and hydro load, etc.)? Or is there absolutely going to be an arid band around the subsolar pole?

I haven't seen or done studies, so I don't actually know.

In the slowly-rotating limit (as illustrated in Merlis & Schneider (op. cit.)) there is a fairly simple pattern of winds and temperatures which seems to be readily understood in terms of a thermally-direct toroidal cell of circulation that is absolutely dictated by the fact that the subsolar region is hotter than what surrounds it. So you're definitely going to get a convergence and humid upwelling at the subsolar point, with a ring of downdrafts some way terminator-wards of it, and a pattern of brisk, relatively dry winds towards the subsolar region. In the case where the torrid zone is too hot for agriculture and permanent human habitation then the band of equable temperatures will be comparatively dry. But that doesn't necessarily mean that all the land will be non-arable. I live in the horse latitudes at 31°5' south, and the area around here is usually pretty green and well-vegetated. The is a bit of semi-arid grassland to the west, and an awful lot of desert west of that, but geographical features interrupt the straightforward patterns of winds and ocean currents — the South Equatorial Current that ought to flow to the west is diverted to the south by a continent in its way, the resulting East Australian current makes the seas here warmer and the air more humid than would be if the entire world were covered by a uniform ocean 5m or 4 km deep.

So I suspect that in the slowly-rotating case you start on a planet 30 K cooler than Earth with an "eyeball" world: only the subsolar region is free from ice. Then as you consider gradually warmer worlds the ice around the subsolar optimum gives way to an expanding ring of arid to semi-arid that is streaked and speckled with fortunate lands where local and geographical effects produce adequate rainfall. Once you get to a point a little warmer than Earth the subsolar region starts to become uninhabitable owing to heat — but there remain some favoured areas in the torrid zone that are cooled by diverted ocean currents etc and remain habitable and wet. At that stage, the zone of habitation is like, say, Australia: largely arid to semi-arid but with some good bits. Eventually the zone of equable temperatures gets pushed towards the terminator. In the twilight zone it isn't as windy, but the low slanting sunlight gets too dim to support agriculture.

Worlds that rotate rapidly enough for the Coriolois effects to be significant (and I don't know how fast that has to be) are more fortunate. The pattern of the winds and rainfall is a lot more complicated, yielding a warm/wet patch that is shaped like a lobster with its tail stretching towards, perhaps even across, the terminator east of the subsolar point. Screw that pattern up with a few mountain ranges and you have an excellent chance of finding pretty much any climate you need, albeit probably in a small area. Take a look at the map of surface temperature on p.4 of the paper by Merlis & Schneider, and the map of potential evaporation minus precipitation on p.5. The fast-rotating cases (on the right of Fig. 1 and Fig. 2 respectively) are charmingly quirky.

[Daigoro]

Originally Posted by Agemegos

⁸ I consider it unfortunate that the cases they published involved (a) one rotation per 24 hours, which is a lot faster than we expect for any habitability candidate, and (b) one rotation per 8 640 hours, which is a good deal slower than we expect for any habitability candidate.

One rotation per 24 hours is fair enough- it allows them to easily compare to Earth, isolating from other factors. And it might be close enough to good candidates. Looking at TRAPPIST-1e, chosen because it's terrestrial size and in the habitable zone (albeit lacking water), it has a 6 day period, so it's Coriolis effect would be within an order of magnitude of Earth's.

Actually, 1d appears it might have watery oceans, despite being inside the habitable zone (however that happens), and it has a 4 day period.

ETA: Another planetary heating mechanism that wasn't mentioned is geothermal heating. This Wikipedia entry suggests that 1d might have geothermal activity to due tidal locking with the primary, so that could be another mechanism for warming the nightside.

ETA2: Agemegos, I refer you to Carone, et al's paper, Stratosphere circulation on tidally locked ExoEarths, for some in-depth discussion of different polar-equatorial transport regimes. It might give you some useful insights.