[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.