An Apple Pie from Scratch, Part VIc: Climate: Climate Zones of Exotic Worlds

ESO/M. Kornmesser

(Note: I've been running a series of simulations with ExoPlaSim, the "Climate Explorations" in the Site Map, some of which have turned up rather different results than I've seen here. I'll leave this page here for now and probably update some of it in the future, but check against those models for comparison).

This post is essentially a synthesis of the last two. We’ve seen how atmospheric and oceanic circulation produce climate zones on Earthlike worlds, and we’ve discussed how these factors might change under different forcings, so now let’s see how climate zones will shift under different forcings.

To do this, I’ll take the topography of Teacup Ae (with modification where necessary) and work through the process I laid out last time under a number of different climate scenarios. I won’t hit every possibility (there are so many variations of day length, obliquity, and eccentricity), and in the interest of getting this post out in a reasonable time I also won’t be spending as much time on fine detail as for the “true” map in Part VIb and taking a few shortcuts—in particular, I’m going to largely ignore the effect of elevation on temperature and mark off everywhere above 2000 meters as “highland” to avoid getting lost in all the details there. The idea here is not so much to produce complete climate maps of these worlds, but rather to demonstrate the key differences between them and out “baseline” case in the last post.

And, once again, tidal-locked worlds will get their own discussion in the next section after this one.

Finally, just bear in mind while I’m doing my best to apply lessons from existing research, for many of these cases there’s either been no attempt to model the climate in this level of detail, or what attempts have been made used incomplete or now-outdated models. They also tend to use only Earth’s geography, so it can be hard to judge how generalizable the results area. As such, some of my intuitions regarding how the climate should work may be wrong in these cases for reasons that haven’t been explored by existing models.

 

• Climate State
• Hotter
• Colder
  
• Rotation Rate
• Longer Day
• Shorter Day
  
• Seasonal Forcings
• No Obliquity or Eccentricity
• High Eccentricity
• High Obliquity
  
• Less Water
• In Summary
• Notes 

Back to Part VIb

Climate State

For our first couple of cases, let’s look at what happens to Teacup Ae when greater or lower greenhouse heating causes a shift in average temperatures, with all other factors held constant. In essence, how does our baseline icehouse interglacial world compare with a hothouse or icehouse glacial world, like those that have appeared in Earth’s—and Ae’s—past (incidentally, these simulations might prove useful in determining some of the details of the modern terrain and geology)?

Hotter

First, a hotter world, with no ice caps and comfortable temperatures extending to the poles.

Conveniently, Clima-Sim has a preset in the “Terrestrial” settings designed to replicate such a hothouse state, such as existed during the Cretaceous. This didn’t give me quite the boost in warmth I wanted; remember that Earth’s shifts in climate have been related to orbital changes as well (it may also have been generally more humid in the Cretaceous for reasons not accounted for by Clima-Sim, and generally there are still some open questions about the mechanisms of a hothouse climate). But in this case I don’t want to change too many things at once, so I left most of the orbital settings the same as the baseline “modern” climate (obliquity of 25°, eccentricity of 0.1, axis of obliquity at 120°) and just reduced the simulation’s semimajor axis (which, I remind you, has no effect except to alter total insolation) from 1.01 to 1.00, which produced a global average temperature of 26 °C; warmer by 10 °C compared to the baseline.

Now, if we actually wanted to see what Teacup Ae looked like after a transition to a hothouse, we’d have to adjust sea level as well and introduce some inland seas that would have influence precipitation patterns, but for the purposes of direct comparison I’m ignoring that here.

 

Temperature maps of the hothouse scenario; based on the raw outputs from Clima-sim with no adjustment for currents, topography, or resolution, and sides not filled in. Northern winter on top and summer on bottom, as per usual.

 

The result is, unsurprisingly, a hot world with summer temperatures above 30 °C across most of the planet and even exceeding 50 °C in Agassiz, the southern continent. There are no permanent ice caps, nor even any lowland tundra (I imagine there still might be some mountain glaciers in Steno, the northwest continent, but again I’m not really bothering with highland climates). We can still expect high latitudes to experience winter snows and pack ice on the oceans, but at mid latitudes snow will be rare.

Currents should be broadly similar to the baseline case, but because the temperature gradient between equator and poles is smaller, the actual effect of the currents on temperature will be fairly minor in comparison.

In terms of wind patterns, I've mentioned before that warmer global temperatures will generally widen the Hadley cells, but in extreme hothouse cases like this they’ll dramatically shrink again. As such, I’ve place the subtropical highs further equatorward here in both seasons. This should pull the polar fronts equatorward as well (I’ve seen some suggestions that ice-free poles would cause an additional circulation cell in each hemisphere or just totally disorganized circulation past the subtropical highs, but these are unconfirmed so I won’t overcomplicate matters here). Despite this, the ITCZ is not constrained in its seasonal motion; Earth’s monsoons in hothouse periods appear to have moved as far poleward as our modern ones.

 

ITCZ in blue, subtropical highs in red, winds in grey, and fronts in purple for each season.

Regarding precipitation, I’ve also mentioned that warmer conditions appear to generally increase contrasts between warm and dry regions, but as we reach extreme hothouse conditions the overall humidity will be much higher. Climate simulations of the Cretaceous Earth show a very wet world with few proper deserts.

 

Model of Earth's climate in the Cretaceous. Hay et al. 2018

However, the world at that time had more scattered continents, more inland seas, and fewer tall mountain ranges, so it’s not exactly a “fair” comparison to today. Higher temperatures also means more evaporation, which is accounted for in the Koppen system by defining the precipitation threshold for arid zones to depend in part on average temperature. I don’t expect these factors to all neatly balance out, but I won’t dramatically alter my process for drawing out precipitation.

 

The final result:

 

Naturally, the tropical band is much wider, reaching to 40 ° latitude and now straddling both sides of the desert belts. This eliminates much of the pockets of Mediterranean climate that usually appear on the poleward sides of the deserts, though portions of the tropical zones there can be expected to have similar rainfall patterns (dry summers, wet winters). Humid subtropical zones, on the other hand, extend across the continents to create broad stretches of lush, evergreen forest.

One consistent issue we’ll run into throughout this post is that the positions of Teacup Ae’s major deserts is determined as much by the distribution of mountains as by wind patterns. So though the deserts have moved a bit equatorward here, the difference is not obvious.

Finally, describing this world as having “polar rainforests” may be a bit misleading, but there are indeed oceanic zones reaching to over 70 ° latitude, which could host temperate rainforests, and the fairly moderate humid continental zones reach all the way to the poles. Tundra and subarctic zones may still exist in highlands, but in essence there are no barren polar regions without vegetation as we see on Earth today.

Colder

Next, let’s go in the opposite direction, to a glacial episode in Teacup’s current icehouse state. Not a full snowball, but a period with larger icecaps—on both poles, this time—and less greenhouse heating.

Again, Clima-Sim has a preset for this, to replicate conditions at the Last Glacial Maximum, and with the simulation semimajor axis left at 1.01 this neatly produces a global average temperature of 11 °C; cooler by 5 °C compared to the baseline.

Again, an actual glacial episode would be accompanied by lower sea levels, but I’m ignoring that for the sake of comparison.

 

 

This world has permanent subfreezing temperatures extending far over the northern continents (Steno and Hutton), and an additional ice cap appearing in the south on Agassiz. There should also be large glaciers covering several major mountain ranges, though I’m not showing those here. But there are still substantial warm regions around the equator; an “ice age” does not imply cold, snowy conditions across the whole planet (unless it further degrades to a snowball state).

Currents should have a somewhat stronger effect on temperature here than in the baseline case, but warm currents at high latitudes will be partially blocked by ice sheets.

Regarding wind patterns, there’s not much to write home about on this one; the Hadley cells might be slightly thinner, but not enough to be worth adjusting for here, and some of the winter subtropical highs should be shifted around to the colder regions created by the large glaciers.

 

Overall this should be a drier world, but again this will be somewhat offset by less evaporation. I might be a bit stingier with spreading around precipitation here.

The tropical band has of course shrunk, but not by a lot; the most substantial differences here are in higher latitudes where the ice caps have expanded and mid latitudes where the continental band has moved equatorward and there are substantially more areas of oceanic climate, due to cooler summers. Mediterranean zones have also expanded into many areas formerly occupied by tropical savanna.

Rotation Rate

For our next couple of cases, let’s consider how shifting the day length—and thus, altering global air and ocean circulation—might alter the climate maps. This will only have a minor impact on temperature patterns, but it will profoundly alter precipitation. The external temperature forcings are all roughly the same, so for both of these I’ve just taken the climate bands from the original tutorial and altered them to match the new current patterns, without bothering to redo the temperature maps.

Longer Day

Teacup already has a somewhat longer day (34 hours) than Earth, but not enough to change global circulation much. In this case, let’s increase it to 96 hours—4 Earth days—which will push the boundaries of the Hadley cells far poleward and eliminate the polar cells entirely.

The ocean currents will look similar to the baseline near the equator, but once currents turn equatorward they’ll continue most of the way to the poles rather than crossing the ocean at mid-latitudes, creating large ocean gyres.

This causes heating and cooling effects at mid-latitudes somewhat reversed from that in the baseline case, which shifts around our climate bands somewhat.

I’ve mentioned before that longer days with Earthlike obliquity should cause the ITCZ to move much further with the seasons, so in each season I’ll intentionally place it around 10° latitude further poleward of where I normally would based on the thermal equator. The subtropical highs will similar move much further poleward and equatorward in their expanded ocean gyres.

One might naively expect that in this scenario that moving the subtropical highs further polewards would just shift the position of the desert belts, but these highs are also moving much further with the seasons and have broader fronts between them, so many coastal areas receive rain at least part of the year. Inland, however, there are no longer polar fronts to bring winter rain, so there are much more substantial deserts at high latitudes.

The tropics are also fairly dry, because the ITCZ only briefly passes over in spring and fall. Overall the climate here seems to be marked by sharp contrasts in seasonal rain patterns to a greater extent than on Earth, with widespread dry winters.

Shorter Day

In this case, let’s shorten the days to just 12 hours, half their length on Earth. This should cause the formation of two additional convection cells in each atmosphere.

This will cause the currents to swap coasts a couple more times than the baseline, creating a mess of small and sometimes very elongated gyres. This probably implies weaker transport of heat towards the poles, but I’m not sure exactly how much so I’m not taking it into account too much here.

In reverse to the previous case, I’m purposely constraining the seasonal movement of the ITCZ here, and the subtropical highs similarly don’t have far to move in their small gyres. There are also now two distinct belts of subtropical highs, with a long front dividing them in each hemisphere—an additional convergence zone between the ITCZ and polar fronts. It’s possible such a world would actually have much more chaotic circulation patterns

Given these conditions, precipitation patterns should generally be pretty consistent year-round; there will be smaller regions with torrential rains in one season and droughts in the other.

Looking at the final result, we can see hints of a second desert belt forming near the second belt of subtropical highs, but these belts are pretty thin and there’s enough movement of the neighboring fronts between seasons to keep much of the mid-latitudes somewhat wet, and even the tropics to some extent. The result is more of a patchwork landscape of small deserts rather than large belts.

Seasonal Forcings

For our next few cases, I want to alter the major drivers of seasonal change: obliquity and eccentricity. Each of these has infinite variations over a broad spectrum, and can be combined together to produce an unimaginable range of variation. I remind you that a shift of just 2° in obliquity has pushed the Earth into and out of glacial states, and a shift of just 0.05 in eccentricity has shifted the Sahara from lush grassland to a lifeless desert.

I can’t even begin to account for all the possibilities, so I’ll settle for just tackling a few straightforward “extreme” cases. You can explore various other combinations of your own in Clima-Sim, but it has its limits; I can’t give you hard numbers, but substantially higher obliquity or eccentricity than we see on Earth seems to produce results that strongly diverge from more complex models designed to handle these cases.

No Obliquity or Eccentricity

First, I want to look at a planet with no obliquity and eccentricity; in essence, a planet without seasons.

Clima-Sim models 0 obliquity and 0 eccentricity just fine, but the result is a much cooler planet; so to keep the global average temperate at 16 °C, same as the baseline, I chose the “Cretaceous” preset for greenhouse heating and lowered the simulated semimajor axis to 0.995.

Given that there are no seasons, we only have to look at the temperature and precipitation at one point of time, which could be anywhere in the planet’s circular orbit. What we see is that despite having the same average temperature as the baseline, this world has huge ice caps, but also huge tropical regions. The greater thermal inertia of the oceans no longer factors in to surface temperatures, so there’s also a smoother temperature gradient between equator and poles (though that’ll be less true when I take currents into account)

Wind patterns are fairly straightforward. The ITCZ follows the equator pretty closely; without seasons, thermal inertia doesn’t come into play, and the only differences in temperature there are due to elevation and albedo. I put the subtropical highs firmly in the center of the ocean gyres, at 30° latitude, and I haven’t included any continental highs.

 

Precipitation can be marked out as normally, but without any seasonal shift in winds the desert belts are wider, and there are more areas that are permanently rainshadowed by mountains.

 

The final result is rather bizarre. The continental band has been eliminated entirely, as well as the, Mediterranean, humid subtropical, and tropical monsoon zones. The tropical savanna zones are also smaller, and technically impossible at average temperatures above 29 °C (as the threshold of precipitation for arid zones rises above the threshold between rainforest and savanna zones). Vast rainforests and deserts dominate the tropics and much of the mid latitudes, but at the same time the ice caps are larger than in our “ice age” case.

 

(Note that, while my tutorial in the previous post implied that tropical savanna zones require a seasonal shift in rain, this was a necessary simplification of the formal definition, which uses a set of formulas relating seasonal and annual rain patterns to define monsoon, savanna, steppe, and desert zones; run through the numbers and you'll see that savanna zones can still exist even with no seasonal shift in rains. I didn't account for this in tutorial because with Earthlike seasons, essentially the entire tropical band sees seasonal shifts in rain and there are few areas just moderately wet year-round; but that assumption doesn't apply quite so well here so we have to adjust the methods accordingly.)

Temperature and rain aren’t totally constant, of course; there will be some random variation day-to-day, and longer-term movement of atmospheric waves along the convection cell boundaries, both of which should form some transitional areas between temperate climates and tundra, and between rainforest and steppe. But overall this is a planet of stark contrasts: Dominated by rainforests, deserts, and glaciers, with only thin temperate, tundra, and savanna regions dividing them.

High Eccentricity

Next, I want to look at a case with no obliquity, but higher eccentricity—nothing extreme, but enough to drive global seasons on its own. These seasons will be the same in both hemispheres, and apply pretty uniformly from the tropics to the poles.

In Clima-Sim, I reduced obliquity to 0 and, after some testing, found 0.3 eccentricity to produce reasonable seasonal temperature ranges at mid-latitudes (I also moved periapsis to July 1st, to make it easier to keep track of the seasons). To keep global average temperate at 16 °C, like the baseline, I used the “Last Glacial Maximum” preset to greenhouse heating and increased the simulation semimajor axis to 1.05.

But though the global average temperature across the whole year matches the baseline, it swings from 11 °C in winter to 23 °C in summer.

 

Local temperatures too see some extreme variation. Clima-Sim predicts tropical temperatures exceeding 60°C in summer, though it’s hard to say how accurate the simulation is in this case. In contrast to the previous scenario, thermal inertia now plays a much bigger role in this world, as it keeps coastal areas in the tropics warm through the globally cool winters.

Because the seasons are not symmetric across the hemispheres, the ITCZ doesn’t shift much as a whole with the seasons, but sections of it move more towards the center of the tropical landmasses in summer and then out to sea in winter. In presuming that the subtropical highs are still shifting poleward and equatorward with the seasons due to shifts in heat transport and the strength of the Hadley cell, but I could very well be wrong about that.

 

There really isn’t much detailed modelling on the climate of this sort of world, so I’m advancing under the shaky assumption that precipitation works the same as in our baseline. It may be that the extreme summer heat in the tropics would actually inhibit rainfall in ways I’m not accounting for here.

 

This scenario can be largely summed up as a world with stronger seasons at the equator but weaker seasons at the poles, compared to Earth. At low latitudes, rainforests dominate many of the coasts, but inland regions swing between hellish summers and cool winters. Mid-latitudes are actually fairly similar to Earth, but at high latitudes there are thinner continental bands dividing the temperate regions from the vast icecaps.

High Obliquity

For our final look at seasonal forcings, let’s increase obliquity all the way to 90°, tipping the planet “on its side”. The poles will spend half the year with constant sunlight, facing directly towards the star in midsummer, but then spend the other half of the year in total darkness. Meanwhile, the equator will pass through an odd twice-yearly seasonal cycle of its own.

 

Unfortunately, Clima-Sim can’t give us much help here. It simply wasn’t designed to handle the intense and rapidly shifting heat transport involved here. We’ll have to work out the temperature and precipitation patterns based entirely on intuition and guesswork, though I’ll be using this paper as a rough guide to follow. But I’ll be honest from the outset: There’s a lot more guesswork going into this one compared to the previous scenarios.

As I’ve mentioned before, there are multiple ways that the ice distribution and atmospheric circulation of such a planet might play out, depending on the history and subtle factors of heating and rotation. To keep things interesting, let’s say in this case that there’s at least some permanent ice at the equator, and that atmospheric circulation is flipped, carrying heat out from the poles towards the equator.

Ocean circulation on high-obliquity planets hasn’t been modelled in detail, but the reference seems to show warm currents from the poles following the eastern coasts of the continents, and cold currents from the equator following the western coasts, together forming one large gyre in each hemisphere in each ocean; as we’ll see in a bit this makes some sense given the nature of air circulation in this scenario. 

 

Without Clima-Sim data, I’ll skip over temperature and jump straight to marking out the climate bands, using the reference and these currents to guess at where the critical boundaries in seasonal temperature will be.

 

I’ll start out by marking out the ice belt, which will dominate the continents near the equator and reach towards large mountains at mid-latitudes, but remain fairly thin over the oceans, with some gaps in the permanent ice cover.

 

 

Next, we’ll mark out the temperate band (yes I’m going in a weird order), which occurs as a thin coastal strip at high latitudes, extending further equatorward near warm currents and occupying large areas near the poles, but never extending too far inland. Most of this will be hot-summer temperate—few areas that stay above freezing in winter don’t reach over 22 °C in summer—though small areas of cold-summer temperate can exist in highlands and near cold currents at lower latitudes.

 

 

Based on some models, it is actually possible for areas near the poles to remain above 18 °C throughout the winter, season-long night be damned, and thus you can add some tropical zones if you like. But it appears difficult to get both that and an equatorial ice belt at the same time, so I’ve excluded the tropical band in this case.

 

Finally, remaining interior areas can be filled in with continental zones; mostly humid continental at high latitudes and substantial strips of subarctic near the tundra.

 

 

Based on the reference paper, atmospheric circulation seems to be a lot simpler than on Earth, with no subtropical anticyclones or ITCZ: Instead, there is a broad band of westerly winds at low latitudes that oscillates north and south with the seasons, reaching roughly 45° latitude in summer and 15° in winter—moving somewhat further poleward in summer and equatorward in winter over land, but not by much. Easterly winds dominate in higher latitudes, though they’re much weaker; a convergence zone exists between these bands of winds, though it only has a significant impact in summer. There should probably also be winds converging on the poles, but those cover a pretty small area of the globe.

 

 

The precipitation patterns here are also a little different from the baseline, not really so easily categorized as just “summer” or “winter” rains, but I’ll try to summarize them in this way just for convenience.

 

First, there is an early-summer “pulse” of rain, occurring mostly equatorward of the large convergence zone in the summer hemisphere, downwind of the onshore westerly winds there but not extending too far inland.

 

Then, there is a late-summer pulse, bringing heavy rains to areas mostly poleward of the convergence zone, downwind of onshore easterly winds and extending far inland.

 

 

Finally, there are some rains in the winter hemisphere on coasts with onshore easterly winds, extending further inland near the poles but never very far.

 

 

The result doesn’t so much create desert belts, but moreso allows for deserts to more readily appear nearer to the equator.

 

 

The final result gives us a pretty odd world with temperate coastlines—overwhelmingly humid subtropical and continental or desert interiors. Extreme seasons are the norm, and really this climate scheme just can’t really express just how bizarre the conditions are—especially in the continental interiors near the poles, which cycle between hellish summers, biblical fall downpours, and bitterly cold winters. I imagine that complex life and civilization would be much more concentrated towards the coasts on such a world.

Less Water

For the last of these cases, let’s look at what happens to a typical “desert” planet; one where, instead of disconnected continents in a global ocean, there are disconnected seas in a global landmass.

To do this I took my Teacup map, shifted all the existing land up by 1 kilometer, and then drew in some rough coastlines for smaller seas in each of the major ocean basins.

 

I then ran this new topography through Clima-Sim, with some quick guesses at the surface biomes, the same orbital parameters as the baseline, and the “Cretaceous” preset for greenhouse heating to keep the global average at 16 °C despite the greater global albedo (though it seasonally swings between 12 °C and 22 °C, because its lower thermal inertia makes it more sensitive to forcing by Teacup’s eccentricity).

 

The lack of any big oceans to moderate temperature has a pretty clear impact; There are huge seasonal temperature swings across the planet, and this scenario actually has the coldest winters at the poles of any we’ve discussed today, plunging to below -60 °C in the highlands formerly known as Steno.

As a result we can expect the ITCZ to move pretty far with the seasons, and for the Hadley cells to be pretty broad and strong. Past that, it’s hard to say much about atmospheric circulation patterns, because our established “rules” presume the presence of large oceans. The largest seas will probably have at least some internal circulation, forming gyres that can form subtropical highs.

Other than that, I have to assume that cool areas formed by smaller lakes in summer or highlands in winter will cause the formation of other subtropical highs

There’s still enough water on this world to form substantial areas of high precipitation, but overall I expect they won’t extend as far inland here as in the baseline case, even near the large oceans; moist air coming off the seas will be more rapidly diluted as it mixes in with dry air from the vast deserts.

 

The result is a world in which deserts dominate, but aren’t ubiquitous; patches of rainforest still exist, and there are pretty broad areas near all the major seas that could host forests, complex ecologies, farmland, and developed societies. Indeed, the total area of hospitable land isn’t all that much less here than in the baseline case. The main difference is that these areas are now mostly divided by deserts, not seas, which will have implications for the development of life, the rise of civilizations, and later patterns of trade and power; It’s generally easier to send a trade expedition or an invasion force across an ocean than across a desert this broad. The narrow corridors of savannah and steppe connecting these seas may become strategic chokepoints in much the same way that ocean straights are for Earth.

In Summary

• Climate states shift the boundaries between all zones, but have a bigger impact closer to the poles.
• Hothouse worlds have broad tropical bands, temperate zones reaching almost to the poles, and no ice caps.
• Ice age worlds have thing but still present tropical bands and large ice caps.
• Rotation Rate primarily shifts the position of major deserts, and alters the pattern of heating by ocean currents.
• Slow-rotating worlds have broad deserts at high latitude.
• Fast-rotating worlds have patches of desert scattered across the world.
• Seasonal forcings can dramatically increase or decrease the contrasts between areas at different latitudes.
• Non-seasonal worlds have vast rainforests, deserts, and ice caps, with fairly small transitional zones between them.
• Eccentric-only worlds have stronger seasonal shifts in the tropics, but relatively weaker seasons at the poles.
• High-obliquity worlds have warmer conditions at the poles than at the equator, and extreme weather conditions in the interiors of continents.
• Dry worlds have vast deserts, but can still have large wet areas with hospitable conditions.

These scenarios have all been put together with a lot of guesswork (especially the last two) and it’s very possible that better climate modelling of exotic worlds in the future will show much of my work here to be completely wrong. They’re best taken as suggestions of the effects various climate forcings might have, rather than strict templates. That said, I think there are a couple of robust patterns to bear in mind when working on any sort of even vaguely Earthlike world:

• Habitable worlds are all fairly diverse in their surface conditions, and in general you can’t describe any such world with a single adjective: Hot worlds still have cool areas, cold worlds have hot areas, and dry worlds have wet areas.
• The climate patterns we see on Earth are the result of the very specific climate forcings it currently experiences. Subtle shifts in these forcings can shift climate patterns around in dramatic ways, contrary to any intuitions we may have about how climate zones “must” be arranged.

Notes

“Getting this post out in a reasonable time” is something I wrote a long ways backs and looks too hilarious in hindsight not to have included in the final draft (If you’re reading this far in the future and confused, check the publishing dates for this and the previous post).

 

The header image is actually a depiction of iron rain on a brown dwarf, but it looked cool and nonspecific enough that I figured it would work.

 

Buy me a cup of tea (on Patreon)

Part VId